IAHR MONOGRAPH

Methods for the Quantitative Assessment of Channel Processes in Torrents (Steep Streams) INTERNATIONAL ASSOCIATION FOR HYDRO-ENVIRONMENT ENGINEERING AND RESEARCH

Dieter Rickenmann IAHR MONOGRAPH

An important part of the risk management of natural hazards in mountain regions concerns the
hazard assessment and the planning of protection measures in steep headwater catchments, i.e.
torrent control and slope stabilization. Torrent processes in steep channels have their rightful

Methods for the Quantitative Assessment

in Torrents (Steep Streams)

Methods for the Quantitative Assessment of Channel Processes
place among the various alpine natural hazards and the corresponding control measures
have a long tradition in the European alpine countries. In the planning and execution of such
measures, professional experience has been of paramount importance. This experience was
based primarily on observations made during and after hazardous events, as well as on regular of Channel Processes in Torrents
field visits in the catchments of a steep headwater stream. Quantitative measurements, e.g. of
the discharge and of the eroded and deposited solid materials, have been increasingly carried (Steep Streams)
out only in the last decades. This set the basis to develop and improve quantitative methods to
predict flow hydraulics, bedload transport and debris flows in torrent catchments.

This publication presents an overview of methods to quantify channel processes in steep

catchments. The understanding and the quantitative description of channel processes provides
an essential basis for the planning of protection measures.

Rickenmann
Dieter Rickenmann

an informa business
 Methods for the Quantitative
Assessment of Channel Processes
in Torrents (Steep Streams)

RICKENM-Book.indb i 2/23/2016 7:38:56 AM

 IAHR Monograph

Series editor
Peter A. Davies
Department of Civil Engineering,
The University of Dundee,
Dundee,
United Kingdom

The International Association for Hydro-Environment Engineering and Research (IAHR), founded
in 1935, is a worldwide independent organisation of engineers and water specialists working in fields related to
hydraulics and its practical application. Activities range from river and maritime hydraulics to water resources
development and eco-hydraulics, through to ice engineering, hydroinformatics and continuing education and
training. IAHR stimulates and promotes both research and its application, and, by doing so, strives to contribute to
sustainable development, the optimisation of world water resources management and industrial flow processes.
IAHR accomplishes its goals by a wide variety of member activities including: the establishment of working groups,
congresses, specialty conferences, workshops, short courses; the commissioning and publication of journals,
monographs and edited conference proceedings; involvement in international programmes such as UNESCO,
WMO, IDNDR, GWP, ICSU,The World Water Forum; and by co-operation with other water-related (inter)national
organisations. www.iahr.org

RICKENM-Book.indb ii 2/23/2016 7:38:56 AM

 Methods for the
Quantitative Assessment
of Channel Processes in
Torrents (Steep Streams)

Dieter Rickenmann
Research Unit Mountain Hydrology and Torrents,
Swiss Federal Research Institute WSL, Birmensdorf,
Switzerland

RICKENM-Book.indb iii 2/23/2016 7:38:56 AM

 CRC Press/Balkema is an imprint of the Taylor & Francis Group, an informa business
© 2016 Taylor & Francis Group, London, UK
Typeset by V Publishing Solutions Pvt Ltd., Chennai, India
Printed and Bound by CPI Group (UK) Ltd, Croydon, CR0 4YY
All rights reserved. No part of this publication or the information contained
herein may be reproduced, stored in a retrieval system, or transmitted in any
form or by any means, electronic, mechanical, by photocopying, recording or
otherwise, without prior permission in writing from the publisher. Innovations
reported here may not be used without the approval of the authors.
Although all care is taken to ensure integrity and the quality of this publication
and the information herein, no responsibility is assumed by the publishers nor
the author for any damage to the property or persons as a result of operation
or use of this publication and/or the information contained herein.
Published by: CRC Press/Balkema
P.O. Box 11320, 2301 EH Leiden, The Netherlands
e-mail: [email protected]
www.crcpress.com – www.taylorandfrancis.com
Library of Congress Cataloging-in-Publication Data
Applied for

ISBN: 978-1-138-02961-3 (Hbk)

ISBN: 978-1-4987-7662-2 (eBook PDF)

RICKENM-Book.indb iv 3/4/2016 10:32:15 AM

 About the IAHR
Book Series

An important function of any large international organisation representing the

research, educational and practical components of its wide and varied membership is
to disseminate the best elements of its discipline through learned works, specialised
research publications and timely reviews. IAHR is particularly well-served in this
regard by its flagship journals and by the extensive and wide body of substantive
historical and reflective books that have been published through its auspices over the
years. The IAHR Book Series is an initiative of IAHR, in partnership with CRC Press/
Balkema – Taylor & Francis Group, aimed at presenting the state-of-the-art in themes
relating to all areas of hydro-environment engineering and research.
The Book Series will assist researchers and professionals working in research and
practice by bridging the knowledge gap and by improving knowledge transfer among
groups involved in research, education and development. This Book Series includes
Design Manuals and Monographs. The Design Manuals contain practical works, the-
ory applied to practice based on multi-authors’ work; the Monographs cover refer-
ence works, theoretical and state of the art works.
The first and one of the most successful IAHR publications was the influential
book “Turbulence Models and their Application in Hydraulics’’ byW. Rodi, first pub-
lished in 1984 by Balkema. I. Nezu’s book “Turbulence in Open Channel Flows’’,
also published by Balkema (in 1993), had an important impact on the field and, dur-
ing the period 2000–2010, further authoritative texts (published directly by IAHR)
included Fluvial Hydraulics by S. Yalin and A. Da Silva and Hydraulicians in Europe
by W. Hager. All of these publications continue to strengthen the reach of IAHR and
to serve as important intellectual reference points for the Association.
Since 2011, the Book Series is once again a partnership between CRC Press/
Balkema – Taylor & Francis Group and the Technical Committees of IAHR and
I look forward to helping bring to the global hydro-environment engineering and
research community an exciting set of reference books that showcase the expertise
within IAHR.
Peter A. Davies
University of Dundee, UK
(Series Editor)

RICKENM-Book.indb v 2/23/2016 7:38:57 AM

This page intentionally left blank
 Table of contents

Abstract xi
Preface xiii

1 Introduction 1
1.1 Torrent processes and hazard assessment 1
1.2 On the contents of the present publication 2

2 Flow resistance in gravel-bedded streams and torrents 5

2.1 Logarithmic flow law 5
2.2 Empirical flow resistance laws (power laws) 8
2.3 Variable power law 10
2.4 Partitioning of the flow resistance 15

3 Fluvial bedload transport 21

3.1 Characterization of torrents and mountain rivers 21
3.2 Sedimentological parameters 24
3.3 Determination of grain size distribution and
of characteristic grain sizes 25
3.3.1 Volume analysis of sediment 25
3.3.2 Analysis of surface bed material 27
3.3.3 The Line by Number Analysis (LNA) of the surface
layer in streams with coarse bed material 27
3.4 Bedload transport in steep channels 30
3.4.1 Fluvial bedload transport 32
3.4.2 Start of mobilization 34
3.4.3 Approaches for calculating bedload transport 40
3.4.4 Consideration of energy losses 44
3.4.5 Transition to debris flood and debris flow 49
3.4.6 Deposition slope behind check dams 51
3.4.7 Comments on the estimation of the solids transport 51

RICKENM-FM.indd vii 3/7/2016 2:33:19 PM

 viii Table of contents

3.5 Driftwood in torrents and mountain rivers 53

3.5.1 Flood hazards associated with driftwood 53
3.5.2 Origin and amount of large wood in stream channels 54
3.5.3 Transport of driftwood and logjam hazards 55
3.6 Critical channel cross-sections and potential deposition 56
3.7 Numerical simulation models 57

4 Debris flows 59
4.1 Properties of debris flows 59
4.2 Important elements of the process and hazard assessment 65
4.3 Occurrence of debris flows 66
4.3.1 Predisposition for debris flow occurrence 66
4.3.2 Triggering conditions 67
4.4 Empirical approaches to characterize the flow
and deposition behavior 72
4.4.1 Maximum discharge 73
4.4.2 Flow velocity 75
4.4.3 Total runout distance 76
4.4.4 Deposition length of the fan 78
4.4.5 Impact forces 78
4.5 Models for the simulation of debris flows 80
4.5.1 Empirical approaches 80
4.5.2 Simple analytical methods 80
4.5.3 Numerical simulation models 81
4.6 Scenarios and deposition in the area of the fan 86
4.6.1 Uncertainty and scenarios 86
4.6.2 Traces of earlier deposits on the fan 86
4.6.3 Simple assessment of depositional behavior 87
4.7 Final remarks 88

5 Magnitude and frequency of torrent events 91

5.1 Empirical approaches to estimate the magnitude of an event 93
5.2 Field-based estimate of the event magnitude 95
5.3 Combined method for estimating the event magnitude 97
5.4 Flood runoff and debris-flow occurrence 98
5.5 Frequency of torrent events 98
5.5.1 Debris-flow events 99
5.5.2 Fluvial sediment transport events 102
5.6 General remarks on the estimate of sediment loads
in torrent catchments 102

RICKENM-Book.indb viii 2/23/2016 7:38:57 AM

 Table of contents ix

6 General remarks on hazard assessment of channel processes

in torrents 105
6.1 Reproducibility of the process analysis and hazard assessment 105
6.2 Procedure of hazard assessment and important aspects 105
6.2.1 Basic data 106
6.2.2 Preparation of hazard map 106

References 111
List of symbols 131
Subject index 135

RICKENM-Book.indb ix 2/23/2016 7:38:57 AM

This page intentionally left blank
 Abstract

An important part of the risk management of natural hazards in mountain regions

concerns the hazard assessment and the planning of protection measures in steep
headwater catch-ments, i.e. torrent control and slope stabilization. This publication
presents an overview of methods to quantify channel processes in steep catchments.
The understanding and the quantitative description of channel processes provides an
essential basis for the planning of protection measures. In the European Alps, chan-
nel processes are mostly triggered by rainfall events and associated runoff processes.
Apart from possible flood hazards during an intense rainstorm event, a lot of damage
is often caused by fine and coarse sediment which is entrained either in the form of
fluvial bedload transport or of a debris flow. Typically, the damage increases with the
total amount of sediment transported to the fan during an event, particularly if the
water and the sediments leave the channel on the fan. This document mainly discusses
the topics of flow resistance, bedload transport, debris flows and the relation between
magnitude and frequency of torrential sediment events.
A first focus is put on the calculation of flow resistance in steep channels. Flow
resistance is shown to increase considerably in steep channels which are often char-
acterized by very irregular bed morphology and large-sized sediment particles such as
boulders and pebbles. Together with limited runoff in small catchments, this produces
relatively shallow flow depths. For these conditions, some flow resistance approaches
used for flatter streams and rivers are not valid. A recently-developed, new flow resist-
ance law is presented, and a quantitative procedure is introduced which allows to
account for high flow resistance in bedload transport calculations.
A second key aspect concerns fluvial bedload transport in steep streams. Steep
torrent channels show differences from flatter mountain rivers. Grain size analysis is a
prerequisite for the calculation of bedload transport. Several formulae are introduced
which may be used for the prediction of bedload transport for a given hydrograph.
The quantification of three main elements is discussed, namely initiation of particle
motion, transport rate, and accounting for high flow resistance. A serious complica-
tion during a flood event may be the entrainment of large woody debris, which may
lead to clogging at critical channel locations. Erosion and aggradation of sediment
may also become a crucial process during a flood event.
As a third core area, debris flows and important elements for its hazard assess-
ment are pre-sented. The occurrence of debris flows is discussed in terms of the pri-
mary mechanisms and of triggering rainfall conditions. Empirical and semi-empirical
equations are introduced to es-timate the main parameters characterizing the flow

RICKENM-Book.indb xi 2/23/2016 7:38:57 AM

 xii Abstract

and deposition behavior of debris flows. Simulation tools are presented, which may
be used primarily to estimate the potentially-affected areas on the fan as well as the
flow dynamics. A geomorphic assessment of the natural fan surface can provide indi-
cations about the process behavior including, for example, the runout distance of
former events.
The last focus is on the magnitude and frequency of torrential sediment events.
Apart from historic documents, a field-based geomorphic assessment is recommended
to arrive at a good estimate of a future event magnitude. A recently-developed pro-
cedure is introduced which combines a field assessment with a GIS-based analysis of
other factors that may be relevant for sediment supply to channel system and for sedi-
ment entrainment along the channels during a rainstorm event. A study from several
Swiss headwater catchments is pre-sented which identified typical patterns in the rela-
tions of the magnitude and frequency of torrential sediment events.

RICKENM-Book.indb xii 2/23/2016 7:38:57 AM

 Preface

This book is based on lecture notes which were developed for the courses “Natural
Hazards” at the University of Natural Resources and Life Sciences (BOKU), Vienna,
Austria, and “Management of Torrents and Hillslopes” at the Swiss Federal Insti-
tute of Technology, Zurich, Switzerland. The quantitative assessment of channel pro-
cesses in torrents has always been an important topic of my professional activities,
both in research and in teaching, while I was employed at the University of Natural
Resources and Life Sciences in Vienna and at the Swiss Federal Institute for For-
est, Snow and Landscape Research (WSL) in Birmensdorf, Switzerland. Without the
support of these institutions, this book would not have been realized. Some parts of
earlier versions of the German course notes were developed in collaboration with
my BOKU colleagues Dr. Michael BRAUNER and Dr. Roland KAITNA, and the
most recent German versions were checked by my WSL colleagues Christian RICKLI
and Christoph GRAF. This book is essentially a translation of the following docu-
ment published in German, with some updates and minor modifications: “Methoden
zur quantitativen Beurteilung von Gerinneprozessen in Wildbächen”, Swiss Federal
Institute for Forest, Snow and Landscape Research, WSL Berichte, Nr. 9, 105p.
(http: www.wsl.ch/publikationen/pdf/13549.pdf). I also would like to acknowledge
the translation work of Edward G. PRATER, Bubikon, Switzerland. Peter DAVIES,
University of Dundee, UK, supported the publication in his role as IAHR Monograph
Series Editor and made a last check of the English text.

RICKENM-Book.indb xiii 2/23/2016 7:38:57 AM

This page intentionally left blank
 Chapter 1

Introduction

1.1 TORRENT PROCESSES AND HAZARD ASSESSMENT

In European countries the term “torrent” refers typically to steep channels in Alpine
headwater catchments, with channels steep enough that debris flows can occur apart
from fluvial sediment transport. According to this definition, such catchments are
associated with drainage areas of less than about 25 km2 and, typically, with mean
channel gradients steeper than 10% (Rickenmann & Koschni 2010). This term
has not often been used in the recent English scientific literature. As pointed out by
Slaymaker (1988), the English expression “debris torrent” (originally used in the US
Pacific Northwest) was associated with a debris-flow event, which is in contrast to the
European meaning of the term referring to a steep stream channel.
Torrent processes in steep channels have their rightful place among the various
alpine natural hazards and the corresponding control measures have a long tradition
in the European alpine countries. In the planning and execution of such measures,
professional experience has been of paramount importance. This experience was
based primarily on observations made in earlier torrent events as well as on regular
field visits in the catchments of a steep mountain stream. Quantitative measurements,
e.g. of the discharge and of the eroded and deposited solid materials, have been
increasingly carried out from about the 1990s. Thus, from before that time, there are
very limited data on quantitative methods to describe channel processes.
In the meantime, the assessment of torrent processes is based increasingly and
primarily on quantitative approaches but also on numerical simulation models. The
quantitative description of the channel processes in torrents is based in many cases
on earlier and comprehensive investigations to describe similar processes in relatively
flat channels or larger catchments. Thus, for example, for about one hundred years
systematic investigations on bedload transport have been carried out based on
measurements in hydraulics testing laboratories and in natural channels. In more
recent times, systematic measurements of bedload transport in steep channels have
and are being carried out both in the laboratory and in the field. These studies help
improve our knowledge of the processes involved in torrents and show to what extent
earlier methods can be adopted or adapted. In this context it was shown, for example,
that the flow behavior or the hydraulics in steep channels exhibit differences compared
with the behavior in flatter channels, which have to be taken into account in the
analysis of bedload transport in steep channels. Likewise, in studying the behavior

RICKENM-Book.indb 1 2/23/2016 7:38:57 AM

 2 Methods for the quantitative assessment of channel processes in torrents (steep streams)

of debris flows, systematic quantitative measurements using automatic monitoring

devices have only been performed in Europe since the 1990s.
In Switzerland, new legal requirements came into force in the 1990s to deal with
natural hazards (Buwal/Bww/Brp, 1997; Bww/Brp/Buwal, 1997). To produce, for
example, hazard maps and flood protection concepts, the process intensities and
potential hazard areas have to be quantified. An examination of technical reports
on hazard maps and flood protection concepts in the case of torrents and bedload
transport processes showed that a comparative assessment is often difficult since,
for the production of hazard maps (and, especially, for the process assessment), very
different methods are sometimes employed. Furthermore, the methods applied and the
basic data, such as the reported input or model parameters used, are, unfortunately,
sometimes insufficiently documented. After Kienholz (1999) and Kienholz et al.
(2002) technical correctness and clarity in reporting (i) the selected methods and (ii) the
assumptions made are the two most important requirements of technical reports with
regard to the assessment of natural hazards. The choice and documentation of the
procedures adopted is admittedly difficult, since there are few comprehensive surveys
of existing methods for the process assessment of natural hazards in torrents–above
all, regarding the processes of bedload transport and debris flow. The aim of the
present publication is to support the quantitative description of torrent processes as
well as the determination of the key parameters and the choice and documentation
of the methods.

1.2 ON THE CONTENTS OF THE PRESENT

PUBLICATION

Chapter 2 considers the flow resistance in gravel-bedded streams and torrents. It is

shown, thereby, that, in the case of steep channels with small flow depths compared
with a characteristic size of the surface bed sediment (including clasts up to boulder
size), the flow resistance increases considerably and the well-known approaches for
flat channels (Manning-Strickler equation, logarithmic flow law) are no longer
applicable. The chapter also report show the additional flow resistance can be taken
into account quantitatively with regard to bedload transport calculations.
In chapter 3, which considers fluvial bedload transport, the behavior of torrents
is introduced and the differences compared with mountain streams are pointed out.
Methods are then presented for the characterization of the granular material in the
stream bed, such as its grain size distribution. In the analysis of the bedload transport
formulae the quantification of the following three main elements is discussed: start
of transport, bedload transport rate and consideration of energy losses due to the
additional flow resistance in steep channels. In addition, the problem of driftwood
debris is discussed, as well as possible hazard locations as a result of erosion or
deposition.
Chapter 4 considers firstly the properties of debris flows and presents important
elements of process and hazard assessment. The conditions under which debris flows
can occur are then discussed. Quantitative empirical or semi-empirical methods
are presented, with the determination of important parameters of the flow and
depositional behavior. In the presentation of simulation models, the focus is on

RICKENM-Book.indb 2 2/23/2016 7:38:57 AM

 Introduction 3

the flow and depositional behavior on the torrent`s fan. Finally, the importance of
scenarios of torrent processes and of traces of deposition on the fan is indicated.
In chapter 5 the magnitude and frequency of torrent events are considered. The
importance of a field-based estimate of the size of the event is emphasized. Following
this, a combined method for estimating the size of an event is presented. Finally, in the
discussion of the frequency of torrent events, attention is drawn to the considerable
importance of historical data, with a presentation of an important study of debris
flow activity in typical Swiss torrent catchments.
Chapter 6 contains a short summary of important aspects that should be taken
into account in the hazard assessment of torrent processes.

RICKENM-Book.indb 3 2/23/2016 7:38:57 AM

This page intentionally left blank
 Chapter 2

Flow resistance in gravel-bedded

streams and torrents

Flow resistance is a measure of the friction between the water and the base of the
channel and the slopes of the banks. For a given flow depth or discharge, flow
resistance laws allow the determination of the mean flow velocity in the channel as
a function of the channel geometry and the bed roughness. The quantification of the
flow resistance is also important for the calculation of bedload transport.

2.1 LOGARITHMIC FLOW LAW

For the determination of the flow behavior in open channels, the universal logarithmic
flow law, developed by Colebrook-White for the hydraulics of pipe flow, and the
Darcy-Weisbach friction coefficient can be applied (Chanson, 2004). In the universal
flow resistance law Eq. 2.1 with the von Karman constant, the logarithmic part
contains a quotient of an integration constant times the flow depth and the equivalent
roughness ks. As Fig. 2.1 shows, the vertical velocity distribution in the case of steep
and rough channels (left) sometimes deviates substantially from the logarithmic
velocity law (right). Different modifications to the original distribution law attempt
to take this phenomenon into account.

8 1 ⎛ a h⎞
= ln (2.1)
f κ ⎜⎝ ks ⎟⎠

V V 8
= = (2.2)
v* ghS f

Here f = friction coefficient after Darcy-Weisbach as defined by Eq. 2.2, κ = von

Karman coefficient (= 0.4), h = flow depth, ks = equivalent roughness height (“sand
roughness”), a = coefficient (frequently a = 12), V = mean flow velocity, g = gravita-
tional acceleration, S = channel slope (or friction slope) (in all equations in this publi-
cation S is expressed in the unit [m/m] and not in [%]), v* = (ghS)0.5 = shear velocity,
ks = const. Dx, and Dx = characteristic grain size for which x% of the material is finer.
The characteristic grain size of the channel bed refers in all flow formulae to the grain
size distribution of the surface material or of the armor layer. In gravel-bed streams
the relative flow depth is typically defined as h/D84.

RICKENM-Book.indb 5 2/23/2016 7:38:57 AM

 6 Methods for the quantitative assessment of channel processes in torrents (steep streams)

a) Roaring River (USA) b) EPF-Lausanne flume

S = 0.035 m/m S = 0.03 m/m
D84 = 223 mm D84 = 28.8 mm
D84/D50 = 2.05 D84/D50 = 1.3
h/D84 = 2.46 h/D84 = 1.74

1
Relative flow depth

0.5

0
0 1 2 3 0 0.5 1 1.5
Flow velocity (m/s)

Figure 2.1 Velocity distribution in steep channels for shallow flows (small relative flow depth).
(a) Non-uniform grain size distribution, Roaring River, Colorado, USA. (b) Uniform grain
size distribution, laboratory flume at EPF Lausanne. Modified from BATHURST (1993).

Table 2.1 Empirical derivation of the equivalent sand roughness ks for sand/gravel beds, gravel beds
and for rough channels.

ks = D65 EINSTEIN (1942) sand/gravel bed

ks = D90 GARBRECHT (1961) sand/gravel bed
ks = 3.5 D84 HEY (1979) gravel bed
ks = 4.5 D50 THOMPSON & CAMPELL (1979) cobble/boulder bed

As equivalent roughness ks, the natural surface roughness is understood as being

given by “grains of constant size with the densest possible packing” (Schröder
1994). It should be noted that, according to this definition, the actual grain size is
only of secondary importance for the determination of the grain roughness. Some
empirically-determined relations between ks and Dx are listed in Table 2.1.
The flowing water separates at the roughness elements of the bed and the turbu-
lence that is thereby initiated finally disappears due to energy dissipation. The bed
structure is also of importance for the determination of the flow resistance, besides the
height of the roughness elements ks. Thus, a derivation of ks from the characteristic

RICKENM-Book.indb 6 2/23/2016 7:38:57 AM

 Flow resistance in gravel-bedded streams and torrents 7

grain size diameter can only be determined from empirical, functional relationships
and subsumed values of geometrical roughness, packing and arrangement of the
roughness elements. These functional relationships can only, therefore, give an indi-
cation of the actual physical behavior.
For the determination of the flow resistance, based on the logarithmic resistance
law, various formulations have been proposed for gravel-bed streams (Keulegan
1938; Hey 1979; Bathurst 1985; Smart & Jäggi 1983).

V ⎛ h⎞
= 6.25 + 5.62 log ⎜ ⎟ (2.3)
v* ⎝ ks ⎠ Keulegan (1938)

V ⎛ h ⎞
= 4 + 5. log ⎜ (2.4)
v* ⎝ D84 ⎟⎠ Bathurst (1985), natural gravel,
plane bed, 0.4% < S < 9%, h/D84 < approx. 7–10

⎡ h ⎤
V ⎢ −0.05
⎢⎣
⎥
S ⎥⎦ ⎛ h ⎞
= 5 .7 D
e log ⎜ 8.2 (2.5)
v* ⎝ D90 ⎟⎠ Smart & Jäggi (1983),
flume, plane bed, S < 20%

As many investigations have shown, there is a functional relationship between

flow depth, equivalent roughness (ks) and channel slope. In particular, the resistance
behavior changes at a relative flow depth of about 4 to 5. After Dittrich (1998),
a layer with a reduced velocity distribution develops in the channel-water contact
region above very rough beds. The layer is called the roughness sub-layer. This situ-
ation results in an S-shaped deformation of the vertical velocity distribution, which
results in an increased stabilizing velocity reduction in the region of the bed and thus
a reduction of shear stress, but at a certain distance from the bed the velocity increases
(Rosport 1998). Bezzola (2002) attempted to improve the description of the resist-
ance behavior of the bed through an increased influence of the height yR. of the lower
roughness layer After Bezzola (2002), the lower roughness layer depends on the
shape, density and exposure of the roughness elements, but not on the relative flow
depth; the author proposed to quantify its thickness as a function of D90 (Table 2.2).
In general, in the case of narrow channel cross-sections with W/h < approx. 10
(W = channel width), the flow depth is often replaced by the hydraulic radius R,
such that the wall or bank friction is taken into account. The hydraulic radius R is

Table 2.2 Thickness of the lower roughness layer after BEZZOLA (2002).

Roughness element Height of the lower roughness layer yR

Uniform size, spherical 0.5 D90

Uniform size, natural particles 1 D90
Variabel size, natural particles 2 D90

RICKENM-Book.indb 7 2/23/2016 7:38:58 AM

 8 Methods for the quantitative assessment of channel processes in torrents (steep streams)

defined as R = A/P, with A = flow cross-section and P = wetted perimeter of the flow
cross-section. In the flow laws presented in chapter 2, the hydraulic radius R is then
used instead of the flow depth, and in this case, the shear velocity is calculated as
v* = (gRS)0.5.

2.2 EMPIRICAL FLOW RESISTANCE LAWS

(POWER LAWS)

The most widely-known empirical flow formula is that of Gauckler-Manning-

Strickler (Hager 2001), based on studies of Gauckler (1867), Manning (1890),
and Strickler (1923). The associated flow resistance laws are in the form of power
laws, thus being similar to the definition of the Darcy-Weisbach friction coefficient
(Eq. 2.2). The Strickler coefficient kSt is dimensional and is usually used for the
characterization of the total resistance, which takes into account grain and form
roughness as well as any additional roughness. In the case of a large influence of the
grain roughness (i.e. a straight channel with a plane bed surface), there is a strong
correlation with the equivalent sand roughness ks.

2 1
V kSt R 3 S 2
(2.6)
Strickler (1923)
21.1
kst = (2.7)
6 ε0 Strickler (1923)

26
kst = (2.8)
6 D90 Meyer-Peter & Müller (1949)

1 21.1
kst g ≈ (2.9)
6 ks 6 ε0
after Strickler (1923)

1/ 6
V 8 ⎛ R ⎞
c= = = 6 .7 ⋅ ⎜ (2.10)
v f ⎝ D90 ⎟⎠
Strickler (1923) with Eq. 2.9 and ks = D90

The Strickler formula was obtained from tests with relative flow depths h/D > 10.
If εo = D90 is inserted into Eq. 2.9 as the roughness height, then Eq. 2.6 can be trans-
formed into Eq. 2.10. A comparison of Eq. 2.10 with the logarithmic flow law in
Fig. 2.2 shows that, for shallow flow depths, the Strickler formula clearly exhibits
a different trend from that of the logarithmic flow law. For Eq. 2.6, Eq. 2.7, Eq. 2.8,
Eq. 2.9, the value R is in [m], εo, D90 and ks are in [m], kSt is in [m1/3/s] so that V has the
unit [m/s]. Eq. 2.10 describes a medium flow resistance in alpine gravel-bed streams
with relative flow depths R/D90 > approx. 10 (cf. also Eq. 2.21 in chapter 2.4).

RICKENM-Book.indb 8 2/23/2016 7:38:58 AM

 Flow resistance in gravel-bedded streams and torrents 9

V/v* log. law, eq. 2.3, ks = D90

25 log. law, eq. 2.3, ks = 2D90
log. law, eq. 2.3, ks = 3D90
Manning-Strickler, eq. 2.10
20
Different behavior for small
values R/D90. This range is
typical for steep headwater
15 streams.

10

0
0.1 1 10 100 1000
R/D90

Figure 2.2 Comparison of different flow resistance formulae, presented as normalized flow velocity
V/v* as a function of the relative flow depth, R/D90. The empirical STRICKLER formula Eq. 2.10
exhibits a very different behavior for small relative flow depths from that for the logarithmic
flow laws such as, e.g., KEULEGAN (1938) (Eq. 2.3). Modified from BEZZOLA (2005).

Table 2.3 STRICKLER coefficients kSt for the total roughness of natural channels after
ZELLER & TRÜMPLER (1984).

kSt [m1/3/s]

Torrents (steep headwater streams)

Coarse gravel bed with cobbles, straight 20–25
Coarse gravel bed with cobbles, winding 15–20
Stone bed with individual boulders, straight 12–17
Boulder bed, step-pool or rapid-pool, irregular 8–15
Boulder bed, step-pool or rapid-pool, irregular with vegetation 5–12
Mountain rivers
Gravel and cobble bed, straight 20–33
Cobble bed with boulders, straight 14–25
Boulder bed, straight 10–15

For torrents, with typically small relative flow depths, little is known about suit-
able values of the equivalent sand roughness ks in the logarithmic flow law or any
reasonable modifications of this flow law. The Strickler formula was, therefore,
often used in the past with typical values of the Strickler coefficients kSt as given in
Table 2.3.
Many field studies (Jarret 1984; Hodel 1993; Bathurst 1985; Ruf 1990;
Rickenmann 1994, 1996; Zeller 1996; Palt 2001) and also laboratory flume studies

RICKENM-Book.indb 9 2/23/2016 7:38:59 AM

 10 Methods for the quantitative assessment of channel processes in torrents (steep streams)

(Meyer-Peter & Müller 1948; Rosport 1998) showed that a marked change in
the resistance behavior is observed for a channel slope of more than approximately
1–3%. This phenomenon is due to the effect of a greater presence of distinct morpho-
logical bed structures (macro-roughness) at steeper slopes, as well as to the influence
of smaller relative flow depths. In natural channels, for channel slopes of more than
approximately 1–3%, discharges frequently occur with relative flow depths h/D or
R/D smaller than 3–5.
Rickenmann (1994, 1996) developed formulae for flow velocity as a function of
discharge, channel slope and a characteristic grain size, based on 373 field measure-
ments (see Eq. 2.11, Eq. 2.12, with Q = channel discharge). The formulae are dimen-
sionally correct and apply to natural channel sections with bed slopes between 0.01
and 63%. The partitioning into two domains with the dividing point S = 0.8% reflects
the situation mentioned above, that, above approximately 1% channel slope, there
is an increased flow resistance due to pronounced bed structures. The equations are
analogous to approaches following hydraulic geometry theory (Griffiths 2003;
Singh 2003, Parker et al., 2007; Eaton 2013). Flow velocities calculated accord-
ing to Eq. 2.11 and Eq. 2.12 are compared further below with independent velocity
measurements (Fig. 2.6), along with other more recent approaches discussed in the
following subchapter.

0 37 g 0 33Q0.34 S 0 20
V= 0 35
(2.11)
D90 Rickenmann (1996), 0.8% ≤ S < 63%

0 96 g 0 36Q0.29 S 0 35
V= 0 23
(2.12)
D90 Rickenmann (1996), 0.01% < S < 0.8%

2.3 VARIABLE POWER LAW

In the domain of relative flow depths of h/D84 and R/D84 (or R/D90) below a value
of approximately10, the Strickler formula exhibits a rather different form from
the logarithmic flow law (Fig. 2.2, Fig. 2.3). On the other hand, simple logarithmic
flow laws result partially in too small or even negative flow velocities for h/D84
< approx. 1.
Rickenmann & Recking (2011) compared six flow formulae with a total of
2890 worldwide field measurements of flow velocities in gravel-bed streams. This
data set also includes many measurements for steep streams. The best description of
the average trend of all data was achieved using the variable power equation (VPE)
of Ferguson (2007):

V 8 a1 a2 (h / D84 )
= = (2.13)
v f a + a22 (h / D84 )5 / 3
2
1 Ferguson (2007)

RICKENM-Book.indb 10 2/23/2016 7:39:00 AM

 Flow resistance in gravel-bedded streams and torrents 11

V/v*
V/v*
100
14 MS
MS/RL
12 log
Katul
10
10 VPE

8
1

6
MS
Rick
4
0.1 log
TC
Smart 2
VPE

0.01 0

0.1 1 10 100 0.1 1 10 100

(a) (b) R/D84
R/D84

Figure 2.3 (a) Double logarithmic plot and (b) Semi-logarithmic plot of (8/f)0.5 (= V/v*) versus rela-
tive flow depth R/D84, with 376 field measurements. The colored lines represent different
flow laws: MS = Manning-Strickler Eq. (valid for large h/D84 values), RL = roughness layer
Eq. (valid for small h/D84 values), MS/RL = “best” combination, log = logarithmic Eq. HEY
(1979), Katul = KATUL et al. (2002), Rick = RICKENMANN (1991), Smart = SMART et al. (1992),
TC = THOMPSON & CAMPELL (1979), VPE = variable power equation of FERGUSON (2007).
Modified from FERGUSON (2007).

Here, the coefficient values a1 = 6.5 and a2 = 2.5 were used. With the aid of two
dimensionless parameters for the flow velocity, U**, and for the unit discharge, q**,
Eq. 2.13 with a1 = 6.5 and a2 = 2.5 can be presented alternatively as follows:

−0.2435
⎡ ⎛ q ** ⎞ 0.8214 ⎤
⎢1 + ⎜
0 .6 0
U 1 443 q ** ⎟ ⎥ (2.14)
⎢⎣ ⎝ 43.78⎠ ⎥⎦ Rickenmann & Recking (2011)

V
U ** = (2.15)
gSD84
q
q ** = (2.16)
3
gSD84

In the range of relative flow depths h/D84 smaller than approximately10, the
majority of the measurements from torrent-like channel reaches (mostly data of

RICKENM-Book.indb 11 2/23/2016 7:39:01 AM

 12 Methods for the quantitative assessment of channel processes in torrents (steep streams)

V/v*
100
all data (2890 values)
VPE eq. (2.13), with a1=6.5, a2=2.5
Smart-Jäggi (1983), 2890 values
Manning-Strickler, with Kst = 25 m^0.33/s
Manning-Strickler, with Kst = 2.5 m^0.33/s

10

0.1
0.1 1 10
h/D84

Figure 2.4 Measurements of the flow velocity in natural channels, plotted as V/v* vs. h/D84, for relative
flow depths smaller than 20. VPE = variable power equation, Eq. 2.13. The data points
connected with solid lines refer to observations of DAVID et al. (2010) for step-pool and cas-
cade reaches with channel slopes S = 0.06 to 0.18; they are from torrent-like channel reaches
and generally lie below the VPE line. Based on data from RICKENMANN & RECKING (2011).

David et al. (2010) in Fig. 2.4) exhibit much larger Darcy-Weisbach friction coeffi-
cients f (or a smaller coefficient a2) than the average trend of the remaining data accord-
ing to Eq. 2.13 (Fig. 2.4) or Eq. 2.14 (Fig. 2.5). The change in the flow velocity (V/v*)
with relative flow depth (h/D84) for these data (shown by the colored connecting lines
per channel reach) fits better with Eq. 2.13 than with the Strickler formula. For rela-
tive flow depths h/D84 smaller than approximately 4, Eq. 2.13 can be approximated by:

V 8 h
= = 2 .2 (2.17)
v f D84

According to the data of David et al. (2010) the coefficient a2 in very rough chan-
nels can be reduced to about 0.4; i.e. the flow velocity compared with the average

RICKENM-Book.indb 12 2/23/2016 7:39:02 AM

 Flow resistance in gravel-bedded streams and torrents 13

U**
all data (2890 values)
eq. 2.14 (=VPE, eq. 2.13 with a1=6.5, a2=2.5)
Smart & Jäggi (1983), data David et al. (2010)
Rickenmann (1996), data David et al. (2010)

10

0.1
0.1 1 10
q**

Figure 2.5 Measurements of the flow velocity in natural channels, plotted as U** vs. q**, for smallrela-
tive flow depths (q** < 20). VPE = variable power equation in the form of Eq. 2.14. The data
points connected with solid lines refer to observations of DAVID et al. (2010) for step-pool
and cascade reaches with channel slopes S = 0.06 to 0.18; they are from torrent-like channel
reaches and generally lie below the VPE line. The flow laws of SMART & JÄGGI (1983) (Eq. 2.5) as
well as of RICKENMANN (1996) (Eq. 2.11) tend to over- and partly under-estimate the observations
of DAVID et al. (2010). Based on data data from RICKENMANN & RECKING (2011).

trend of the other data can be reduced approximately by a factor 5 to 6. A reason

for this could be that, in the case of the channels investigated by David et al. (2010),
large wood pieces were caught in the stream’s bed structures, thereby increasing the
roughness. When the flow laws of Smart & Jäggi (1983) (Eq. 2.5), as well as of
Rickenmann (1996) (Eq. 2.11), are applied to the data from the torrent-like channel
reaches of David et al. (2010) presented in Fig. 2.5, they show a tendency to overesti-
mate the flow velocity in such conditions. A further comparison of the flow formulae
discussed here was made using independent measurements of the flow velocity in
mountain rivers in the Himalayas (Palt 2001); it is likewise shown that generally the
best agreement is obtained with the VPE solutions Eq. 2.13 and Eq. 2.14 (Fig. 2.6).

RICKENM-Book.indb 13 2/23/2016 7:39:03 AM

 Vcal / Vobs
eq. 2.14; Rickenmann & Recking (2011)
3.0
eq. 2.11, 2.12; Rickenmann (1996)

2.5

2.0

1.5

1.0

0.5

0.0
0 2 4 6
(a) R/D84

Vcal / Vobs
eq. 2.13 or 2.22; Ferguson (2007)
3.0 eq. 2.5; Smart & Jäggi (1983)

2.5

2.0

1.5

1.0

0.5

0.0
0 2 4 6
(b) R/D84

Figure 2.6 Comparison of the calculated flow velocity (Vber) with different flow laws and with inde-
pendent measurements of the flow velocity (Vgem) in mountain rivers in the Himalayas (PALT
2001): (a) RICKENMANN (1996), Eq. 2.11 and Eq. 2.12; RICKENMANN & RECKING (2011), Eq. 2.14;
(b) FERGUSON (2007), Eq. 2.13 or Eq. 2.22, calculated with R; SMART & JÄGGI (1983), Eq. 2.5,
calculated with R.

RICKENM-Book.indb 14 2/23/2016 7:39:04 AM

 Flow resistance in gravel-bedded streams and torrents 15

The VPE flow resistance formula Eq. 2.13 with the coefficients a1 = 6.5 and
a2 = 2.5 was also applied to some steep torrents (channel sections without wood
debris) in Switzerland and resulted in quite good agreement with field measurements
for mean flow velocity (Nitsche et al., 2012a).
In summary, more recent investigations revealed that in steep and rough chan-
nels (with h/D84 < 4) the flow resistance can be better described with the power law
based on Ferguson (2007) than with the Strickler formula or with a logarithmic
flow law. However, the roughness coefficient a2 in this power law is also strongly
dependent on the channel morphology. It is well known that the Strickler value
for torrents can vary within a range of about 2 m1/3/s up to about 30 m1/3/s (c.f.
Table 2.3); however, the Manning-Strickler formula does not correctly predict
the increase of V/v* with increasing h/D84 (Fig. 2.4). It should also be taken into
account that (i) the flow behavior in torrents may be difficult to approximate by a
one-dimensional approach, (ii) frequently there may be local changes between sub-
and supercritical flow and (iii) the flow conditions generally depend strongly on
the discharge or the relative flow depth. Many of these aspects still require further
detailed study.

2.4 PARTITIONING OF THE FLOW RESISTANCE

Total flow resistance of surface runoff in a channel consists of the skin friction of
the water flowing along the individual grains of the bed (grain roughness) and the
friction losses resulting from bed forms, large immobile grains and irregular channel
geometry (macro-roughness). In steep channels, the additional friction losses (apart
from grain roughness) may be due to form drag, local flow accelerations and hydrau-
lic jumps and the formation of multiple flow paths in shallow flows characterized by
low relative flow depths h/D. Earlier concepts of flow resistance partitioning distin-
guished between grain and form roughness. A correction term to account for rough-
ness losses in the calculation of bedload transport was introduced, for example, in the
approaches of Meyer-Peter & Müller (1948) and Palt (2001), where the relation
between grain roughness (kr) and total roughness (kSt) is important. A similar cor-
rection is introduced here by firstly using the Manning coefficient (n), which is the
reciprocal of the Strickler coefficient (n = 1/kSt).
Based on 373 measurements of mean flow velocity in steep channels (Ricken-
mann, 1994, 1996), a correction factor was determined to partition the flow resist-
ance into the base-level roughness (no) and the total roughness (ntot). This correction
factor is expressed as a function of either total discharge Q in Eq. 2.18 after Ricken-
mann et al. (2006a) or of flow depth h in Eq. 2.19 after Chiari et al. (2010) (see also
Fig. 2.7). These equations for the partitioning of flow resistance are implemented in
the sediment transport simulation program SETRAC (Chiari & Rickenmann 2011)
and in the first version of the follow-up program TOMSED (www.bedload.at).

⎛ no ⎞ 0.131 Q0 19
=
⎜⎝ n ⎟⎠ g 096 S 0.19 D0 47
0 (2.18)
tot 90 Rickenmann et al. (2006a)

RICKENM-Book.indb 15 2/23/2016 7:39:04 AM

 16 Methods for the quantitative assessment of channel processes in torrents (steep streams)

no/n
1

eq. 2.18
data of Rickenmann (1996)
eq. 2.19
0.8
trend line for data of Palt (2001)

0.6

0.4

0.2

0.001 0.01 0.1 1

S

Figure 2.7 The correction factor (no/ntot) as a function of the channel slope S, to take into account
energy losses due to high flow resistance in steep channels. Modified from CHIARI et al.
(2010).

0 33
⎛ no ⎞ ⎛ h ⎞
⎜⎝ n ⎟⎠ = 0.092 ⎜⎝ D ⎟⎠ S −0 35 (2.19)
tot 90 Chiari et al. (2010)

The partitioning of the flow resistance can also be carried out using the Darcy-
Weisbach coefficient f by dividing the total friction (ftot) into a base-level friction (fo)
and an additional friction component (fadd):

ftot f + fadd (2.20)

Most bedload transport equations are based on laboratory flume experiments,

in which the form or macro-roughness losses were negligible and the flow conditions
were associated with relative flow depths mostly larger than about 7–10. Thus the
Manning-Strickler equation such as Eq. 2.10 or Eq. 2.21 provides a good quanti-
fication of the mean flow resistance in deeper flows, which is considered here as the
base-level resistance (fo). However, additional flow resistance (fadd) is present in steep
streams with coarse roughness elements such as immobile boulders and step-pool
sequences and small relative flow depths less than about 10 (e.g. torrent channels).
Rickenmann & Recking (2011) used the VPE approach of Ferguson (2007)
(Eq. 2.13) to develop a method to partition the flow resistance for conditions with
medium to large-scale roughness (in the sense of Bathurst et al., 1981; here for

RICKENM-Book.indb 16 2/23/2016 7:39:04 AM

 Flow resistance in gravel-bedded streams and torrents 17

V/v*
100
logarithmic flow law, Hey (1979)
VPE Ferguson (2007), eq. 2.22
Manning−Strickler, large values of
h/D84, eq. 2.21
Manning-Strickler, base-level
friction for small values of h/D84

10

Additional energy losses in steep and

shallow flows with low relative flow
1 depths (h/D84). Bedload transport
equation were not developed for such
conditions.
-> Correction through a partitioning of the
total flow resistance (ftot) into a base-level
resistance (fo) and additional flow
resistance (fadd)

0.1
0.1 1 10 100
h/D84

Figure 2.8 Partitioning of the flow resistance for medium and large-scale roughness conditions (here
h/D84 < approx. 10). Based on a formula of the type MANNING-STRICKLER, a base level for the
flow resistance is determined (dashed purple line), corresponding to the flow conditions
in the case of small-scale (here h/D84 > approx. 10) roughness (solid purple line), after the
approach of RICKENMANN & RECKING (2011).

h/D84 < approx. 7). Based on a Manning-Strickler type formula, a base-level flow
resistance is calculated (Fig. 2.8), which corresponds to flow conditions with small-
scale roughness (here for h/D84 > approx. 7):

1/ 6
Vo 8 ⎛ h ⎞
= = 6 .5 ⎜ (2.21)
v fo ⎝ D84 ⎟⎠

If Eq. 2.21 is used for medium- to large-scale roughness conditions, an appropriate

level for the base-level friction can be calculated for fo or for the virtual flow velocity
Vo. The total flow resistance is determined with the VPE as follows:

Vtot 8 ( .5) (2.. ) (h / D84 )

= = (2.22)
v ftot 6.52 + 2.52 (h / D84 )5 / 3

RICKENM-Book.indb 17 2/23/2016 7:39:05 AM

 18 Methods for the quantitative assessment of channel processes in torrents (steep streams)

(fo/ftot)0.5
1

0.9
additional flow
resistance(fadd)
resistance(
0.8

0.7

0.6

0.5 base-level flow

resistance(fo)
resistance(

0.4

0.3
all data, ftot from eq. (2.14) and
0.2 fo from eq. (2.24)
ftot from eq. (2.14) and fo from
eq. (2.24), h calculated
0.1 (ftot from eq. (2.22) and fo from
eq. (2.21), h observed
0
0.1 1 10 100
h/D84

Figure 2.9 Partitioning of the flow resistance (fo/ftot)0.5, based on 2890 measurements in gravel-bed
streams and partly in torrents. The values (fo/ftot)0.5 are essentially a function of the relative
flow depth.The dark red line corresponds to calculation according to Eq. 2.25, while the light
purple line corresponds to calculation according to Eq. 2.23. Modified after RICKENMANN &
RECKING (2011).

Thus, the ratio of base-level to total flow resistance can be calculated as:

fo V (h)
= tot (2.23)
ftot Vo (h)

The above proposed partitioning of the flow resistance is basically a function of the
relative flow depth (Fig. 2.9). It is an all-inclusive and empirical approach, but it
implicitly contains information about an average increase of roughness in steep chan-
nels with irregular bed morphology and shallow flows. Instead of calculating (fo/ftot)
as a function of flow depth with Eq. 2.21 to Eq. 2.23, (fo/ftot) can be determined as a
function of the unit discharge q as follows:

RICKENM-Book.indb 18 2/23/2016 7:39:06 AM

 Flow resistance in gravel-bedded streams and torrents 19

Vo
= 3. 04 ( q*
q *)
0 .4
U o ** = (2.24)
gSD84

Together with Eq. 2.16, Vo(q) is then obtained. For the calculation of the total resist-
ance ftot or Vtot Eq. 2.14 together with Eq. 2.15 and Eq. 2.16 is used, which then gives
Vtot(q). The ratio of base-level to total flow resistance is then given as:

1 .5
fo ⎛ V (q) ⎞
= tot (2.25)
ftot ⎜⎝ Vo (q) ⎟⎠

The ratio (fo/ftot) represents a reduction factor, which is a measure of that part
of the total flow energy (or total bed shear stress) available for bedload transport.
The partitioning of the flow resistance by means of the value (fo/ftot)0.5 is the basis for
determining a reduced energy slope Sred, which is then introduced into the calculation
of the bedload transport (see chapter 3.4.4).

RICKENM-Book.indb 19 2/23/2016 7:39:07 AM

This page intentionally left blank
 Chapter 3

Fluvial bedload transport

3.1 CHARACTERIZATION OF TORRENTS

AND MOUNTAIN RIVERS

In mountain rivers and torrents, hydrologic and hydraulic processes are characterized
by extreme variability, both in the spatial and the temporal domains. The main causes
of this lie in the strong interaction of geological constraints, earth surface processes,
and the channel network in alpine catchments (Hassan et al., 2005a; Comiti & Mao
2012; Church 2013). This results in a high variability of the following parameters:

• sediment supply or availability and sediment transport

• composition of the grain size distribution of the stream bed and of the source
areas
• channel geometry along the stream and in the lateral direction
• highly variable (but generally low) runoff depths
• flow behavior in the transition regions subcritical-supercritical-subcritical flow

Typical torrent channels are greatly influenced by these parameters so that a con-
sideration of the geological and morphological conditions in and along the channel is
of great importance. Sediment is often supplied to the channel by colluvial processes,
while the channel bed may consist partly of bedrock and, thus, may have only a semi-
alluvial character. Mountain rivers, in contrast, typically have an alluvial streambed
that reflects a single dominant formation process. In European countries, torrents refer
typically to Alpine catchments with channels steep enough that debris flows can occur
in addition to fluvial sediment transport. According to this definition, such catch-
ments are associated with drainage areas of less than about 25 km2 (Rickenmann &
Koschni 2010; Marchi & Brochot 2000; Marchi & D’agostino 2004). Typical
differences between torrents and mountain rivers are summarized in Fig. 3.1.
An important difference between torrents and alluvial mountain rivers concerns
both sediment supply or sediment availability and the runoff conditions. Typically,
torrents have sediment supply-limited conditions, whereas bedload movement in
mountain rivers is mostly limited by the hydraulic transport capacity (Montgomery &

RICKENM-Book.indb 21 2/23/2016 7:39:07 AM

 22 Methods for the quantitative assessment of channel processes in torrents (steep streams)

TORRENT MOUNTAIN RIVER

Process debris fl
fflows
ows
fluvial
f uvial sediment transport
fl

Bedslope steeper than 5 ... 10 % appr. 10 ... 1 %

boulders, bedrock armour layer, residual blocks
Bed
morphology irregular geometry more uniform geometry
wide grain size distribution (g.s.d.) less wide g.s.d.
cascades, step-pools rapids, riffles, pools, bars

Sediment storage
in bed

Lateral sediment
input

Figure 3.1 Overview of some differences between torrents and alluvial mountain rivers.The gray areas
signify, from left to right, the increasing and decreasing importance, respectively, of sediment
storage in the bed and of lateral sediment input with increasing catchment size.

cascade plane-bed dune-ripple

Sediment delivery QS
Transport capacity QB

step-pool pool-riffle

supply limited transport limited

QB QS

Catchment area

Figure 3.2 Transport-limited and sediment supply-limited conditions in mountain catchments.

With increasing drainage area the channel slope decreases. The typical morphological
structures also vary with the channel slope. Modified after MONTGOMERY & BUFFINGTON (1997).

Buffington 1997; Fig. 3.2). In torrent catchments, peak water runoff and also
the formation of debris flows is often triggered by convective rainstorms with high
precipitation intensities and short durations. These storms are associated with a rapid
increase in discharge, and flood events typically lasting less than a few hours. Longer

RICKENM-Book.indb 22 2/23/2016 7:39:08 AM

 Fluvial bedload transport 23

duration rainfall events may be more important in mountain rivers, resulting in a

more gradual increase in runoff conditions.
Three separate morphological elements can be differentiated along a stream
channel: streambed, bank and the adjoining hillside. The morphology of the bed
and the banks of a mountain river is closely associated with the runoff and bedload
transport processes. In its lower parts a mountain river in its natural environment
can adjust its course or total width by meandering, bar formation and braiding,
reflecting the interactions between runoff, bedload transport, grain sizes and the
general valley slope (Montgomery & Buffington 1997). In contrast, if a torrent
or mountain river is constrained laterally by the valley slopes or bedrock outcrops,
morphological structures in the vertical dimension (e.g. step-pools and cascades)
become more prominent (Fig. 3.1, Fig. 3.2). The formation of these structures may
be favored by large immobile boulders or large wood fragments, and they are associ-
ated with a high energy dissipation of flowing water. A destruction of steps may lead
to a temporary increase of bedload transport (Turowski et al., 2009). The typical
morphological structures generally change with mean channel slope or with the size
of the catchment area (Montgomery & Buffington 1997), which is also indicated
in Fig. 3.1 and Fig. 3.2.
The channel morphology can be described by the three elements: channel geome-
try, bed form and grain shape (De Jong & Ergenzinger 1995; Morvan et al., 2008).
Considering these structures, previous authors have distinguished between grain
resistance and form resistance in bedload transport calculations for gravel-bed rivers
(Meyer-Peter & Müller 1948; Carson 1987; Gomez & Church 1989). How-
ever, distinguishing between these resistances is only really possible in channels with
sandy beds; in gravel-bed streams (and especially in torrent channels), such a distinc-
tion is questionable (Zimmermann 2010; Rickenmann & Recking 2011).
Colluvial and fluvial stream bed types can be distinguished according to the
relative importance of hillslope sediment delivery and hydraulically-forced evolu-
tion of the bed sediments (Montgomery & Buffington 1997). In colluvial stream
reaches, sediment delivery from talus slopes is dominant, and along–channel sedi-
ment transfer occurs mainly due to debris flows or very rare and extreme flood
events. Colluvial reaches are, therefore, characterized by angular, unsorted particles.
Reaches dominated by fluvial bedload transport are characterized by rounded, well-
sorted sediment particles, and often a coarse armor layer is formed with a preferred
orientation of the particles depending on flow direction (Aberle & Nikora 2006).
A torrent channel is often composed of both colluvial and fluvial sediments, and
an armor layer is not necessarily formed, as is often the case in mountain rivers
with purely fluvial sediments. It is, therefore, questionable to what extent concepts
regarding the formation and breakup of an armor layer are also applicable in torrent
channels.
Further important characteristics of steeper channel reaches in mountain
streams and torrents are longitudinal bed structures that are developed. Typical
morphologies for stream beds with slopes steeper than about 3% are step-pool
sequences (Fig. 3.3) and cascades (Montgomery & Buffington 1997; Grant et
al., 1990). Empirical studies show that such structures may be stable for a flood
event with return periods on the order of about 50 years (Chin 1989; Lenzi et al.,
2004).

RICKENM-Book.indb 23 2/23/2016 7:39:08 AM

 24 Methods for the quantitative assessment of channel processes in torrents (steep streams)

rock step

riffle step

pool
boulder step
rock step

pool

pool

Figure 3.3 Step-pool sequences as typical bed structures in steeper streams and torrent channels.
Adapted from HAYWARD (1980).

3.2 SEDIMENTOLOGICAL PARAMETERS

The grain size distribution and the grain shape have a large influence on the transport
of sediments in alpine watercourses. Due to the large spatial and temporal variability
of the sediment distribution, it is a challenging task to determine a representative
grain size distribution for a given channel reach.
For the computation of channel discharge and sediment transport, characteristic
grain diameters are required to determine the flow resistance (e.g. D84, D90), the start
of mobilization (e.g. D50, D65, and possibly grain shape) and the transport efficiency
as a function of grain size distribution parameters (D16, D30, D84, D90). In terms of a
cross-section, the directly-measured grain size distribution represents only a snapshot,
the result of the immediate hydraulic conditions (i.e. those that have just taken place)
and of the current bedload transport. Such a static snapshot, therefore, only gives a
partial picture of the dynamically-changing sedimentological conditions during an
event or over a specific period of time.
The spatial variability results from the often very heterogeneous transport and
deposition conditions as well as the various sediment sources. Thus, the sedimentology
of sediment sources and the channel geometry (separate for bed and bank) should be
distinguished. To determine the grain size distribution of the surface layer (armor layer)
over a reach, different morphological elements such as steps, pools, rapids and gravel
banks should be taken into account, ideally proportionately to their spatial occurrence.
However, it is often not easy to clearly separate zones of different stream-bed morphol-
ogy. The temporal variability depends on the process dynamics during the transport
event and can only be taken into account more thoroughly by a comparison of the situ-
ation after several transport events. A possible approach is to consider exposed earlier
deposits or the comparative evaluation of the bed and the adjacent bank.
Among other factors, the grain shape influences the initial mobilization and the
transport process itself. Thus, with an increasing plate-like shape of the individual
grain a greater alignment similar to a roof-tile structure of the fluvially-formed bed
is possible and, thus, a reduction of the effective grain roughness for the same stable

RICKENM-Book.indb 24 2/23/2016 7:39:08 AM

 Fluvial bedload transport 25

high sphericity

low sphericity

very angular sub-rounded well

angular rounded
sub-angular rounded

Figure 3.4 Subdivision of the grain shape according to roundness (very angular to very round) and
according to sphericity. Modified from: [http://homepage.usask.ca/∼mjr347/prog/geoe118/
geoe118.017.html]

bedding structure. A general classification according to shape and roundness can be

carried out following, for example, Schreiner (1997). Fig. 3.4 gives a description
of grain shape according to roundness, and it is also possible to distinguish between
differences in sphericity.

3.3 DETERMINATION OF GRAIN SIZE DISTRIBUTION

AND OF CHARACTERISTIC GRAIN SIZES

Various analysis methods may be found in the literature to determine the grain size
distribution (Bunte & Abt 2001). These methods have been developed either for
a specific grain size spectrum or for specific deposition conditions. In the case of
steep channels, typically with a broad grain size spectrum, different methods have
to be combined for a suitable analysis. In using such statistical methods it has to be
recognized that they were usually derived for different sedimentological conditions
and that they have not yet been verified systematically for use in torrent channels.
For the purpose of a better comparison, all methods should be based on a unified
grain size classification; an overview of common classification schemes is provided
in Table 3.1.

3.3.1 Volume analysis of sediment

A certain volume of sediment is collected from the stream bed. In this way the grain
size can be determined without the influence of alignment and layering, but the site has
to be easily accessible and the sample must consist of granular material. Samples can be
taken not only near the surface but also at different depths. The bigger the maximum
grain size, the bigger is the volume that has to be tested (Church et al., 1987).
The volumetric analysis can be carried out by means of sieving and weighing the
material within each individual grain size class so that different characteristic grain
sizes such as the median axis (b-axis) are determined. Sieving can be either wet or
dry, but weighing should always be carried out on dry material, since the influence
of the weight of water in the case of fine-grained material can be substantial. A few
approaches to estimate the necessary sample size are listed in Table 3.2.

RICKENM-Book.indb 25 2/23/2016 7:39:09 AM

 26 Methods for the quantitative assessment of channel processes in torrents (steep streams)

Table 3.1 Classification schemes to enable determination of grain size distributions. The typical range
of application of the various methods of analysis is depicted by gray bars. ISO: International
Organization for Standardization; VSS: Swiss Association of Road and Traffic Engineers;
USCS: Unified Soil Classification System; ÖNORM: Austrian Standards Institute. LNA: line
by number analysis (see below in this chapter).

Table 3.2 Necessary sample size for volume samples.

VolumeSample(m³) = 2.5 Dmax (m) FEHR (1987a, b)

MassSample(kg) = 0.1 10b ρs Dmax³ (m) BUNTE & ABT (2001)

ρs = sediment density; accuracy: high (b = 5), medium (b = 4), low (b = 3).

RICKENM-Book.indb 26 2/23/2016 7:39:09 AM

 Fluvial bedload transport 27

Basically, a preliminary fractioning of the coarse components in the terrain is pos-

sible. This involves very coarse material with a grain size D > 32 mm that is measured
individually (b-axis), grading these components according to the corresponding grain
sizes and directly weighing them using weighing scales or otherwise by means of water
displacement. That is, the weight is determined from the displaced volume of water
and the unit weight (or specific gravity) of the grains. However, the influence of a
subjective selection using a shovel is considerable. A possibility here would be taking
a sample using an excavator shovel, and carrying out a complete measurement of the
whole sample. Below a grain size D < 32 mm sieving should be used. If an armor layer
is present, then, ideally, a sample of the armor layer should be taken together with a
sample of the underlying layer. The thickness of the armor layer may then be assumed
to be approximately D90 to Dmax.

3.3.2 Analysis of surface bed material

The grain size distribution (GSD) of the surface bed material is measured with the
aid of different methods of random sampling (grid, line, purely random), allowing
measurements to be carried out for the surface of the bed or a gravel bank or a sedi-
ment exposure (Bunte & Abt 2001). New methods of surface analysis are based on
photographs of the streambed. With the so-called photo-sieving technique the limiting
parameter is the grain size, since the projected area of the largest grain should take
up at most 1% of the area of analysis. As a result, the apparent b-axis or the mean
diameter of the exposed grain surface is determined by an automatic demarcation of
the grain boundary. For this purpose there are already various programmed proce-
dures that can be resorted to (Warrick et al., 2009; Buscombe et al., 2010; Graham
et al., 2010; Detert & Weitbercht 2012a, b). However, the use of photo-sieving
in torrent channels is problematic due to the very large granular components (and
the required vertical distance of the camera). An alternative method is to use cameras
capable of spatial, high-resolution distance measurement (Nitsche et al., 2010).
In order to obtain the whole grain size distribution, the results of surface sam-
pling methods (typically involving a minimum cut-off grain size) have to be converted
into a standard recording method and combined with statistically-determined rela-
tionships for the finer material. These functional relationships depend on the grain
shape, macrostructure, grading and porosity and are, therefore, only representative
for some conditions of deposition. In mountain rivers and torrents of the Alps, the
line by number analysis (LNA) developed at the Laboratory of Hydraulics, Hydrology
and Glaciology, ETH, Zurich, is often used (Anastasi 1984, Fehr 1987a, b). The
method has mainly been tested in mountain rivers.

3.3.3 The Line by Number Analysis (LNA)

of the surface layer in streams with
coarse bed material
In streams with coarse bed material, sieve analysis of the surface layer often cannot
be carried out, since either the sieve set for large grain diameters is unsuitable or the
amount of bedload cannot be handled economically (see also Table 3.2). Therefore,

RICKENM-Book.indb 27 2/23/2016 7:39:09 AM

 28 Methods for the quantitative assessment of channel processes in torrents (steep streams)

simpler methods have been developed that take into account the conditions in coarse
bedload streams. The most frequently used techniques for analyzing armor layers
involve taking samples from a specific area using grids and lines (for an overview
see Bunte & Abt 2001). Here, the line by number sampling method LNA (Fehr
1987a, b) is explained in more detail.

3.3.3.1 Procedure for the execution of a line sample

A piece of string is stretched over the surface layer to be analyzed, helpingto avoid
systematic errors in the choice of the particles to be investigated. For all particles
that are greater than 1–2 cm and lie under the string, the middle axis (b-axis) is
measured. The particles are divided into diameter classes (fractions) and counted. The
line samples should include least 150 particles. If, at the same time, a volume-weight
analysis of the fine material of the subsurface layer is carried out, then the classes of
the line sample in the overlapping region should be adjusted using a volume sample
of the latter.

3.3.3.2 Analysis of the line sample

The result of an LNA is a frequency distribution of the coarse fraction (>1 cm) of the
surface particles on the streambed. To compare and combine the result with other
methods, the proportions as expressed in percentages of the LNA values have to
be converted to equivalent weight fractions in a volume analysis. Fehr (1987a, b)
developed the following empirical conversion formula (for hydraulically loaded
samples):

Δqi Dmi 0 8
Δpi = (3.1)
ΣΔqi Dmi 0 8

where Δpi = (weight of fraction i)/(weight of whole sample), Δqi = (number of stones
in the fraction i)/(number of stones in the whole sample), Dmi = mean grain diameter
of the fraction i. Since the fines are underestimated by the LNA method, the grain
size distribution curve (or grading curve) still has to be adjusted. It is noted that the
theoretical exponent in Eq. 3.1 is 1.0 (Kellerhals & Bray, 1971) rather than the
experimentally-determined value of 0.8 (Fehr1987a, b). For the conversion of an
LNA into the grading curve of the subsurface layer, after Fehr (1987a, b) a proportion
of fines of 0.25 for the sediments <1 cm may be assumed:

ic 0 25 0 75 ∑ Δp
pi (3.2)
1 Conversion of LNA to GSD of subsurface layer

If a conversion of an LNA into the grading curve of the surface layer (armor layer)
is to be carried out, a smaller proportion of sediments <1 cm should be assumed.
According to a summary by Recking (2013) based on measurements in 78 gravel-bed

RICKENM-Book.indb 28 2/23/2016 7:39:10 AM

 Fluvial bedload transport 29

streams, this proportion varies between about 0.01 and 0.2 and is, on average, 0.11.
Therefore, for a conversion of an LNA sample into a surface layer grading curve it
may be assumed that:

i
pic 0 11 0 89 ∑ Δp
pi (3.3)
1 Conversion of LNA to GSD of subsurface layer

where pic = corrected cumulative frequency (relative proportion) of the grains with
D ≤ Di (Di = grain diameter of the grain size class i). To complete the grading
curve for the fraction <1cm, either a volume sample is used (sieve, sedimentation
analysis) or a distribution after Fuller is assumed (Fehr 1987a, b). Based on river
sediments and assuming an optimum packing density, Fuller & Thompson (1907)
developed a single parametric synthetic distribution based on Dmax of the grading
curve. The associated Fuller curve describes the grain size distribution of well-
graded fluvial fine sediments. According to the investigations of Meyer-Peter &
Müller (1948), the Fuller curve provides a good approximation for mountain riv-
ers in Switzerland and in the alpine region. Thus, a theoretical grading curve for the
fines may be calculated as follows:

Di
pFU i = (3.4)
Dmax

where pFUi = cumulative frequency of the grains with D ≤ Di according to the Fuller
curve. This grading curve after Fuller is then merged with the grading curve for the
coarse part obtained using the LNA, where Dmax is the maximum grain diameter of
the fine sediments.

3.3.3.3 Statistical methods for the combination procedure

To combine a surface analysis with a volume analysis, the approach of Anastasi
(1984) was calibrated and verified by Fehr (1987a) for mountain rivers. Fehr
(1987b) also recommended the rigid or flexible adjustment of the synthetic volume
distributions after Fuller and subsequent combinationwith the converted surface
distribution obtained from the LNA.
In the determination of characteristic grain sizes, a relatively large variability
of the resulting grain size distribution must be expected due to the broad grain
size distribution and the poor grading. Statistical methods for the LNA or similar
analysis methods involve some uncertainty when used for torrents, since they were
developed primarily for mountain rivers and have to be tested for torrent condi-
tions. Nevertheless, it is usually found that, in the case of coarse torrent beds, the
important characteristic grain sizes (D50, D84, D90) are determined using surface
analysis and are only moderately influenced by the uncertain combination with a
sieve analysis.

RICKENM-Book.indb 29 2/23/2016 7:39:10 AM

 30 Methods for the quantitative assessment of channel processes in torrents (steep streams)

3.4 BEDLOAD TRANSPORT IN STEEP CHANNELS

To understand bedload transport processes in mountain rivers and torrents requires con-
sideration of several important factors, including the spatially-variable mobilization of
the solids, the active sediment input, the formation or the mobilization of an armor
layer or other stable morphological structures, sediment availability, and the transitional
regime with debris-flow like sediment transport. It follows that there is not necessarily
a functional relationship between channel discharge and bedload transport, the latter of
which can vary considerably in a given channel as shown in Fig. 3.5 and Fig. 3.6.
To determine bedload transport quantitatively, the following aspects must be
considered in particular: (i) initiation of transport, (ii) bedload transport function,
(iii) partitioning of the flow resistance (additional energy losses), (iv) possible armor
layer, (v) sediment availability.

PBIS data 1986−1999

Qc = 0.2 m3/s
Qc = 0.5 m3/s
Qc = 0.8 m3/s
100

QB =1.5 (Q-QC)S1.5
(eq. 3.23)

10
QB (m3/min)

0.1

0.01

0.001
0.1 1 10
Q (m3/s)

Figure 3.5 Bedload transport measurements in Erlenbach (Switzerland) using piezoelectric bedload
impact sensors (PBIS) (RICKENMANN & MCARDELL 2007, eq. 7), and comparison with a labo-
ratory-based bedload transport formula (Eq. 2.23) calculated for a channel slope S = 0.105.

RICKENM-Book.indb 30 2/23/2016 7:39:10 AM

 Q (m3/s) Erlenbach: flood of 14 July 1995 QB (m3/min)
10 6

bedload transport (Qb)

discharge (Q) 5
8

4
6

4
2

2
1

0 0
0 60 120 180 240
(a) Time (min)

Q (m3/s) QB (m3/min)
Erlenbach: flood of 9 July 1987
5 4
bedload transport (Qb)
discharge (Q)
4
3

1
1

0 0
0 40 80 120 160
(b) Time (min)

Figure 3.6 Bedload transport measurements in Erlenbach (Switzerland) using piezoelectric bedload
impact sensors PBIS (RICKENMANN & MCARDELL 2007, eq. 7), using examples (a), (b) from two
flood events.

RICKENM-Book.indb 31 2/23/2016 7:39:12 AM

 32 Methods for the quantitative assessment of channel processes in torrents (steep streams)

3.4.1 Fluvial bedload transport

In the case of fluvial bedload transport, the solid components are moved by the
medium water. The material does not move continuously but intermittently with
phases of inactivity, the grains sliding, rolling, or saltating, with little transport in the
bottom layer and with intensive transport but decreasing concentration in the upper
layers (Schmidt & Ergenzinger 1992; Smart & Jäggi 1983; Rickenmann 1990).
With increasing discharge, larger grains are moved; thus, after the start of motion,
so-called selective bedload transport prevails (Fig. 3.7). Due to the selective bedload
transport, mountain rivers and torrents can develop an armor layer. In the process
the bed becomes coarse through washing out of the fines, and coarse components can
nestle themselves into the bed, resulting in a higher stability with regard to renewed
erosion. The bed structures thereby formed are an important feature of steep gravel
bed channels (Montgomery & Buffington 1997; Rosport 1998; Wohl 2000;
Schächli 1991; Grant et al., 1990).
In torrents, bedload transport occurs often in the form of pulses or sediment
waves (Fig. 3.6). This behavior can be traced back to spatially- and temporally-varia-
ble hydraulic conditions during the sediment transport or to discontinuous sediment
availability and mobilization mechanisms that may be caused by turbulence peaks
or retrogressive bed erosion after the scouring of coarse components. Hereby, the
following transport processes are distinguished:

100

90 water discharge (m3/s)

at which bedload
sample was
80 collected
1.1 2.6
70 0.3 4.0
2.3
60
Percent finer

50

40

30

20
bed material
bedload size size distribution
10 distributions

0
0.1 1 10 100 1000
Size of median axis (mm)

Figure 3.7 Composition of the sieve samples of the transported bedload for different discharges and
stream bed material. The coarsening of the grain size distribution with increasing discharge
is the result of selective bedload transport. The measurements are from the Roaring River
(Colorado, USA), and the bed slope at the measuring section was approximately 5%.
Modified from BATHURST (1987).

RICKENM-Book.indb 32 2/23/2016 7:39:12 AM

 Fluvial bedload transport 33

• Transport over an intact armor layer.

• Transport over an armor layer with limited sediment exchange.
• Transport with permanent sediment exchange between the transported solids and
the bed sediment after breaking up the armor layer (Rosport 1998, Hunziker
1995, Günter 1971).

Whereas at the beginning there is a selective, grain-size-dependent mobility of the

solid material, all grains are essentially moving (equal mobility) after the complete
break-up of the armor layer. This greatly affects the transported material in the grain
size spectrum and it must be taken into account in the case of long-term transport
simulations in which the first two transport processes play an important role (Webb
et al., 1997; Hunziker 1995).
A further parameter that is difficult to measure is the dynamic change of the
flow resistance during the transport process. The transitions between grain, form and
channel roughness are smooth. At the start of the mobilization of the bed, moreover,
the surface structure changes and, therefore, the grain and form roughness also
(Ergenzinger & Schmidt 1995; Ergenzinger et al., 1994). Characteristic grain size
parameters (e.g. D30, D50, D90) in mountain streams often exhibit fluctuation ranges of
up to 30% (Reid & Dunne 1996; Jäggi 1992). In particular, the temporal variability
is difficult to take into account and causes a corresponding uncertainty in the analysis.
The solids transport can be expressed as a functional relationship between the
following different independent parameters controlling the process:

• The critical shear stress or the critical discharge at the initiation of mobilization.
These parameters are mainly relevant immediately after the start of bedload trans-
port when the critical values are exceeded. It is important in the evaluation of the
dynamics of the armor layer and also in the case of the selective transport of only
a few sub-fractions of the entire spectrum of grain sizes present on the bed.
• The bedload transport capacity of the channel flow. After the onset of general
bedload transport (over the whole grain size distribution) this parameter pre-
dominates. Bedload transport capacity is primarily expressed by the parameters
hydraulic radius R or flow depth h, and channel slope S, or in simple approaches
the first two may be replaced by the (unit) discharge q.
• Correction factor or calculation procedure to take into account the additional
energy losses due to high flow resistance.

Natural channel beds of mountain rivers and torrents are typically characterized
by a broad grain size spectrum. These conditions favor selective bedload transport
during the rising or falling flood hydrograph, just after the initiation of mobilization
or before the end of mobilization, and cause separation processes that can lead to the
formation of an armor layer. For increasing discharge this selective bedload transport
is followed by general bedload transport, during which the whole grain size spectrum
is mobilized and transported due to stronger hydraulic forces (Fig. 3.8).
The range of weaker bedload transport between the initial mobilization and the
general start of mobilization (discharge Qc < Q < QD in Fig. 3.8) is also described in
the literature as phase-1 transport and the range with increasing bedload transport for
Q > QD as phase-2 transport. Phase-1 transport also corresponds to the transport of

RICKENM-Book.indb 33 2/23/2016 7:39:13 AM

 34 Methods for the quantitative assessment of channel processes in torrents (steep streams)

measurements
Bedload transport QB

ion
lat
po
tra
ex
weak
transport

QC QD Discharge Q

Figure 3.8 Increase of the bedload transport with discharge. In the figure Qc denotes the initial mobilization,
QD the general start of mobilization after the breaking up of the armor layer, implying the mobi-
lization of the coarser grains of the surface layer. Modified from BEZZOLA (2005).

fines over a coarse armor layer. A thorough discussion of measurements and analyses
on transport in phases 1 and 2 can be found in Jackson & Beschta (1982), Ryan
et al. (2005) and Bathurst (2007).

3.4.2 Start of mobilization

The start of mobilization is a key factor for the evaluation of bedload transport pro-
cesses. At the beginning the mobilization is more of a random process in which individ-
ual grains separate and move from the bed matrix due to the action of shear stress peaks
near the bed caused by turbulence (Rosport 1998). Based on investigations of the equi-
librium stability between a flow-induced action (shear stress, form resistance, hydrody-
namic uplift) and the resisting force of the grain (friction, tilting action, self-weight) in a
sloping channel, Shields (1936) formulated the concept of the dimensionless Shields
number θ. This number depends on the bed shear stress τ, the grain diameter D, the
channel slope S, and the ratio of the sediment density to the density of water s = ρs/ρ.

τ ρ ghS (3.5)
τ hS hS
θ= = = (3.6)
g (ρ ρ) D (ρ ρ− )D ( − )D

RICKENM-Book.indb 34 2/23/2016 7:39:13 AM

 Fluvial bedload transport 35

The critical Shields number at the start of transport is denoted by θc. Many
bedload transport formulae are based on this assumption of a constant limiting shear
stress for a given grain size, even if the spatial shear stress distribution near to the
bed should be considered as a random parameter in a more exact analysis due to
the effects of turbulence. With increasing channel slope, the erosional resistance of
grains on the bed is theoretically reduced due to the slope-parallel weight component
favoring destabilization (Chiew & Parker 1994). However, this effect is superim-
posed by reduced flow forces on the bed in the case of small relative flow depths at
steeper channel slopes. Measurements show that the second effect predominates; that
is, increasing values of θc were also determined for increasing channel slopes (Eq. 3.7,
Eq. 3.8; Fig. 3.9).

θc = 0 15 S 0 25 (3.7)
(Lamb et al., 2008)
−γ
⎛ D ⎞
θci = ( .32S + 0. )⎜ i ⎟ (3.8)
⎝ D50 ⎠ (Recking 2009)

For a natural bed with a broad grain size distribution, the calculation of the
bedload transport can also be carried out separately for the different grain fractions
(classes of grain size). For this calculation of fractional bedload transport rates, a
so-called hiding function is often used to modify the Shields criterion compared
to simpler calculations for a standard representative grain size only (Parker 2008).
In Eq. 3.8 the dependence on both the slope and also the hiding function are taken
into account for the initiation of particle motion; the index i refers to a percentile of
the grain size distribution. The exponent γ lies typically in the range of 0.64 to 1.0
(Recking 2009). An exponent γ = 1 signifies that all grain sizes start moving at the
same absolute critical bed shear stress, while with an exponent γ = 0 the absolute

Flume data 101

Field data 0.25 Di=D50
S
0.15 Di=D84
Critical Shileds strees τ*c

10 -1 τ *c = Di=D16
τ*c = 0.06 100 a=3.24, b=0.092, r2=0.56
a=1.32, b=0.037, r2=0.80
a=0.56, b=0.021, r2=0.75
θC

τ*c = 0.03

10−1

10-2

10−2 −4
10−4 10−3 10−2 10−1 100 10 10−3 10−2 10−1 100
(a) Channel slope S = tan β (b) S

Figure 3.9 Variation of the critical SHIELDS number θc with channel slope S and empirical equations.
(a) data from LAMB et al. (2008) with Eq. 3.7; where τ*c = θc; (b) data from RECKING (2009)
with lines obtained using the function θc = aS + b. Eq. 3.8 with the exponents γ = 0.93
shows good agreement of +/−50% with the data of Fig. 3.9(b). Both data sets include flume
and field data. Figure (a) from LAMB et al. (2008), Figure (b) from RECKING (2009), both with
permission from Wiley/American Geophysical Union.

RICKENM-Book.indb 35 2/23/2016 7:39:13 AM

 36 Methods for the quantitative assessment of channel processes in torrents (steep streams)

critical bed shear stress varies linearly with grain diameter (i.e. the Shields number is
constant and independent of the grain size to be moved).
In mountain rivers and torrents the shape of the bed, the low relative flow depth
and the broad grain size spectrum exert a significant influence on the bedload mobi-
lization (Palt 2001). For these channels, therefore, the following aspects should be
considered with regard to the bedload mobilization:

• Ability to determine a representative grain size distribution, as well as the grain

packing and the compactness of the grain structure
• Type of dependence between channel slope, characteristic grain size and relative
flow depth
• Selective start of transport for different grain fractions
• Start of transport with or without armor layer
• Influence of the bed and channel geometry (deformed channel bed—plane bed
after mobilizing all bed shapes)

For torrents, it is difficult to differentiate between the armor layer and the under-
lying sediment (subsurface layer), as a result of the heterogeneous origin of the bed
sediments (heterogeneous mixture of various sediment sources). Thus, in a torrent,
a grain size analysis of the coarse surface layer along the stream channel may reflect
more the spatially-different lateral sediment inputs of scree material rather than the
hydraulic sorting process. Due to the large grain size spectrum, determining complete
grain size distributions is only possible using a statistical combination of different
analysis methods (Fehr 1987a).
Many empirical investigations exhibit a relationship between channel slope, rela-
tive flow depth and start of mobilization. Palt (2001) traces the apparent relationship
between channel slope and relative flow depth back to higher flow velocities at steeper
slopes. Smaller relative flow depths are associated with a reduction of the near-bed
flow velocity and, thus, they result in smaller shear stresses near the bed (Bezzola
2002). The apparent increased resistance of steeper channels is explained by Palt
(2001) by the deformation of the bed due to the formation of bed structures that are
initiated above a slope of 0.01. In steep channels there may be also additional flow
resistance caused by the presence of large immobile boulders (e.g. Yager et al., 2007;
Nitsche et al., 2012a).
After Shields (1936), the critical shear stress for hydraulically-rough beds has
the constant value of approximately 0.05. For non-uniform grain size distributions,
Meyer-Peter & Müller (1948) calculated a critical Shields number θc of 0.047,
which was determined for a range of slopes up to 2.3% and relative flow depths > 10.
For S smaller than approximately 2%, θc lies typically in the range 0.03 to 0.06, for
D = D50 (Buffington & Montgomery 1997). Bezzola (2002) considered the influ-
ence of grain shape and estimated the associated range of variation of θc to be about
40%, whereby θc would lie between 0.028 and 0.066 (c.f. also Fig. 3.9).
Since the determination of a representative mean flow depth is often difficult
in torrents, the use of an alternative mobilization criterion is attractive, whereby
a critical discharge per unit channel width, qc, is determined instead of a critical
Shields number. Based on flume measurements for channel slopes in the range

RICKENM-Book.indb 36 2/23/2016 7:39:15 AM

 Fluvial bedload transport 37

0.025 ≤ S ≤ 0.20, Bathurst et al. (1987) proposed an approach (Eq. 3.9), which has
been slightly modified by Rickenmann (1991), namely (Eq. 3.10):

qc 0 15 g 0.5 D50
15
S −1.12 (3.9)
(Bathurst et al., 1987)
qc 0 065 (s − 1)1.67 g 0 5 D50
1 .5
S −1.12 (3.10)
(Rickenmann 1991)

A more recent empirical equation by Bathurst (2013) (Eq. 3.11; Fig. 3.10) is
also based on flume measurements and differs insignificantly from the earlier Eq. 3.9:

qc 0 13 g 0.5 D50
15
S −1.146 (3.11)
(Bathurst 2013)

An armor layer may be present in mountain rivers if the fine bed material is
washed out during the receding limb of a flood hydrograph. For the critical discharge
at the break-up of the armor layer (i.e. the start of transport of grains from the sub-
surface layer) Jäggi (1992), based on the investigations of Günter (1971), proposed
the following relationship for the dimensionless bed shear stress θc,D:

1000
Bathurst (2013), eq. 3.11

100
qc / [g0.5 D1.5]

10

0.1
0.001 0.01 0.1 1
S

Figure 3.10 Normalized critical unit discharge qc at initiation of bedload motion as a function of the
channel slope S. Shown are (i) data from flume experiments performed with relatively
uniform grain sizes, as compiled by BATHURST (2013) and (ii) Eq. 3.11 developed in the same
study. Modified from BATHURST (2013).

RICKENM-Book.indb 37 2/23/2016 7:39:15 AM

 38 Methods for the quantitative assessment of channel processes in torrents (steep streams)

2/3 2/3
⎡ Dm, D ⎤ ⎡D ⎤
θc,D θc ⎢ ⎥ ≈ θc ⎢ 90 ⎥ (3.12)
⎣ Dm ⎦ ⎣ Dm ⎦ (Jäggi 1992)

where Dm,D is the mean grain diameter in the armor (or surface) layer and Dm is the
mean grain diameter of the subsurface layer; Dm,D can be approximated and replaced
by the D90 value of the subsurface layer (Jäggi 1992). Using the Manning-Strickler
equation, it can be shown that the discharge per unit width has the following propor-
tionality: q ∼ h5/3 ∼ θ5/3. Thus, based on Eq. 3.12, the critical discharge per unit width
at the break-up of the armor layer qc,D can be expressed as follows (where qc may be
determined using Eq. 3.9 or Eq. 3.11 (Badoux & Rickenmann 2008)):

5/ 3
⎡⎛ D ⎞ 2 / 3 ⎤ ⎡D ⎤
10 / 9

qc , D qc ⎢⎜ 90 ⎟ ⎥ = qc ⎢ 90 ⎥ (3.13)
⎢⎣⎝ Dm ⎠ ⎥⎦ ⎣ Dm ⎦ (Badoux & Rickenmann 2008)

Making a similar transformation for steeper channels, it is preferable to base the criti-
cal discharge on the (simplified) VPE Eq. 2.17 for small relative flow depths, where
the discharge per unit width is q ∼ h5/2 ∼ θ5/2. The critical discharge per unit width at
the break-up of the armor layer qc,D can then be expressed as follows:

5/ 2
⎡⎛ D ⎞ 2 / 3 ⎤ ⎡D ⎤
5/ 3

qc , D qc ⎢⎜ 90 ⎟ ⎥ = qc ⎢ 90 ⎥ (3.14)
⎢⎣⎝ Dm ⎠ ⎥⎦ ⎣ Dm ⎦

If it is assumed that, on average, D90/Dm ≈ 2, then, using Eq. 3.13, the ratio (qc,D/qc) is
given by (qc,D/qc) = 2.2 and with Eq. 3.14 the ratio is (qc,D/qc) = 3.2; that is, there is a
considerable increase of the critical discharge for the break-up of an armor layer for
steeper channel slopes.
Further approaches to account for the formation of an armor layer are described
in Porto & Gessler (1999), Hunziker & Jaeggi (2002) and Weichert & Bezzola
(2002). Another equation to determine the critical discharge qc,B on the break-up up or
destruction of a “pavement-type layer” consisting of coarse granular material is based
on investigations of the stability of block ramps with large blocks by Witthaker &
Jäggi (1986) for ramp slopes in the range 0.05 ≤ S ≤ 0.25:

qc , B 0 (s − 1)0.5 g 0 5 D65
1 .5
S −1.17 (3.15)
(Witthaker & Jäggi 1986)

Here, the grain size corresponds to D65, a “mean” diameter for the large blocks,
and, in a torrent channel, this quantity could be approximated roughly by the grain
size D90. Investigations into bedload transport carried out for flood events in the
Valais in 2000 showed that the start of transport in flatter channels with S smaller
than approximately 5% is sometimes clearly overestimated by Eq. 3.15 (Badoux &
Rickenmann 2008).

RICKENM-Book.indb 38 2/23/2016 7:39:15 AM

 Fluvial bedload transport 39

Field measurements for the start of transport were made in mountain rivers in the
Himalayas by Palt (2001). These data are compared with discharge-based approaches
in steep channels in Fig. 3.11. This comparison suggests that the wide (grain size-
dependent) range of values for the start of bedload transport in steep channels can be
approximated by assuming lower and upper limits for the start of mobilization. As a
lower limit, Eq. 3.9 or Eq. 3.11 are proposed, after Bathurst (2013). As an upper
limit (transition to the general mobilization of all fractions), Eq. 3.14 or Eq. 3.15 after
Whittaker & Jäggi (1986), for example, could be used.
Another interesting comparison of these discharge-based approaches for steep
channels can be made with data from laboratory or controlled field environments
where rock riprap was subjected to overtopping flow conditions on steep slopes.
In the compilation of Abt et al. (2013), a total of 96 experiments are reported in

100
Subsurface material I

Subsurface material II

Armor layer material

Whittaker & Jäggi (1986)

10
Palt & Dittrich (2002), fit
armor layer material
qc/[g0.5 D1.5]

Linford, Saunders, Olivier

(in Knauss 1979)

Bathurst (2013)

1 Knauss (1979)

Palt & Dittrich (2002), fit

subsurface material

0.1
0.01 0.1 1
S

Figure 3.11 Discharge-based approaches for bedload mobilization compared with field data from
mountain rivers in the Himalayas (PALT 2001), in the form of plots of normalized critical
unit discharge qc as a function of the channel slope S.The different equations proposed for
qc are given in PALT & DITTRICH (2002) and in ABT et al. (2013). The field data refer to finer
bedload material from subsurface layer transported over an armor layer, and to bedload
material from the armor layer. Eq. 3.11 represents approximately a lower limit criterion,
whereas Eq. 3.15 represents approximately an upper limit criterion.

RICKENM-Book.indb 39 2/23/2016 7:39:16 AM

 40 Methods for the quantitative assessment of channel processes in torrents (steep streams)

data in Abt et al. (2013)

Bathurst (2013), eq. 3.11
100 Whittaker & Jäggi (1986), eq. 3.15

10
qc/[g0.5 D1.5]

0.1
0.01 0.1 1
S

Figure 3.12 Discharge-based approaches for bedload mobilization compared with flume and field data
for the failure of rock riprap layers at steep slopes (data reported in ABT et al. (2013),
shown as normalized critical unit discharge qc as a function of the channel slope S. Eq. 3.11
represents approximately a lower limit criterion, whereas Eq. 3.15 represents approxi-
mately an upper limit criterion.

which discharge was increased until the riprap layer started to fail. Here, the failure
unit discharge is set equal to the critical unit dischare qc, and these data are compared
with Eq. 3.11 and Eq. 3.15 in Fig. 3.12. The data from the riprap experiments tend to
be closer to the upper limit criterion (Eq. 3.15) for the very steep slopes.

3.4.3 Approaches for calculating bedload transport

In the evaluation of bedload transport processes one must differentiate between
approaches for calculating bedload transport capacity and approaches for calculating
or estimating the probable actual bedload transport. The bedload transport capac-
ity indicates a maximum transport level for given hydraulic conditions. The second
type of approach should give a result closer to the actual bedload transport rates

RICKENM-Book.indb 40 2/23/2016 7:39:16 AM

 Fluvial bedload transport 41

(e.g. taking account of additional energy losses), which for mountain rivers are up to
almost an order of magnitude lower (Jäggi 1992), and for torrents are up to several
orders of magnitude lower (Rickenmann 1997a, 2001a) than the transport capac-
ity. The approaches to calculate the bedload transport capacity are based on flume
experiments (often with a plane bed without armor layer) with an adequate sediment
input from upstream, and, mostly, with relative flow depths greater than about 7
(Meyer-Peter & Müller 1949; Smart & Jäggi 1983; Rickenmann 1990). Modi-
fied approaches were derived and tested, based generally on field data from chan-
nels with coarse stream bed structures, which partly included an armor layer and a
limited sediment availability (Palt 2001; Rickenmann 2005a; Chiari et al., 2010;
Rickenmann & Recking 2011; Nitsche et al., 2011; Schneider et al., 2014).
The equations for calculating the bedload transport capacity can be presented in
a standard way using the Shields number θ (Eq. 3.6) and the dimensionless bedload
transport rate Φb according to Eq. 3.16. The unit bedload transport rate qb (per meter
channel width) is then calculated using Eq. 3.17.

qb qb
Φb = = (3.16)
⎛ ρs − ρ ⎞ 3 (s 1)gD3 dimensionsless unit bedload transport rate
⎜⎝ ρ ⎟⎠ gD

qb Φ b (s − )gD3 (3.17)
unit bedload transport rate

Based on flume investigations, Meyer-Peter & Müller (1948) developed the

well-known formula for bedload transport in gravel-bed streams, which is valid for
channel slopes in the range 0.0004 ≤ S ≤ 0.023:

15
⎡⎛ k ⎞ 1 5 ⎤
= 8 ⎢⎜ stt ⎟ θ − θc ⎥ = 8 [ ]
15
b − c (3.18)
⎢⎣⎝ ko ⎠ ⎥⎦ Meyer-Peter & Müller (1948)

The expression (kst/ko)1.5 in Eq. 3.18 reduces θ to θ ′ when taking into account
the energy losses due to form roughness. In Eqs. 3.6, 3.16 and 3.17, D = Dm is to be
used to determine θ and Φb, where Dm is the arithmetic mean value of the grain size
distribution. After Hunziker (1995), the bed resistance was underestimated in these
investigations and he proposed a reduction of the coefficient from 8 to 5 (see also
Hunziker & Jäggi 2002). A similar reduction of the coefficient was also proposed
by Wong & Parker (2006).
Smart & Jäggi (1983) extended the investigations of Meyer-Peter & Müller
(1948) to steep channel slopes and developed a slightly modified equation valid for
0.0004 ≤ S ≤ 0.20:

θ ( s ) Dm ⎞
02
⎛ D90 ⎞ ⎛
qb 4 (s − )−1 ⎜ qS1 6 ⎜ 1 − c (3.19)
⎝ D30 ⎟⎠ ⎝ hS ⎟⎠
Smart & Jäggi (1983)

RICKENM-Book.indb 41 2/23/2016 7:39:16 AM

 42 Methods for the quantitative assessment of channel processes in torrents (steep streams)

Here, the ratio (D90/D30) represents an empirical correction of the transport efficiency
related to the width of the grain size distribution, where, according to the experi-
ments, the maximum ratio (D90/D30) was 10. This ratio is often exceeded in torrents
and is subject to large fluctuations. The correction factor increases the bedload trans-
port and is qualitatively in agreement with the increased bedload transport of the
gravel fractions and larger sediments with increasing sand content in the surface layer,
as proposed by Wilcock & Crowe (2003).
Rickenmann (1990, 1991, 2001a) analyzed the data of Meyer-Peter & Müller
(1948) and Smart & Jäggi (1983), together with data from further flume tests taking
into account increased concentrations of fine material in a clay suspension. He proposed
the following equation to account properly for the effect of a changed fluid density:

02
⎛ D90 ⎞
Φb = 3 1(s − )−0 5 ⎜ θ 0.0 5 ( − ) Fr1.1 (3.20)
⎝ D30 ⎟⎠
c
Rickenmann (1990, 1991)

Eq. 3.20 is valid for channel slopes in the range 0.0004 ≤ S ≤ 0.20, and Fr = V/(gh)0.5
is the Froude number. Eq. 3.20 was simplified by Rickenmann (2001a) to Eq. 3.21
by approximating the exponent of Fr as 1.0, setting (D90/D30)0.2 = 1.05 as for uniform
bed material after Smart & Jäggi (1983), and taking s = 2.68 for the density ratio
of quartz sediment to water. For the development of Eq. 3.20, D = Dm was used in
Eq. 3.6 to determine θ and Φb, whereas, in later applications of Eq. 3.20 and Eq. 3.21,
D = D50 was used (Nitsche et al., 2011; Heimann et al., 2015b). In addition, using
the definitions for Φb and θ as well as the continuity equation q = Vh, Eq. 3.21 was
transformed into the discharge-based equation Eq. 3.23 (Rickenmann 2001a).

Φb = 2 5 θ 0 5 ( − c ) Fr (3.21)
Rickenmann (2001a)
02
⎛ D90 ⎞
qb 3 1(s − ) −1 5
⎜⎝ D ⎟⎠ (q qc ) S1 5 (3.22)
30 Rickenmann (2001a)

qb 1 5 (q − qc ) S1 5 (3.23)
Rickenmann (2001a)

According to the mathematical transformation, the term qc in Eq. 3.22 and Eq. 3.23
has to be multiplied by V/Vc (with the critical flow velocity Vc, corresponding to the
discharge for θc). However, this is neglected, since for qc, mostly empirical functions
are used. Eq. 3.22 is valid for channel slopes in the range 0.0004 ≤ S ≤ 0.10; for higher
slopes in the flume tests, due to larger bedload concentrations, the flow depth was sig-
nificantly increased, which, in Eq. 3.20, is taken into account by θ but is not considered
in Eq. 3.22. Eq. 3.22 has the advantage that, even without detailed information on the
flow hydraulics, a comparison with field measurements is possible, provided that the dis-
charge is known or can be estimated. The comparison of some formulae with the flume
data of the hydraulics laboratory at ETH Zurich (VAW-ETH) is shown in Fig. 3.13.

RICKENM-Book.indb 42 2/23/2016 7:39:17 AM

 Φb, calculated
100
perfect agreement
data Meyer-Peter & Müller
10 data Smart & Jäggi
data Rickenmann

0.1

0.01

0.001

0.0001
0.0001 0.001 0.01 0.1 1 10 100
Φb, observed
(a)

Φb
100

10

0.1
eq. 3.18 (MPM)

eq. 3.20 (Ri), Fr=3

eq. 3.20 (Ri), Fr=2

0.01
eq. 3.20 (Ri), Fr=1

eq. 3.20 (Ri), Fr=0.5

0.001 data Meyer-Peter Müller

data Smart & Jäggi

data Rickenmann

0.0001
0.01 0.1 1 10
(b) θ [Ri, SJ] or θ’

Figure 3.13 (a) Comparison of the transport rates measured in laboratory flume tests (VAW-ETH
data) with the transport rates calculated using Eq. 3.20, expressed as dimensionless
transport rate. (b) Comparison of Eq. 3.18 after MEYER-PETER & MÜLLER (1948) (MPM) and
Eq. 3.20 after RICKENMANN (1990) (Ri) with the same flume data as in (a). In figure (b) the
MPM data were corrected using θ′ in Eq. 3.18.

RICKENM-Book.indb 43 2/23/2016 7:39:18 AM

 44 Methods for the quantitative assessment of channel processes in torrents (steep streams)

For the flume data of VAW-ETH with channel slopes in the range 0.03 ≤ S ≤ 0.20
the following equation was determined:

02
12.6 ⎛ D90 ⎞
qb = ⎜⎝ D ⎟⎠ (q qc ) S 2 (3.24)
(s − )
16
30 Rickenmann (1990, 1991)

Eq. 3.24 shows better agreement with the data for S ≥ 0.10 than Eq. 3.22. Other
bedload transport investigations for steep channel slopes (Mizuyama 1981; Ward
1986) led to an equation similar to Eq. 3.24 with an exponent of 2 for the channel slope
factor. Fig. 3.14a compares measured bedload transport rates and rates calculated
with Eq. 3.24 using the VAW-ETH data for steep channels. In the same figure two
other independent data sets are also shown: the flume tests of Aziz & Scott (1989)
in a conventional flume were obtained for channel slopes in the range 0.03 ≤ S ≤ 0.10
and with sand grain sizes from 0.29 to 1 mm; the flume tests of Nnadi & Wilson
(1992) were carried out in a closed horizontal channel under pressure, with pressure
gradients equivalent to 0.013 ≤ S ≤ 0.206, sand grains of 0.7 mm size and nylon
particles of 4 mm size.
For steep slopes, the weight component of sediment grains parallel to the channel
slope contributes to the bedload transport, and Abrahams et al. (2001) or Abrahams
(2003) based on Schoklitsch (1914) proposed the following correction, which
incorporates an increased slope factor Sk as follows:

⎛ sin φs ⎞
Sk S⎜ = Sak (3.25)
⎝ sin(φs − β) ⎟⎠ (Abrahams 2003)

where ϕs is the natural slope angle (friction angle) of the bedload particles under
water and β is the angle of the channel slope. Eq. 3.25 in combination with Eq. 3.22
was applied to the VAW-ETH data as well as to those of Aziz & Scott (1989).
The results are shown in Fig. 3.14(a, b), together with the bedload transport rates
for the data of Nnadi & Wilson (1992), calculated using Eq. 3.22. The two differ-
ent approaches lead to a similarly good agreement with the observed values. With
the slope correction, Eq. 3.25, an equation of the type Eq. 3.22 or Eq. 3.23 can be
applied over a very large range of slopes 0.0004 ≤ S ≤ 0.20. At high transport rates,
the mixture flow depth, as compared with the purely water-flow depth, is increased
significantly, resulting in a large bed shear stress. This effect is considered implicitly
in Eq. 3.20, since the mixture flow depth was used in its derivation.

3.4.4 Consideration of energy losses

The transport formulae for steep channel slopes, e.g. Eq. 3.20 or Eq. 3.22, can be
extended using the efficiency factor α (i.e. by multiplying a “constant” coefficient
of e.g. 1.5 in Eq. 3.23 by the factor α), which can adopt values between 0.001
and 1 (Rickenmann 2001a). In mountain rivers and torrents with bed slopes

RICKENM-Book.indb 44 2/23/2016 7:39:18 AM

 qb, calculated
(kg/s/m)
100

10

perfect agreement

0.1 VAW-ETHZ: MPM

VAW-ETH: SJ+Ri
Nnadi & Wilson (1992)
Aziz & Scott (1989)

0.01
0.01 0.1 1 10 100

(a) qb, observed (kg/s/m)

qb, calculated (kg/s/m)

100
perfect agreement
VAW-ETH: SJ+Ri
Nnadi & Wilson (1992)
Aziz & Scott (1989)
10

0.1

0.01
0.01 0.1 1 10 100
(b) qb, observed (kg/s/m)

Figure 3.14 (a) Comparison of the measured bedload transport rates with the values calculated using
Eq. 3.24 for the VAW-ETH data for steep channel slopes and two independent data sets
from AZIZ & SCOTT (1989) and NNADI & WILSON (1992). (b) Comparison of the measured
bedload transport rates with the values calculated using Eq. 3.22. For the VAW-ETH data
(MPM: MEYER-PETER & MÜLLER; SJ: SMART & JÄGGI; Ri: RICKENMANN) of all channel slopes and
the data of AZIZ & SCOTT (1989) the slope correction using Eq. 3.25 was taken into account
(after RICKENMANN 2005a).

RICKENM-Book.indb 45 2/23/2016 7:39:19 AM

 46 Methods for the quantitative assessment of channel processes in torrents (steep streams)

steeper than about 3 to 5%, observed values of bedload transport were found,
in general, to be smaller than values calculated with bedload transport
equations (Rickenmann & Koschni 2010). Here the following aspects should be
considered:

• The influence of form losses or macro-roughness effects on flow resistance is dif-

ficult to quantify.
• The start of substantial bedload transport is difficult to quantify, since, with a
broad distribution of grain sizes in a river bed, the flow conditions and the start
of movement are influenced in a complex manner.
• The bedload transport is often limited by the sediment availability and not by the
transport capacity.
• For bed slopes steeper than about 10%, according to Eq. 3.20 or Eq. 3.24 high
solids concentrations result, which are more plausible for debris floods (transitional
regime) or debris flows than for fluvial bedload transport (Rickenmann 2005a,
2012).

To determine bedload transport taking into account energy losses due to coarse
roughness elements (so-called macro-roughness), Eq. 3.20 or Eq. 3.22 (or any other
transport equation) can be applied in combination with a method for the partitioning
of the flow resistance (see chapter 2.4). Rickenmann (2005a) introduced an empirical
function to estimate a reduced slope of the energy line for large-scale roughness condi-
tions in steep channels. With this method a better agreement between bedload volumes
observed in nature and calculated could be achieved for various flood events (e.g.
Chiari et al., 2010; Chiari & Rickenmann 2009, 2011; Badoux & Rickenmann
2008). This approach was modified by Rickenmann & Recking (2011), who used an
extended database. The partitioning of flow resistance is based on an earlier proposal
by Meyer-Peter & Müller (1948) and later tested by Palt (2001). The reduced
energy slope Sred is calculated with reference to a base level of the flow resistance (for
a basic roughness of the bed material) and determines the energy that is available for
the bedload transport:

e e
⎛ f ⎞ ⎛ n ⎞
Sred S⎜ o ⎟ =S⎜ o ⎟ (3.26)
⎝ tot ⎠
f ⎝ ntot ⎠ (Rickenmann & Recking 2011)

According to the r Darcy-Weisbach flow law (Eq. 2.2), the slope of the energy
line S is proportional to the friction coefficient f or, according to the flow law
after Manning-Strickler, Eq. 2.10, it is proportional to the Manning coefficient
n squared, and the exponent e should have the value 2. Meyer-Peter & Müller
(1948), based on theoretical considerations, showed that e can also take on smaller

RICKENM-Book.indb 46 2/23/2016 7:39:19 AM

 Fluvial bedload transport 47

values (down to 1.33), and based on their experimental results they proposed an
empirically-determined value e = 1.5. Rickenmann et al. (2006a) suggested that plau-
sible values for e may lie in the range 1 ≤ e ≤ 2.
For the calculation of bedload transport the reduced energy slope Sred is either
introduced via θ in Eq. 3.20 or directly in Eq. 3.22. The values for θc are determined
empirically and refer in general to the total flow resistance (or the total bed shear
stress). Thus, in the use of Eq. 3.20 θc also has to be reduced. The reduced value of θc
can be determined such that θc,r = hc Sred(hc) [(s−1) D50]−1 corresponds to the discharge
conditions at the start of bedload transport, i.e. Sred(hc), and thus θc,r is constant for
a given channel slope and a given grain size distribution (D50, D84) (Nitsche et al.,
2011, 2012b). Alternatively, θc can also be reduced as follows, using a discharge-
dependent value of Sred: θc,r = θc (Sred/S). It is difficulty to verify which of the two
approaches is more plausible.
The approach of Rickenmann & Recking (2011) given in chapter 2.4 for the
partitioning of flow resistance is basically a function of the relative flow depth. However,
as a general empirical approach, it implicitly contains information about a mean increase
in roughness in steep and rough channels. In the study of Nitsche et al. (2011), other
ways of partitioning the flow resistance were investigated, including, for example, con-
sideration of the additional energy losses caused by large immobile boulders (Yager
et al., 2007; Whittaker et al., 1988) or by step-pool sequences (Egashira & Ashida
1991). All these approaches were combined with Eq. 3.21 and Eq. 3.26 with an expo-
nent e = 1.5, and the calculated bedload transports were compared with observations
of the transported bedload volumes (flood events in Switzerland in 2005; flood events
in canton Valais, Switzerland, in 2000; long-term discharge and bedload measure-
ments of the WSL institute in torrents in Switzerland). Overall, for all channel types
(stream bed morphologies), the best results were obtained with the empirical approach
of Rickenmann & Recking (2011) and with the more physically-based approach of
Yager et al. (2007). A summary of these results is presented in Fig. 3.15.
Table 3.3 shows the combinations of equations used for the bedload transport
calculation and for the partitioning of the flow resistance, according to which the
results are arranged in Fig. 3.15. For detailed investigations for a given channel type
(e.g. influence of large boulders in different concentrations), specific approaches
should be preferred (Yager et al., 2007; or Whittaker et al., 1988 for channel
slopes not exceeding about S ≈ 0.07); however, these approaches require more exact
investigations of the riverbed morphology. To have some idea about the uncertainty
of the estimates of bedload transport, the most suitable approaches for partitioning
the flow resistance for a given channel type can be used to examine the range of
possible results.
The exponent e = 1.5 used in Nitsche et al. (2011, 2012b) and in Schneider
et al. (2014) is near the range of the best exponents according to simulations with
the software SETRAC using an earlier approach for flow resistance partitioning
(Chiari & Rickenmann 2009, 2011).

RICKENM-Book.indb 47 2/23/2016 7:39:19 AM

 Long term data Event data All data
1
2.4 3.2 2.7
Ri−no

0.5
0

1
2.1 1.0 1.4
Ri−PC

0.5
0

1
1.1 0.4 0.9
Ri−W

0.5
0
Fraction of estimates

1
1.7 0.6 1.1
Ri−Y

0.5
0

1
0.6 0.2 0.3
Ri−EA

0.5
0

1
0.9 1.5 1.0
Ri−RR

0.5
0

1
10.2 3.2 7.0
P−Y

0.5
0
<0.3 0.3−3 >3 <0.3 0.3−3 >3 <0.3 0.3−3 >3
Vpred/Vmeas

Figure 3.15 Ratio of the calculated and measured bedload volumes (Vpred/Vmeas), calculated with different
combinations of equations (rows; defined in Table 3.3) and differentiated according to data
groups. The (Vpred/Vmeas) ratios are presented in three classes, of which the middle class rep-
resents all calculations within a factor 10 of the measured bedload transports. The gray
numbers indicate the mean value of the (Vpred/Vmeas) ratios. The group «Long-term Data»
consists of 207 transport events, while the group «Event Data» consists of 9 transport
events. The group «All Data» consists of the summed bedloads of the individual channels,
in order to weight each stream independently of the number of events in the same way.
The approach of WHITTAKER et al. (Ri-W) was not used for 4 channels. From NITSCHE et al.,
2011, with permission from Wiley/ American Geophysical Union.

RICKENM-Book.indb 48 2/23/2016 7:39:19 AM

 Fluvial bedload transport 49

Table 3.3 Abbreviations used for the combination of equations for bedload transport and for the
partitioning of flow resistance leading to the results shown in Fig. 3.15.

Bedload transport Equation Partitioning of flow resistance Equation Abbreviation

RICKENMANN (2001a) Eq. (22) No reduction – Ri-no

RICKENMANN (2001a) Eq. (27) PAGLIARA & CHIAVACCINI (2006) Eq. (10) + (11) Ri-PC
RICKENMANN (2001a) Eq. (27) WHITTAKER et al. (1988) Eq. (3) + (4) Ri-W
RICKENMANN (2001a) Eq. (27) EGASHIRA & ASHIDA (1991) Eq. (7) + (9) Ri-EA
RICKENMANN (2001a) Eq. (27) YAGER et al. (2007) Eq. (13) + (14) Ri-Y
RICKENMANN (2001a) Eq. (27) RICKENMANN & RECKING (2011) Eq. (15) + (16) Ri-RR
+ (17)

3.4.5 Transition to debris flood and debris flow

The mechanical behavior of debris flows is complex and depends on different factors
such as viscosity and turbulence of the mixture, dispersive forces due to collision of
the coarse components, friction forces between the (coarser) grains and the shear
strength of the matrix (consisting of fine components and water). Depending on the
dominance of these factors, a rough classification into mudflows and (granular) debris
flows is possible. This simplified presentation does not, however, take into account
the property that the solids concentration often decreases in the rearward region of
a mudflow or debris flow surge. Debris-flood conditions of solids transport can be
caused by the sudden input of solids, by the liquefaction of larger streambed reaches,
or after breaking through logjams and dams (Takahashi 1991).
For a volume proportion of more than about 5% silt or clay, debris-floods or hyper-
concentrated flows exhibit an increasing viscous behavior. From a volume proportion of
solids in total of 45 to 55% a debris flow or mudflow develops (Costa 1984; Julien &
O’Brien 1997). Empirical investigations in Switzerland indicate that debris flows initiate
typically in steep scree with slopes between 40 and 58%, in the contact zone between
bedrock/scree with slopes between 45 and 70%, in gullies with slopes between 45 and
70% and in stream channels with slopes between 23 and 65% (Haeberli et al., 1991;
Rickenmann & Zimmermann 1993). These results are in agreement with flume inves-
tigations with uniform sediments for which a change can be observed from the usual
mobilization mechanism to a sliding-type en mass instability of the channel bed for slopes
larger than 20% (Smart & Jäggi 1983). Thus in torrent channels with bed slopes of
more than ca. 20% and in the absence of stabilizing bed structures, transport processes
with the characteristics of debris flows can be expected (Jäggi 1992). The transport
formulae of Smart & Jäggi (1983) and Rickenmann (1990) predict that for channel
slopes steeper than about 10 to 15% and high flow intensities, solids concentrations
occur that are typical for debris-flood or debris-flow conditions. In the case of high flow
resistance (structured torrent channels) and discharges close to the start of transport such
formulae tend to overestimate the observed transport rates considerably if no correction
for the effective shear stress is made (such as discussed in chapter 3.4.4).
The flume experiments of Smart & Jäggi (1983), Rickenmann (1990) as well
as similar laboratory and field observations indicate that very high sediment con-
centrations occur for channel slopes steeper than about 20%. An extrapolation to
even steeper slopes leads to similarly high transport rates, as observed in field tests

RICKENM-Book.indb 49 2/23/2016 7:39:21 AM

 50 Methods for the quantitative assessment of channel processes in torrents (steep streams)

on the formation of debris flows. Further, above limiting slopes of approximately

20–25%, a general instability of the bed has to be considered, together with a con-
tinuous transition from fluvial bedload transport to debris floods and debris flows.
Likewise, comparisons of discharge criteria for the initiation of bedload transport
and for debris-flow formation (as well as of empirical flow resistance laws for water
runoff in steep channels and debris flows) indicate that a continuous transition is to be
expected (Rickenmann 2012). A fairly continuous transition of transported sediment
loads was observed between fluvial sediment transport and debris flows for a large
rainstorm event in Switzerland (Fig. 3.16).

Bedload volume (GF) / water runoff volume (Vw)

10
debris flow
fluvial transport
Rickenmann, eq. 3.23 (mod.)
MPM-HJ upper, eq. 3.18 (mod.) Rotlauibach
MPM-HJ lower, eq. 3.18 (mod.)
Haldibach
1

Glyssibach Acherlibach

0.1

0.01

0.001
GF = 1.95 VwS1.5

0.0001
0.1 1 10 100
Channel slope S (%)

Figure 3.16 Data from different types of transport processes indicate a relatively continuous tran-
sition from fluvial transport to debris flows. The data comes from the flood events of
August 2005 in Switzerland. The purple line corresponds to the integration of Eq. 3.23 for
fluvial bedload transport over the flood period, wherein a pore volume (voids content)
of the deposited material of 30% is considered, and the channel slope S was determined
upstream of the deposited material. The modified Eq. 3.23 (MPM-HJ) refers to the equa-
tion of MEYER-PETER & MÜLLER (1948) but using a coefficient of 5 according to HUNZIKER &
JÄGGI (2002), and also accounting for 30% pore volume. With the 4 debris-flow data points
denoted by the names of the streams, much sediment entered the channel due to large
landslides. Modified from RICKENMANN & KOSCHNI (2010).

RICKENM-Book.indb 50 2/23/2016 7:39:21 AM

 Fluvial bedload transport 51

3.4.6 Deposition slope behind check dams

The expected deposition slope upstream of check dams or in sediment retention basins
is an important parameter for the design of protection measures. In designing a series
of check dams, the deposition slope between successive check dams has to be known
or assumed to determine the exact locations (spacing) and heights of the individual
check dams. In the case of a sediment retention basin, the expected deposition slope
essentially defines (together with the width of the basin) the sediment volume that can
be stored to reduce the sediment load that will be transported further downstream
during an event.
According to bedload transport equations, the deposition slope will increase
with increasing sediment input from upstream and with coarsening of the trans-
ported solids. However, both the expected runoff hydrograph and the sediment
input from upstream are difficult to determine. In addition, in steeper channels and
headwater catchments, it is possible that not only fluvial bedload transport occurs
but also debris flows. These factors all complicate a “theoretical” estimation of the
deposition slope.
Therefore, in engineering practice, deposition slopes Sdep are mostly estimated
based on experience. A frequently-taken assumption is that Sdep may vary in the
range from (1/2)So to (2/3)So, where So is the original stream bed slope (Ikeya 1979;
Romang 2004; Planat 2008; Piton & Recking 2015a). In Japanese design guide-
lines for the construction of sediment retention basins it is recommended to use the
same range of expected deposition slopes (Pwri 1988). Published data on deposi-
tion slopes are very rare (e.g. Romang 2004); some observations made in steep
streams in Italy (Porto & Gessler 1999) and in Iran (Nameghi et al., 2008) are
illustrated in Fig. 3.17.

3.4.7 Comments on the estimation of the solids

transport
In torrents, due to the very variable delivery of solid material, a very broad grain size
distribution and a spatially-variable sediment availability, one must expect a very
large fluctuation of the solid material transport rate and a very noticeable phase of
selective transport. This implies that:

• The solids transport rate, especially for small to medium flow intensities, only
has a limited functional relationship with the discharge. As an upper threshold,
the calculated transport capacity (for quasi-plane bed conditions and unlimited
sediment availability), has the highest relative accuracy. The lower range can be
estimated from the (often limited) availability of solids, the use of an armor layer
criterion or a higher critical shear stress at the start of mobilization, or from a
consideration of energy losses.
• Owing to the broad grain size distribution and very heterogeneous sedi-
ment availability, selective bedload transport takes place for small to medium
flow intensities. This favors the formation of an armor layer and results in
spatially strongly variable grain size distributions. The macro-roughness
elements of the bed reduce the effective shear stress acting near the
stream bed.

RICKENM-Book.indb 51 2/23/2016 7:39:21 AM

 0.2
Aspromonte group (n = 91)
Sila group (n = 41)
Sdep = (2/3) So
0.15
Sdep = (1/3) So
Deposition slope Sdep

0.1

0.05

0
0 0.05 0.1 0.15 0.2
(a) Original bed slope So

0.3

check dams in Iran (n = 60)

Sdep = (2/3) So
0.25
Sdep = (1/3) So

0.2
Deposition slope Sdep

0.15

0.1

0.05

0
0 0.05 0.1 0.15 0.2 0.25 0.3
(b) Original bed slope So

Figure 3.17 Deposition slopes behind a series of check dams in comparison with the original slope of
the stream bed for (a) data from two Calabrian streams in Italy with gravel bed sediments
with D50 ranging from about 5 to 20 mm (PORTO & GESSLER 1999), (b) data from a stream
in Iran with mostly sandy sediments with D50 typically ranging from about 0.4 to 2 mm
(NAMEGHI et al., 2008).

RICKENM-Book.indb 52 2/23/2016 7:39:21 AM

 Fluvial bedload transport 53

3.4.7.1 Critical bed shear stress or critical specific discharge

Besides the characteristic grain diameter and the channel slope, the basis for analyzing
the bedload transport is primarily the ratio of actual to critical shear stress. For steep
channels, the critical shear stress reacts directly and very sensitively to variations in
flow depth, which is why some calculation methods also use the simpler approach of
specifying a critical specific discharge. More recent investigations show the tendency
of an increase in the critical dimensionless bed shear stress θc with increasing channel
slopes (Lamb et al., 2008; Recking 2009; Bunte et al., 2013), where by θc is deter-
mined using the total bed shear stress.

3.4.7.2 Transport reduction effect of the armor

layer—selective transport
Calculating θc or qc according to the armor layer criteria may result in considerably
reduced transport rates. This may be appropriate if no extreme events have to be con-
sidered, and when relatively small discharges occur over longer time periods. For the
range of fluctuation of the start of transport, with or without an armor layer, a factor
of about 2 to 3 with regard to θc or qc can be expected.

3.4.7.3 Calculation of solids transport

In torrents with typically steep channel slopes, only a few transport formulae have
been tested with field data. The two extreme cases consist of determining the transport
capacity (maximum possible transport rate) and a reduced transport rate due to high
flow resistances (considering additional energy losses, e.g. with Eq. 3.26). If bedload
transports are calculated according to the transport capacity, then these conditions
may be more representative for debris floods or debris flows. Calculations taking into
account a reduced energy slope are valid for conditions with fluvial bedload transport.

3.5 DRIFTWOOD IN TORRENTS AND MOUNTAIN

RIVERS

In mountainous and forested catchments wood can find its way into the stream chan-
nel through landslides, debris flows, erosion, snow avalanches or storms (windthrow)
(Rudolf-Miklau et al., 2011). Depending on the type and origin, one speaks also
of dead wood, old wood, fresh wood and wood from trees uprooted by avalanches.
For water-related transport and in-channel deposits of logs and rootstocks, the term
“large woody debris” was in use for some decades but has been replaced more recently
by “large wood” (e.g. Wohl et al., 2010; Jackson & Wohl 2015).

3.5.1 Flood hazards associated with driftwood

An overview of the problem of floating woody debris in mountain rivers and torrents
during high discharges is given by Hartlieb & Bezzola (2000), Mazzorana
et al. (2009, 2011), Rudolf-Miklau et al. (2011), Comiti et al. (2012), and

RICKENM-Book.indb 53 2/23/2016 7:39:22 AM

 54 Methods for the quantitative assessment of channel processes in torrents (steep streams)

Ruiz-Villanueva et al. (2014). In flood events large pieces of wood often cause
problems due to logjams at bridges, culverts or even natural constrictions like gorges.
The most important effects are: (i) logjams or temporary blockages obstructing water
flow and bedload transport in natural channel sections, which can favor the forma-
tion of debris flows in steep reaches, (ii) overtopping of water and sediment out of
the channel onto the fan or banks can lead to large amounts of deposited material
and debris. Another frequent and undesirable consequence of excessive large wood
is the partial or complete clogging of open check dams of sediment retention basins,
whereby the desired regulation effect (dosage) regarding bedload transport during a
flood event is impaired or completely inhibited (Piton & Recking 2015b). Further,
it can also lead to the destruction of bridges, or large wood pieces can cause impact
damage to buildings.

3.5.2 Origin and amount of large wood in stream

channels
Information on the type and origin of the wood in the channels may be found in
Hartlieb & Bezzola (2000), Rimböck (2003), Hassan et al. (2005b), Mazzorana
et al. (2009), Kasprak et al. (2012), and Gurnell (2013). Once the wood lies in the
channel, the following aspects are important: the shape and the dimensions of the
individual elements, whether the pieces have branches, and, particularly, the pres-
ence of and the proportion of rootstock. In addition, the type of wood and the water
absorption, and thus the density of the wood, influence the degree of mobilization in
the channel during floods.
Concerning possible amounts of wood, it is possible todifferentiate between
the effective amount of driftwood that is transported during a flood and the
amount of potential wood that can be mobilized from within the channel or
supplied from areas near the river. Investigationsbased on data from the Swiss
Alps, Japan and North America show that both the amount of transported
wood debris and the wood debris potential can be correlated roughly with the
catchment size (Rickenmann 1997b; Waldner et al., 2008). Furthermore, the
transported amount of driftwood also depends on the integrated water runoff or
on the transported bedload volume of a flood event (Rickenmann 1997b). To
make more exact statements about the potential amount of large wood, detailed
investigations in a given catchment are necessary, whereby factors like state of
the forest, erosion processes, channel profile and bed slope need to be considered.
Rimböck (2003) developed a method for estimating the wood debris potential
based on aerial photographs.
For the floods of 2005 in Switzerland a budget of large wood was determined
for selected catchments, including a quantitative assessment of wood recruitment
processes, namely of landslides, debris flows, bank erosion, and entrainment of in-
channel wood (Waldner et al., 2007, 2008). The contribution of in-channel wood
was based additionally on an earlier study of wood in torrent channels (Rickli &
Bucher 2006). A GIS procedure was developed for estimating potential wood debris
contributions due to landslides (Mächler, 2009). Essentially, the new data confirm
the approximate relationship between the amount of transported large wood in flood
events as a function of catchment area (Fig. 3.18).

RICKENM-Book.indb 54 2/23/2016 7:39:22 AM

 Fluvial bedload transport 55

floods August 2005 in Switzerland

100000
floods 1977−1996 in Switzerland

10000
Transported large wood (m3)

1000

100

10

1
0.1 1 10 100 1000 10000
2
Catchment area (km )

Figure 3.18 Transported volumes of large wood as a function of the catchment area, as observed after
flood events (mainly in Switzerland) (WALDNER et al., 2008).

3.5.3 Transport of driftwood and logjam hazards

The initial transport of the pieces of wood lying in the channel depends mainly on the
flow conditions, the type of wood (density, presence of branches, roots) as well as the
location with respect to the flow action (Bezzola et al., 2002; Braudrick & Grant
2000, 2001). Flume investigations indicate that the transport of logs begins at relative
flow depths h/d ∼ 0.5 to 1.5, where d is the log diameter. The critical values of h/d
increase in the above-mentioned range with increasing Froude number (as well as
with the increasing number of branches or with the presence of rootstocks), whereas
the critical values for h/d with simultaneous bedload transport are more likely to lie
in the lower range.
The transport distances of pieces of wood increase if the lengths of logs are smaller
than the mean channel width (Lange & Bezzola 2006; Schmoker & Hager 2011;
Lassettre et al., 2012; Lucia et al., 2015). Generally, driftwood floats at the surface.
After Rimböck (2003) coniferous wood (softwood), rootstocks and old dried wood
usually float on the surface due to their low density, whereas the heavy oak and beech
woods seldom float. Bulky, very branched wood, and rootstocks only float in the case
of large flow depths, and, otherwise, are often transported by rolling over the stream
bed. During the transport process the position of the wood continually changes due
to turbulent flow. Often the wood is aligned parallel to the direction of flow. During
transport the wood debris becomes smaller in size. Zollinger (1983) reports that a
whole tree with the crown and rootstock can have its branches removed and its bark
peeled off, and that the tree may be broken into 1 to 5 m long pieces within relatively
few meters during transport in a torrent channel.

RICKENM-Book.indb 55 2/23/2016 7:39:22 AM

 56 Methods for the quantitative assessment of channel processes in torrents (steep streams)

The deposition of driftwood takes place under natural conditions as soon as the
discharge decreases and the buoyancy and the force of flow for further transport are
no longer sufficient. After deposition of individual pieces and with the accumulation
of additional wood, fairly shallow heaps may be formed. With the transport of short
individual logs there is only a relatively small risk of logjams, since the logs can easily
align themselves in the flow direction and thus pass a constriction.
An overview of possible measures to reduce the risk of logjams is given by
Lange & Bezzola (2006). The risk of logjams near bridge cross-sections was inves-
tigated in flumes with hydraulic model tests (Bezzola et al., 2002). For a batch-wise
delivery of wood mixtures the probability of logjams pv (number of tests with logjams
in relation to all tests of the same category) reached values of 0.2 to 1.0. Noticeably,
pv was clearly larger if rootstocks were present. The probability of logjams of indi-
vidual pieces of wood depends mainly on their dimensions relative to the width of the
critical cross-section. In the case of individual logs, an increase of up to pv = ca. 0.4
was obtained in the range 0.5 < LW/B < 2, where LW = length of individual log and
B = width of the opening. In the case of rootstocks a striking increase of up to pv = 1.0
was obtained in the range 0.6 < dW∗/H < = 1, where H = clear height of the critical
cross-section, dW∗ = (dWmax dWmin Lh)1/3, with dWmax, dWmin = maximum and minimum,
respectively, of the dimension of the root plate and Lh = length of trunk extension.
To reduce the risk of logjams it is recommended that the bed width of the channel
should be about twice the dimension of the expected maximum length of the logs, and
that the clear height under the bridge should be at least 1.7 times the critical dimen-
sion of the expected rootstocks. The tests also show that the amount of driftwood
is primarily important for the temporal evolution of the logjam process. Whether
logjams occur depends, in the first instance, on the dimension and shape of the largest
components (Lange & Bezzola 2006).

3.6 CRITICAL CHANNEL CROSS-SECTIONS

AND POTENTIAL DEPOSITION

The assessment of the flood hazard along torrents and mountain rivers carrying bed-
load basically requires consideration of three questions: (i) Is the hydraulic conveyance
capacity of the existing channel adequate to discharge the flood without damage?; (ii)
When and where can intensive bedload transport lead to deposition with associated
flow overtopping?; (iii) When and where can high flows with little bedload lead to
erosion that could endanger the stability of banks and the foundations of structures?
The assessment of these three aspects can be carried out at two different levels of
detail: (a) with simple estimates of the hydraulic conditions and of the bedload trans-
port at critical locations (cross-sections), combined with an integrative assessment
of possible effects for the entire duration of the flood event; or (b) using numerical
models to simulate the hydraulics and the bedload transport, though experience with
these models is limited thus far for steeper slope ranges and for flows overtopping a
channel on the fan (e.g. Chiari & Scheidl 2015).
For the procedure with the simple estimates (a), the analysis methods presented in
the previous sections can be applied. If significant sediment depositions occur in the
channel area, flow overtopping is to be expected, together with deposition of bedload

RICKENM-Book.indb 56 2/23/2016 7:39:22 AM

 Fluvial bedload transport 57

outside the channel (e.g. on the fan). Especially prone to critical depositions are sud-
den concave changes in the longitudinal profile (decrease in channel slope without
increase of discharge). If there are also bridges at such critical locations, the risk of a
complete blocking of the flow cross-section is especially high in the case of driftwood
in addition to bedload transport. Analytical methods for predicting the depositional
behavior for a sudden change of slope are described in Bezzola et al. (1996) and in
French et al. (2001).
In calculating the bedload transport capacity of the channel on the fan, special
attention is required in the case of an artificially-paved (or concreted) and smooth
channel bed. In this case, the bedload transport capacity is considerably higher than
in a natural channel with a movable bed and modified calculation approaches are
necessary (Hunzinger & Zarn 1996; Smart & Jäggi 1983). If the bedload-carrying
flow leaves the channel, it is necessary to predict the flow paths and areas of deposi-
tion on the fan. This can be done mainly based on the fan topography, but structures
(buildings, roads) can also influence the flow and deposition behavior considerably.
Therefore, especially in populated areas, different scenarios of the flooding process
may have to be considered, depending on the depositional process (which can be
influenced in addition by the amount of driftwood). Basically, in the case of a spread-
ing out of fluvial deposition on the fan, the entire bedload volume (deducting the por-
tion deposited in the channel) has to be distributed along the flow path. The average
thicknesses of such deposits are likely to be less than for debris-flow deposits on the
fan. In the case of fluvial transport, the coarser bedload grains tend to be deposited
in steep zones, whereas the finer grains may be transported further downstream to
flatter zones.

3.7 NUMERICAL SIMULATION MODELS

There are several sophisticated hydraulic-sedimentologic numerical simulation models

for gravel-bed and sand-bed streams with limited channel slopes. For steep channels,
however, only relatively few simulation models have been developed, e.g. SHESED
(Wicks & Bathurst 1996), ETC (Mathys et al., 2003), SETRAC (Rickenmann
et al., 2006a), PROMAB (Rinderer et al., 2009), and sedFlow (Heimann et al.,
2015a). These models are similar to numerical sediment transport models that were
applied primarily in flatter mountain rivers, e.g. MORMO (Schilling & Hunziker
1995) and BASEMENT (Vetsch et al., 2005). However, experience with the use of
such simulation models for the steeper channels (and especially for torrents) has been
very limited. Especially regarding the two-dimensional simulation of bedload deposi-
tion on torrent fans, there is scarcely any experience, except for the recent study of
Chiari and Scheidl (2015).
The one-dimensional bedload transport model SETRAC (Rickenmann et al.,
2006a; Chiari et al., 2010) was tested for its suitability in the case of steep channels
by means of flume experiments (Kaitna et al., 2011) and well-documented bedload-
transporting flood events in August 2005 in Switzerland and in Austria (Chiari &
Rickenmann 2009, 2011). The simulation model was developed especially for use in
torrent catchments and in mountain rivers, taking into account a reduced transport
capacity due to high-energy losses caused by macro-roughness elements. TomSed is

RICKENM-Book.indb 57 2/23/2016 7:39:22 AM

 58 Methods for the quantitative assessment of channel processes in torrents (steep streams)

the follow-up model of SETRAC; it is freely available from http://www.bedload.at.

To take into account increased energy losses for bedload transport a method for par-
titioning flow resistance was developed for SETRAC. This method is based on (only)
373 flow velocity measurements. It leads to a similar reduction of the energy slopes
as with the other method (Rickenmann & Recking (2011) presented in chapter 2.4
for partitioning the flow resistance and based on 2890 flow velocity measurements.
In a new version of TomSed the new approach of Rickenmann & Recking (2011) to
partition flow resistance was also implemented. The one-dimensional bedload trans-
port model sedFlow also includes the new approaches to calculate flow resistance and
bedload transport in steep channels. Being based on rectangular cross-sections and
with an option for using simplified hydraulics without flow routing, it requires only
short calculation times. Thus the program sedFlow allows simulating many different
scenarios or conducting sensitivity analyses with the variation of different input and
model parameters in a relatively short time. The model sedFlow was calibrated with
observations of bedload transport in several Swiss mountain rivers (Heimann et al.,
2015b; Rickenmann et al., 2014, 2015). The model sedFlow is freely available from
http://www.wsl.ch/sedFlow.

RICKENM-Book.indb 58 2/23/2016 7:39:22 AM

 Chapter 4

Debris flows

4.1 PROPERTIES OF DEBRIS FLOWS

In steep headwater catchments in the Alps debris flows occur every year, and fre-
quently such events cause considerable damage. A debris flow is a rapidly flowing
mixture of soil with different amounts of water. At the front of a debris flow there
is a high concentration of solids, and these flows are characterized by an unsteady
and surging flow behavior, clearly distinguishable from a typically more steady water
discharge in a stream channel.
The grain composition of debris flows can vary considerably. In the Alps coarse
blocks frequently collect at the front of a debris flow. Coarse particles are often also
transported in the other parts of a debris-flow surge. In the case of channel overtop-
ping on the fan, a lot of coarse debris may be deposited there. This type is also termed
a granular debris flow (Fig. 4.1). In the case of mudflows (Fig. 4.2) the fine material
and the water dominate, whereas generally the coarse stones and blocks are missing
or they have a negligible influence on the flow behavior. In the rearward part of a
debris flow (or mudflow) the solids concentrations are usually smaller than in the
front part (Fig. 4.3). The deposition conditions are then similar to those caused by the
processes of fluvial bedload transport outside of a channel (Costa 1988; Hübl et al.,
2002; Pierson 2005).
Depending on the material composition, different theoretical approaches were
proposed to describe the flow behavior. However, a partitioning into various flow
types based on field observations is often only possible in a rudimentary way. This
difficulty of a simple identification of different flow types is also reflected in the termi-
nology and the classification of debris flows, and in different languages there are some
differences in meaning. A rough correspondence of the terms in German, French, Ital-
ian and English is presented in Table 4.1.
Hillslope debris flows (Hungr et al., 2001, 2014; Hürlimann et al., 2015) are
distinguished from debris flows basically by the place where they occur (terrain con-
ditions with weakly or no predetermined lateral limits to the flow path) and often by
relatively short flow distances, while the latter flow type typically runs in a channel
or a gully. Generally, hillslope debris flows do not occur several times at the same
place and also do not exhibit multiple surges. In the early stages, hillslope debris
flows (Fig. 4.4) can be compared with spontaneous shallow landslides, after a larger

RICKENM-Book.indb 59 2/23/2016 7:39:22 AM

 60 Methods for the quantitative assessment of channel processes in torrents (steep streams)

Figure 4.1 Front of a granular debris flow, Kamikamihori valley, Japan (photo H. SUWA).

Figure 4.2 Front of a mudflow, Jiangjia valley, China (photo Z. WANG).

flow direction

large wood
onset of turbulence
coarse particles in suspension
bouldery front
tail head

Transition fully developed debris flow

(debris flood)
boulder accumulation
normal streamflow (not liquefied)

Figure 4.3 Typical longitudinal section through a debris flow with decreasing solids concentration from
the front to the rearward part. Adapted from PIERSON (1986).

RICKENM-Book.indb 60 2/23/2016 7:39:22 AM

 Debris flows 61

Table 4.1 Terms for debris flow in a few languages.

German French Italian English

(Granularer) Lave torrentielle Colata detritica, Debris flow, granular or stony

Murgang lava torrentizia debris flows
Schlammstrom Coulée de boue, Colata di fango Mud flow
lave torrentielle
boueuse
Hangmure Coulée de boue Colata detritica Hillslope debris flow
de versant di versante (debris avalanche)

Figure 4.4 Example of hillslope debris flows (Sachseln, Switzerland). These often occur in non-forested
areas (photo: Oberforstamt Obwalden).

displacement distance the flow behavior may be similar to that of debris flows. In
rainstorm events shallow landslides may transform into hillslope debris flows.
The typical debris flows in the Alps can be considered in a simplified way to
be a mixture of the three main components water, fine material and coarse granu-
lar material. Based on their composition and flow behavior, debris flows are a
mixture of floodwaters, landslides and rock slides or debris avalanches. Fig. 4.5
shows the relative proportions of the three main components for such rapid mass
movements. Therefore, the physical processes in the formation, the flow and the
deposition of debris flows are correspondingly complex and are only partially
understood.
In comparison with floods with fluvial bedload transport in torrent channels,
debris flows have greater flow depth, may cause greater erosion and entrainment
of solid material, thus often transporting large amounts of debris to the fan or
confluence area. During a flood, particles are moved along the channel by the
driving force of the water. Debris flows with high solids concentrations typically
exhibit a much greater viscosity or frictional resistance than just water alone. For
the triggering of debris flows, a minimum amount of granular material is required
in addition to water, as well as steep slopes. The most important properties of
debris flow and of traces that are left behind on the terrain are summarized in
Table 4.2.

RICKENM-Book.indb 61 2/23/2016 7:39:24 AM

 62 Methods for the quantitative assessment of channel processes in torrents (steep streams)

water

river torrent

en r t in

be nspo
tra
n

dlo rt
sio
su nspo

ad
sp
t ra
mu ahar
ow
dfl
L

debris flow

rock
landslide avalanche

fine particles coarse particles

Figure 4.5 Main components of a debris flow in a three-phase diagram, in comparison with other rapid
mass movements. Modified from PHILLIPS & DAVIES (1991).

Table 4.2 Characteristic properties of debris flows (after RICKENMANN 1996).

Material composition and flow behavior:

• wide range of grain sizes, +/− uniformly distributed over flow depth
• transport of very large blocks possible
• highest bedload concentration at the front, further back often a more fluid mixture
• high bulk density of the water-solid mixture (ca. 1.6 ... 2.4 t/m3 at the front)
• high “viscosity” (non-Newtonian fluid)
• mostly discharge in waves, i.e. intermittent advance of one or more surges
• formation of debris walls (levées) in absence of already existing boundary of the flow
cross-section, e.g. by steep rock wall
• deposition especially at locations with widening of stream bed or with sudden decrease
of channel slope; in Alpine channels often at slopes of 5% ... 18%
Characteristic traces in terrain, which point to debris flow activity:
• levées (lateral debris walls)
• residual debris tongues, on the fan or in the case of local channel widening
• unsorted deposition (all grain sizes well mixed, no layering)
• coarse blocks and fine-grained matrix (if not yet washed out) in the deposits
• polished and scarred areas of rock
• clear boundary of the deposits
• often little damage to vegetation in the deposition areas outside of the channel
• erosion or flow cross-section usually U-shaped

If the banks of a channel are shallow, then natural levées tend to form due to
material deposition along the path of the debris flow, in a manner that createstheir
own lateral boundary of the flow cross-section. The debris is deposited typically in
flatter terrain and in a non-uniform way. During the depositional phase, the high

RICKENM-Book.indb 62 2/23/2016 7:39:25 AM

 Debris flows 63

viscosity or the large grain-to-grain friction of the decelerating mixture leads to a

relatively abrupt stop of the frontal part of the debris flow. Thus, the front part of the
deposits is usually demarcated clearly from the old terrain. The irregular deposition
of multiple surges results in a rough terrain surface of a debris-flow fan. Observers of
debris flows report that such events are often accompanied by a loud noise, ground
vibrations and sometimes also by a sulfuric smell. Occasionally these phenomena
have also already been noticed shortly before the arrival of the debris flow. A more
detailed description of the debris flow process may be found, e.g., in Costa (1984,
1988).
Typical debris-flow parameters are summarized in Table 4.3, as estimated for the
biggest events during two major rainstorms of summer 1987 in Switzerland. Table 4.4
gives an overview of characteristic properties of debris flows compared with floods
with sediment transport and with debris flow-like discharges in the transition zone.
The debris load is usually estimated from the volume of the deposits for a whole event
and, thus, also contains the pore volume.
The maximum discharge usually occurs near the debris flow front. With alpine
debris flows, the maximum discharge may be 100 m3/s to 1000 m3/s, and, there-
fore, is about 10 to 100 times greater than a comparable peak flood discharge
in the same torrent channel (Table 4.3). The height of the debris flow front can
amount to 10 m, and flow velocities up to 15 m/s (54 km/h) have been estimated
for alpine debris flows. With larger debris flow events in the Alps, a few 10,000
m3 to several 100,000 m3 of sediment could be deposited on the fan. A substantial
amount of material is eroded sometimes in the area of the fan. The total runout
distance depends, among other things, on the amount of material transported for
each surge.

Table 4.3 Typical debris-flow parameters of the biggest events in the summer of 1987 in Switzerland
(after ZIMMERMANN & RICKENMANN 1992). DQ = data quality: **** = very good, *** = reliable,
** = rough estimate, * = very rough estimate/uncertain traces.

Val Varuna Val da Plaunca Val Zavragia Minstigertal

18.7.87 18.7.87 18.7.87 24.8.87
Debris flow event,
Date: Value DQ Value DQ Value DQ Value DQ

Debris load [m3] 200’000 **** 250’000 **** 30’000 ** 30’000 ***
Maximum flow velocity 8 ** 10 * 8 *** 14 **
at the fan apex [m/s]
Flow depth at the fan 6 *** ? 6 **** 10 ***
apex [m]
Maximum discharge [m3/s] 400–800 ** 400–900 * 500–700 **** 150–250 ***
Peak discharge of water 7 * 9 * 30 ** 17 **
only, estimated [m3/s]
Number of surges ca. 10 ** >5 * <=6 ** 1 ***
Max. load per surge [m3] 50’000 ** 80’000 ** <30’000 ** <30’000 ***
Max. erosion depth [m] 11 **** 12 **** 2 ** 4 **
Max. erosion cross-section 650 **** 550 **** 20 ** 55 **
[m2]
Historical events ca.10 in **** none *** ca. 7 in *** uncertain, *
150 years known 150 years unknown

RICKENM-Book.indb 63 2/23/2016 7:39:26 AM

RICKENM-Book.indb 64

Table 4.4 Overview of the properties of characteristic displacement processes in torrents (after PIERSON & COSTA 1987; COSTA 1988; HUNGR et al., 2001;
PIERSON 2005; HÜBL et al., 2006).

Process type Flood Debris flow

Terms (German) Hochwasser Fluvialer Murgangartiger Murgang

Feststofftransport Feststofftransport
Terms (English) Flood Bedload transport Debris flood (hyper-concentrated Debris flow
flow, immature debris flow)
Process type Discharge of Weak bedload transport Strong bedload transport Debris flow
water only
Flow behavior Newtonian Newtonian Approx. Newtonian Non-Newtonian
Vol. solids concentration Per mill range 0–20% 20–40% >40%
(approx. range)
Max. grain size mm–cm dm m M
Density (approx. range) 1000 kg/m³ 1300 kg/m³ <1300–1700 kg/m³ >1700 Kg/m³
Viscosity (approx.) 0.001–0.01 Pas 0.01–0.2 Pas 0.2–2 Pas >2 Pas
Shear strength None None None Present
Relevant acting forces Turbulence, bed Turbulence, bed shear stress Buoyancy, turbulence, bed shear Buoyancy, dispersive
shear stress stress, dispersive pressure pressure, viscous
and frictional forces
Vertical distribution of solid Coarser particles near the bed Solids and suspended sediment Solids distributed
particles over flow depth (rolling, hopping, jumping) distributed in cross-section in cross-section
and suspended sediment
distributed in cross-section
Deposition characteristics Horizontal or inclined Weak horizontal stratification; Terminal debris lobes;
stratification; coarser mostly grain supported marginal levées and
clasts may be imbricated tongue-shaped deposits;
grain or matrix
supported; usually clear
boundary of deposits,
U-shaped channels
Sorting of the deposited solids (yes) Sorting moderate to good Moderate to poor Non-stratified, extremely
within individual bedding sorting poorly sorted
units
2/23/2016 7:39:26 AM
 Debris flows 65

4.2 IMPORTANT ELEMENTS OF THE PROCESS

AND HAZARD ASSESSMENT

Standardized procedures and regulations for the management of natural hazards were
introduced, for example, in Europe during the last two decades (e.g. Hübl et al.,
2002; Petrascheck & Kienholz 2003; Greminger 2003; Fuchs et al., 2008;
Hürlimann et al., 2008). These regulations require the determination of hazard dan-
ger levels for potentially affected areas such as a fan. The hazard danger levels are
a function of process intensity and probability of occurrence. For debris flows and
floods, process intensities are typically defined as a function of flow velocity and flow
depth, both of which vary spatially and depend on the magnitude or peak discharge
of the process (Hürlimann et al., 2008). The sediment volume or the sediment-water
volume of a debris-flow event or of a single surge is typically taken as a measure of
the magnitude (Jakob 2005).
Thus, for the process and hazard assessment of debris flows—similar to other
gravitational natural hazards—two key aspects need to be investigated: (a) the proba-
bility of occurrence (or return period) and the magnitude of the event (magnitude-fre-
quency relationship), and (b) the flow and deposition behavior. The most important
elements and existing dependencies are presented schematically in Fig. 4.6. The topic
“magnitude-frequency” of torrent events is considered in chapter 5. Other important
aspects are discussed below, where a range of methods and approaches are presented.
A further section is devoted to a brief overview of GIS-based and numerical simula-
tion models. At the end of chapter 4 the depositional behavior on the fan is discussed,
which is often very important for the hazard assessment.
The most important questions in relation to the hazard assessment are:

• What event magnitude has to be expected?

• What probability of occurrence has to be considered?
• Which are the endangered areas?

Starting zone, Flow

initiation propagation

Event
magnitude

Frequency Deposition
of events behaviour

Figure 4.6 Most important elements in the assessment of debris flow events, and the dependencies
between the elements (after RICKENMANN 2001b).

RICKENM-Book.indb 65 2/23/2016 7:39:26 AM

 66 Methods for the quantitative assessment of channel processes in torrents (steep streams)

In a somewhat simplified way, the two key aspects can be grouped as follows:

a) Magnitude and Frequency

Here, the following three primary elements have to be considered:
• Initiation (location, type of triggering mechanism)
• Event magnitude
• Event frequency

b) Flow and depositional behavior

Here, the following three primary elements have to be considered:
• Event magnitude
• Flow behavior in the channel
• Depositional behavior on the fan
As may be seen from a detailed process assessment of debris flow events, the
above subdivision into two key aspects represents a simplification. The debris flow
can entrain and accumulate additional material during the flow process (from the
stream bed, the banks, and the side slopes) or also deposit the material again. With
the present state of knowledge it is difficult to quantify these processes reliably. Thus,
in an initial step an event magnitude is frequently estimated for the location of the fan
apex, and this value becomes an important input quantity in the assessment of the
flow behavior further downstream. This simplification may be a reasonable approxi-
mation if the debris-flow parameters are primarily needed for estimating potential
hazards in the area of the fan where, in many cases, no significant material entrain-
ment takes place.
For the hazard assessment many different methods and approaches are available.
This variety reflects, on the one hand, the different characteristics of various debris
flow types and, on the other, the limited state of knowledge. A tabular overview of
the available methods to determine the important elements is given in Rickenmann
(2001b, 2015).

4.3 OCCURRENCE OF DEBRIS FLOWS

4.3.1 Predisposition for debris flow occurrence

The assessment of the possibility of a debris flow occurring in a torrent should be
based primarily on an interpretation of the fan area as well as on the traces of earlier
events and/or historical information. If these produce no clues, some general charac-
teristics of the catchment area may be used to allow a rough assessment (Table 4.5).
A minimum streambed slope and a sufficiently large bedload potential are the neces-
sary requirements for debris flows to occur at all. These two factors are also the most
important criteria to assess the hazard potential of a torrent. For the formation of a
debris flow from the channel bed or from a hillslope, the minimum slope is approxi-
mately 25–30%. With the presence of other factors promoting the development of
a debris flow (e.g. channel constrictions, driftwood), a debris flow can also form in
the case of slopes of approximately 15–25%. With channel slopes less than 15% the
development of debris flows is not likely.

RICKENM-Book.indb 66 2/23/2016 7:39:27 AM

 Debris flows 67

Table 4.5 Influence of channel slope (S) and bedload potential (F) on debris-flow hazard. Significance
of hazard classes: A1: high debris-flow hazard, A2: medium debris-flow hazard, B: low debris-
flow hazard, C: practically no debris-flow hazard (from RICKENMANN 1995).

Triggering zone: slope

of stream bed or hill-slope Channel features and bedload potential F (channel + hillslopes) Hazard class

S > 25% Channel in granular material, potentially larger slope A1

instabilities (F > 10 000 m³)
Channel mainly in granular material A2
(F = 1 000 – 10 000 m³)
Channel mainly in bedrock (F < 1 000 m³) B
15% < S < 25% Channel in schistose, flysch-like rocks, potential slope A1
instabilities (F > 10 000 m³)
Other rock types, channel with possible log jams A2
(F > 10 000 m³)
Channel without possible log jams (F = 1 000 – 10 000 m³) B
Channel mainly in bedrock (F < 1 000 m³) C
S < 15% Not relevant C

The classification of the debris flow potential in Table 4.5 is based on the analysis
of the debris flows in the Swiss Alps in 1987 (Rickenmann & Zimmermann 1993)
as well as on a semi-quantitative assessment of the debris flow hazard after Aulitzky
(1973) and Nakamura (1980). The hazard classes in Table 4.5 correspond to a mix
of a probable intensity of the events and a possible frequency of occurrence. The state-
ment, however, refers primarily to the expected maximum intensity of an event. Due
to the large bedload potential and the steep slopes in hazard class A, for example,
small debris flows could also occur. This outcomeleads to a higher overall frequency
than, for instance, for class C, where only small debris flows are to be expected. (The
significance of the hazard classes is given in the legend in Table 4.5.)
A rough differentiation between the processes of debris flow and bedload trans-
port can be made, based on the morphometric parameters of the catchment and of the
fan. Therein, the mean channel slopes on the torrent fan, Sf, are plotted as a function of
the Melton number, Me, defined as the difference between the highest and the lowest
elevation values, normalized with the square root of the catchment area (Marchi &
Brochot 2000; Bardou 2002; Rickenmann & Scheidl 2010). Larger values for Sf
and Me define the range of occurrence of debris flows, while smaller values define the
occurrence of bedload transport. However, the demarcation between the two ranges
is not very clear, but there is quite a wide transition range (Fig. 4.7). Other classifica-
tion schemes were proposed for example in terms of catchment length and Melton
number (Wilford et al., 2004).

4.3.2 Triggering conditions

The triggering of debris flows can take place in the form of hillslope instabilities or
of channel destabilization. With more intensive rainfall there is an increased prob-
ability for shallow landslides to occur in steep terrain. There are quite a large number
of studies which focused on (critical) rainfall conditions necessary for the occurrence

RICKENM-Book.indb 67 2/23/2016 7:39:27 AM

 68 Methods for the quantitative assessment of channel processes in torrents (steep streams)

fluvial bedload transport AUT

debris floods AUT
debris flows AUT
debris flows CH87
0.5 debris flows CH05
debris flows ITST
Bardou (2002)

0.4

C
Mean fan slope Sf

0.3

0.2

0.1
A
B

0
0 1 2 3 4 5
Melton number

Figure 4.7 Rough demarcation of torrents capable of debris flow and those with fluvial bedload trans-
port (modified from SCHEIDL & RICKENMANN 2010), based on the mean fan slope Sf and the
Melton number Me. The data come from Switzerland (CH), Austria (AUT) and South Tirol
in Italy (ITST). The zones A (fluvial transport), B (transition range) and C (debris flow) cor-
respond to the classification of BARDOU (2002), who also used data from MARCHI & BROCHOT
(2000).

of shallow landslides or hillslope debris flows. Regarding debris-flow initiation from

massive channel erosion, however, there are only relatively few studies that attempted
to define a limiting discharge (analogous to the start of bedload transport) based on
laboratory flume investigations as well as on simple theoretical estimates. These meth-
ods, however, usually provide only a rough estimate of the triggering rainfall condi-
tions or the limiting discharge for the formation of the debris flow, since the influence
of the properties of the hillslope material or of the streambed material is generally not
taken into account.
In the case of more intensive and prolonged snow melt, the tendency for debris-
flow formation is increased with the increasing ground saturation. For bigger debris-
flow events in the alpine regions, however, more intensive rainfall is often needed in
addition. Water plays an important part in the initiation process. In the high alpine
scree the destabilization can already be brought about by an underground satura-

RICKENM-Book.indb 68 2/23/2016 7:39:27 AM

 Debris flows 69

tion of the granular material. Since the surface discharge is not the only factor for
the triggering process, not only is the rainfall intensity of importance, but also the
extent of ground saturation due to prolonged rainfall. In the case of storm rainfall
in Switzerland, for example, a minimum triggering intensity of around 30 mm/h and
a minimum total rainfall of around 40 mm have to be reached at the same time for
the formation of debris flows (Zimmermann et al., 1997). In the inner-alpine regions
of Switzerland somewhat smaller triggering rainfalls are necessary for debris-flow
formation; this requirementcould be connected with the smaller annual rainfall com-
pared with the areas bordering the Alps (Zimmermann et al., 1997). Estimates for
critical rainfall conditions are frequently expressed in terms of mean rainfall intensity
I [mm/h] and the duration DR [h] of the triggering rainfall event. These limiting con-
ditions can vary strongly both regionally and locally. Critical rainfall conditions for
debris-flow occurrence in Switzerland are shown in Fig. 4.8 with data from Austria in
comparison with threshold lines for related processes in regions close-by. Threshold
rainfall conditions for slope instabilities have been determined for many regions in the
world (e.g., Caine 1980, Guzzetti et al., 2007, 2008).
Debris flows can occur in torrent channels if sufficient solids material is available
in the bed. For the formation of debris flow, a minimum amount of granular material
needs to start moving with a relatively high solids concentration. This process occurs
primarily in steep channels and in constricted places with previous obstruction of the
material flow, possibly with a temporary clogging of the flow cross-section, or with
an abrupt increase of the erosion in the channel. Fig. 4.9 shows the difference between
the process of debris flow and other types of solids displacement in and near steep
channels and its connection with the formation of debris flows.
Average rainfall intensity I [mm/h]

1= 43 DR–0.89(Switzerland)
I = 42 DR–0.77
100 (Carinthia + E Tyrol,
Austria)

1= 20 DR–0.55
10
(Lombardy, N Italy)

Occurrence of debris flows

1 events
threshold events
non-events
0.1
0.1 1 10 100
Rainfall duration DR [h]

Figure 4.8 Empirical relationships for critical rainfall conditions for the trigger-
ing of debris flows and landslides. The data for debris-flow occurrence in
Switzerland as well as the corresponding limiting criterion are taken from
ZIMMERMANN et al. (1997).The threshold lines for southern and eastern Austria (soil slips) as
well as for northern Italy (landslides) are taken from a compilation of GUZZETTI et al. (2007).
The three geographic regions are all nearby.

RICKENM-Book.indb 69 2/23/2016 7:39:27 AM

 70 Methods for the quantitative assessment of channel processes in torrents (steep streams)

Flood flow Precipitation

channel, low to channel, medium to

medium inclination hillslopes
high inclination
(≤15 ... 30%) initiation:
(≥15 ... 30%)
initiation: critical discharge reduction of
initiation: critical discharge or
shear strength
temporary blockage of flow

Bed load Debris Landslide

transport flow

Figure 4.9 Significant bedload delivery processes in torrents and the role played in the formation of
debris flows. Modified from RICKENMANN (1996).

A simple analysis of slope stability leads to a theoretical limiting slope angle of

about 12° to 17° for typical conditions in the stream bed, assuming a friction angle of
the material in the range 33° to 37° (Takahashi 1987). In steep headwater channels
(as also in rivers) a minimum discharge is necessary for bedload to be transported (see
critical discharge in Fig. 4.9). The combined loading of the bed due to the discharge
and the bedload in motion can suffice in steep slopes to cause enough solids to move
suddenly due to a channel-bed destabilization. The resulting debris-flow formation
by channel erosion appears to depend both on the slope at the initiation area and the
discharge, as indicated by the data in Fig. 4.10. An example of the formation of a
debris flow from the channel is shown in Fig. 4.11.
If shallow landslides are sufficiently large and fluid or occur very near to chan-
nels, they may transform into hillslope debris flows, get into the torrent channel and
continue moving downstream as a debris flow. Shallow landslides that occurred in
Switzerland in the period 1997 to 2005, exhibit typical slope angles from about 24°
to 43° and volumes of 50 to 100 m3 (Raetzo & Rickli 2007). The rainfall conditions
for triggering shallow landslides in Switzerland were investigated by Rickli et al.
(2008).
A continuous transition from fluvial bedload transport to debris floods to debris
flows may be expected where channel slopes become steeper than about 20–25%,
according to investigations of Smart & Jäggi (1983) and Rickenmann (1990) and
other flume and field observations. A similar conclusion may be drawn by comparing
critical discharge criteria for bedload-transport initiation and debris-flow formation.
Analogous to the limiting discharge at the start of fluvial bedload transport, a critical
dimensionless discharge qc* for the formation of debris flows can be defined:

q*
c
= qc/[g0.5 D1.5] = ag/Sαg (4.1)

RICKENM-Book.indb 70 2/23/2016 7:39:27 AM

 Debris flows 71

1.0
hillslope instability
channel erosion

0.8
Slope at initiation area

0.6

0.4

0.2

0.0
0.01 0.1 1 10 100
Catchment area upstream of initiation point (km2)

Figure 4.10 Hillside and channel slopes in the initiation area of debris flows that formed either due to
hillslope instability or channel erosion (channel destabilization) are shown as a function
of the catchment area above the triggering zone (as indicator for the water discharge).
Data taken from Swiss investigations (VAW 1992; ZIMMERMANN et al., 1997).

Figure 4.11 (a) Channel above the triggering zone (corresponds to the situation in the fig-
ure on the right before the initiation of debris flow); (b) Channel at the triggering
zone after the initiation of a debris flow. The channel slope is 51% (S = 0.51). (Photos
M. ZIMMERMANN).

Here qc is the critical specific discharge per meter channel width, D is a characteristic
grain size in the channel bed, S = sinβ represents the channel slope, g = gravitational
acceleration, ag = empirical coefficient, and αg = semi-empirical exponent.

RICKENM-Book.indb 71 2/23/2016 7:39:28 AM

 72 Methods for the quantitative assessment of channel processes in torrents (steep streams)

100
Bathurst et al. (1987): initiation
bedload transport, eq. 3.9
Whittaker & Jaeggi (1986): block
ramp stability, eq. 3.15
Takahashi (1987): beginning of
‘immature debris flow’
Tognacca et al. (2000): debris-flow
formation in channels
Takahashi (1987): debris-flow
qc* = qc/(g0.5 D1.5)

formation in channels

10

1
0 0.1 0.2 0.3 0.4 0.5 0.6
S = tan

Figure 4.12 Relations based on flume experiments to determine the critical unit discharge for the ini-
tiation of different process types. The criteria for the formation of debris flow by channel
erosion have not yet been checked against field data.

Such relations to determine the critical unit discharge for the initiation of different
process types are shown in Fig. 4.12. The relations of Tognacca et al. (2000), Whit-
taker & Jaeggi (1986), and Bathurst et al. (1987) are all based on laboratory flume
tests. Fig. 4.12 also show a limiting condition for the start of “immature debris flows”
after Takahashi (1987), a state which corresponds roughly to intensive bedload trans-
port or “debris flood” conditions; this relation is also based on flume experiments. The
figure shows a large range of possible limiting discharges for the formation of debris
flows in channels that still have to be checked against field observations.

4.4 EMPIRICAL APPROACHES TO CHARACTERIZE

THE FLOW AND DEPOSITION BEHAVIOR

To assess the flow and deposition behavior either empirical approaches including estimate
formulae or numerical simulation models can be used. The main objectives are to deter-
mine critical locations where there is the possibility of flow overtopping a channel and to
delineate areas of the fan that are likely to be inundated and covered by solid deposits.
A summary of the debris flow parameters estimated for events in 1987 in Switzerland is
given in Table 4.3. More exact measurements have been made for several years at several
debris flow monitoring stations in the Alps (e.g. Genevois et al., 2000; Marchi et al.,
2002; Rickenmann et al., 2001; Hürlimann et al., 2003; McArdell et al., 2007).
The sequence of some simple empirical calculations using important parameters
for characterizing debris-flow behavior is illustrated schematically in Fig. 4.13. If, for
a future event, the debris load M is estimated, an approximate maximum discharge

RICKENM-Book.indb 72 2/23/2016 7:39:29 AM

 Debris flows 73

Debris-flow volume M

Maximum discharge QP

Flow velocity V

Flow cross-sectional area A

Total runout distance L

Runout distance on fan Lf

Figure 4.13 Calculation sequence of empirical approaches to estimate the most important flow param-
eters of debris flows. Modified from RICKENMANN (1999).

Qp can be then be calculated. This value together with the channel slope determines
essentially the flow velocity V. The maximum required flow cross-section A is then
given by A = Qp/V. A comparison with the existing channel cross-section gives indi-
cations of possible places of channel overtopping. Also, the total runout distance of
a debris flow from the point of initiation to the lowest deposition point, L, or the
deposition length on the fan, Lf, can be estimated roughly based on the debris load, if
a more exact determination by means of simulation models is not possible.
The observed values of debris load usually contain both the bedload and also the pore
or water volumes. If the data were obtained from a measuring station, then M is typically
determined by the integration of the mixed discharge over time. In other cases M is deter-
mined from the observed deposition area and the (mean) deposition thickness, whereby
the pore volume is included, which corresponds approximately to the water content.
Regarding the erosion behavior along the transit stretch, relatively few quantita-
tive observations were made in the field (Hungr et al., 2005; Schürch et al., 2011;
Berger et al., 2011a; McCoy et al., 2012). Besides using modelling concepts based
on soil mechanics (Iverson 2012; McCoy et al., 2012), analogies were made with
approaches developed for bedload transport (Egashira et al., 2001; Rickenmann
et al., 2003; Cao et al., 2004). In estimating the debris load using geomorphologic
methods, therefore, possible solids input from the transit stretch are included, if the
estimation procedure is applied in the fan area according to Fig. 4.13.

4.4.1 Maximum discharge

Field observations indicate an empirical relation between maximum discharge Qp
[m3/s] and debris load M [m3] (Rickenmann 1999), as illustrated in Fig. 4.14. Here
one distinguishes between granular and muddy debris flows (Mizuyama et al., 1992):

RICKENM-Book.indb 73 2/23/2016 7:39:30 AM

 74 Methods for the quantitative assessment of channel processes in torrents (steep streams)

Qp = 0.135 M0.78 (granular debris flow) (4.2)

Qp = 0.0188 M0.79 (muddy debris flow) (4.3)

This distinction is based on observations made in Japan, and it has been con-
firmed partly by debris flow data from other parts of the world. The classification,
however, is not always easy. Thus, the data of the Rio Moscardo in Fig. 4.14 possibly
lies nearer to the line of Eq. 4.3, since the debris flow discharges may have been rich
in water. The debris flows in the Jiangjia ravine are very rich in fine material and one
would expect therefore that the data points would scatter around Eq. 4.3 rather than
Eq. 4.2. The debris flows in the Illgraben are also rich in fine material, but contain,
in general, little cohesive sediment. When using the equations for predictions, the
data range of the observed values should be considered as well as the fact that the
maximum discharge should be correlated preferably with the volume of the biggest
single surge than with the total debris load of an event. In the case of the data of the
debris flows “Switzerland 1987” plotted in Fig. 4.14, the individual surge volume is
not known. In alpine regions the assumption that M is greater than about 50,000 m3

Kamikamihori (Japan)
Jiangjia ravine (China)
Rio Moscardo (Italy)
10000 Switzerland (1987)
Illgraben (CH, 2000, 2008, 2009)
Schipfenbach (CH, 2000)
Chemolgan (Kazakhstan)
granular debris flows, eq. 4.2
muddy debris flows, eq. 4.3
1000
Maximum discharge Qp (m3/s)

100

10

1
100 1000 10000 100000 1000000
3
Debris-flow volume M (m )

Figure 4.14 Maximum discharge Qp of the water-solids mixture as a function of the debris load M.
Data sources are indicated in RICKENMANN (1999), and some more recent data are from
MARCHI et al. (2002), HÜRLIMANN et al. (2003), RICKENMANN et al. (2003), and BERGER et al.
(2011b).

RICKENM-Book.indb 74 2/23/2016 7:39:30 AM

 Debris flows 75

is rather implausible, as shown in the analysis of the debris-flow surges of the 1987
events in Val Varuna (VAW 1992).

4.4.2 Flow velocity

To estimate the mean flow velocity V [m/s] two different equations are proposed here
(RICKENMANN 1999):

V = 2.1 Q0.33 S0.33 (4.4)

V = kSt h0.67 S0.5 (4.5)

where Q [m3/s] is the discharge, S the channel slope [m/m] in the considered stream reach,
kSt a pseudo Strickler coefficient [m1/3/s] and h [m] the flow depth. The use of Eq. 4.4
and Eq. 4.5 for data from debris flows and water discharges are shown in Fig. 4.15 and
Fig. 4.16. In these figures only some datasets from Rickenmann (1999) were used, i.e.
datasets A and B (debris-flow data with directly measured flow velocities) and dataset G.
For the debris-flow data shown in Fig. 4.16, an average friction value is about
kSt = 10 m1/3/s. For granular debris flows in natural channel reaches, kSt values are
obtained in the range of 6 m1/3/s (Rickenmann & Weber 2000). For debris flow dis-
charges in artificial (canalized) channels, the pseudo Manning-Strickler coefficients
could be up to 50% higher (Pwri 1988). Eq. 4.4 is valid for natural channel reaches;
thus, in artificial channels an approach such as Eq. 4.5 is preferable.

100
perfect agreement
debris flows
water flows

10
Vcalculated (m/s)

0.1
0.1 1 10 100
Vobserved (m/s)

Figure 4.15 Use of Eq. 4.4 for debris flows and water discharges. Comparison of calculated and
observed flow velocities. Modified from RICKENMANN (1999).

RICKENM-Book.indb 75 2/23/2016 7:39:30 AM

 76 Methods for the quantitative assessment of channel processes in torrents (steep streams)

100
perfect agreement
debris flows
water flows

10
Vcalculated(m/s)

0.1
0.1 1 10 100
Vobserved(m/s)

Figure 4.16 Use of Eq. 4.5 for debris flows and water discharges. For both sets of data in each case,
a mean STRICKLER coefficient kSt is assumed. Comparison of calculated and observed flow
velocities. Modified from RICKENMANN (1999).

4.4.3 Total runout distance

If there are significant deposition or redistribution reaches, a small debris flow could
come to a standstill there. The estimate of the possible deposition volumes in such a
reach and the comparison with the estimated expected debris load permits an assess-
ment of whether discontinuation of the debris flow is probable. Depositions and over-
flowing of the banks may also take place upstream of narrow flow cross-sections as
a result of retrogressive aggradation. In the case of large debris-flow events (which
are relevant for the hazard assessment), it is generally very likely that debris flows
reach to the fan. The analysis of 82 debris-flow events in the summer of 1987 in
Switzerland showed that: (a) A minimum general slope of less than 19% was nowhere
found. The general slope denotes the mean slope of the total flow path from the point
of initiation of the debris flow down to the lowest deposition point; (b) In general,
the runout distance L [m] is dependent on the debris load M [m3]. The analysis of
further data from other regions suggested the inclusion, in addition, of the height dif-
ference He [m] between the uppermost point of initiation and the lowest deposition
point (Rickenmann 1999; Rickenmann & Scheidl 2010), resulting in the following
equation for the mean runout distance:

L = 1.9 M0.16 He0.83 (4.6)

RICKENM-Book.indb 76 2/23/2016 7:39:30 AM

 Debris flows 77

To estimate an upper limit of the runout distance Lmax the following relation may
be used:

Lmax = 5 M0.16 He0.83 (4.7)

In contrast to Eq. 4.2 and Eq. 4.3 for the derivation of Eq. 4.6 and Eq. 4.7 the total
debris load was used and, thus, it has to be input here. The use of Eq. 4.6 and Eq. 4.7
for data from debris flows is shown in Fig. 4.17. For predictive estimates, a further rela-
tion between L and He is necessary. This is the longitudinal profile of the expected flow
path, whereby Eq. 4.6 and Eq. 4.7 can be solved either mathematically or graphically.
Other empirical equations for the total travel distance of debris flows were pro-
posed by Corominas (1996), Legros (2002), Toyos et al. (2008), and Prochaska
et al. (2008). In most of these approaches the runout length is essentially a function of

100000
Swiss debris-flow events
Kamikamihori (Japan)
other events
mean L according to eq. 4.6
maximum L according to eq. 4.7

10000
Total runout distance L (m)

1000

100
100 1000 10000
0.16 0.83
M *H

Figure 4.17 Estimation of the total runout distance with Eq. 4.6 and Eq. 4.7. The field debris-flow data
are from RICKENMANN (1999). Modified from RICKENMANN (1999).

RICKENM-Book.indb 77 2/23/2016 7:39:31 AM

 78 Methods for the quantitative assessment of channel processes in torrents (steep streams)

the volume and angle of reach or the longitudinal profile of the expected flow path.
Another empirical approach to estimate the total travel distance is based on a sedi-
ment budget along the flow path (Cannon 1993; Fannin & Wise 2001).

4.4.4 Deposition length of the fan

For the deposition length Lf on the fan the data used by Rickenmann (1999) shows
only a very weak dependence on the debris load M. Thus, an empirical estimation for-
mula is not recommended. Nevertheless, for a particular torrent channel and a given
fan topography, it may be expected that with similar material properties larger debris
flows flow further than smaller ones. Larger debris flows have a tendency for larger
maximum discharges (see Eq. 4.2 and Eq. 4.3), together with larger flow velocities
and/or larger flow cross-sections. Based on a momentum consideration of the flow of
the water-debris mixture on a uniformly sloping surface, the deposition length Lf can
theoretically be estimated as follows (Hungr et al., 1984):

Lf = AV2/G (4.8)
AV = Vu cos(βu – β) [1 + (g hu cosβ u)/(2 Vu2)] (4.9)
G = g (SR cosβu – sinβ) (4.10)

where β = slope of the deposition reach, βu = slope of the steeper inflow chan-
nel, Vu = flow velocity in the inflow channel, hu = flow depth in the inflow chan-
nel, SR = friction slope (sliding friction only), assumed to be constant in the runout
reach, and g = gravitational acceleration. Hungr et al. (1984) assumed that
SR = 0.176 = tan(10°) and obtained thereby good agreement between observed values
of Lf for five debris flows in Western Canada and values calculated using Eq. 4.8.
On the other hand, the use of Eq. 4.8 for debris flows in the Kamikamihori valley in
Japan (with measured discharge parameters) resulted in better estimated values for
Lf, provided SR ≈ 1.1 tanβ is chosen instead of SR = tan(10°). This holds likewise for
the application of Eq. 4.8 to some debris flows in Switzerland in 1987, if flow depths
obtained from field observations are determined and the flow velocities are calculated
with a Chezy equation or with Eq. 4.2 and Eq. 4.4 (Rickenmann 2005b). Fig. 4.18
presents a comparison of calculated and observed deposition lengths.
It is interesting to note that the values calculated with Eq. 4.8 to Eq. 4.10 assume
a minimum for flow velocities Vu between about 2 m/s and 4 m/s (Rickenmann
2005b). A possible dependence of SR on β is not surprising since the friction slope
depends on the material properties of the debris flow, which is reflected roughly also
in the fan slope. Other methods to estimate the runout distance of debris flows are
discussed in Rickenmann (2005b) and Rickenmann & Scheidl (2010).

4.4.5 Impact forces

Several studies suggested that the impact pressure of debris flows on obstructions
may be estimated in a similar way to that for the dynamic pressure for Newtonian
water flows. In the region of the front of the debris flow, stones and large blocks of
up to several meters diameter can be transported. In this case, locally higher pressures
are to be expected. Based on measurements of the debris flow impact force on rigid
obstructions, it is estimated that the pressure is, on average, about a factor 2 to 4

RICKENM-Book.indb 78 2/23/2016 7:39:31 AM

 Debris flows 79

1000

800

600
Lf, calculated (m)

400

perfect agreement
200
Kamikamihori (Japan): Vu(hu)
Switzerland (1987): Vu(hu)
Switzerland (1987): Vu(Qp), Qp(M)
0

0 200 400 600 800 1000

Lf, observed (m)

Figure 4.18 Comparison between calculated (Eq. 4.8 to Eq. 4.10) and observed deposition lengths
Lf, using SR = 1.1 tanβ. With the events in Switzerland in 1987 the flow velocity (Vu) was
estimated two ways: firstly as a function of the flow depth estimated in situ (hu) with a
CHEZY equation for Vu (after RICKENMANN & WEBER 2000), and secondly by means of the
observed debris load (M) and Eq. 4.2 and Eq. 4.4. Adapted from RICKENMANN (2005b).

higher than the hydrostatic pressure (Geo 2000). Thus, the following formula for the
dynamic impact pressure pd [N/m2] due to debris flow is proposed:

pd = αd ρM V2 sinβd (4.11)

where ρM [kg/m3] = density of the debris flow mixture, V = flow velocity [m/s],
βd = impact angle (often βd = 90°) and αd = empirical coefficient (for debris flow of
about αd = 2 to αd = 4). In flume tests the impact forces of both viscous and granular
debris flows were measured by Scheidl et al. (2013), and values for αd ranging from
about 1.5 to 12 were determined. A recent experimental study on the impact force of
viscous debris flow indicated that the empirical coefficient αd may be a function of the
approach flow Froude number Fr, with αd = 5.3 Fr1.5 (Cui et al., 2015). Interestingly,
in a flume study by Scheidl et al. (2015), a similar dependence on the Froude num-
ber was found for the correction coefficient in debris-flow velocity estimates using the
superelevation equation for Newtonian flows.

RICKENM-Book.indb 79 2/23/2016 7:39:31 AM

 80 Methods for the quantitative assessment of channel processes in torrents (steep streams)

To estimate the impact force of granular debris flows, a theoretically-derived for-

mula of Coussot (1997) resulted in a similar form to that for the dynamic pressure
of Newtonian fluids, and the multiplication coefficient αd may take on higher values
than 2 to 4. To take into account the impact of individual large blocks on a structure,
a computational method is described in Egli (2005). Further approaches for calculat-
ing the impact force of debris flows were proposed by Armanini & Scotton (1993),
Zanuttigh & Lamberti (2006), and Ancey & Bain (2015). Rickenmann (2008)
compared these methods for the range of flow depths and flow velocities that can be
expected on the torrent fan with debris flows. It was shown that for Froude numbers
Fr much smaller than 1 the hydrostatic component is important, whereas for Fr > 1
the hydrodynamic component dominates (see also Eq. 4.11).

4.5 MODELS FOR THE SIMULATION OF DEBRIS FLOWS

4.5.1 Empirical approaches

Some empirical methods or simple models to estimate the one-dimensional runout dis-
tance or deposition length on the fan are discussed in chapter 4.4.3 and 4.4.4. To delin-
eate potentially endangered areas in more detail, the runout pattern or the surface area
of potential debris-flow deposits on the cone should be known. A simple topography-
based empirical approach was developed by Iverson et al. (1998), in which the depo-
sition area of Lahars is correlated by means of an empirical scaling function with the
event volume. With additional assumptions and the aid of a Geographic Information
System (GIS), potentially endangered areas can be demarcated in a simple way. Similar
methods were tested and implemented for debris flows by Hofmeister et al. (2003),
Crosta & Agliardi (2003), Berti & Simoni (2007), Oramas Dorta et al. (2007),
Griswold & Iverson (2008), and Scheidl & Rickenmann (2010). The key element
of all approaches is an empirical scaling relation between the planimetric deposition
area of debris flows and the event volume. As a modification, in the model TopRunDF
much greater consideration is given to the fan topography using a random algorithm to
determine a distribution of the potential flow paths (Scheidl & Rickenmann 2010).
The model TopRunDF was tested using numerous debris flow events in Switzerland,
in Austria and in South Tirol (Rickenmann & Scheidl 2010). TopRunDF can be
downloaded from the web page: http://www.debris-flow.at.

4.5.2 Simple analytical methods

Zimmermann et al. (1997) and Gamma (2000) describe an automated application of
simple models using a GIS technique. This method was used in several areas in Switzer-
land to produce hazard index maps. The zones of initiation of debris flows are deter-
mined basically from the limiting slope or limiting range of slopes. To estimate the
flow velocity and the total runout distance, a mass point model for the flow behavior
of a Voellmy fluid is used. The two friction parameters of the Voellmy approach must
be estimated on the basis of earlier events (Rickenmann 2005b). The spreading on the
fan is simulated by means of a stochastic algorithm, which also has to be calibrated
beforehand. The bedload potential is estimated in a simplified way by accounting for
potential sediment delivery areas. Depending on the expected debris load, a large num-

RICKENM-Book.indb 80 2/23/2016 7:39:31 AM

 Debris flows 81

Figure 4.19 Comparison of the deposition of (a) a debris flow (event of 19.7.1987; Photo A. Godenzi,
Chur) and (b) a snow avalanche (Photo R. Godenzi, Poschiavo, photo date 8.5.1978), which
both occurred in the Val Varuna catchment in the neighborhood of Poschiavo (Canton
Grisons, Switzerland). The two events have roughly comparable deposition volumes.

ber of simulation runs are calculated, and thereby the spreading over the fan is coupled
implicitly with the event magnitude. The resulting model is called DFWalk. In Switzer-
land the parameter of the Voellmy model DFWalk was investigated with the help of
the back-calculation of a total of 75 debris-flow events with volumes in the range of
3000 m3 to 450,000 m3 (Zimmermann et al., 1997; Gamma 2000; Genolet 2002).
The Voellmy model was originally developed for the analysis of the flow behavior
of snow avalanches (Bartelt et al., 1999), and is based partly on hydraulics methods.
A similarity between the deposition of debris flows and of snow avalanches is illus-
trated in Fig. 4.19.
Hungr et al. (1984) and Takahashi (1991) present a simple analytical method
to describe the flow distance of a constant debris flow stream in the outflow region
on the fan. The method is based on the momentum equation and the assumption of
constant friction losses along the runout zone (see Eq. 4.8 to Eq. 4.10); it was first
developed for snow avalanches and then applied to debris flows. The main difficulty
lies in the selection of an appropriate friction coefficient (Rickenmann 2005b). This
method was implemented in TopFlowDF, which otherwise exhibits similarities to
TopRunDF (Scheidl & Rickenmann 2011). In contrast to TopRunDF an empiri-
cal surface-volume relationship is not necessary, but instead with TopFlowDF the
(empirically determined) friction coefficient is important.

4.5.3 Numerical simulation models

Interestingly, to describe the general kinematic flow characteristics of rapid gravi-
tational mass movements such as snow avalanches, debris flows, rock avalanches
and shallow landslides, to some extent similar modeling approaches were proposed.

RICKENM-Book.indb 81 2/23/2016 7:39:31 AM

 82 Methods for the quantitative assessment of channel processes in torrents (steep streams)

This is particularly evident for the simpler dynamic approach determining the
flow behavior of a single-phase bulk mixture represented by a mass-point model
(e.g. Scheidl et al., 2013), discussed for debris-flow application in the previous
chapter 4.5.2.
Kinematic flow parameters like flow velocities or dynamic impact forces are often
needed for a more detailed hazard assessment. This typically requires the application
of numerical simulation models, which represent a more physically-based description
of the flow behavior of gravitational mass movements of solids-water mixtures. The
kinematic flow characteristics of a debris flow depend, for example, on the topo-
graphical and surface friction conditions, the water content, the sediment size and
sorting and the dynamic interaction between the solid and fluid phases of the debris-
flow mixture (Iverson 1997). Debris flows with high flow velocities often exhibit
a fluid-like displacement behavior, whereas, during the initiation and deposition
phases, soil mechanics aspects are more important.
To describe the material and flow behavior of debris flows, various approaches
were proposed and implemented in numerical simulation models. An important
element of many proposed models is an appropriate formulation for the constitutive
behavior of debris flows. The main problem for practical hazard assessment is that
there are no clear criteria as to which methods (or constitutive equations) can be best
applied to the various debris-flow types encountered in nature.
Initially, one group of simulation models considered the debris-flow mixture to
be a quasi-homogeneous fluid as a first approximation, enabling the flow behavior to
be described by a rheological model. A rheological model provides a relation between
the shear rate γs (= change in flow velocity/change in flow depth) and the applied
shear stress τ. The laminar flow behavior of water is defined as Newtonian flow (see
Fig. 4.20) and can be described with the formula:

τ = μγs (4.12)

where μ = the dynamic viscosity. The simplest model to describe the flow behavior of
viscous debris flow is the so-called Bingham model (see Fig. 4.20):

τ = τB + μγs (4.13)

The variable τB stands here for the shear strength—a second material parameter—
which has to be overcome by the driving forces before a fluid deformation (flow)
can occur. As is evident from Fig. 4.20, a series of further models were proposed to
describe the flow behavior of debris flows. A “pseudo-plastic” rheology is shear thin-
ning (i.e. the effective viscosity decreases with increasing shear stress); a “dilatant”
rheology is shear thickening (i.e. the effective viscosity increases with increasing shear
stress). A number of models are partly- or fully-based on a rheologic formulation for
a Bingham or viscoplastic fluid (Choi & Garcia 1993; Laigle & Coussot 1997;
Fraccarollo & Papa 2000; Imran et al., 2001; Malet et al., 2004). The rheological
properties of a real debris-flow mixture are difficult to determine. For the fine mate-
rial of a debris flow, the rheological parameters were determined from laboratory
measurements for some model applications (e.g. Laigle & Coussot 1997; Imran
et al., 2001; Malet et al., 2004). However, it is much more challenging to deter-

RICKENM-Book.indb 82 2/23/2016 7:39:33 AM

 Debris flows 83

pseudoplastic with
yield stress, n < 1

Shear stress τ
Bingham

viscoplastic with yield

stress (dilatant), n > 1

pseudoplastic, n < 1

Newtonian
Yield stress τ0

viscoplastic (dilatant),
n>1

Shear rate γs

Figure 4.20 Rheological characterization of different fluids. The shear strength τB or τo is also termed
yield stress in some studies. Modified from KAITNA (2006).

mine the effect of the coarser components on the rheology (Phillips & Davies 1991;
Coussot et al., 1998).
In several applications to natural debris flows, the pure Bingham model was
modified by adding a friction term accounting for channel roughness and turbulence
(O’Brien et al., 1993; Han and Wang 1996; Jin and Fread 1999). The model FLO-
2D (O’Brien et al., 1993) has probably been the most widely applied, commercially-
available, two-dimensional simulation program for debris flows. The constitutive
equations consist of a rheological model that combines the Bingham rheology with
an inertial friction term after Bagnold/Takahashi (1991) as well as a turbulent
friction term; the effects of the last two friction terms are lumped into an empiri-
cally determined pseudo Manning coefficient (O’Brien et al., 1993). An example
of the application of FLO-2D to simulate the deposition area on the fan of a debris-
flow event in Switzerland is shown in Fig. 4.21, which also illustrates the effect of
buildings, which can be considered optionally with this model. As a somewhat more
complex alternative for a viscoplastic fluid, a Herschel-Bulkley model was imple-
mented in another simulation program for debris flows (Laigle & Coussot 1996;
Rickenmann et al., 2006b).
With the second group of simulation models, the mass continuity for the
water and the solids is considered separately, i.e. two-phase models are considered.
The erosion and deposition of solids are taken into account using simple approaches.
Such models were developed especially in Japan (e.g. Nakagawa et al., 2000). With

RICKENM-Book.indb 83 2/23/2016 7:39:33 AM

 84 Methods for the quantitative assessment of channel processes in torrents (steep streams)

deposition
0.01-0.1
0.1-0.2
0.2-0.3
0.3-0.4
0.4-0.5
0.5-0.75
0.75-1
1-1.25
1.25-1.5
1.5-2

Figure 4.21 Simulation of the area and thickness of debris-flow deposits of the event of 24 August
1987 on the fan of the Minstiger stream (Switzerland) with the program FLO-2D.The area
with dark-red points corresponds to the observed debris-flow depositions. In the figure
on the right the effect of houses on the flow were taken into account in the simulation,
but not in the figure on the left. In both cases the calculation was carried out with the
same pseudo MANNING-STRICKLER value, but with different BINGHAM-parameters. The area
with yellowish-green points in the lower fan region indicates fluvially redistributed finer
sediment due to subsequent flooding.

the two-phase models, a discharge hydrograph can be used as input so that the result-
ing solids concentration depends basically on the channel slopes and the properties
of the bed material. The deposition of the solids is obtained using similar methods as
applied to fluvial bedload transport.
The modeling approach of Iverson & Denlinger (2001) takes account of
basal pore water pressures and other soil mechanics aspects. The two phases of
granular solids and a viscous fluid are coupled using mixture theory (Iverson &
Denlinger 2001; Denlinger & Iverson 2001). The model is based on a generali-
zation of the approach of Savage and Hutter (1989) for dry granular avalanches.
A further development is the D-Claw model (Iverson & George 2014; George &
Iverson 2014), which combines continuum conservation laws with concepts from
soil mechanics, fluid mechanics, and grain–fluid mixture mechanics. An important
aspect of this model is that both the solid volume fraction and basal pore-fluid pres-
sure can evolve over time. It has been successfully applied to the 2014 landslide/debris
flow event near Oso, Washington, USA (Iverson et al., 2015).

RICKENM-Book.indb 84 2/23/2016 7:39:33 AM

 Debris flows 85

The DAN model was derived from the work of Hungr (1995) for the analysis
of the one-dimensional flow behavior of mass movements, with an option to select
different rheological “friction” approaches. Similar simulation models were devel-
oped by Rickenmann & Koch (1997) and Näf et al. (2006). The DAN model was
later extended to two-dimensional analyses and is designed to be an efficient tool
for practical application (McDougall & Hungr 2004; Hungr & McDougall
2009).
The Voellmy approach is well known in Switzerland, above all due to its
application to snow avalanches. It involves a base (Coulomb) parameter and a
“turbulent” friction parameter (Bartelt et al., 1999). Numerical models with
Voellmy rheology were successful in back-calculating shallow landslides,
hillslope debris flows and channelized debris flows (Rickenmann & Koch 1997;
Hürlimann et al., 2003; Chen & Lee 2003; Swartz et al., 2003; McArdell
et al., 2003; Revellino et al., 2004). The model RAMMS is also based on the
Voellmy rheology; the module for debris-flow simulation is available both in a 1D
version and in a 2D version (Scheuner et al., 2009; Christen et al., 2012). The
latter version of RAMMS was also adapted for the simulation of hillslope debris
flows (Christen et al., 2012).
Numerical simulation models applied in case studies to real debris flows include
RAMMS (Christen et al., 2010, 2012), DAN or DAN-3D (Ayotte & Hungr 2000;
McDougall & Hungr, 2004, 2005), FlatModel (Medina et al., 2008), MassMov2D
(Begueria et al., 2009), RASH-3D (Pirulli & Sorbino, 2008), and TRENT-2D
(Armanini et al., 2009). Typically, appropriate values for the rheologic or friction
parameters were assumed or back-calculated from field observations (Hungr 1995;
Rickenmann & Koch 1997; Ayotte & Hungr 2000; Revellino et al., 2004; Naef
et al., 2006; Rickenmann et al., 2006; Tecca et al., 2007; Hungr 2008; Hürlimann
et al., 2008; Pirulli 2010).
Basically, continuum mechanics simulation models provide the most accurate
description of the flow processes, including the deformation of the moving mass
along its path as well as detailed spatial and temporal information on the flow
parameters. Knowledge of the spatial distribution of the parameters flow velocity
and flow depth is important for the production of hazard maps (e.g. BWW/BRP/
BUWAL 1997). It must be stressed, however, that generally the rheological model or
friction parameters cannot be determined directly (i.e. from samples), but must be
assumed on the basis of experience or ideally be “calibrated” from past events in the
same region (Rickenmann et al., 2006b; Rickenmann, 2016). Theoretically, based
on sediment samples and laboratory tests, the rheological parameters of viscoplastic
fluids can be determined for a Bingham or Herschel-Bulkley model; however,
this approach typically cannot account for the influence of sediment particles greater
than several mm in size.
Various investigations showed that, for the depositional behavior of debris flows
on the fan, the topography is a very important and governing factor (Rickenmann
et al., 2006b; Scheidl & Rickenmann 2010; Rickenmann & Scheidl 2010). Thus,
an appropriate digital terrain model (DTM) must include an accurate representation
of the channel and other depressions on the fan For simulations in the context of a
hazard assessment, appropriate scenarios for the input conditions have to be defined
for example at the fan apex, including assumptions of the total volume of debris flow

RICKENM-Book.indb 85 2/23/2016 7:39:39 AM

 86 Methods for the quantitative assessment of channel processes in torrents (steep streams)

in an event, the number of surges, and possible depositions on the channel bed due to
smaller surges.

4.6 SCENARIOS AND DEPOSITION IN THE AREA

OF THE FAN

4.6.1 Uncertainty and scenarios

The term scenario is used here in the context of various assumptions that typically
have to be made for the hazard assessment of a future debris-flow event. These
assumptions refer, for example, to the number and size of individual surges, the
rheologic properties of the mixture (e.g. water content), and possible clogging or
log-jam locations that could induce flow overtopping out of the channel. All of
these parameters are difficult to predict, yet they can have a big effect on the flow
and deposition behavior on the fan by influencing the sequence of flow processes
over the duration of a debris-flow event. The establishment of such scenarios is
the more challenging the scarcer the information on past events. It is also difficult
because the estimation of the frequency of events of a given magnitude is typically
imprecise. Thus, one has to work with scenarios that are assigned approximate
recurrence periods.
Debris flows generally occur in steep headwater catchments, where often there is
a strong interaction between different processes. Hillslope processes can lead to sig-
nificant sediment delivery to the channels. A larger landslide can rapidly deposit solid
material in the channel leading to a (partial) blockage (retention) of the channel dis-
charge with the danger of a collapse of the “dam” thus formed. In steep channels the
processes of “flooding”, “bedload transport” and “debris flows” often occur in close
combination. Here, of interest are not only possible sediment supply sources but,
above all, the critical places where an obstruction or complete blockage of the water
and sediment discharge may induce an overflowing of the channel banks (e.g. due to
a log jam at a bridge cross-section), thereby influencing the potential area affected by
debris-flow or bedload deposition.
For the assessment of sediment transfer processes in headwater catchments, rela-
tively large uncertainties still exist with the use of both empirical methods and of
numerical simulation models. A proper assessment is also hampered by the difficulty
to distinguish between various debris-flow types (granular debris flows, mudflows).
Thus, in a technical report on hazard assessment, it is important to state the assump-
tions made, to point out uncertainties and possibly to perform sensitivity analyses on
the chosen input and boundary conditions.

4.6.2 Traces of earlier deposits on the fan

The dominant sediment transfer process (i.e. fluvial bedload transport or debris flow)
influences the type of deposit. In the region of debris-flow deposits, all particle sizes
are distributed more or less uniformly over the area of deposition, but large boulders
at the front of individual debris-flow surges are often transported to the distal end of
the deposit. On the other hand, in floods with fluvial bedload transport the coarser
components are likely to be deposited on steeper slopes, whereas the finer particles

RICKENM-Book.indb 86 2/23/2016 7:39:39 AM

 Debris flows 87

are transported to flatter parts of the fan. In addition, finer particles of debris-flow
deposits may be partly re-entrained and re-distributed due to the subsequent runoff
that is less sediment-laden.
If there are traces of earlier debris flows on the fan and/or historical documents
are available, the assessment of the depositional behavior should be partly based on
this information (see also Table 4.2). In comparison with earlier deposits, the size
(areas) of the endangered zones may have to be adjusted according to the expected
debris load (see also chapter 4.5). Generally steeper fans of irregular fan topography
with a rough surface (in the case of a natural fan) point to debris-flow activity. In
the evaluation of the traces on the fan (old deposits) primarily the following factors
should be taken into account.

4.6.2.1 Old deposits

An accumulation of boulders bigger than about 0.5 m to 1 m diameter points to old
debris flow deposits, especially if these deposits lie outside the channel. Levées or debris
lobes are (if they are still recognizable) also pointers to areas endangered by debris
flow. However, the present fan situation may not necessarily be the same in future. For
example, a changed debris-flow topography at the fan apex (after sediment deposition)
may induce a new direction of the course of the stream, and, thus, another side of the
fan may become endangered. Or a changed future discharge regime with more fluid
discharges containing less bedload may lead to a more incised channel at the fan apex,
and areas covered earlier by debris may no longer be endangered. Inactive discharge
channels and relicts of old stream courses generally point to earlier debris flow activity
over the fan, which may have been caused by both debris flows and “normal” flood
events with fluvial sediment transport. It should also be noted that due to human use
or construction on a fan, old debris traces may no longer be visible.

4.6.2.2 Vegetation cover

If, in the region of the fan, there are areas with tree populations of clearly differing
ages, then these may be associated with corresponding debris flow events going back
a long time (if a similar influence of snow avalanche events can be excluded).

4.6.2.3 Outcrops in old deposits

If outcrops are present the type of layering may be an indication of earlier debris-flow
activity and the individual layer thickness may provide information on the magnitude
of the event. If future potential deposition areas are inferred from old debris-flow
deposits, possible changes due to channel protection works or new buildings should
also be considered. Further, traces of old deposition may have been obliterated by
agricultural or other land use on the fan.

4.6.3 Simple assessment of depositional behavior

If there are no traces or indications of earlier events, the potential deposition zones
have to be estimated solely using the expected debris load, the topography of the fan,
and with assumptions concerning the geometry of the deposition.

RICKENM-Book.indb 87 2/23/2016 7:39:39 AM

 88 Methods for the quantitative assessment of channel processes in torrents (steep streams)

An estimate can be made of possible locations of outflow from the channel by

comparing the required flow cross-section with the existing channel cross-sections
along the fan (see also Fig. 4.13). Hereby, scenarios should also consider, for exam-
ple, that log jams may form at cross-sections under bridges as a result of driftwood
transport (see also chapter 4.6.1). Essentially, the depositional behavior of a debris
flow on the fan is controlled by its material properties, in addition to the topography.
In the case of a simple estimate of affected areas, first that part of the debris load is
determined which will probably be deposited in the channel itself. This part can vary,
depending on the location where overflowing on the riverbanks is expected. Even
with an adequate channel capacity, filling of the channel is possible; for example, if, in
the region of the confluence, there is insufficient room for the deposition of the mate-
rial or if the receiving stream is not able to transport the material any further. Then,
in the second step, the remaining debris load has to be “distributed” over the fan.
If the length of debris-flow deposition on the fan is estimated using empirical
approaches, this value should be compared to historical deposits (if any) and if neces-
sary adjusted accordingly. If there is no evidence of old deposits, as a rough approxi-
mation it may be assumed that the width of the deposit is about ten times that of the
debris-flow width at the fan apex (Ikeya 1981). The mean thickness of the deposit of
debris on the fan is often in the range 0.5 m to about 3 m. With these approximate
guide values it is possible with some plausibility to distribute the debris load over the
fan. Hereby, it is always necessary to keep in mind that the possible overflow loca-
tions and the topography as well as structures (buildings, traffic structures) can influ-
ence considerably the propagation and depositional behavior.

4.7 FINAL REMARKS

Besides the approaches presented here in chapter 4, other documentations on esti-

mating the various debris-flow parameters can also be found for example in Hungr
et al. (1984), Ikeya (1981, 1987), Pwri (1988), Rickenmann (1995, 1999),
Takahashi (1981), VanDine (1985, 1996) and Heinimann et al. (1998). Regarding
the hazard assessment of torrents prone to debris-flow occurrence, mainly several
empirical methods have been discussed in detail here. Overall, there is a variety of
models to describe the initiation, flow and depositional behavior of debris flows. Each
of these models, however, applies to specific material mixtures and boundary condi-
tions. In nature, the material composition and the water content of debris flows can
vary greatly. This makes the selection of different types of debris-flow models diffi-
cult, because appropriate criteria for that are still largely missing. Therefore empirical
approaches continue to have an important part to play.
Regarding the prediction of debris flows, it is possible, in principle, to deter-
mine critical values of rainfall for regionally-limited areas, provided that a sufficient
number of observations are available concerning past events. However, it is scarcely
possible to make a prediction of the exact place (stream channel) of occurrence. Early
warning systems are primarily useful to close endangered traffic routes in case of
a debris-flow event with the aid of appropriate monitoring devices (Chang 2003;
Badoux et al., 2009; Stähli et al., 2015).
Often the information on earlier debris-flow events is very limited, and estimat-
ing the frequency of occurrence of events of a particular magnitude is only possible in

RICKENM-Book.indb 88 2/23/2016 7:39:39 AM

 Debris flows 89

approximate terms. Nevertheless, information on earlier debris-flow activity is very

important in addition to the assessment of the potential sediment supply during a
future event. Such information is not only the basis for the magnitude-frequency rela-
tionship, but can also provide important clues for the flow behavior. Therefore it is
very important to document past events (e.g. Hübl et al., 2002) as well as to establish
event catalogues of landslides, debris flows and floods.
In engineering practice, the process assessment of debris flows depends mainly on
the analogy to earlier events, on an integral assessment of the catchment, on empirical
approaches to estimate the most important debris flow parameters, and on simula-
tion models. The most important parameter to estimate the debris-flow hazard is the
possible debris load. The use of Geographic Information Systems (GIS) has become
more important to estimate the erosion potential and potential debris loads. Simple,
semi-empirical models to determine the extent of the debris flow are used for example
in combination with GIS, for the preparation of hazard index maps.
Detailed process assessments are necessary for the preparation of hazard maps at
a more refined scale. This is often accomplished with numerical models for the simula-
tion of the debris-flow behavior. An uncertainty in this process is the determination or
estimation of the model parameters to characterize the material or flow behaviors. In
general, these parameters cannot be determined directly using samples of debris mate-
rial, since the necessary rheological testing equipment for this is not available. There-
fore, these parameters have to be estimated by back-calculations of similar events. In
assessing the hazards also a possible combination of processes should be taken into
account. An example of sediment deposits from a debris flow that were partially
eroded again by a subsequent flood on the fan and transported further downstream is
illustrated in Fig. 4.21 and Fig. 4.22.

Figure 4.22 Debris-flow deposits of the event of 24 August 1987 on the Minstiger stream fan in Swit-
zerland (photo M. Zimmermann, Thun). The debris flow consisted of a single surge that
took place at the beginning of the afternoon, while the fluvial deposits are the result of
a flood that occurred in the evening of the same day and also re-entrained finer material
from the debris-flow deposits. (Compare also Fig. 4.21).

RICKENM-Book.indb 89 2/23/2016 7:39:39 AM

This page intentionally left blank
 Chapter 5

Magnitude and frequency

of torrent events

Although the relation between the magnitude and the frequency of a debris-flow event
is essential for any hazard or risk analysis, it is often difficult to assess. The magnitude
of a debris flow event forms an input or basis both for simple empirical relations to
estimate important flow parameters (chapter 4.4) and for numerical models simulat-
ing flow propagation and deposition (chapter 4.5).
It is very challenging to determine accurately the probability of occurrence of
debris flow events of a given magnitude in torrent catchments, because historic data
are generally approximate and a detailed assessment of sediment deposits by strati-
graphic analysis is typically very expensive (Jakob 2012). This statement is also valid
partly also for bedload transporting flood events in torrents. While, in this case, exist-
ing rainfall data may facilitate the frequency of rainfall-runoff events of different mag-
nitudes, the sediment supply or sediment availability is much more difficult to assess.
If historical data about earlier events are available, they often provide very important
information, even if, in general, no statistical evaluation can be made with them in a
narrow sense. The traditional concept of extreme value analysis of flood discharges
cannot be transferred directly to torrent events; in case of a limited bedload potential,
for example, the probability of a future event may largely depend also on the actual
stock of movable sediments.
The most important factors in connection with the occurrence of debris flows are
the identification of possible triggering zones and sediment supply sources (and, thus,
of the event magnitude) as well as the estimation of the frequency of events (Jakob
2005). In torrents prone to debris flow occurrence, the debris load is generally (much)
bigger than the sediment load of a flood event with only fluvial bedload transport, and
thus is more relevant for the hazard assessment. Here, therefore, the question of appro-
priate methods to determine the debris load is discussed primarily in detail. The debris
loads reported for past events usually include both the volume of solids and pores.
In addition, empirical values of debris loads are often based on the solids deposition
of a whole event, possibly including multiple debris-flow surges and fluvial bedload
transport. In this publication, the event magnitude is assumed to equal the debris load.
To estimate the debris load or the bedload of torrent events the hydrological and
geomorphological characteristics of a catchment are important. The sediment supply
to the channel network in steep headwater catchments and the total amount of sedi-
ment transferred to the fan is controlled both by mass movements on the adjacent
hillslopes and erosion and deposition of sediments along the channel (bed and banks)
during a flow event (Fig. 5.1).

RICKENM-Book.indb 91 2/23/2016 7:39:41 AM

 92 Methods for the quantitative assessment of channel processes in torrents (steep streams)

precipitation

hillslope
sediment delivery to
channel network

surface runoff fluvial erosion

subsurface water landslides

lateral deposition outside

erosion of channel

Sediment
channel runoff sediment transfer delivery to
fan area

channel in-channel
erosion deposition
channel

Figure 5.1 Simplified process system for torrents, after LIENER (2000) and GERTSCH (2009). The total
amount of sediment transferred to the fan during a torrent event depends on the sediment
supply from the hillslopes as well as on the erosion and deposition of sediments along the
channel. Modified from LIENER (2000) and GERTSCH (2009).

The fundamental importance of the geologic and geomorphologic aspects of tor-

rent catchments for erosion processes and mass movements is summarized by Bunza
et al. (1976). After Stiny (1931), a torrent catchment with relatively young alluvial
deposits (“young debris” torrent) is characterized by sediment supply sources con-
sisting mainly of recently weathered rock material. By contrast, according to Stiny
(1931), in a torrent catchment with abundant residual colluvium (“old debris” tor-
rent) there is generally almost unlimited sediment availability (e.g. due to glacial mate-
rial consisting of moraine deposits from the various ice ages). The moraine materials,
especially those with much fine material, belong to the most hazardous sediment
sources of the Alps for torrent floods and debris flows (Luzian et al., 2002). The
importance of hillslope processes for the sediment supply to channel processes in steep
catchments is also discussed for example in Benda et al. (2005), Wichmann et al.
(2009), Geertsema et al. (2010), and Theler et al. (2010).
In principle, potential future sediment delivery processes may be modeled for
entire headwater catchments to estimate possible event magnitudes for example based
on rainfall scenarios and slope stability analyses (e.g. Montgomery & Dietrich,
1994; Baum et al., 2010; von Ruette et al., 2013). However, when applying such
models in practical hazard assessment, one of the main challenges is that the char-
acteristics of the soil and subsurface layers are often heterogeneous and unknown.
Therefore, the focus here is more on primarily field-based approaches relying on an

RICKENM-Book.indb 92 2/23/2016 7:39:41 AM

 Magnitude and frequency of torrent events 93

assessment of hillslope sediment delivery and empirically derived channel erosion

rates to be expected during a defined triggering event (in terms of rainfall conditions).

5.1 EMPIRICAL APPROACHES TO ESTIMATE THE

MAGNITUDE OF AN EVENT

Some empirical approaches for estimating the debris load or solids load of a torrent
event are summarized in Table 5.1. Such approaches usually include simple catchment
parameters. They allow an estimate of either an upper limit or of a mean value of the
possible debris load or the bedload volume. Only two approaches account for geo-
logic characteristics. In the equation of Kronfellner-Kraus (1984), the coefficient K
varies with the geology and the catchment area, with values between about 250 (tor-
rents of the alpine foothills in Austria) and about 1750 (torrents with large sediment
sources in residual colluvium [see also chapter 5.5 below]). The value for the geologic-
lithologic index IG after D’Agostino & Marchi (2001) can take on values in the
range 0.5 to 5, depending on the susceptibility to weathering of the surface material.
A comparison between observed bedload as a function of the size of the catch-
ment area exhibits a large scatter of several orders of magnitude (Fig. 5.2); thus, these
formulae can only provide very rough estimates. For the development of more reliable
approaches, in particular the special geological, geomorphological and hydrological
features of a torrent catchment must be taken into account, requiring more detailed
investigations.
It may be helpful for the hazard assessment of a particular torrent catchment
to compare the estimated event load with the range of values from earlier observa-
tions. As an example, for Switzerland there is a compilation of the specific event
loads per unit catchment area, grouped according to the predominant geology (i.e. for
torrents in the alpine limestone regions, in the crystalline rocks, in the Molasse and
Flysch areas) (Spreafico et al., 2005; Grasso et al., 2007) (Fig. 5.3 and Fig. 5.4).

Table 5.1 Simple empirical equations for a rough estimate of the event load of a debris-flow event
or a bedload-transporting flood in a torrent; N = number of events as a basis to derive a
formula. Definition of the parameters: M = “maximum” event load [m3]; Ma = mean event
load [m3]; Ac = catchment area [km2]; Sc = mean channel slope [−]; Sf = mean fan slope [−];
Lc = length of the active channel [m]; K = torrentiality factor; IG = geologic-lithologic index.
(*) This relationship was first derived for event loads in the case of bedload transport, and
the coefficients were then adjusted for 15 larger debris flow events in Austria.

Equation N Source

M = K Ac 100 Sc 1420 KRONFELLNER-KRAUS (1984);

KRONFELLNER-KRAUS (1987)
M = 27000 Ac0.78 ∼65 ZELLER (1985); RICKENMANN (1995)
Ma = 150 Ac (100 Sf − 3)2.3 15 (*) HAMPEL (1980)
M = Lc (110 − 250 Sf) 82 RICKENMANN & ZIMMERMANN (1993)
Ma = 13600 Ac0.61 551 TAKEI (1984)
Ma = 29100 Ac0.67 64 D’AGOSTINO et al. (1996)
Ma = 70 Ac Sc1.28 IG 84 D’AGOSTINO & MARCHI (2001)

RICKENM-Book.indb 93 2/23/2016 7:39:41 AM

 94 Methods for the quantitative assessment of channel processes in torrents (steep streams)

Zeller (1985): extreme conditions

D’Agostino et al. (1986): Ma (Italy)
Takei (1980): Ma (Japan)
Switzerland
Canadian Cordillera
1000000
Austria
Italy (D’Agostino et al., 1986)
Italy (Marchi & Tecca, 1996)

100000
Debris load M (m3)

10000

1000
0.01 0.1 1 10 100
Catchment area Ac (km2)

Figure 5.2 Observed event load (magnitude) of debris-flow events, mainly for Switzerland and North-
ern Italy, as a function of the size of the catchment area. Also shown are some estimating
formulae from Table 5.1.

100000
mean annual load
maximum annual load
maximum load of single event

10000
specific sediment load [m3/km2]

1000

100

10

1
conglomerate flysch limestone cristalline

Figure 5.3 Range of observed specific sediment load (normalized by catchment area) subdivided
according to geology, based on observed deposition volumes in sediment retention basins
in Switzerland. Data from GRASSO et al. (2007) in Hydrological Atlas of Switzerland.

RICKENM-Book.indb 94 2/23/2016 7:39:41 AM

 Magnitude and frequency of torrent events 95

100000

cristalline
limestone
flysch
10000 conglomerate

maximum annual specific sed. load (m3/km2/yr)

1000

100

10

1
0.1 1 10 100 1000

catchment area (km2)

Figure 5.4 Specific annual sediment load (normalized by catchment area) versus catchment area,
and subdivided according to geology. There is a tendency for the specific annual event
load to decrease with increasing size of the catchment. Data from GRASSO et al. (2007) in
Hydrological Atlas of Switzerland.

Further evaluations of these data from sediment retention basins indicate that the
shape of the catchment could also have some influence (Grasso et al., 2010).

5.2 FIELD-BASED ESTIMATE OF THE EVENT

MAGNITUDE

For a more exact assessment of the potential event magnitude (debris load or bedload),
which could be mobilized during a rainstorm event, a geological-geomorphological
assessment of the catchment is performed in many cases, whereby the use of a Geo-
graphic Information System (GIS) may be helpful. The main potential triggering areas
of debris flows are steep channels or gullies with abundant regolith (colluvium, allu-
vium) or unstable hillslopes. The latter can also be important for the formation of
hillslope debris flows and sediment delivery to the channel network.
A method used frequently in engineering practice to assess the potential event
magnitude is to estimate average erosion cross-sections (“channel debris yield rates”
according to Hungr et al., 1984) for more or less homogeneous channel reaches.
The resulting erosion volumes are then summed over the whole length of the chan-
nel network thought to be affected by sediment entrainment during the event. Typi-

RICKENM-Book.indb 95 2/23/2016 7:39:42 AM

 96 Methods for the quantitative assessment of channel processes in torrents (steep streams)

cal values of channel debris yield rates as a function of the channel properties and
geological-lithological conditions are given in Table 5.2, according to Hungr et al.
(1984). Similar observations on specific channel erosion rates for debris flows and
torrent events are reported in Spreafico et al. (1996), Zimmermann & Lehmann
(1999), Marchi & D’Agostino (2004), and Hungr et al. (2005). According to an
investigation of debris-flow events in Switzerland (Rickenmann & Zimmermann
1993) the mean specific channel erosion rates varied between 40 m3/m and 90 m3/m,
and locally values of 500 m2 to 650 m2 were observed. Such large values were also
reported after the outbreak of water from water pockets in glacial areas (HAEBERLI
1983). Outburst flows of water from glacial lakes can lead to very hazardous debris
flows, since below the dam breach large discharge peaks can occur and further down-
stream there are typically steep channels within morainic material that can often be
eroded easily (Clague & Evans 2000; Chiarle et al., 2007).
Based on the limited number of observations in Switzerland and Austria an
approximate empirical formula was proposed to estimate the “maximum” erosion
depth Te [m] in function of the local channel slope S [m/m] (Vaw1992):

Te = 1.5 + 12.5 S (5.1)

However, as mentioned in chapter 4.3, only very limited field observations are
available to document the erosion of debris-flows along the flow path. Therefore,
methods for a practical estimation of debris entrainment are largely lacking, despite
its importance for the hazard assessment. An example of strong bed and bank erosion
along the channel during debris-flow events is shown in Fig. 5.5.
As a method to estimate a potential event magnitude for a torrent catchment
based on specific channel erosion rates, in Switzerland the field-based approach of
Lehmann (1993) was further developed (Frick et al., 2008, 2011; Kienholz et al.,

Table 5.2 Typical values for channel debris yield rates in function of the channel properties and geologic-
lithologic conditions from a Canadian investigation of HUNGR et al. (1984). Catchments with
areas of 1 to 3 km2 were investigated.The stability condition (*) refers to the situation prior
to the expected event.

Channel Gradient Stability Channel debris

type [°] Bed material Hillslopes condition (*) yield rate [m3/m]

A 20–35 bedrock non-erodible stable, practically 0–5

bare of soil cover
B 10–20 thin debris or non-e Stable 5–10
loose soil over rodible
bedrock (bedrock)
C 10–20 deep talus or less than Stable 10–15
moraine 5 m high
D 10–20 deep talus or talus, over hillslopes at 15–30
moraine 5 m high repose
E 10–20 deep talus or talus, over hillslopes up to 200
moraine 20 m high potentially (consider as
unstable point source)
(landslide area)

RICKENM-Book.indb 96 2/23/2016 7:39:42 AM

 Magnitude and frequency of torrent events 97

Figure 5.5 Strong erosion along the stream channel during the two debris flow events of 18 July and
24 August 1987 in Val Varuna (near Poschiavo, Canton Grisons), Switzerland. (a) Situation
before the events (Photo Kraftwerke Brusio AG), (b) Situation after the first event (Photo
U. Eggenberger, 29 July 1987), (c) Situation after the second event (Photo G. Paravicini,
29 August 1987). The blue circles mark the positions of an old masonry torrent check dam.

2010). This method called “SEDEX” allows for a more systematic assessment of the
sediment contributions of individual channel reaches to estimate a total event volume
to be expected at the fan apex. Hereby, several possible event scenarios (e.g. typical
rainstorms with a given return period) as well as uncertainties are considered sys-
tematically. An important goal of this approach is to ensure the reproducibility and
transparency of the assessments.

5.3 COMBINED METHOD FOR ESTIMATING THE EVENT

MAGNITUDE

In addition to the method SEDEX, a somewhat more complicated approach for

estimating a potential event magnitude for debris flows and fluvial bedload trans-
port in torrents was developed subsequently (Gertsch 2009; Gertsch et al., 2010;
Kienholz et al., 2010). This method is called “Gertsch” method. In addition to field
or map-based estimates of specific channel erosion rates, it includes expert knowledge
to modify these first estimated based on a number of catchment characteristics. The
method was developed from the analysis of 58 large torrent events in the Swiss Alps
(mainly debris flow events) with a recurrence period of at least about 100 years. The
method was programmed as a kind of expert system in the form of an Excel file with
an assessment of both hillslope and channel processes, and it was validated using
43 test catchments with large torrent events. The method is suitable for catchments
smaller than 10 km2 in alpine and pre-alpine torrent systems with a mean channel
slope well above 10%, and in which active bedload transfer processes can be expected.
The method consists of a system-based approach. It is assumed that the extent
of sediment entrainment along a channel reach is given by the characteristics of
the given reach, by the conditions and expected processes in the upstream channel
reaches, and by threshold processes affecting the entire torrent system as a whole.

RICKENM-Book.indb 97 2/23/2016 7:39:42 AM

 98 Methods for the quantitative assessment of channel processes in torrents (steep streams)

The basic catchment parameters can be determined using GIS. Many analysis steps
are partly automated and executed in an Excel file. The method does not necessar-
ily require fieldwork, but then the assessment is expected to be less reliable. When
using field data additionally, this part of the approach is very similar to the method
SEDEX: The results of the method are erosion and deposition loads and thus a sedi-
ment budget along the entire flow path. Finally, the expected sediment load at the fan
apex can be determined for 100 to 300 year events. A further option is to consider
pessimistic scenarios with an assumed return period of more than 300 years.

5.4 FLOOD RUNOFF AND DEBRIS-FLOW OCCURRENCE

For the triggering of debris flows a minimum amount of water is necessary. Debris-
flow formation is not only influenced by the surface runoff, which may be largely
controlled by rainfall intensity in steep headwater catchments, but also by the
degree of soil saturation which is also controlled by rainfall duration. If, through
a hillslope instability, a larger sediment volume is moved to a channel, a minimum
water input (into the pores of the soil and/or as channel discharge) is required, so
that the solids-water mixture is able to reach the fan. Modeling the rainfall-runoff
response can be useful in small catchments to estimate a potential water input vol-
ume that may limit the maximum emerging debris load that may transform into a
debris flow.
Some approaches were proposed to derive a possible debris flow hydrograph based
on a pure water hydrograph (Gostner et al. 2003). The amount of the entrained sol-
ids may then be estimated based on the sediment transport capacity of the water dis-
charge, or more simply by assuming a (constant) bulking factor (Gallino & Pierson
1985; Pierson 1995; Breien et al., 2008; Gartner et al., 2008; Santi et al., 2008).
However, these approaches are subject to large uncertainties. In particular, in this
way the maximum discharge of a debris flow can be greatly underestimated, because
often the maximum of a debris flow does not correspond just to a simple increase of
the peak water discharge by bulking the hydrograph with the additional sediment
volume. Estimates show that the maximum discharge of a debris flow may be as
much as 10 to 100 greater than the peak discharge of a flood in the same area for the
same rainfall conditions (Zimmermann & Rickenmann 1992; see also Chapter 4
and Table 4.3).

5.5 FREQUENCY OF TORRENT EVENTS

The location and the type of triggering influence primarily the magnitude of the event.
The two elements together are important for the flow behavior in the channel. In an
area with a limited sediment source potential (“young debris” torrent; see also begin-
ning of chapter 5) the frequency and the magnitude of future debris flows depend on
material removal (location, extent) by previous debris flows (Zimmermann et al.,
1997a, b).
The most reliable estimate of the possible future frequency of events is based
on information of past events. An important source is historical documents (also by

RICKENM-Book.indb 98 2/23/2016 7:39:43 AM

 Magnitude and frequency of torrent events 99

spoken communications), while another possibility is in the analysis and interpreta-

tion of geomorphological traces. Often the information concerning earlier events is
very limited, and the estimate of the frequency of occurrence of events of a certain
magnitude is only approximate. Therefore, one has to work with scenarios which
are assigned to approximate recurrence periods. A very rough idea, however, can be
obtained from the frequency of peak rainfall conditions.
Geomorphological methods to determine debris-flow frequency and, to some
extent also, magnitude include dendrochronology (e.g. Schneuwly-Bollschweiler
et al., 2013) and lichenometry (Innes, 2006). Radiocarbon dating may be applied
where natural exposures or test trench sediments provide organic materials for dating
(Chiverrell & Jakob 2011). Mapping of dated events along with determination of
the thickness of the respective deposits can yield magnitude estimations. But even if a
comparatively large set of historical data on past debris-flow occurrence is available,
the estimation of magnitude-frequency relation with statistical methods may not be
straightforward (Jakob 2012; Nolde & Joe 2013; Rickenmann & Jakob 2015).

5.5.1 Debris-flow events

Characteristic patterns regarding the torrent activity were identified in a study based
on 189 historically documented events with mainly debris flows which had occurred
in 17 torrent catchments in Switzerland (Zimmermann et al., 1997a, b). These pat-
terns appear to depend mainly on the supply of bedload or the geology and lithology
of the catchment. Comparing geomorphological characteristics of the catchments and
historical event information resulted in the definition of the following four types of
torrent activity (Zimmermann et al., 1997a, b).

1 Torrents with a more or less regular occurrence of debris flows (Fig. 5.6). The
time interval of the inactive periods between events amounts to around 15 to 30
years, most bedload is eroded diffusely along the channel, and the sediment load
of the whole event is typically smaller than 100,000 m3. The sediment sources
frequently consist of relatively young weathered material (mainly “young debris”
torrents after Stiny 1931), which is eroded along the flow path.
2 Torrents with a rather irregular occurrence of debris flows (Fig. 5.7). After a
relatively active period lasting from years to a few decades there may be a longer
period of several decades with little to no activity. These torrents run mainly in
relatively weak rock of variable strength such as Bündner schist. Debris flow
events can transport large amounts of sediments of clearly more than 100,000 m3,
which is mainly eroded along the flow path. It may be expected that a major
debris flow results in a destabilization of the bed and banks.
3 The occurrence of debris flows is irregular (Fig. 5.8). The catchment is char-
acterized by large abundant debris mainly in moraines and talus slopes (“old
debris” torrents after Stiny 1931), and the main sediment sources are in the
upper part of the catchment. The sediment load is variable and may amount to
several 100,000 m3.
4 Torrent catchments for which there are no historical parallels for the occurrence
of debris flows (Fig. 5.9). In this category two torrent catchments were identified.
An example is the event of 24 August 1987 in the Minstiger stream (Canton

RICKENM-Book.indb 99 2/23/2016 7:39:43 AM

 100 Methods for the quantitative assessment of channel processes in torrents (steep streams)

XXL
Steinlouibach (Lungern, OW)
XL
Size of event

1900 1920 1940 1960 1980 2000

XXL
Leimbach (Frutigen, BE)
XL
Size of event

1820 1840 1860 1880 1900 1920 1940 1960 1960 2000

XXL
Ri di Foioi (Faedo, Bavonatal, TI)
XL
Size of event

1860 1880 1940 1920 1940 1960 1980 2000

Figure 5.6 Event magnitudes and their frequency in torrent catchments. Here Type 1 is illustrated for
three torrents in Switzerland (ZIMMERMANN et al., 1997a, b). The lighter shaded parts of the
bars denote an uncertainty regarding the estimated event magnitude or the event type
(possibly flood with bedload transport). Adapted from ZIMMERMANN et al. (1997a, b).

Valais, Switzerland) when a single debris-flow surge reached down to the village
of Münster. For a long time, debris-flow activity only occurred in the upper part
of the catchment, but for a time period of nearly 300 years before 1987 no simi-
lar event has taken place. The event could be related to changes in the subglacial
runoff in a warmer climate (Zimmermann et al., 1997a, b).

RICKENM-Book.indb 100 2/23/2016 7:39:43 AM

 Magnitude and frequency of torrent events 101

XXL
Buochser Rübi (Buochs, NW)
XL
Size of event

1720 1760 1800 1840 1880 1920 1960 2000

Figure 5.7 Event magnitudes and their frequency in torrent catchments. Here Type 2 is
illustrated for one torrent in Switzerland (ZIMMERMANN et al., 1997a, b). The lighter
shaded parts of the bars denote an uncertainty regarding the estimated event magnitude
or the event type (possibly flood with bedload transport). Adapted from ZIMMERMANN et al.
(1997a).

XXL
Ritigraben (St. Niklaus,VS)
XL
Size of event

1860 1880 1900 1920 1940 1960 1980 2000

XXL
Dorfbach Randa
XL
Size of event

1860 1880 1900 1920 1940 1960 1980 2000

Figure 5.8 Event magnitudes and their frequency in torrent catchments. Here Type 3 is
illustrated for two torrents in Switzerland (ZIMMERMANN et al., 1997a, b). The lighter
shaded parts of the bars denote an uncertainty regarding the estimated event magnitude
or the event type (possibly flood with bedload transport). Adapted from ZIMMERMANN et al.
(1997a, b).

RICKENM-Book.indb 101 2/23/2016 7:39:44 AM

 102 Methods for the quantitative assessment of channel processes in torrents (steep streams)

XXL
Minstigerbach
XL
Size of event

1720 1760 1800 1840 1880 1920 1960 2000

Figure 5.9 Event magnitudes and their frequency in torrent catchments. Here Type 4 is illustrated for
one torrent in Switzerland (ZIMMERMANN et al., 1997a, b). The lighter shaded parts of the
bars denote an uncertainty regarding the estimated event magnitude or the event type
(possibly flood with bedload transport). Adapted from ZIMMERMANN et al. (1997a, b).

5.5.2 Fluvial sediment transport events

For torrent events with fluvial bedload transport (or solids transport), the frequency
is determined mostly using the estimated frequency of a flood event or of the cor-
responding rainfall event. Calculations of fluvial bedload transport are appropriate,
for example, in torrent channels with limited bed slopes, where the probability of
occurrence of debris flows is small. In this case, estimates of event sediment loads can
be made by integrating the bedload transport over the period of a flood hydrograph.
With calculations of fluvial bedload transport it has to be taken into account that
transport formulae based on total bed shear stress give a maximum solids transport
capacity, which is determined from the discharge and the flow hydraulics. However,
in steep headwater streams the transported bedload can also be highly dependent on,
or limited by, the sediment supply or sediment availability.

5.6 GENERAL REMARKS ON THE ESTIMATE

OF SEDIMENT LOADS IN TORRENT CATCHMENTS

The estimation of sediment loads in torrent catchments for events of different return
periods is an important task within the framework of a hazard assessment but it is
subject to considerable uncertainty. These quantitative approaches are helpful only to
a limited extent, and the estimates are based frequently on documented loads from
earlier events and are often strongly dependent on expert and field-based evaluations
of the conditions in the catchment. Differences in the estimated sediment loads of a
factor 2 (between different expert reports) are not uncommon.
A frequent problem is the question to what extent existing or planned protective
measures can or should be taken into account in estimating the sediment load and
the rainfall-runoff response in the case of torrent catchments. Here too, quantitative
statements are often difficult to make. A simpler case is to determine the effect of a
sediment retention basin: Basically the expected sediment load downstream can be

RICKENM-Book.indb 102 2/23/2016 7:39:44 AM

 Magnitude and frequency of torrent events 103

reduced by the capacity (volume) of the retention basin, if the channel below it is not
subject to significant erosion. More difficult to estimate, perhaps, is the influence of
a series of torrent check dams on the amount of sediment retained along the chan-
nel; this possible sediment retention depends also on the state of the check dams (e.g.
type of construction, age, wear and tear or damage) and potential risk of their failure.
A detailed discussion of the effectiveness of protection measures in torrent catchments
in the context of hazard assessment may be found in Planat (2008).

RICKENM-Book.indb 103 2/23/2016 7:39:45 AM

This page intentionally left blank
 Chapter 6

General remarks on hazard

assessment of channel processes
in torrents

6.1 REPRODUCIBILITY OF THE PROCESS ANALYSIS

AND HAZARD ASSESSMENT

According to Kienholz (1998, 1999, 2002) and Planat (2000) the hazard assess-
ment should not only fulfill the requirement of factual correctness, but also guaran-
tee the best possible reproducibility. Securing good reproducibility of this process
requires a certain effort regarding the detailed documentation of the assumptions
and the methods used in a technical report, but it helps the quality control and
simplifies the technical discussion as well as the comparison of different hazard
assessments.
The requirement of reproducibility involves ensuring that the selected method of
hazard assessment is transparent. In this way, the procedure, the applied approaches
and methods, together with the interpretation if the compiled data can be more easily
checked. This is also important because an exact evaluation of the process assessment
is difficult even after the occurrence of a (larger) event. After Kienholz (1999), there-
fore, the following basic rules should be adhered to:

• comprehensive documentation (cartographic presentation) of all the relevant

perimeters (areas of initiation and of impact of hazard, i.e. the torrent catchment
area and the alluvial fan)
• clear choice of methods and their combination and their reporting
• clear-cut decision criteria in the final evaluation
• provision of evidence for identified sub-processes (whose existence can be quali-
fied through traces of earlier events either as “proven”, “assumed” or “potential”)

6.2 PROCEDURE OF HAZARD ASSESSMENT

AND IMPORTANT ASPECTS

In Switzerland, the national platform for natural hazards Planat (2000) published
recommendations for quality assurance in the assessment of natural hazards. Accord-
ing to this document, the most important elements of the hazard assessment are:

• event documentation/analysis of causes

• map of the phenomena

RICKENM-Book.indb 105 2/23/2016 7:39:45 AM

 106 Methods for the quantitative assessment of channel processes in torrents (steep streams)

• hazard map
• risk analysis
• specific hazard assessment for selected critical locations
• protective measures/early warning systems

With regard to and based on the procedure for the hazard assessment, the
following sub-steps can be derived. These sub-steps are briefly summarized in connec-
tion with the recommendations produced in Switzerland (BWW/BRP/BUWAL 1997;
BUWAL/ BWW/BRP 1997):

6.2.1 Basic data

The event documentation and the analysis of causes (especially after large and impor-
tant events) represent important tasks in the compilation of the basic data. They
include investigations of earlier events in the catchment, the dominant processes, the
areas affected, the damage observed, and the triggering factors such as the meteoro-
logical conditions. A detailed documentation with the analysis of causes also exam-
ines how the event evolved and why damage was caused. Potential hazard processes
are documented with the map of the phenomena, which can be derived primarily on
the basis of characteristics observed in the field and indicators (traces). This map also
represents an important part of the fundamentals.

6.2.2 Preparation of hazard map

The main step in the hazard assessment is the preparation of the hazard map. Within a
clearly defined perimeter it contains comprehensive information on the hazard poten-
tial. The most important elements in the preparation of the hazard map concern, for
a torrent catchment, the determination of the dominant hazardous processes (e.g.
importance of debris flows, slope instabilities, fluvial bedload transport etc.), the
analysis of triggering conditions and the assessment of the probability of occurrence
of an event. This is followed by the calculations and modeling of the processes with
respect to the spatial spreading out, i.e. especially the process intensities within the
potentially endangered areas for a given probability of occurrence. Finally, for each
process and the associated event frequency this determines the hazard level (red, blue,
yellow, yellow-hatched).
The hazard map serves as the foundation for the risk analysis, the land use plan-
ning, the conception and design of protective measures as well as of preventive meas-
ures as part of the event management (early warning, emergency planning, etc.). From
this perspective it is clear that the creation of a hazard map and the preparation of
the corresponding technical report should be as comprehensive as possible and well
documented.
The evaluation of the different processes in torrents with a view to the hazard
assessment is based mainly on investigations and calculations according to the main
steps summarized in this chapter. The most important aspects of this procedure as
well as of the sub-processes to be considered are summarized in a compact form in
Table 6.1.

RICKENM-Book.indb 106 2/23/2016 7:39:45 AM

 Table 6.1 The evaluation of channel processes in torrents and mountain rivers with regard to hazard
assessment depends basically on investigations and analyses of the following important
aspects or sub-processes.

A. Compilation of basic data

• Aim of the investigation (repair, new construction, planning of hazard zones, etc.)
• Event documentation, chronicle, event register, possibly analysis of triggering processes
• Geomorphologic consideration of the catchment (incl. maps of the processes or maps
of the phenomena)
• Geographic parameters of the catchment
• Existing protective measures
B. Magnitude-frequency relationship
• Rainfall (intensity, duration)
• Discharge
• Triggering mechanisms and place of origin in the case of debris flows
• Sediment supply potential/expected sediment load
• Driftwood
To be assessed for different probabilities of occurrence according to BWW/BRP/BUWAL (1997)
Here it should be noted that this step is not independent of steps (C) and (D). In the case of
fluvial sediment transport the material entrainment and deposition is typically dependent on
the discharge along the channel. With debris flows local initiation zones (e.g. slope instabilities)
and sediment inputs in the channel can dominate the total solids transport, but also the material
entrainment along the flow path can be very important. Essentially steps (B), (C) and (D)
represent an iterative process.
C. Considerations concerning process sequences and scenarios
• Possible Interaction between different processes
• Scenarios in the case of difficult determination of process sequences and of difficult
quantification of the corresponding probabilities of occurrence (cf. chapter 4.6)
• Influence of protective measures
D. Analyses of flow, transport and depositional behavior
Important to determine the spatial distribution of the process intensities for given probability
of occurrence and event magnitude
D1. Flood formation
• Determination of discharge (peak discharge, hydrograph)
• Hydraulic calculations
D2. Driftwood
• Mobilized load of large wood and dimensions of individual pieces
• Transport capacity
• Possible clogging or log-jam places, identification of endangered areas
D3. Fluvial sediment transport
• Discharge and hydraulics (cf. D1)
• Mobilized volume of solids
• Transport capacity
• Erosion and deposition
• Influence of driftwood (cf. D2)

(Continued )

RICKENM-Book.indb 107 2/23/2016 7:39:45 AM

 108 Methods for the quantitative assessment of channel processes in torrents (steep streams)

Table 6.1 (Continued)

D4. Debris flow

• Mobilized volume of solids (slope instabilities and channel erosion)
• Possibly, comparison of expected sediment loads with water runoff volumes
• Solids concentration and sediment properties (flow properties/rheology)
• Debris-flow “hydrograph” at the fan apex
• Flow and depositional behavior
• Influence of driftwood (cf. D2, D3)
• Any process superposition (e.g. fluvial reworking of debris-flow deposits)
E. Uncertainties
• Quantification of uncertainties
• Sensitivity analyses on uncertainties regarding input or model parameters
• Partial consideration of uncertainties by means of scenarios (cf. C)

6.2.2.1 Magnitude-frequency relationship

Determining the probability of occurrence and the corresponding magnitude of an
event is often a key step, since essentially in many cases the event magnitude deter-
mines the process intensity. Thus it is important to give clear information on how the
probability of occurrence of a given process and the corresponding event magnitude
(i.e. magnitude-frequency relationship) were assessed.

6.2.2.2 Calculations and process modeling

Another important step is the prediction of the temporal development and spatial
displacement of a sub-process (“modeling”), for example, of a debris flow or a flood
with bedload transport. Here (numerical) simulation models or, additionally, empiri-
cal approaches can be used. To reproduce this sub-process, exact information on the
assumptions and basic principles of the calculations are necessary, especially with
regard to the selected model parameters. For example, for the modeling of debris
flows the selected model parameters should ideally be based on the back-calculation
of earlier events. If this is not possible, other assumptions have to be made, e.g. in the
use of model parameters as determined for similar areas and material compositions of
debris flows based on back-calculations.

6.2.2.3 Formulation of scenarios

The meaning of the term “scenario formulation” as used here (see also chapter 4.6)
developed against the background that, above all, the interaction between different
processes is often difficult to quantify, particularly in relation to the probability of
occurrence of such a combined effect. Examples of scenarios in this sense include the
influence of driftwood with regard to a possible (or not occurring) log jam in the
case of a bridge cross-section, or the occurrence of a larger landslide supplying sedi-
ment to a channel, which depending on the place and type of input can have different
effects in the hazard zone. Also generally important for scenario formulation is the

RICKENM-Book.indb 108 2/23/2016 7:39:45 AM

 General remarks on hazard assessment of channel processes in torrents 109

investigation of possible critical channel locations (e.g. regarding flow overtopping

from channel), since for a given probability of occurrence and event magnitude,
depending on the place of channel blockage and outflow, different hazard situations
can arise.

6.2.2.4 Uncertainties
The uncertainties should be explicitly mentioned in a technical report and, as far as
possible, quantified. This applies especially to the determination of the magnitude-
frequency relationship, cf. also (i). As already mentioned, the estimate of the sediment
loads in torrents for events of different return periods within the framework of a
hazard assessment is associated particularly with considerable uncertainty. Differ-
ences in the estimated sediment load by a factor 2 are definitely within the range of
uncertainty. Further, the uncertainty should be specified especially in relation to the
process modeling and the assumed model parameters (ii), and also in relation to the
scenarios (iii).

RICKENM-Book.indb 109 2/23/2016 7:39:45 AM

This page intentionally left blank
 References

Abrahams, A.D. (2003): Bedload transport equation for sheet flow. Journal of Hydraulic
Engineering, 129, 159–163.
Abrahams, A.D., Li, G., Krishnan, C., Atkinson, J.F. (2001): A sediment transport equation
for interrill overland flow on rough surfaces. Earth Surface Processes and Landforms, 26,
1443–1459.
Abt, S.R., Thornton, C.I., Scholl, B.A., Bender, T.R. (2013): Evaluation of Overtopping
Riprap Design Relationships. Journal of the American Water Resources Association, 49(4),
923–937.
Anastasi, G. (1984): Geschiebeanalysen im Felde unter Berücksichtigung von
Grobkomponenten. Mitteilungen der Versuchsanstalt für Wasserbau, Hydrologie und
Glaziologie, Eidgenössische Technische Hochschule Zürich, Nr. 70, 99p.
Ancey, C., Bain, V. (2015): Dynamics of glide avalanches and snow gliding. Reviews of
Geophysics, 53, 745–784.
Armanini, A., Scotton, P. (1993): On the Dynamic Impact of a Debris Flowon Structures.
Proceedings of XXV IAHR Congress, Tokyo, Tech. Sess. B III, pp. 203–210.
Armanini, A., Fraccarollo, L., Rosatti, G. (2009): Two-dimensional simulation of debris
flows in erodible channels. Computers & Geosciences, 35, 993–1006.
Ashmore, P.E. (1982): Laboratory modelling of gravel braided stream morphology. Earth
Surface Processes and Landforms, 7, 201–225.
Aulitzky, H. (1973): Berücksichtigung der Wildbach- und Lawinengefahrengebiete
als Grundlage der Raumordnung von Gebirgsländern. In: 100 Jahre Hochschule für
Bodenkultur, Band IV, Teil 2, pp. 81–117.
Ayotte, D., Hungr, O. (2000): Calibration of a runout prediction model for debris flows and
avalanches. In: Wieczorek, G.F. and Naeser, N.D. (eds), Debris-Flow Hazards Mitigation:
Mechanics, Prediction, and Assessment, Proceedings 2nd International Conference, Taipei,
Taiwan, pp. 505–514. Balkema, Rotterdam.
Aziz, N.M., Scott, D.E. (1989): Experiments on sediment transport in shallow flows in high
gradient channels. Hydrological Sciences Journal, 34, 465–478.
Badoux, A., Graf, C., Rhyner, J., Kuntner, R., McArdell, B.W. (2009): A debris-flow alarm
system for the Alpine Illgraben catchment: design and performance. Natural Hazards, 49(3),
517–539.
Bardou, E. (2002): Méthodologie de diagnostic et prévision des laves torrentielles sur un
bassin versant alpin. Thèse no. 2479, Ecole Polytechnique Fédérale de Lausanne. 187 pp.
+ annexes.
Bartelt, P., Salm, B., Gruber, U. (1999): Calculating dense-snow avalanche runout using a
Voellmy-fluid model with active/passive longitudinal straining. Journal of Glaciology, 45,
242–254.

RICKENM-Book.indb 111 2/23/2016 7:39:45 AM

 112 References

Bathurst, J.C. (1985): Flow Resistance Estimation in Mountain Rivers. Journal of Hydraulic
Engineering, 111, 625–643.
Bathurst, J.C. (1987): Measuring and modelling bedload transport in channels with coarse bed
materials. In K. Richards (ed.), River Channels – Environment and Process, pp. 272–294.
Blackwell, Oxford.
Bathurst, J.C. (1993): Flow resistance through the channel network. In K. Beven & M. J.
Kirkby (eds), Channel Network Hydrology, pp. 69–98, John Wiley, New York.
Bathurst, J.C. (2007): Effect of coarse surface layer on bed-load transport. Journal of
Hydraulic Engineering, 133, 1192–1205.
Bathurst, J.C. (2013): Critical conditions for particle motion in coarse bed materials of
nonuniform size distribution. Geomorphology, 197, 170–184.
Bathurst, J.C., Graf, W.H., Cao, H.H. (1982): Initiation of sediment transport in steep
channels with coarse bed material. In: Proc. of Euromech 156, Mechanics of Sediment
Transport, pp. 207–213.
Bathurst, J.C., Graf, W.H., Cao, H.H. (1987): Bed load discharge equations for steep
mountain rivers. In Thorne, C.R., Bathurst, J.C., & Hey, R.D. (eds), Sediment transport in
gravel bed rivers, pp. 453–477. Wiley, Chichester.
Baum, R.L., Godt, J.W., Savage, W.Z. (2010): Estimating the timing and location of shallow
rainfall-induced landslides using a model for transient, unsaturated infiltration. Journal of
Geophysical Research, 115, F03013. doi:10.1029/2009 JF001321.
Beguería, S., Asch, T.W.J., Malet, J.-P., Gröndahl, S. (2009): A GIS-based numerical model
for simulating the kinematics of mud and debris flows over complex terrain. Natural
Hazards and Earth System Sciences, 9, 1897–1909.
Benda, L., Hassan, M.A., Church, M., May, C.L. (2005): Geomorphology of steepland
headwaters: the transition from hillslopes to channels. Journal of the American Water
Resources Association, 41, 835–851.
Berger, C., McArdell, B.W., Schlunegger, F. (2011a): Direct measurement of channel
erosion by debris flows, Illgraben, Switzerland. Journal of Geophysical Research – Earth
Surface, 116, F01002. doi:10.1029/2010 JF001722.
Berger, C., McArdell, B.W., Schlunegger, F. (2011b): Sediment transfer patterns at the
Illgraben catchment, Switzerland: Implications for the time scales of debris flow activities.
Geomorphology, 125, 421–432.
Berti, M., Simoni, A. (2007): Prediction of debris flow inundation areas using empirical
mobility relationships. Geomorphology, 90, 144–161.
Bezzola, G.R. (2002): Fliesswiderstand und Sohlenstabilität natürlicher Gerinne unter
besonderer Berücksichtigung des Einflusses der relativen Überdeckung. Mitteilungen der
Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie, Eidgenössische Technische
Hochschule Zürich, Nr. 173, 258p.
Bezzola, G.R. (2005): Vorlesungsmanuskript Flussbau, Fassung Wintersemester 2005/2006.
Professur für Wasserbau an der Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie,
Eidgenössische Technische Hochschule Zürich.
Bezzola, G.R., Gantenbein, S., Hollenstein, R., Minor, H.-E. (2002): Verklausung von
Brückenquerschnitten. Int. Symposium Moderne Methoden und Konzepte im Wasserbau.
In: Mitteilungen der Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie,
Eidgenössische Technische Hochschule Zürich, Nr. 175, pp. 87–97.
Bezzola, G.R., Schilling, M., Oplatka, M. (1996): Reduzierte Hochwassersicherheit durch
Geschiebe. Schweizer Ingenieur und Architekt, 41, 886–892.
Braudrick, C., Grant, G.E. (2000): When do logs move in rivers. Water Resources Research,
36, 571–584.
Braudrick, C., Grant, G.E. (2001): Transport and deposition of large woody debris in
streams: A flume experiment. Geomorphology, 41, 263–284.

RICKENM-Book.indb 112 2/23/2016 7:39:45 AM

 References 113

Brauner, M. (1999): Modelling the sediment budget of an alpine catchment within a GIS
environment. In: Proceedings of the 28th IAHR Congress (1999), Graz, CD-ROM, 6p.
Brauner, M. (2001): Aufbau eines Expertensystems zur Erstellung einer ereignisbezogenen
Feststoffbilanz in einem Wildbacheinzugsgebiet. Dissertation am Institut für Alpine
Naturgefahren und Forstliches Ingenieurwesen, Universität für Bodenkultur, Wien.
Bray, D.I. (1979): Estimating average velocity in gravel-bed rivers. Journal of the Hydraulic
Division, 105(HY9), 1103–1122.
Breien, H., De Blasio, F.V., Elverhøi, A., Høeg, K. (2008): Erosion and morphology of a
debris flow caused by a glacial lake outburst flood, western Norway. Landslides, 5, 271–280.
Buffington, J.M., Montgomery, D.R. (1997): A systematic analysis of eight decades of
incipient motion studies, with special reference to gravel-bed rivers. Water Resources
Research, 33, 1993–2029.
Bunte, K., Abt, S.R. (2001): Sampling surface and subsurface particle size distributions in
wadable gravel- and cobble-bed streams for analysis in sediment transport, hydraulics, and
streambed monitoring. U.S. Dep. of Agric., For. Serv., Rocky Mt. Res. Stn., Fort Collins,
Colorado, USA.Gen. Tech. Rep. RMRS-GTR-74, 428p.
Bunte, K., Abt, S.R., Swingle, K.W., Cenderelli, D.A., Schneider, J.M. (2013): Critical
Shields values in coarse-bedded steep streams. Water Resources Research, 49, 1–21.
doi:10.1002/2012 WR012672.
Bunza, G., Karl, J., Mangelsdorf, J. (1982): Geologisch-morphologische Grundlagen der
Wildbachkunde. Schriftenreihe des Bayerischen Landesamtes für Wasserwirtschaft, Heft 17,
München, 128p.
Buscombe, D., Rubin, D.M., Warrick, J.A. (2010): A universal approximation of grain size
from images of noncohesive sediment. Journal of Geophysical Research – Earth Surface,
115, F02015. doi:10.1029/2009JF001477.
BUWAL/BWW/BRP (1997): Empfehlungen: Berücksichtigung der Massenbewegungsgefahren
bei raumwirksamen Tätigkeiten. Herausgeber: Bundesamt für Umwelt, Wald und Landschaft
(BUWAL), Bundesamt für Wasserwirtschaft (BWW), Bundesamt für Raumplanung (BRP),
Bern, 42p.
BWW/BRP/BUWAL (1997): Empfehlungen: Berücksichtigung der Hochwassergefahren bei
raumwirksamen Tätigkeiten. Herausgeber: Bundesamt für Wasserwirtschaft (BWW), Bundesamt
für Raumplanung (BRP), Bundesamt für Umwelt, Wald und Landschaft (BUWAL), Biel, 32p.
Caine, N. (1980): The rainfall intensity-duration control of shallow landslides and debris
flows. Geografiska Annaler, 62 A, 23–27.
Cannon, S.H. (1993): An empirical model for the volume- change behavior of debris flows. In:
H.W. Shen, S.T. Su & F. Wen (eds), Hydraulic Engineering 93,Vol. 2, American Society of
Civil Engineers, New York, pp. 1768–1773.
Carson, M.A. (1987): Measures of flow intensity as predictors of bed load. Journal of
Hydraulic Engineering, 113, 1402–1421.
Cao, Z., Pender, G., Wallis, S., Carling, P. (2004): Computational dam-break hydraulics
over erodible sediment bed. Journal of Hydraulic Engineering, 130, 689–703.
Chang, S.Y. (2003): Evaluation of a system for detecting debris flows and warning road
traffic at bridges susceptible to debris flow hazard. In: D. Rickenmann & C.L. Chen (eds),
Debris-Flow Hazards Mitigation: Mechanics, Prediction, and Assessment, pp. 731–742.
Millpress, Rotterdam.
Chanson, H. (2004): Hydraulics of Open Channel Flow. 2nd Edition, Elsevier, 585p.
Chen, H., Lee, C.F. (2003): A dynamic model for rainfall-induced landslides on natural slopes.
Geomorphology, 51, 269–288.
Chiari, M., Rickenmann, D. (2009): Modellierung des Geschiebetransportes mit dem Modell
SETRAC für das Hochwasser im August (2005) in Schweizer Gebirgsflüssen. Wasser,
Energie, Luft, 101, 319–327.

RICKENM-Book.indb 113 2/23/2016 7:39:45 AM

 114 References

Chiari, M., Rickenmann, D. (2011): Back-calculation of bedload transport in steep channels

with a numerical model. Earth Surface Processes and Landforms, 36, 805–815.
Chiari, M., Scheidl, C. (2015): Application of a new cellular model for bedload transporting
extreme events at steep slopes. Geomorphology, 246, 413–419.
Chiari, M., Friedl, K., Rickenmann, D. (2010): A one dimensional bedload transport model
for steep slopes. Journal of Hydraulic Research, 48, 152–160.
Chiarle, M., Iannotti, S., Mortara, G., Deline, P. (2007): Recent debris flow occurrences
associated with glaciers in the Alps. Global and Planetary Change, 56, 123–36.
Chiew, Y.M., Parker, G. (1994): Incipient sediment motion on non-horizontal slopes. Journal
of Hydraulic Research, 32(5), 649–660.
Chin, A. (1989): Step–pools in stream channels. Progress in Physical Geography, 13, 391–408.
Chiverrell, R., Jakob, M. (2011): Radiocarbon dating: alluvial fan/debris cone evolution
and hazards. In: M. Schneuwly-Bollschweiler, M. Stoffel, F. Rudolf-Miklau (eds), Dating
torrential processes on fans and cones - methods and their application for hazard and risk
assessment; Advances in Global Change Research, 47, 265–282.
Choi, S.U., Garcia, M.H. (1993): Kinematic wave approximation for debris flow routing.
Proc. XXV IAHR Congress, Technical Session B, Vol. III, pp. 94–101.
Christen, M., Bartelt, P., Kowalski, J. (2010): Back calculation of the In den Arelen avalanche
with RAMMS: interpretation of model results. Annals of Glaciology, 51, 161–168.
Christen, M., Bühler, Y., Bartelt, P., Leine, R., Glover, J., Schweizer, A., Graf, C.,
McArdell, B.W., Gerber, W., Deubelbeiss, Y., Feistl, T., Volkwein, A. (2012):
Integral hazard management using a unified software environment: numerical simulation
tool “RAMMS” for gravitational natural hazards. In: Proc. International Symposium
Interpraevent, Grenoble, France, Vol. 1, pp. 77–86.
Church, M. (2013): Steep Headwater Channels. In: Shroder, J., Wohl, E. (eds), Treatise
on Geomorphology. Academic Press, San Diego, CA, vol. 9, Fluvial Geomorphology,
pp. 528–549.
Church, M., McLean, D., Wolcott, J.F. (1987): River bed gravels: Sampling and analysis.
In Thorne, C.R., Bathurst, J.C., Hey, R.W. (eds), Sediment transport in gravel bed rivers,
Wiley, Chichester, pp. 43–79.
Clague, J.J., Evans, S.G. (2000): A review of catastrophic drainage of moraine-dammed lakes
in British Columbia. Quaternary Science Reviews, 19, 1763–1783.
Comiti, F., Mao, L. (2012): Recent advances on the dynamics of steep channels. In:
Church, M., Biron, P., Roy, A.G. (eds), Gravel Bed Rivers 7: Processes, Tools, Environments.
Wiley-Blackwell, Chichester, pp. 353–377.
Comiti, F., D’Agostino, V.D., Moser, M., Lenzi, M.A., Bettella, F., Agnese, A.D., Rigon,
E., Gius, S., Mazzorana, B. (2012): Preventing wood-related hazards in mountain basins:
from wood load estimation to designing retention structures. Proceedings, 12th Congress
INTERPRAEVENT 2012, Grenoble, France; 651–662.
Corominas, J. (1996): The angle of reach as a mobility index for small and large landslides.
Canadian Geotechnical Journal, 33, 260–271.
Costa, J.E. (1984): Physical geomorphology of debris flows. In: J.E. Costa & P.J. Fleischer
(eds), Developments and applications of geomorphology, Springer, Berlin, 268–317.
Costa, J.E. (1988): Rheologic, geomorphic, and sedimentologic differentiation of water floods,
hyperconcentrated flows, and debris flows. In: V.R. Baker, R.C. Kochel & P.C. Patton (eds),
Flood Geomorphology, John Wiley & Sons, New York, pp. 113–122.
Coussot P. (1997): Mudflow rehology and dynamics. IAHR Mongraph Series. Rotterdam:
Balkema. 255p.
Coussot, P., Laigle, D., Arattano, M., Deganutti, A., Marchi, L. (1998): Direct
determination of rheological characteristics of debris flow. J Hydraul Eng, 124, 865–868.

RICKENM-Book.indb 114 2/23/2016 7:39:45 AM

 References 115

Cui, P., Zeng, C., Lei, Y. (2015): Experimental analysis on the impact force of viscous debris
flow. Earth Surface Processes and Landforms, 40, 1644–1655.
David, G.C.L., Wohl, E.E., Yochum, S.E., Bledsoe, B.P. (2010): Controls on spatial variations
in flow resistance along steep mountain streams. Water Resources Research, 46, W03513.
doi:10.1029/2009WR008134.
D’Agostino, V., Cerato M., Coali, R. (1996): Il trasporto solido di eventi estremi nei torrenti
del Trentino Orientale. [Sediment transport of extreme events in torrents of eastern Trentino],
Proc. International Symposium Interpraevent, Garmisch-Partenkirchen, Germany, Bd. 1,
pp. 377–386 [in Italian].
D’Agostino, V., Marchi, L. (2001): Debris flow magnitude in the Eastern Italian Alps: data
collection and analysis. Physics and Chemistry of the Earth (C), 26, 657–663.
De Jong, C., Ergenzinger, P. (1995): The interrelations between mountain valley form and
river-bed arrangement. In: E.J. Hickin (ed.), River Geomorphology, pp. 55–91. John Wiley &
Sons Ltd.
Detert M., Weitbrecht V. (2012a): Automatic object detection to analyze the geometry of
gravel grains – a free stand-alone tool. In: Proc. of the 6th International Conference on
Fluvial Hydraulics, River Flow 2012, San José (Costa Rica), pp. 595–600. Boca Raton,
London.
Detert M., Weitbrecht V. (2012b): BASEGRAIN 1.0. Wasser, Energie, Luft, 104, 334.
Dittrich, A. (1998): Wechselwirkung Morphologie/Strömung naturnaher Fliessgewässer.
Mitteilungen des Institutes für Wasserwirtschaft und Kulturtechnik, Universität Karlsruhe,
Heft 198, 208p.
Eaton, B.C. (2013): Hydraulic geometry: empirical investigations and theoretical approaches.
In: Shroder, J., Wohl, E. (eds), Treatise on Geomorphology. Academic Press, San Diego, CA,
vol. 9, Fluvial Geomorphology, pp. 313–329.
Egashira, S., Ashida, K. (1991): Flow resistance and sediment transportation in streams
with step-pool bed morphology. In: Fluvial Hydraulics of Mountain Regions, pp. 45–58.
Springer: Heidelberg.
Egashira, S., Honda, N., Itoh, T. (2001): Experimental study on the entrainment of bed
material into debris flow. Physics and Chemistry of the Earth, Part C, 26(9), 645–650.
Egli, T. (2005): Wegleitung Objektschutz gegen gravitative Naturgefahren. Vereinigung
Kantonaler Feuerversicherungen VKF, Bern, Kapitel 5 Murgänge, pp. 77–87.
Einstein, A. (1942): Formulas for the transportation of bed load. Transactions of the American
Society of Civil Engineers, 107, 561–597.
Fannin, R.J., Wise, M.P. (2001): An empirical-statistical model for debris flow travel distance.
Canadian Geotechnical Journal, 38, 982–994.
Fehr, R. (1986): A method for sampling very coarse sediments in order to reduce scale effects in
movable bed models. In: Proc. Symposium on Scale effects in modelling sediment transport
phenomena, Toronto, IAHR, Delft, pp. 383–397.
Fehr, R. (1987a): Geschiebeanalysen in Gebirgsflüssen – Umrechnung und Vergleich von
verschiedenen Analyseverfahren. Mitteilungen der Versuchsanstalt für Wasserbau,
Hydrologie und Glaziologie, Eidgenössische Technische Hochschule Zürich, Nr. 92, 139p.
Fehr, R. (1987b): Einfache Bestimmung der Korngrössenverteilung von Geschiebematerial mit
Hilfe der Linienzahlanalyse. Schweizer Ingenieur und Architekt, 38, 1104–1109.
Ferguson, R. (2007): Flow resistance equations for gravel- and boulder-bed streams. Water
Resources Research, 43, W05427. doi:10.1029/(2006)WR005422.
Fraccarollo, L., Papa, M. (2000): Numerical Simulation of Real Debris-Flow Events. Physics
and Chemistry of the Earth, Part B: Hydrology, Oceans and Atmosphere, 25(9), 757–763.
French, R.H., Miller, J.J., Curtis, S. (2001): Estimating the depth of deposition (erosion) at
slope transitions on alluvial fans. Journal of Hydraulic Engineering, 127, 780–782.

RICKENM-Book.indb 115 2/23/2016 7:39:45 AM

 116 References

Frick, E., Kienholz, H., Roth, H. (2008): SEDEX – eine Methodik zur gut dokumentierten
Abschätzung der Feststofflieferung in Wildbächen. Wasser, Energie, Luft, 100, 131–136.
Frick, E., Kienholz, H., Romang, H. (2011): SEDEX (SEDiments and EXperts), Anwender-
Handbuch. Geographica Bernensia P42, Geographisches Institut der Universität Bern,
128p., ISBN 978-3-905835-27-4.
Fuchs, S., Kaitna, R., Scheidl, S., Hübl, J. (2008): The Application of the Risk Concept to
Debris Flow Hazards. Geomechanik und Tunnelbau, 1(2), 120–129.
Fuller, W., Thompson, S.E. (1907): The laws of proportioning concrete. Transactions of the
American Society of Civil Engineers, 59, 67–143.
Gallino, G.L., Pierson, T.C. (1985): Polallie Creek debris flow and subsequent dam-break
flood of 1980. East Fork Hood River Basin. Oregon. U.S. Geological Survey, Water-Supply
Paper, 2273, 22pp.
Gamma, P. (2000): dfwalk – Ein Murgangsimulationprogramm zur Gefahrenzonierung.
Geographica Bernensia, G66, Geographisches Institut der Universität Bern, 144p.
Garbrecht, G. (1961): Abflussberechnung für Flüsse & Kanäle. Die Wasserwirtschaft, 51, (2):
40–45, (4): 72–77.
Gartner, J.E., Cannon, S.H., Santi, P.M., deWolfe, V.G. (2008): Models to predict debris
flow volumes generated by recently burned basins. Geomorphology, 96, 339–354.
Gauckler, P.G. (1867): Etudes Théoriques et Pratiques sur l’Ecoulement et le Mouvement des
Eaux. Comptes Rendues de l’Academie des Sciences, Paris, France, Tome 64, pp. 818–822.
Geertsema, M., Schwab, J.W., Jordan, P., Millard, T.H., Rollerson, T.P. (2010): Hillslope
Processes. In R.G. Pike et al. (editors), Compendium of forest hydrology and geomorphology
in British Columbia. B.C. Min. For. and Range, For. Sci. Prog., Victoria, B.C. and FORREX
Forum for Research and Extension in Natural Resources, Kamloops, B.C. Land Management
Handbook 66, pp. 213–273.
Genevois, R., Tecca, P.R., Berti, M., Simoni, A. (2000): Debris-flow in the Dolomites:
Experimental data from a monitoring system. In: G.F. Wieczorek & N.D. Naeser (eds),
Debris-Flow Hazards Mitigation: Mechanics, Prediction, and Assessment; Proceedings
2nd International DFHM Conference, Taipei, Taiwan, August 16–18, 2000, pp. 283–291.
Rotterdam: Balkema.
Genolet, F. (2002): Modélisation de laves torrentielles – Contribution à la paramétrisation
du modèle Voellmy-Perla. Postgraduate thesis, Ecole Polytechnique Fédérale de Lausanne,
Suisse, 70p.
GEO (2000): Review of natural terrain landslide debris-resisting barrier design. Geotechnical
Engineering Office (GEO), Civil Engineering Department, The Government of the Hong
Kong Special Administrative Region, Special Project Report SPR 1/(2000).
George, D.L., Iverson, R.M. (2014): A depth-averaged debris-flow model that includes the
effects of evolving dilatancy: II. Numerical predictions and experimental tests. Proc. R. Soc.
Lond. Ser. A470, 20130820. http://dx.doi.org/10.1098/rspa.2013.0820.
Gertsch, E. (2009): Geschiebelieferung alpiner Wildbachsysteme bei Grossereignissen –
Ereignisanalysen und Entwicklung eines Abschätzverfahrens. Dissertation am Geographischen
Institut der Universität Bern, 203p. [http://www.zb.unibe.ch/download/eldiss/09 gertsch_e.pdf]
Gertsch, E., Kienholz, H., Spreafico, M. (2010): Projektbericht. Geschiebelieferung
alpiner Wildbachsysteme bei Grossereignissen – Ereignisanalysen und Entwicklung eines
Abschätzverfahrens. Hydrologie und Wasserbewirtschaftung, 54, 310–315.
Gomez, B., Church, M. (1989): An assessment of bedload sediment transport formulae for
gravel bed rivers. Water Resources Research, 25, 1161–1186.
Gostner, W., Bezzola, G.R., Schatzmann, M., Minor, H.E. (2003): Integral analysis of
debris flow in Alpine torrent – the case study of Tschengls. In: D. Rickenmann & C.L.
Chen (eds), Debris-Flow Hazards Mitigation: Mechanics, Prediction, and Assessment,
Proceedings of 3rd International DFHM Conference, Davos, Switzerland, September 10–12,
2003, pp. 1129–1140. Rotterdam: Millpress.

RICKENM-Book.indb 116 2/23/2016 7:39:45 AM

 References 117

Graf, W.H., Suszka, L. (1987): Sediment transport in steep channels. Journal of Hydroscience
and Hydraulic Engineering, 5, 11–26.
Graham, D.J., Rollet, A.J., Piegay, H., Rice, S.P. (2010): Maximizing the accuracy of
image-based surface sediment sampling techniques. Water Resources Research, 46, W02508.
doi:10.1029/2008WR006940.
Grasso, A., Dobmann, J., Jakob, A. (2007): Hydrological Atlas of Switzerland, Plate 7.8:
Bed-Material Loads in Selected Catchments.
Grasso, A., Jakob, A., Spreafico, M., Bérod, D. (2010): Monitoring von Feststofffrachten in
schweizerischen Wildbächen. Wasser, Energie, Luft, 102, 41–45.
Grant, G.E., Swanson, F.J., Wolman, M.G. (1990): Pattern and origin of stepped-bed
morphology in high gradient streams, Western Cascades, Oregon. Geological Society of
America Bulletin, 102, 340–352.
Greminger, P. (2003): Managing the risks of natural hazards. In D. Rickenmann & C.L.
Chen (eds), 3rd Int. Conf. on Debris-Flow Hazards Mitigation. Millpress, Rotterdam,
The Netherlands, pp. 39–56.
Griffiths, G.A. (2003): Downstream hydraulic geometry and hydraulic similitude. Water
Resources Research, 39(4), 1094. doi: 10.1029/2002 WR001485.
Günter, A. (1971): Die kritische mittlere Sohlenschubspannung bei Geschiebemischungen
unter Berücksichtigung der Deckschichtbildung und der turbulenzbedingten Sohlenschubs-
pannungsschwankungen. Mitteilungen der Versuchsanstalt für Wasserbau, Hydrologie und
Glaziologie, Eidgenössische Technische Hochschule Zürich, Nr. 3, 69p.
Gurnell, A.M. (2013): Wood in fluvial systems. In: Shroder, J., Wohl, E. (eds), Treatise on Geo-
morphology. Academic Press, San Diego, CA, vol. 9, Fluvial Geomorphology, pp. 163–188.
Guzzetti, F., Peruccacci, S., Rossi, M., Stark, C.P. (2007): Rainfall thresholds for the
initiation of landslides. Meteorology and Atmospheric Physics, 98, 239–267.
Guzzetti, F., Peruccacci, S., Rossi, M., Stark, C.P. (2008): The rainfall intensity–duration
control of shallow landslides and debris flows: an update. Landslides, 5, 3–17.
Haeberli, W. (1983): Frequency and characteristics of glacier floods in the Swiss Alps. Annals
of Glaciology, 4, 85–90.
Haeberli, W., Rickenmann, D., Zimmermann, M. (1991): Murgänge. Ursachenanalyse der
Unwetterereignisse (1987): Mitteilungen des Bundesamtes für Wasserwirtschaft, Bern,
Schweiz, Nr. 4, pp. 77–87.
Hager, W.H. (2001): Gauckler and the GMS formula. Journal of Hydraulic Engineering,
127(8), 635–638.
Hampel, R. (1980): Grundlagen für Gefahrenzonen in Wildbächen. Proc. International
Symposium Interpraevent, Bad Ischl, Austria, Bd. 3, pp. 83–91.
Han, G., Wang, D. (1996): Numerical modeling of Anhui debris flow. Journal of Hydraulic
Engineering, 122(5), 262–265.
Hartlieb, A., Bezzola, G.R. (2000): Ein Überblick zur Schwemmholzproblematik. Wasser,
Energie, Luft, 92, 1–5.
Hassan, M.A., Church, M., Lisle, T.E. (2005a): Sediment transport and channel morphology of
small, forested streams. Journal of the American Water Resources Association, 41, 853–876.
Hassan, M.A., Hogan, D.L, Bird, S.A, May, C.L., Gomi T., Campbel, D. (2005b): Spatial and
temporal dynamics of wood in headwater streams of the pacific northwest. Journal of the
American Water Resources Association, 41(4), 899–919.
Hayward, J.A. (1980): Hydrology and stream sediments in a mountain catchment, in Tussock
Grasslands and Mountain Lands Institute Special Publ. no. 17 (Ph. D. dissertation, Lincoln
College, Canterbury), New Zealand, 235p.
Hegg, C., Rickenmann, D. (1999): Comparison of bedload transport in a steep mountain
torrent with a bedload transport formula. In: Hydraulic Engineering for Sustainable Water
Resources Management at the Turn of the Millenium. Proceedings 28th IAHR Congress,
22–27 August (1999) in Graz, Austria. [CD-ROM] Graz, Technical University, 7p.

RICKENM-Book.indb 117 2/23/2016 7:39:45 AM

 118 References

Heimann, F.U.M., Rickenmann, D., Turowski, J.M., Kirchner, J.W. (2015a): sedFlow –
a tool for simulating fractional bedload transport and longitudinal profile evolution in
mountain streams. Earth Surface Dynamics, 3, 15–34.
Heimann, F.U.M., Rickenmann, D., Böckli, M., Badoux, A., Turowski, J.M., Kirchner,
J.W. (2015b): Calculation of bedload transport in Swiss mountain rivers using the model
sedFlow: proof of concept. Earth Surface Dynamics, 3, 35–54.
Heinimann, H.R., Hollenstein, K., Kienholz, H., Krummenacher, B., Mani, P. (1998):
Methoden zur Analyse und Bewertung von Naturgefahren. Umwelt-Materialien Nr. 85.
Bundesamt für Umwelt, Wald und Landschaft, Bern, 248p.
Hey, R.D. (1979): Flow Resistance in Gravel Bed Rivers. Journal of the Hydraulics Division,
105 (HY4), 365–379.
Hodel, H. (1993): Untersuchung zur Geomorphologie, der Rauheit des Strömungswiderstandes
und des Fliessvorganges in Bergbächen. Dissertation Nr. 9830, Eidgenössische Technische
Hochschule Zürich, 289p.
Hofmeister, R.J., Miller, D.J. (2003): GIS-based modeling of debris-flow initiation,
transport and deposition zones for regional hazard assessments in western Oregon, USA.
In: D. Rickenmann & C.L. Chen (eds), Debris-Flow Hazards Mitigation: Mechanics,
Prediction, and Assessment, Proceedings 3rd International DFHM Conference, Davos,
Switzerland, pp. 1141–1149. Rotterdam: Millpress.
Hübl, J. (2006): Vorläufige Erkenntnisse aus 1:1 Murenversuchen: Prozessverständnis
und Belastungsannahmen. In: FFIG, G. Reiser (Hrsg.), Geotechnik und Naturgefahren:
Balanceakt zwischen Kostendruck und Notwendigkeit. Institut für Geotechnik, BOKU
Universität Wien, Geotechnik und Naturgefahren, 19.10.2006, Wien.
Hübl, J., Kienholz, H., Loipersberger, A. (eds) (2002): DOMODIS-Documentation of
Mountain Disasters, State of Discussion in the European Mountain Areas. International
Research Society Interpraevent, Klagenfurt, Austria. http://wasser.ktn.gv.at/interpraevent.
Hungr, O. (1995): A model for the runout analysis of rapid flow slides, debris flows, and
avalanches. Canadian Geotechnical Journal, 32, 610–623.
Hungr, O. (2008): Numerical Modelling of the Dynamics of Debris Flows and Rock
Avalanches. Geomechanik und Tunnelbau, 1(2), 112–119.
Hungr, O., McDougall, S. (2009): Two numerical models for landslide dynamic analysis.
Computers & Geosciences, 35, 978–992.
Hungr, O., Morgan, G.C., Kellerhals, R. (1984): Quantitative analysis of debris torrent
hazards for design of remedial measures. Canadian Geotechechnical Journal, 21, 663–677.
Hungr, O., McDougall, S., Bovis, M. (2005): Entrainment of material by debris flows. In:
M. Jakob & O. Hungr (eds), Debris-Flow Hazards and Related Phenomena, pp. 135–158.
Heidelberg: Praxis-Springer.
Hungr O., Leroueil, S., Picarelli, L. (2014): The Varnes classification of landslide types, an
update. Landslides 11, 167–194.
Hungr, O., Evans, S.G., Bovis, M.J., Hutchinson, J.N. (2001): A review of the classification
of landslides of the flow type. Environmental & Engineering Geoscience, 7, 221–238.
Hunziker, R.P. (1995): Fraktionsweiser Geschiebetransport. Mitteilungen der Versuchsanstalt
für Wasserbau, Hydrologie und Glaziologie, Eidgenössische Technische Hochschule Zürich,
138, 191p.
Hunziker, R.P., Jaeggi, M.N.R. (2002): Grain sorting processes. Journal of Hydraulic
Engineering, 128, 1060–1068.
Hunzinger, L., Zarn, B. (1996): Geschiebetransport und Ablagerungsprozesse in
Wildbachschalen. Proc. International Symposium Interpraevent, Garmisch-Partenkirchen,
Germany, Bd. 4, pp. 221–230.
Hürlimann, M., Rickenmann, D., Graf, C. (2003): Field and monitoring data of debris-flow
events in the Swiss Alps. Canadian Geotechechnical Journal, 40, 161–175.

RICKENM-Book.indb 118 2/23/2016 7:39:45 AM

 References 119

Hürlimann, M., McArdell, B.W., Rickli, C. (2015): Field and laboratory analysis of the
runout characteristics of hillslope debris flows in Switzerland. Geomorphology, 232, 20–32.
Hürlimann, M., Rickenmann, D., Medina, V., Bateman, A. (2008): Evaluation of approaches to
calculate debris-flow parameters for hazard assessment. Engineering Geology, 102, 152–163.
Ikeya, H. (1979): Introduction to SABO works. The Japan Sabo Association, Tokyo, First
English Edition, 1979, 168pp.
Ikeya, H. (1981): A method of designation for area in danger of debris flow. In: Erosion and
Sediment Transport in Pacific Rim Steeplands, International Association of Hydrological
Sciences, Publ. no. 132, pp. 576–588.
Ikeya, H. (1987): Debris flow and its countermeasures in Japan. Bulletin International
Association of Engineering Geologists, 40, 15–33.
Imran, J., Parker, G., Locat, J., Lee, H. (2001): 1D Numerical model of muddy subaqueous
and subaerial debris flows. Journal of Hydraulic Engineering, 127(11), 959–967.
Innes, J.L. (2006): Lichenometric dating of debris-flow deposits in the Scottish Highlands.
Earth Surface Processes and Landforms, 8, 579–588.
Iverson, R.M. (2012): Elementary theory of bed-sediment entrainment by debris flows and
avalanches. Journal of Geophysical Research, 117, F03006. doi:10.1029/2011JF002189.
Iverson, R.M., Denlinger, R.P. (2001): Flow of variably fluidized granular masses across
three-dimensional terrain, 1. Coulomb mixture theory. Journal of Geophysical Research,
106(B1), 537–552.
Iverson, R.M., George, D.L. (2014): A depth-averaged debris-flow model that includes the
effects of evolving dilatancy: I. Physical basis. Proc. R. Soc. Lond. Ser. A470, 20130819.
http://dx.doi.org/10.1098/rspa.2013.0819.
Iverson, R.M., Schilling, S.P., Vallance, J.W. (1998): Objective delineation of lahar-
inundation zones. Geological Society of America Bulletin, 110, 972–984.
Iverson, R.M., George, D.L., Allstadt, K., Reid, M.E., Collins, B.D., Vallance, J.W.,
Schilling, S.P., Godt, J.W., Cannon, C.M., Magirl, C.S., Baum, R.L., Coe, J.A., Schulz,
W.H., Bower, J.B. (2015): Landslide mobility and hazards: implications of the 2014 Oso
disaster. Earth and Planetary Science Letters, 420, 197–208. doi: 10.1016/j.epsl.2014.12.020.
Jackson, W.L., Beschta, R.L. (1982): A model of two-phase bedload transport in an Oregon
Coast Range stream. Earth Surface Processes and Landforms, 7, 517–527.
Jackson, K.J., Wohl, E. (2015): Instream wood loads in montane forest streams of the
Colorado Front Range, USA. Geomorphology, 234, 161–170.
Jäggi, M.N.R. (1992): Sedimenthaushalt und Stabilität von Flussbauten. Mitteilungen der
Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie, Eidgenössische Technische
Hochschule Zürich, 119, 100p.
Jakob, M. (2005): Debris-flow hazard analysis. In Jakob, M., Hungr, O. (eds), Debris-flow
hazards and related phenomena, Praxis and Springer, Heidelberg, pp. 411–443.
Jakob, M. (2012): The fallacy of frequency – Statistical techniques for debris-flow frequency-
magnitude analyses. In: E. Eberhardt, C.A. Froese, K. Turner & S. Leroueil (eds), Landslides
and Engineered Slopes: Protecting society through improved understanding, pp. 741–750.
CRC Press/Balkema.
Jarrett, R.D. (1984): Hydraulics of high-gradient streams. Journal of Hydraulic Engineering,
110, 1519–1539.
Jin, M., Fread, D.L. (1999): 1D modeling of mud/debris unsteady flows. Journal of Hydraulic
Engineering, 125(8), 827–834.
Julien, P.Y., O’Brien, J.S. (1997): On the importance of mud and debris flow rheology in
structural design. In C.L. Chen (ed.), Debris-Flow Hazards Mitigation: Mechanics,
prediction, and Assessment, pp. 350–359. New York: ASCE.
Kaitna, R. (2006): Debris flow experiments in a rotating drum. Dissertation, Institut für Alpine
Naturgefahren, Universität für Bodenkultur Wien, 170p.

RICKENM-Book.indb 119 2/23/2016 7:39:45 AM

 120 References

Kaitna, R., Chiari, M., Kerschbaumer, M., Kapeller, H., Zlatic-Jugovic, J., Hengl, M.,
Huebl, J. (2011): Physical and numerical modelling of a bedload deposition area for an
Alpine torrent, Natural Hazards and Earth System Sciences, 11, 1589–1597.
Kasprak, A., Magilligan, F.J., Nislow, K.H., Snyder, N.P. (2012): A LIDAR derived
evaluation of watershed-scale large woody debris sources and recruitment mechanisms:
coastal Maine, USA. River Research and Applications, 28: 1462–1476.
Katul, G., Wiberg, P., Albertson, J., Hornberger, G. (2002): A mixing layer theory for flow
resistance in shallow streams. Water Resources Research, 38, 1250. doi:10.1029/(2001)
WR000817.
Kellerhals, R., Bray, D.I. (1971): Sampling Procedures for Coarse Fluvial Sediments.
Proceedings American Society of Civil Engineers, Journal of the Hydraulics Division,
97(HY8), 1165–1979.
Keulegan, G.H. (1938): Laws of turbulent flow in open channels. Journal of Research of the
National Bureau of Standards, Vol. 21, Research Paper 1151, 707–741.
Kienholz, H. (1999): Anmerkungen zur Beurteilung von Naturgefahren in den Alpen. In:
Relief, Boden, Paläoklima, Vol. 14, pp. 165–184. Berlin und Stuttgart: Gebr. Borntraeger.
Kienholz, H., Herzog, B., Bischoff, A., Willi, H.P. (2002): Fragen der Qualitätssicherung
bei der Gefahrenbeurteilung. Bündnerwald, 55, 57–67.
Kienholz, H., Frick, E., Gertsch, E. (2010): Assessment tools for mountain torrents:
SEDEX© and bed load assessment matrix. Proc. International Interpraevent Symposium,
Taipei, Taiwan, pp. 245–256.
Kronfellner-Krauss, G. (1984): Extreme Feststofffrachten und Grabenbildungen von
Wildbächen. Proc. International Symposium Interpraevent, Villach, Austria, Bd. 2, 109–118.
Kronfellner-Krauss, G. (1987): Zur Anwendung der Schätzformel für extreme Wildbach-
Feststofffrachten im Süden und Osten Oesterreichs. Wildbach- und Lawinenverbau, 51,
187–200.
Laigle, D., Coussot, P. (1997): Numerical modelling of debris flows. Journal of Hydraulic
Engineering 123, 617–623.
Lamb, M.P., Dietrich, W.E., Venditti, J.G. (2008): Is the critical Shields stress for incipient
sediment motion dependent on channel-bed slope? Journal of Geophysical Research – Earth
Surface, 113, F0(2008). doi:10.1029/(2007)JF000831.
Lange, D., Bezzola G.R. (2006): Schwemmholz, Probleme und Lösungsansätze. Mitteilungen
der Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie, Eidgenössische Technische
Hochschule Zürich, Nr. 188, 125p.
Lassettre, N.S., Kondolf, G.M. (2012): Large woody debris in urban stream channels:
redefining the problem. River Research and Application, 28, 1477–1487.
Legros, F. (2002): The mobility of long-runout landslides. Engineering Geology, 63, 301–331.
Lehmann, C. (1993): Zur Abschätzung der Feststofffracht in Wildbächen – Grundlagen und
Anleitung. Geographica Bernensia G42, Bern.
Lenzi, M., Mao, L., Comiti, F. (2004): Magnitude-frequency analysis of bed load
data in an Alpine boulder bed stream. Water Resources Research, 40, W07201.
doi:10.1029/2003WR002961.
Liener, S. (2000): Zur Feststofflieferung in Wildbächen. Geographica Bernensia G64, Bern.
Lucía, A., Comiti, F., Borga, M., Cavalli, M., Marchi, L. (2015): Dynamics of large wood
during a flash flood in two mountain catchments. Natural Hazards and Earth System
Sciences, 15, 1741–1755.
Luzian, R., Kohl, B., Bichler, I., Kohl, J., Bauer, W. (2002): Wildbäche und Muren – Eine
Wildbachkunde mit einer Übersicht von Schutzmassnahmen der Ära Aulitzky. Forstliche
Bundesversuchsanstalt, Wien, ISBN 3-901347-34-8, 163p.
Mächler, D. (2009): GIS-Modellierung von potentiellen Schwemmholzeinträgen durch
Rutschungen. Semesterarbeit, Umweltingenieurwesen, Zürcher Hochschule für angewandte
Wissenschaften (ZHAW), Wädenswil, 22 p.

RICKENM-Book.indb 120 2/23/2016 7:39:45 AM

 References 121

Malet, J.P., Maquaire, O., Locat, J., Remaître, A. (2004): Assessing debris flow hazards
associated with slow moving landslides: methodology and numerical analyses. Landslides,
1, 83–90.
Manning, R. (1891): On the flow of water in open channels and pipes. Transactions of the
Institution of Civil Engineers of Ireland, 20, 161–207.
Marchi, L., Tecca, P.R. (1996): Magnitudo delle colate detritiche nelle Alpi Orientali Italiane.
Geoingegneria Ambientale e Mineraria, 33(2/3), 79–86 (in Italian).
Marchi, L., Brochot, S. (2000): Les cônes de déjection torrentielles dans les Alpes françaises;
morphométrie et processus de transport solide torrentiel. Revue de géographie alpine, 88,
23–38.
Marchi, L., Arattano, M. & Deganutti, A.M. (2002): Ten years of debris flows monitoring
in the Moscardo Torrent (Italian Alps): Geomorphology, 46(1/2), 1–17.
Marchi, L., D’Agostino, V. (2004): Estimation of debris-flow magnitude in the Eastern Italian
Alps. Earth Surf. Process. Landforms, 29(2), 207–220.
Mathys, N., Brochot, S., Meunier, M., Richard, D. (2003): Erosion quantification in
the small marly experimental catchments of Draix (Alpes de Haute Provence, France):
Calibration of the ETC rainfall-runoff-erosion model. Catena, 50, 527–548.
Mazzorana, B., Zischg, A., Largiader, A., Hübl, J. (2009): Hazard index maps for woody
material recruitment and transport in alpine catchments. Natural Hazards and Earth System
Science, 9, 197–209.
Mazzorana, B., Comiti, F., Volcan, C., Scherer, C. (2011): Determining flood hazard
patterns through a combined stochastic–deterministic approach. Natural Hazards, 59(1),
301–316.
McArdell, B.W., Zanuttigh, B., Lamberti, A., Rickenmann, D. (2003): Systematic
comparison of debris flow laws at the Illgraben torrent, Switzerland. In D. Rickenmann &
C.L. Chen (eds), Debris-Flow Hazards Mitigation: Mechanics, Prediction, and Assessment;
Proceedings 3rd International DFHM Conference, Davos, Switzerland, September 10–12,
2003, pp. 647–657. Rotterdam: Millpress.
McArdell, B.W., Bartelt, P., Kowalski, J. (2007): Field observations of basal forces and fluid
pore pressure in a debris flow. Geophysical Research Letters, 34, L07406. doi:10.1029/
(2006)GL029183.
McCoy, S.W., Kean, J.W., Coe, J.A., Tucker, G.E., Staley, D.M., Wasklewicz, T.A. (2012):
Sediment entrainment by debris flows: In situ measurements from the headwaters of a steep
catchment. Journal of Geophysical Research, 117, F03016. doi:10.1029/2011JF002278.
McDougall, S., Hungr, O. (2004): A model for the analysis of rapid landslide motion across
three-dimensional terrain. Canadian Geotechnical Journal, 41, 1084–1097.
McDougall, S., Hungr, O. (2005): Dynamic modelling of entrainment in rapid landslides.
Canadian Geotechnical Journal, 42, 1437–1448.
Medina, V., Hürlimann, M., Bateman, A. (2008): Application of FLAT Model, a 2D finite
volume code, to debris flows in the northeastern part of the Iberian Peninsula. Landslides,
5, 127–142.
Meyer-Peter, E., Müller, R. (1948): Formulas for bedload transport. In: Proceedings 2nd
meeting Int. Assoc. Hydraulic Structures Res., Stockholm, Sweden, Appendix 2, pp. 39–64.
Meyer-Peter, E., Müller, R. (1949): Eine Formel zur Berechnung des Geschiebetriebs.
Schweizerische Bauzeitung, 67, 29–32.
Mizuyama, T. (1981): An intermediate phenomenon between debris flow and bed load
transport. In: Erosion and Sediment Transport in Pacific Rim Steeplands, International
Association of Hydrological Sciences, Publ. no. 132, pp. 212–224.
Mizuyama, T., Kobashi, S., Ou, G. (1992): Prediction of debris flow peak discharge. Proc.
International Symposium Interpraevent, Bern, Switzerland, Bd. 4, pp. 99–108.
Montgomery, D.R., Dietrich, W.E. (1994): A physically based model for the topographic
control on shallow landsliding. Water Resources Research, 30(4), 1153–1171.

RICKENM-Book.indb 121 2/23/2016 7:39:45 AM

 122 References

Montgomery, D.R., Buffington, J.M. (1997): Channel reach morphology in mountain

drainage basins. Geological Society of America Bulletin, 109, 596–611.
Morvan, H., Knight, D., Wright, N., Tang, X., Crossley, A. (2008): The concept of
roughness in fluvial hydraulics and its formulation in 1D, 2D and 3D numerical simulation
models, Journal of Hydraulic Research, 46, 191–208.
Näf, D., Rickenmann, D., Rutschmann, P., McArdell, B.W. (2006): Comparison of flow
resistance relations for debris flows using a one-dimensional finite element simulation
model. Natural Hazards and Earth System Sciences, 6, 155–165.
Nakagawa, H., Takahashi, T., Satofuka, Y. (2000): A debris-flow disaster on the fan of the
Harihara River, Japan. In G.F. Wieczorek & N.D. Naeser (eds), Debris-Flow Hazards
Mitigation: Mechanics, Prediction, and Assessment; Proceedings 2nd International DFHM
Conference, Taipei, Taiwan, August 16–18, 2000, pp. 193–201. Rotterdam: Balkema.
Nakamura, J. (1980): Investigation manual on prediction of occurrence of dosekiryu [debris
flows], delineation of dangerous zone affected by dosekiryu and arrangement of warning
and evacuation system in mountain torrents in Japan. Proc. International Symposium
Interpraevent, Bad Ischl, Austria, Bd. 3, pp. 41–81.
Nameghi, A.E., Hassanli, A.M., Soufi, M. (2008): A study of the influential factors affecting
the slopes of deposited sediments behind the porous check dams and model development for
prediction. DESERT, 12, 113–119. (http://jdesert.ut.ac.ir)
Nitsche, M., Turowski, J.M., Badoux, A., Pauli, M., Schneider, J., Rickenmann, D.,
Kohoutek, T.K. (2010): Measuring streambed morphology using range imaging. In:
A. Dittrich, K. Koll, J. Aberle & P. Geisenhainer (eds), River Flow 2010, Bundesanstalt für
Wasserbau, pp. 1715–1722. ISBN 978-3-939230-00-7.
Nitsche, M., Rickenmann, D., Turowski, J.M., Badoux, A., Kirchner, J.W. (2011):
Evaluation of bedload transport predictions using flow resistance equations to account
for macro-roughness in steep mountain streams. Water Resources Research, 47, W08513.
doi:10.1029/2011WR010645.
Nitsche, M., Rickenmann, D., Kirchner, J.W., Turowski, J.M., Badoux, A. (2012a):
Macro-roughness and variations in reach-averaged flow resistance in steep mountain
streams, Water Resources Research, 48, W12518. doi: 10.1029/2012WR012091.
Nitsche, M., Rickenmann, D., Turowski, J.M., Badoux, A., Kirchner, J.W. (2012b):
Verbesserung von Geschiebevorhersagen in Wildbächen und Gebirgsflüssen durch
Berücksichtigung von Makrorauigkeit. Wasser, Energie, Luft, 104, 129–139.
Nnadi, F.N., Wilson, K.C. (1992): Motion of contact-load particles at high shear stress.
Journal of Hydraulic Engineering, 118, 1670–1684.
Nolde, N., Joe, H. (2013): A Bayesian extreme value analysis of debris flows. Water Resources
Research, 49, 7009–7022.
O’Brien, J.S., Julien, P.Y., Fullerton, W.T. (1993): Two-dimensional water flood and
mudflow simulation. Journal of Hydraulic Engineering, 119, 244–261.
Pagliara, S., Chiavaccini, P. (2006): Flow Resistance of Rock Chutes with Protruding
Boulders. Journal of Hydraulic Engineering, 132, 545–552.
Palt, S.M. (2001): Sedimenttransporte im Himalaya-Karakorum und ihre Bedeutung für
Wasserkraftanlagen. Mitteilungen des Institutes für Wasserwirtschaft und Kulturtechnik der
Universität Karlsruhe, Heft 209, 257p.
Palt, S.M., Dittrich, A. (2002): Stabilität von Gebirgsflüssen und rauen Rampen.
Österreichische Wasser- und Abfallwirtschaft, 54, 75–86.
Parker, G., Wilcock, P.R., Paola, C., Dietrich, W.E., Pitlick, J. (2007): Physical basis for
quasi-universal relations describing bankfull hydraulic geometry of single-thread gravel bed
rivers. Journal of Geophysical Research, 112, F04005. doi:10.1029/2006JF000549.
Parker, G. (2008): Transport of gravel and sediment mixtures. In: M.H. Garcia (ed.),
Sedimentation engineering: Theories, measurements, modeling, and practice, pp. 165–252.
Reston, USA: ASCE.

RICKENM-Book.indb 122 2/23/2016 7:39:46 AM

 References 123

Petrascheck, A., Kienholz, H. (2003): Hazard assessment and mapping of mountain risks in
Switzerland. In D. Rickenmann & C.L. Chen (eds), 3rd Int. Conf. on Debris-Flow Hazards
Mitigation. Millpress, Rotterdam, The Netherlands, pp. 25–38.
Phillips, C.J., Davies, T.R.H. (1991): Determining rheological parameters of debris flow
material. Geomorphology, 4, 101–110.
Pierson, T.C. (1986): Flow behavior of channelized debris flows, Mount St. Helens,
Washington. In: A.D. Abrahams (ed.), Hillslope Processes, pp. 269–296. Boston, USA: Allen
and Unwin.
Pierson, T.C. (1995): Flow characteristics of large eruption-triggered debris flows at snow-clad
volcanoes: constraints for debris-flow models. Journal of Volcanology and Geothermal
Research, 66, 283–294.
Pierson, T. (2005): Distinguishing between debris flows and floods from field evidence in small
watersheds. USGS Fact Sheet 2004–3142, January 2005.
Pierson, T.C., Costa, J.E. (1987): A rheologic classification of subarerial sediment-water
flows. Geological Society of America, Reviews in Engineering Geology, Vol. VII, 1–12.
Pirulli, M. (2010): On the use of the calibration-based approach for debris-flow
forward-analyses. Natural Hazards and Earth System Sciences, 10, 1009–1019.
Pirulli, M., Sorbino, G. (2008): Assessing potential debris flow runout: a comparison of two
simulation models. Natural Hazards and Earth System Sciences, 8, 961–971.
Piton, G., Recking, A. (2015a): Design of sediment traps with open check dams. I: Hydraulic
and deposition processes. Journal of Hydraulic Engineering. doi:10.1061/(ASCE)HY.1943–
7900.0001048, 04015045.
Piton, G., Recking, A. (2015b): Design of Sediment Traps with Open Check Dams. II: Woody
Debris. Journal of Hydraulic Engineering. doi: 10.1061/(ASCE)HY.1943–7900.0001049.
Planat (2000): Empfehlungen zur Qualitätssicherung bei der Beurteilung von Naturgefahren.
Nationale Plattform Naturgefahren (PLANAT), Bern, 20p.
Planat (2008): Wirkung von Schutzmassnahmen. Nationale Plattform Naturgefahren
(PLANAT), Bern, Strategie Naturgefahren Schweiz, Umsetzung des Aktionsplans PLANAT
2005–2008, Projekt A3, Schlussbericht 2. Phase, Testversion, Dezember 2008. 289p.
Porto, P., Gessler, J. (1999): Ultimate bed slope in Calabrian streams upstream of check
dams: field study. Journal of Hydraulic Engineering, 125, 1231–1242.
Prochaska, A.B., Santi, P.M., Higgins, J., Cannon, S.H. (2008): Debris-flow runout
predictions based on the average channel slope (ACS). Engineering Geology, 98, 29–40.
Pwri (1988): Technical standard for measures against debris flow (Draft). Technical
Memorandum of Public Works Research Institute (PWRI), No. 2632, Ministry of Con-
struction, Japan, 48p.
Raetzo, H., Rickli, C. (2007): Rutschungen. In: G.R. Bezzola & C. Hegg (eds), Ereignisanalyse
Hochwasser (2005), Teil 1 – Prozesse, Schäden und erste Einordnung. Bundesamt für
Umwelt BafU, Bern und Eidg. Forschungsanstalt für Wald, Schnee und Landschaft WSL,
Birmensdorf, pp. 195–209.
Recking, A. (2009): Theoretical development on the effects of changing flow hydraulics on incipi-
ent bed load motion. Water Resources Research, 45, W04401. doi:10.1029/(2008)WR006826.
Recking, A. (2013): An analysis of nonlinearity effects on bed load transport prediction.
Journal of Geophysical Research – Earth Surface, 118, 1264–1281. doi:10.1002/jgrf.20090.
Reid, L.M., Dunne, T. (1996): Rapid evaluation of sediment budgets. Geo-Ecology Texts,
Catena Verlag, Reiskirchen, Germany, 164p.
Revellino, P., Hungr, O., Guadagno, F.M., Evans, S.G. (2004): Velocity and runout
simulation of destructive debris flows and debris avalanches in pyroclastic deposits,
Campania region, Italy. Environmental Geology, 45, 295–311.
Rickenmann, D. (1990): Bedload transport capacity of slurry flows at steep slopes. Mitteilungen
der Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie, Eidgenössische Technische
Hochschule Zürich, 103, 249p.

RICKENM-Book.indb 123 2/23/2016 7:39:46 AM

 124 References

Rickenmann, D. (1991): Hyperconcentrated flow and sediment transport at steep slopes.

Journal of Hydraulic Engineering, 117(11), 1419–1439.
Rickenmann, D. (1994): An alternative equation for the mean velocity in gravel-bed rivers and
mountain torrents. In G.V. Cotroneo & R.R. Rumer (eds), Proceedings ASCE 1994 National
Conference on Hydraulic Engineering, Buffalo N.Y., USA, Vol. 1, pp. 672–676.
Rickenmann, D. (1995): Beurteilung von Murgängen. Schweizer Ingenieur und Architekt, 48,
1104–1108.
Rickenmann, D. (1996): Fliessgeschwindigkeit in Wildbächen und Gebirgsflüssen. Wasser,
Energie, Luft, 88, 298–304.
Rickenmann, D. (1996): Murgänge: Prozess, Modellierung und Gefahrenbeurteilung.
In B. Oddsson (ed.), Instabile Hänge und andere risikorelevante natürliche Prozesse,
Nachdiplomkurs in angewandten Erdwissenschaften, pp. 397–407. Basel: Birkhäuser.
Rickenmann, D. (1997a): Sediment transport in Swiss torrents. Earth Surface Processes and
Landforms, 22, 937–951.
Rickenmann, D. (1997b): Schwemmholz und Hochwasser. Wasser, Energie, Luft, 89, 115–119.
Rickenmann, D. (1999): Empirical relationships for debris flows. Natural Hazards, 19, 47–77.
Rickenmann, D. (2001a): Comparison of bed load transport in torrents and gravel bed streams.
Water Resources Research, 37, 3295–3305.
Rickenmann, D. (2001b): Murgänge in den Alpen und Methoden zur Gefahrenbeurteilung.
In: Mitteilungen des Lehrstuhls und Instituts für Wasserbau und Wasserwirtschaft,
Rheinisch-Westfälische Technische Hochschule Aachen, Nr. 124, pp. 51–77.
Rickenmann, D. (2005a): Geschiebetransport bei steilen Gefällen. In: Mitteilungen der
Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie, Eidgenössische Technische
Hochschule Zürich, Nr. 190, pp. 107–119.
Rickenmann, D. (2005b): Runout prediction methods. In: M. Jakob & O. Hungr (eds),
Debris-Flow Hazards and Related Phenomena, pp. 263–282. Heidelberg: Praxis-Springer.
Rickenmann, D. (2008): Lastfälle aus Murgangprozessen – Bemessungsgrundlagen. Herbstkurs
der Fachleute für Naturgefahren (FAN), Bellinzona, Schweiz, 17.9.2008.
Rickenmann, D. (2012): Alluvial steep channels: flow resistance, bedload transport and
transition to debris flows. In: M. Church, P. Biron & A. Roy (eds), Gravel Bed Rivers:
Processes, Tools, Environment, pp. 386–397. Chichester, England: John Wiley & Sons.
Rickenmann, D. (2016): Debris-flow hazard assessment and methods applied in engineering
practice. International Journal of Erosion Control Engineering (Japan), in press (http://
www.jsece.or.jp/jece/index.html).
Rickenmann, D., Zimmermann, M. (1993): The 1987 debris flows in Switzerland:
documentation and analysis. Geomorphology, 8, 175–189.
Rickenmann, D., Koch, T. (1997): Comparison of debris flow modelling approaches. In
C.L. Chen (ed.), Debris-Flow Hazards Mitigation: Mechanics, Prediction, and Assessment,
Proceedings 1st International DFHM Conference, San Francisco, CA, USA, August 7–9,
1997, pp. 576–585. New York: ASCE.
Rickenmann, D., Weber, D. (2000): Flow resistance of natural and experimental debris
flows in torrent channels. In G.F. Wieczorek & N.D. Naeser (eds), Debris-Flow Hazards
Mitigation: Mechanics, Prediction, and Assessment, Proceedings 2nd International DFHM
Conference, Taipei, Taiwan, August 16–18, 2000, pp. 245–254. Rotterdam: Balkema.
Rickenmann, D., McArdell, B.W. (2007): Continuous measurement of sediment transport
in the Erlenbach stream using piezoelectric bedload impact sensors. Earth Surface Processes
and Landforms, 32, 1362–1378.
Rickenmann, D., Scheidl, C. (2010): Modelle zur Abschätzung des Ablagerungsverhaltens
von Murgängen. Wasser, Energie, Luft, 102, 17–26.
Rickenmann, D., Koschni, A. (2010): Sediment loads due to fluvial transport and debris flows
during the 2005 flood events in Switzerland. Hydrological Processes, 24, 993–1007.

RICKENM-Book.indb 124 2/23/2016 7:39:46 AM

 References 125

Rickenmann, D., Recking, A. (2011): Evaluation of flow resistance equations using a large
field data base. Water Resources Research, 47, W07538. doi:10.1029/2010WR009793.
Rickenmann, D., Jakob, M. (2015): Erosion and sediment flux in mountain watersheds.
In: The High-Mountain Cryosphere, C. Huggel, M. Carey, J.J. Clague & A. Kääb (eds),
Cambridge University Press, pp. 166–183.
Rickenmann, D., Hürlimann, M., Graf, C., Näf, D., Weber, D. (2001): Murgang-
Beobachtungsstationen in der Schweiz. Wasser, Energie, Luft, 93, 1–8.
Rickenmann, D., Weber, D., Stepanov, B. (2003): Erosion by debris flows in field and
laboratory experiments. In: D. Rickenmann & C.L. Chen (eds), Debris-Flow Hazards
Mitigation: Mechanics, Prediction, and Assessment, Proceedings of 3rd International
DFHM Conference, Davos, Switzerland, September 10–12, 2003, pp. 883–894. Rotterdam:
Millpress.
Rickenmann, Chiari, M., Friedl, K. (2006a): SETRAC – A sediment routing model for steep
torrent channels. In R. Ferreira, E. Alves, J. Leal & A. Cardoso (eds), River Flow 2006,
pp. 843–852. London: Taylor & Francis.
Rickenmann, D., Laigle, D., Mcardell, B.W., Hübl, J. (2006b): Comparison of 2D
debris-flow simulation models with field events. Computational Geosciences, 10, 241–264.
Rickenmann, D., Heimann, F.U.M., Böckli, M. Turowski, J.M., Bieler, C., Badoux, A.
(2014): Geschiebetransport-Simulationen mit sedFlow in zwei Gebirgsflüssen der Schweiz.
Wasser Energie Luft (eingereicht).
Rickenmann, D., Böckli, M., Heimann, F.U.M., Badoux, A., Turowski, J.M. (2015): Das
Modell sedFlow und Erfahrungen aus Simulationen des Geschiebetransportes in fünf
Gebirgsflüssen der Schweiz. Synthesebericht. WSL Berichte, Heft 24, 68p. (www.wsl.ch/
publikationen/pdf/14594.pdf).
Rickli, C., Bucher, H.U. (2006): Einfluss ufernaher Bestockungen auf das
Schwemmholzvorkommen in Wildbächen. Projektbericht zuhanden des Bundesamtes für
Umwelt BAFU. Eidgenössische Forschungsanstalt WSL, Birmensdorf, Schweiz, 94p.
Rickli, C., Raetzo, H., McArdell, B., Presler, J. (2008): Hanginstabilitäten. In: G.R.
Bezzola & C. Hegg (eds), Ereignisanalyse Hochwasser 2005: Teil 2 – Analyse von Prozessen,
Massnahmen und Gefahrengrundlagen. Bundesamt für Umwelt BAFU, Bern, Eidgenössische
Forschungsanstalt WSL, Birmensdorf, pp. 97–116.
Rimböck, A. (2003): Schwemmolzrückhalt in Wildbächen. Grundlagen zu Planung und
Berechnung von Seilnetzsperren. Berichte des Lehrstuhls und der Versuchsanstalt für
Wasserbau und Wasserwirtschaft, Nr. 94, Technische Universität München, 163p.
Rinderer, M., Jenewein, S., Senfter, S., Rickenmann, D., Schöberl, F., Stötter, J., Hegg, C.
(2009): Runoff and bedload transport modelling for flood hazard assessment in small alpine
catchments – the model PROMAB-GIS. In E. Veulliet, J. Stötter & H. Weck-Hannemann
(eds), Sustainability in Natural Hazard Management, pp. 69–101. Berlin: Springer-Verlag.
Romang, H. (2004): Wirksamkeit und Kosten von Wildbach-Schutzmassnahmen. Geographica
Bernensia, G 73, Universität Bern, ISBN 3-906151-76-X, 211pp.
Rosport, M. (1998): Fliesswiderstand und Sohlstabilität steiler Fliessgewässer unter
Berücksichtigung gebirgsbachtypischer Sohlstrukturen. Mitteilungen des Institutes für
Wasserwirtschaft und Kulturtechnik der Universität Karlsruhe, Heft 196, 144p.
Rössert, R. (1978): Hydraulik im Wasserbau. 4. Auflage, Oldenbourg, München.
Rudolf-Miklau, F. (2001): Untersuchungen an kohäsionslosen Sedimenten in kalkalpinen
Wildbächen der Steiermark (Österreich). Dissertation, Universität für Bodenkultur Wien.
Rudolf-Miklau F., Hübl J., Schattauer G., Rauch H. P., Kogelnig A., Habersack H.,
Schulev-Steindl E. (2011): Handbuch Wildholz – Praxisleitfaden. Internationale
Forschungsgesellschaft Interpraevent, Klagenfurt, 32p.
Ruf, G. (1990): Fliessgeschwindigkeiten in der Ruetz/Stubaital/Tirol. Wildbach- und
Lawinenverbau, 54(115), 219–227.

RICKENM-Book.indb 125 2/23/2016 7:39:46 AM

 126 References

Ruiz-Villanueva, V., Bladé, E., Díez-Herrero, A., Bodoque, J.M., Sánchez-Juni, M.

(2014): Two dimensional modelling of large wood transport during flash floods. Earth
surface processes and landforms, 39, 438–449.
Ryan, S.E., Porth, L.S., Troendle, C.A. (2005): Coarse sediment transport in mountain
streams in Colorado and Wyoming, USA. Earth Surface Processes and Landforms, 30,
269–288.
Santi, P.M., deWolfe, V.A., Higgins, J.D., Cannon, S.H., Gartner, J.E., Santi, P.M.
(2008): Sources of debris flow material in burned area. Geomorphology, 96, 310–321.
Schälchli, U. (1991): Morphologie und Strömungsverhältnisse in Gebirgsbächen: Ein
Verfahren zur Festlegung von Restwasserabflüssen. Mitteilungen der Versuchsanstalt für
Wasserbau, Hydrologie und Glaziologie, Eidgenössische Technische Hochschule Zürich,
Nr. 113, 112p.
Scheidegger, J. (1970): Theoretical Geomorphology. 2. Auflage, Springer, Berlin.
Scheidl, C., Rickenmann, D. (2010): Empirical prediction of debris-flow mobility and
deposition on fans. Earth Surface Processes and Landforms, 35, 157–173.
Scheidl, C., Rickenmann, D. (2011): TopFlowDF – A simple GIS based model to simulate
debris-flow runout on the fan. In: R. Genevois, D.L. Hamilton & A. Prestininzi, Debris-Flow
Hazards Mitigation: Mechanics, Prediction, and Assessment; Proceedings 5th International
DFHM Conference, Padova, Italy, June 14–17, 2011, pp. 253–262. Roma: Casa Editrice
Università La Sapienza. doi:10.4408/IJEGE.2011-03.B−030.
Scheidl, C., Chiari, M., Kaitna, R., Müllegger, M., Krawtschuk, A., Zimmermann, T.,
Proske, D. (2013): Analysing debris-flow impact models, based on a small scale modelling
approach. Survey of Geophysics, 34, 121–140.
Scheuner T., Keusen, H.R., Mcardell, B.W., Huggel, C. (2009): Murgangmodellierung
mit dynamisch-physikalischem und GIS-basiertem Fliessmodell. Fallbeispiel Rotlauigraben,
Guttannen, August 2005. Wasser, Energie, Luft, 101, 15–21.
Schilling, M., Hunziker, R. (1995): Programmpaket MORMO – Grundlagen. In:
Mathematische Modelle offener Gerinne, ÖWAV-Seminar Konstruktiver Wasserbau –
Landschaftswasserbau, 21. Nov. 1995, Bd. 17, pp. 91–104.
Schmocker, L., Hager, W.H. (2011): Probability of drift blockage at bridge decks. Journal of
Hydraulic Engineering, 137, 470–479.
Schneider, J.M., Rickenmann, D., Turowski, J.M., Bunte, K., Kirchner, J.W. (2014):
Applicability of bedload transport models for mixed-size sediments in steep streams
considering macro-roughness. Water Resources Research, 51, 5260–5283.
Schneuwly-Bollschweiler, M., Corona, C., Stoffel, M. (2013): How to improve dating
quality and reduce noise in tree-ring based debris-flow reconstructions. Quaternary
Geochronology, 18, 110–118.
Schoklitsch, A. (1914): Über Schleppkraft und Geschiebebewegung. Engelmann, Leipzig und
Berlin.
Schreiner, A. (1997): Einführung in die Quartärgeologie. 2. Auflage, E. Schweizerbart´sche
Verlagsbuchhandlung, Stuttgart, 257p.
Schröder, R.C.M. (1994): Technische Hydraulik – Kompendium für den Wasserbau. Springer,
Berlin.
Schürch, P., Densmore, A.L., Rosser, N.J., McArdell, B.W. (2011): Dynamic controls on
erosion and deposition on debris-flow fans. Geology, 39, 827–830.
Shields, A. (1936): Anwendung der Ähnlichkeitsmechanik und der Turbulenzforschung auf
die Geschiebebewegung. Mitteilungen der Preussischen Versuchsanstalt für Wasserbau und
Schiffsbau, Berlin, Heft 26.
Singh, V.P. (2003): On the theories of hydraulic geometry. International Journal of Sediment
Research, 18(3), 196–218.
Slaymaker, O. (1988): The distinctive attributes of debris torrents. Hydrological Sciences
Journal. 33(6), 567–573.

RICKENM-Book.indb 126 2/23/2016 7:39:46 AM

 References 127

Smart, G.M., Jäggi, M.N.R. (1983): Sedimenttransport in steilen Gerinnen. Mitteilungen der
Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie, Eidgenössische Technische
Hochschule Zürich, 64, 188p.
Smart, G.M., Duncan, M.J., Walsh, J.M. (2002): Relatively rough flow resistance equations,
Journal of Hydraulic Engineering, 128, 568–578.
Spreafico, M., Lehmann, C., Naef, O. (1996): Empfehlung zur Abschätzung von
Feststofffrachten in Wildbächen. Teil I: Handbuch, 46p. + Anhang; Teil II: Fachliche
Grundlagen, 113p. Groupe de travail pour l’hydrologie operationelle (GHO), Mitteilung
Nr. 4, Landeshydrologie und –geologie, Bern.
Spreafico, M., Lehmann, Ch., Jakob, A., Grasso, A. (2005): Feststoffbeobachtung in der
Schweiz. Berichte des Bundesamtes für Wasser und Geologie, Serie Wasser, Nr. 8, Bern.
Stähli, M., Sättele, M., Huggel, C., McArdell, B.W., Lehmann, P., Van Herwijnen, A.,
Berne, A., Schleiss, M., Ferrari, A., Kos, A., Or, D., Springman, S.M. (2015): Monitor-
ing and prediction in early warning systems for rapid mass movements. Nat. Hazards Earth
System Sciences, 15, 4, 905–917.
Stiny, J. (1931): Die geologischen Grundlagen der Verbauung der Geschiebeherde in Gewässern.
Springer, Wien.
Strickler, A. (1923): Beiträge zur Frage der Geschwindigkeitsformel und der Rauhigkeitszahlen
für Ströme, Kanäle und geschlossene Leitungen. Sekretariat des Eidg. Amtes für
Wasserwirtschaft, Mitteilungen des Amtes für Wasserwirtschaft, Bern, Nr. 16.
Swartz, M., McArdell, B., Bartelt, P. (2003): Interpretation of the August 2000 Schipfenbach
debris flow event using numerical models. In: Mitteilungen der Versuchsanstalt für
Wasserbau, Hydrologie und Glaziologie, Eidgenössische Technische Hochschule Zürich,
Nr. 184, pp. 51–60.
Takahashi, T. (1981): Estimation of potential debris flows and their hazardous zones; soft
countermeasures for a disaster. Journal of Natural Disaster Science, 3(1), 57–89.
Takahashi, T. (1987): High velocity flow in steep erodible channels. Proc. XXII IAHR
Congress, Lausanne, Switzerland, Technical Session A, pp. 42–53.
Takahashi, T. (1991): Debris Flow. IAHR Monograph Series, Balkema Publishers,
the Netherlands.
Takei, A. (1984): Interdependence of sediment budget between individual torrents and a
river-system. Proc. International Symposium Interpraevent, Villach, Austria, Bd. 2, 35–48.
Tecca, P., Genevois, R., Deganutti, A., Armento, M. (2007): Numerical modelling of two
debris flows in the Dolomites (Northeastern Italian Alps). In C.L. Chen & J.J. Major (eds),
4th Int. Conf. on Debris-Flow Hazards Mitigation. Millpress, Rotterdam, The Netherlands,
pp. 179–188.
Theler, D., Reynard, E., Lambiel, C., Bardou, E. (2010): The contribution of
geomorphological mapping to sediment transfer evaluation in small alpine catchments.
Geomorphology, 124, 113–123.
Thompson. S.M., Campell, P.L. (1979): Hydraulics of a large channel paved with boulders.
Journal of Hydraulic Research, 17, 341–354.
Tognacca, C., Bezzola, G.R., Minor, H.-E. (2000): Threshold criterion for debris-flow
initiation due to channel-bed failure. In G.F. Wieczorek & N.D. Naeser (eds), Debris-Flow
Hazards Mitigation: Mechanics, Prediction, and Assessment, Proceedings 2nd International
DFHM Conference, Taipei, Taiwan, August 16–18, 2000, pp. 89–97. Rotterdam: Balkema.
Toyos, G., Gunasekera, R., Zanchetta, G., Oppenheimer, C., Sulpizio, R., Favalli, M.,
Pareschi, M.T. (2008): GIS-assisted modelling for debris flow hazard assessment based
on the events of May 1998 in the area of Sarno, Southern Italy: II. Velocity and dynamic
pressure. Earth Surface Processes and Landforms, 33, 1693–1708.
Turowski, J.M., Yager, E.M., Badoux, A., Rickenmann, D., Molnar, P. (2009): The impact
of exceptional events on erosion, bedload transport and channel stability in a step-pool
channel. Earth Surface Processes and Landforms, 34, 1661–1673.

RICKENM-Book.indb 127 2/23/2016 7:39:46 AM

 128 References

VanDine, D.F. (1985): Debris flows and debris torrents in the Southern Canadian Cordillera.
Canadian Geotechnical Journal, 22, 44–68.
VanDine, D.F. (1996): Debris flow control structures for forest engineering. Province of British
Columbia, Ministry of Forests Research Program, Working Paper 22/(1996), 75p.
VAW (1992): Murgänge 1987: Dokumentation und Analyse. Unveröffentlichter Bericht,
No. 97.6, Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie (VAW),
Eidgenössische Technische Hochschule Zürich.
Vetsch, D., Fäh, R., Fahrsi, D., Müller, R. (2005): BASEMENT – Ein objektorientiertes
Softwaresystem zur numerischen Simulation von Naturgefahren. In: Mitteilungen der
Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie, Eidgenössische Technische
Hochschule Zürich, Nr. 190, p. 201–212.
von Ruette, J., Lehmann, P., Or, D. (2013): Rainfall-triggered shallow landslides at catchment
scale: Threshold mechanics-based modeling for abruptness and localization. Water
Resources Research, 49, 6266–6285.
Waldner, P., Rickli, C., Köchli, D., Usbeck, T., Schmocker, L., Sutter, F. (2007):
Schwemmholz. In: G.R. Bezzola & C. Hegg (eds), Ereignisanalyse Hochwasser 2005,
Teil 1 – Prozesse, Schäden und erste Einordnung. Bundesamt für Umwelt BAFU, Bern,
Eidgenössische Forschungsanstalt für Wald, Schnee und Landschaft WSL, Birmensdorf,
pp. 181–193.
Waldner, P., Schmocker, L., Sutter, F., Rickenmann, D., Rickli, C., Lange, D., Köchli, D.
(2008): Schwemmholzbilanzen. In: G.R. Bezzola & C. Hegg (eds), Ereignisanalyse
Hochwasser 2005: Teil 2 – Analyse von Prozessen, Massnahmen und Gefahrengrundlagen.
Bundesamt für Umwelt BAFU, Bern, Eidgenössische Forschungsanstalt WSL, Birmensdorf,
pp. 136–143.
Ward, T.J. (1986): Discussion of “Sediment transport formula for steep channels” by G.M.
Smart, Journal of Hydraulic Engineering, 112, 989–990.
Warrick, J.A., Rubin, D.M., Ruggiero, P., Harney, J., Draut, A.E., Buscombe, D. (2009):
Cobble cam: grain-size measurements of sand to boulder from digital photographs and
autocorrelation analyses. Earth Surface Processes and Landforms, 34, 1811–1821.
Webb, B.W., Reid, I., Bathurst, J.C., Carling, P.A., Walling, D.E. (1997): Sediment erosion,
transport and deposition. In: C.R. Thorne, R.D. Hey & M.D. Newson (eds), Applied
Fluvial Geomorphology for River Engineering and Management, pp. 95–135. Chichester:
John Wiley.
Weichert, R., Bezzola, G.R. (2002): Einfluss von Makrorauigkeiten auf die Stabilität alpiner
Gewässer. Wasser, Energie, Luft, 94, 259–264.
Whittaker, J.G., Jäggi, M. (1986): Blockschwellen. Mitteilungen der Versuchanstalt für
Wasserbau, Hydrologie und Glaziologie, Eidgenössische Technische Hochschule Zürich,
Nr. 91, 187p.
Whittaker, J.G., Hickman, W.E., Croad, R.N. (1988): Riverbed Stabilisation with placed
blocks. Report 3-88/3, Central Laboratories, Works and Development Services Corporation,
Lower Hutt, NZ.
Wichmann, V. Heckmann, T., Haas, F., Becht, M. (2009): A new modelling approach to
delineate the spatial extent of alpine sediment cascades. Geomorphology, 111, 70–78.
Wicks, J.M., Bathurst, J.C. (1996): SHESED: a physically based, distributed erosion and
sediment yield component for the SHE hydrological modelling system. Journal of Hydrology,
175, 213–238.
Wilcock, P.R., Crowe, J.C. (2003): Surface-based transport model for mixed-size sediment.
Journal of Hydraulic Engineering, 129, 120–128.
Wilford, D.J., Sakals, M.E., Innes, J.L., Sidle, R.C., Bergerud, W.A. (2004): Recognition of
debris flow, debris flood and flood hazard through watershed morphometrics. Landslides,
1, 61–66.

RICKENM-Book.indb 128 2/23/2016 7:39:46 AM

 References 129

Wohl, E.E. (2000): Mountain Rivers. Water Resources Monograph, American Geophysical
Union, Washington DC, USA, 320p.
Wohl, E., Beckman, ND. (2014): Controls on the longitudinal distribution of channel-spanning
logjams in the Colorado Front Range, USA. River Research and Applications, 30, 112–131.
Wohl, E., Cenderelli, D.A., Dwire, K.A., Ryan-Burkett, S.E., Young, M.K., Fausch, K.D.
(2010): Large in-stream wood studies: a call for common metrics. Earth Surf. Process.
Landf., 35, 618–625.
Wong, M., Parker, G. (2006): Reanalysis and correction of bed-load relation of Meyer-Peter
and Müller using their own database. Journal of Hydraulic Engineering, 132, 1159–1168.
Yager, E.M., Kirchner, J.W., Dietrich, W.E. (2007): Calculating bed load transport in steep
boulder bed channels. Water Resources Research, 43, W07418. doi:10.1029/2006WR005432.
Zanuttigh, B., Lamberti, A. (2006): Experimental analysis of the impact of dry avalanches
on structures and implication for debris flows. Journal of Hydraulic Research, 44, 522–534.
Zeller, J. (1963): Einführung in den Sedimenttransport offener Gerinne. Mitteilungen der
Versuchsanstalt für Wasserbau und Erdbau, Eidgenössische Technische Hochschule Zürich,
Nr. 62, [= Sonderdruck aus Schweizerische Bauzeitung, 81. Jg., Hefte 34, 35, 36].
Zeller, J. (1985): Feststoffmessung in kleinen Gebirgseinzugsgebieten. Wasser, Energie, Luft,
77, 246–251.
Zeller, J. (1996): Der Kstr-Koeffizient in der Geschwindigkeitsgleichung von Strickler und
dessen Problematik. Proc. International Symposium Interpraevent, Garmisch-Partenkirchen,
Germany, Bd. 4, pp. 63–74.
Zeller, J., Trümpler, J. (1984): Rutschungsentwässerungen – Hinweise zur Bemessung
steiler Entwässerungsgräben. Eidgenössische Anstalt für das forstliche Versuchswesen,
Birmensdorf, 276p.
Zimmermann, A. (2010): Flow resistance in steep streams: an experimental study. Water
Resources Research, 46, W09536. doi:10.1029/2009WR007913.
Zimmermann, M., Rickenmann, D. (1992): Beurteilung von Murgängen in der Schweiz:
Meteorologische Ursachen und charakteristische Parameter zum Ablauf. Proc. International
Symposium Interpraevent, Bern, Switzerland, Bd. 2, 153–163.
Zimmermann, M., Mani, P., Gamma, P., Gsteiger, P., Heiniger, O., Hunziker, G. (1997a):
Murganggefahr und Klimaänderung – ein GIS-basierter Ansatz. Schlussbericht NFP 31,
Verlag der Fachvereine, Eidgenössische Technische Hochschule Zürich, Schweiz, 161p.
Zimmermann, M., Mani, P., Romang, H. (1997b): Magnitude-frequency aspects of Alpine
debris flows. Eclogae Geologicae Helvetiae, 90, 415–420.
Zimmermann, M., Lehmann, C. (1999): Geschiebefracht in Wildbächen: Grundlagen und
Schätzverfahren. Wasser, Energie, Luft, 91, 189–194.
Zollinger, F. (1983): Die Vorgänge in einem Geschiebeablagerungsplatz – Ihre Morphologie
und die Möglichkeiten einer Steuerung. Dissertation Nr. 7419, Eidgenössische Technische
Hochschule Zürich.

RICKENM-Book.indb 129 2/23/2016 7:39:46 AM

This page intentionally left blank
 List of symbols

a coefficient in logarithmic flow resistance law (often a = 12)

ag coefficient in equation for critical dimensionless unit discharge qc*
ak correction factor to account for the slope parallel component of the weight of
sediment particles
a1 coefficient in VPE-equation (mean value a1 = 6.5)
a2 coefficient in VPE-equation (mean value a2 = 2.5)
A flow cross-sectional area
AC surface area of a hydrologic catchment
AV factor (numerator) in equation for the runout distance of a debris flow on
the fan
B width of flow cross-section (at an opening)
d stem diameter of large wood pieces
dWmax maximum dimension of the root part (large wood pieces)
dWmin minimum dimension of the root part (large wood pieces)
dW* = (dWmax dWmin Lh)1/3 = mean dimension of the root part (large wood pieces)
Dx characteristic grain size for which x% oft the sediment mixture are finer
Dmi mean grain size of the grain size class i
Dmax maximum grain size
DR duration of a rainfall event
e exponent in equation for the reduced energy slope
f = 8 (v*/V)2 = friction coefficient according to Darcy-Weisbach
fo Darcy-Weisbach coefficient for the grain or base-level resistance
fadd Darcy-Weisbach coefficient for the macro-roughness resistance
ftot Darcy-Weisbach coefficient for the total resistance
Fr = V/(gh)0.5 = Froude number
g gravitational acceleration
G factor (denominator) in equation for the runout distance of a debris flow on
the fan
GF sediment volume (including pore volume of sediment deposits)
h flow depth
hu flow depth in the approach channel (of a debris flow) upstream of the fan apex
H clear height of an opening of a flow cross-section
He elevation difference between starting point and most distal deposition point of
a debris flow

RICKENM-Book.indb 131 2/23/2016 7:39:46 AM

 132 List of symbols

I mean rainfall intensity

K Torrentiality factor (after Kronfellner-Kraus)
ks equivalente roughness heigth (“sand roughness”)
kSt Strickler coefficient (for total flow resistance)
L total runout distance of a debris flow
Lc length of active channel (regarding erosion during a torrential event)
Lf length of debris-flow deposits on a fan
Lmax maximum total runout distance of a debris flow
Lh length of trunk extension of large wood pieces
LW length of log pieces (large woody debris)
M sediment volume of a debris flow (typically estimated from all deposits of an
event; includes also pore volume; may refer to single debris-flow surges in
case of observations from an automatic monitoring system)
Me Melton number
no Manning coefficient for the grain or base-level resistance
ntot Manning coefficient for the total resistance
pd dynamic impact pressure due to debris flow
pFui cumulative frequency (as a fraction) of the Fuller distribution for grains with
D ≤ Di
pi cumulative frequency (as a fraction) of the grain size distribution for grains
with D ≤ Di
pv clogging probability due to large woody debris
Δpi relative proportion of a grain-size class i of all sediments
P wetted perimeter of a flow cross-section
q unit (water) discharge in a channel (per meter channel width)
qb bedload transport rate per meter channel width
qc critical discharge at initiation of bedload motion (per meter channel width)
qc,D critical discharge at breaking up of an armor layer (per meter channel width)
qc,B critical discharge at destabilization of a block ramp (per meter channel width)
qc* critical dimensionless discharge (initiation of debris flow, bedload transport)
q** = q/(gSD843)0.5 = dimensionless discharge (per meter channel width)
Q = qW = total (water) discharge in channel (over entire channel width)
QB = qbW = total bedload transport rate (over entire channel width)
Qc = qcW = critical discharge at initiation of bedload motion
Qc,D = qc,DW = critical discharge at breaking up of an armor layer
Qp maximum discharge (of a debris-flow surge)
QS sediment supply from upstream and/or from tributaries
R = A/P = hydraulic radius
s = ρs/ρ = ratio of sediment density to water density
S channel slope (or friction slope) (in all equations of this document S has to be
used with the units [m/m] und not with the units [%])
Sc mean channel slope upstream of the fan apex
Sf mean channel slope on the fan (or mean fan slope)
Sk = S ak = corrected channel slope to account for the slope parallel component of
the weight of sediment particles
Sred reduced energy slope (or reduced channel slope) to account for macro-
roughness effects in the calculation of bedload transport

RICKENM-Book.indb 132 2/23/2016 7:39:46 AM

 List of symbols 133

SR friction slope of a debris flow on the fan

Te maximum erosion depth in a torrent channel
U** = V/(gSD84)0.5 = dimensionless mean flow velocity
v* = (ghS)0.5 = shear velocity
V mean flow velocity (water or debris flow)
Vo virtual mean flow velocity, related to base-level resistance
Vtot (effectiv) mean flow velocity, related to total resistance
Vu flow velocity in the approach channel (of a debris flow) upstream of the
fan apex
W width of the channel
yR height of the lowest roughness layer (in open channel flow)

GREEK SYMBOLS

αd coefficient for the calculation of the dynamic impact pressure due to debris flow
αg exponent in equation for the critical dimensionless discharge qc*
αο prefactor in a bedload transport equation
β angle of the channel (or angle of a depositional reach)
βd angle of impact for the calculation of the dynamic impact pressure due to
debris flow
βu angle of the channel upstream of the fan apex (debris-flow runout calculation)
γ exponent of the “hiding function”
γs shear rate (change of flow velocity/change of flow depth)
εo roughness height
κ von Karman constant (= 0.4)
μ dynamic viscosity
ρ density of water
ρM density of debris-flow mixture (including solids and water)
ρs density of solids (sediment particles)
θ = hS/[(s − 1)D] = dimensionless bed shear stress
θ’ = reduced dimensionless bed shear stress, accounting for energy losses due to
form or macro-roughness
θc = critical dimensionless bed shear stress at initiation of bedload motion
τ = ρghS = bed shear stress
τΒ shear strength, yield stress (Bingham fluid model)
ϕs natural angle of repose of submerged sediment
Φb = qb/[(s − 1)gD3]0.5 = dimensionless bedload transport rate (per meter channel
width)

RICKENM-Book.indb 133 2/23/2016 7:39:46 AM

This page intentionally left blank
 IAHR MONOGRAPH
Methods for the Quantitative Assessment of Channel Processes in Torrents (Steep Streams) INTERNATIONAL ASSOCIATION FOR HYDRO-ENVIRONMENT ENGINEERING AND RESEARCH

Dieter Rickenmann IAHR MONOGRAPH

An important part of the risk management of natural hazards in mountain regions concerns the
hazard assessment and the planning of protection measures in steep headwater catchments, i.e.
torrent control and slope stabilization. Torrent processes in steep channels have their rightful

Methods for the Quantitative Assessment

in Torrents (Steep Streams)

Methods for the Quantitative Assessment of Channel Processes
place among the various alpine natural hazards and the corresponding control measures
have a long tradition in the European alpine countries. In the planning and execution of such
measures, professional experience has been of paramount importance. This experience was
based primarily on observations made during and after hazardous events, as well as on regular of Channel Processes in Torrents
field visits in the catchments of a steep headwater stream. Quantitative measurements, e.g. of
the discharge and of the eroded and deposited solid materials, have been increasingly carried (Steep Streams)
out only in the last decades. This set the basis to develop and improve quantitative methods to
predict flow hydraulics, bedload transport and debris flows in torrent catchments.

This publication presents an overview of methods to quantify channel processes in steep

catchments. The understanding and the quantitative description of channel processes provides
an essential basis for the planning of protection measures.

Rickenmann

Dieter Rickenmann

an informa business