Methods For The Quantitative Assessment of Channel Processes in
Methods For The Quantitative Assessment of Channel Processes in
Methods for the Quantitative Assessment of Channel Processes in Torrents (Steep Streams) INTERNATIONAL ASSOCIATION FOR HYDRO-ENVIRONMENT ENGINEERING AND RESEARCH
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
Rickenmann
Dieter Rickenmann
an informa business
Methods for the Quantitative
Assessment of Channel Processes
in Torrents (Steep Streams)
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
Dieter Rickenmann
Research Unit Mountain Hydrology and Torrents,
Swiss Federal Research Institute WSL, Birmensdorf,
Switzerland
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
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
References 111
List of symbols 131
Subject index 135
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.
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.
Introduction
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
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.
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.
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
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.
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%
Table 2.2 Thickness of the lower roughness layer after BEZZOLA (2002).
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 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).
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]
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
(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%
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)
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
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
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
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).
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.
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.
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)
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
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):
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
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 ⎟⎠
(fo/ftot)0.5
1
0.9
additional flow
resistance(fadd)
resistance(
0.8
0.7
0.6
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:
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).
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:
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 &
Process debris fl
fflows
ows
fluvial
f uvial sediment transport
fl
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.
Sediment delivery QS
Transport capacity QB
step-pool pool-riffle
QB QS
Catchment area
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
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).
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
high sphericity
low sphericity
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]
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.
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).
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.
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
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.
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.
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.
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.
100
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).
• 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
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).
τ ρ ghS (3.5)
τ hS hS
θ= = = (3.6)
g (ρ ρ) D (ρ ρ− )D ( − )D
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
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.
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:
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
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).
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).
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
Bathurst (2013)
1 Knauss (1979)
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.
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.
(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
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)
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.
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)
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.
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.
10
perfect agreement
0.01
0.01 0.1 1 10 100
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).
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:
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
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).
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.
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.
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).
• 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.
0.1
0.05
0
0 0.05 0.1 0.15 0.2
(a) Original bed slope So
0.3
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).
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).
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.
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).
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).
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
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.
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
Figure 4.1 Front of a granular debris flow, Kamikamihori valley, Japan (photo H. SUWA).
flow direction
large wood
onset of turbulence
coarse particles in suspension
bouldery front
tail head
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).
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.
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
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).
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
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.
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
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).
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:
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).
In a somewhat simplified way, the two key aspects can be grouped as follows:
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).
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).
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).
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)
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.
Figure 4.9 Significant bedload delivery processes in torrents and the role played in the formation of
debris flows. Modified from RICKENMANN (1996).
q*
c
= qc/[g0.5 D1.5] = ag/Sαg (4.1)
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.
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.
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
Debris-flow volume M
Maximum discharge QP
Flow velocity V
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.
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).
is rather implausible, as shown in the analysis of the debris-flow surges of the 1987
events in Val Varuna (VAW 1992).
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).
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).
To estimate an upper limit of the runout distance Lmax the following relation may
be used:
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).
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).
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).
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
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.
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.
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-
pseudoplastic with
yield stress, n < 1
Shear stress τ
Bingham
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
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).
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
in an event, the number of surges, and possible depositions on the channel bed due to
smaller surges.
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.
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).
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).
precipitation
hillslope
sediment delivery to
channel network
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).
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
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.
100000
cristalline
limestone
flysch
10000 conglomerate
1000
100
10
1
0.1 1 10 100 1000
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).
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-
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):
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.
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.
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.
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).
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
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
XXL
Steinlouibach (Lungern, OW)
XL
Size of event
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
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).
XXL
Buochser Rübi (Buochs, NW)
XL
Size of event
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
XXL
Dorfbach Randa
XL
Size of event
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).
XXL
Minstigerbach
XL
Size of event
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).
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
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).
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:
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:
• 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):
(Continued )
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).
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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)
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
Rickenmann
Dieter Rickenmann
an informa business