See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/222406041

What controls relay ramps and transfer faults

within rift zones? Insights from analog models

Article in Journal of Structural Geology · March 2005

Impact Factor: 2.88 · DOI: 10.1016/j.jsg.2004.11.006

CITATIONS READS

25 96

3 authors, including:

Valerio Acocella
Università Degli Studi Roma Tre
142 PUBLICATIONS 3,142 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Valerio Acocella
letting you access and read them immediately. Retrieved on: 18 May 2016
 Journal of Structural Geology 27 (2005) 397–408
www.elsevier.com/locate/jsg

What controls relay ramps and transfer faults within rift zones? Insights
from analogue models
V. Acocella*, P. Morvillo, R. Funiciello
Dip. Scienze Geologiche Università Roma Tre. Largo S.L. Murialdo, 1, 00146, Rome, Italy
Received 7 May 2004; accepted 8 November 2004
Available online 25 January 2005

Abstract
Structures within rift zones exhibit two main types of interaction relevant at the rift scale: relay ramps and transfer faults at high angle to
the rift. Analogue experiments have been performed to investigate whether these types of interaction may be affected by differential
extension along the rift. In these models, sand (brittle crust analogue) overlies two adjacent silicone (ductile crust analogue) layers with
different viscosity, in order to simulate different percentage extension rates (De) along rifts. The experiments show a distinct behaviour as a
function of De. For De!21G3%, extensional structures interact forming relay ramps; for DeO21G3%, the interaction occurs by means of
transfer faults striking subparallel to the extension direction. Experimental data are consistent with the geometries and extension rates of rift
zones. Relay ramps characterize narrow rifts and oceanic ridges, where the mean percentage of extension is low (e!16%). Conversely,
transfer faults are usually found in extensional settings (passive margins, wide rifts, back-arc basins) with significant stretching (eO39%),
where the rift more likely achieves differential extension DeO21%.
q 2005 Elsevier Ltd. All rights reserved.

Keywords: Rift zones; Relay ramps; Transfer faults; Analogue models; Differential extension

1. Introduction et al., 2000a, and references therein). This type of

interaction constitutes a common type of ‘soft linkage’
Rift zones are commonly segmented at various scales (Walsh and Watterson, 1991). Conversely, the interaction
Segments are, however, transient features in the evolution of may develop a transfer fault, which is a subvertical
a rift zone, as they grow and interact and may link to form transtensive fault that strikes at high angle and transfers
larger structures (Macdonald and Fox, 1983; Pollard and displacement between two adjacent crustal sectors under-
Aydin, 1988; Dawers and Anders, 1995; Koukouvelas et al., going differential extension (Fig. 1; Gibbs, 1990; Peacock et
1999). Therefore, the process of interaction is a necessary al., 2000a, and references therein). This type of interaction
step in the evolution of a rift zone over a range of scales. constitutes a common type of ‘hard linkage’ (Walsh and
Two main typologies of interaction between segments Watterson, 1991).
(or groups of segments) that appear to be significant at the Both types of interaction are found in rift zones. Relay
scale of the considered rift (controlling its shape or ramps are widespread; they have been described for
continuity) can be identified. The interaction may develop example in the Rhine Graben (Illies, 1975; Brun et al.,
relay ramps or accommodation zones, which consist of 1991), the Rio Grande Rift (Cordell, 1978; Mack and
broad areas of ductile strain between extensional structures Seager, 1995), the Baikal Rift (Sherman, 1978; Hutchinson
(grabens, normal faults, extensional fractures) usually et al., 1992), the East African Rift System (EARS) (Morley,
characterized by arcuate geometries (Fig. 1; e.g. Peacock 1988; Ebinger, 1989a,b; Ebinger et al., 1989; Morley et al.,
1990; Nelson et al., 1992), East Greenland (Larsen, 1988;
* Corresponding author. Tel.: C39-06-54888043; fax: C39-06-
Peacock et al., 2000b), the British Isles (Peacock and
54888201 Sanderson, 1991; Huggins et al., 1995; Peacock, 2003), the
E-mail address: [email protected] (V. Acocella). Aegean Sea (Gawthorpe and Hurst, 1993), the Suez Rift
0191-8141/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jsg.2004.11.006
398 V. Acocella et al. / Journal of Structural Geology 27 (2005) 397–408

Fig. 1. Types of interaction between extensional structures at various rift zones: relay ramps in the East African Rift System (a) (after Ebinger, 1989b) and the
Rio Grande graben (b) (after Aldrich, 1986); transfer faults in the Suez Rift (c) (after McClay and Khalil, 1998) and the Atlantic margin of Brazil (d) (after
Milani and Davison, 1988); (e) schematic view of relay ramps and transfer faults in extensional domains.

(Moustafa, 1996), the Basin and Range (Anders and Werff, 2000) back-arc margins. In most of the cases, the
Schlische, 1994; Ferrill et al., 1999), the oceanic ridge of transfer faults appear to be relevant at the rift scale (km to
Iceland (Acocella et al., 2000) and the Gulf of Thailand 102 km).
(Kornsawan and Morley, 2002). In most of these cases, the Therefore, while relay ramps are widespread in various
relay ramps appear to control the shape or continuity of the extensional settings, transfer faults are usually observed
rift and therefore they are considered significant at the rift within passive margins, wide rifts and back-arc basins, and
scale (km to 103 km). Also, the geometric and kinematic are essentially lacking within continental narrow rifts
features of relay ramps can be consistent over scales from m (EARS, Rio Grande Rift, Rhine Graben and Baikal Rift).
to 105 m (Acocella et al., 2000). Also, within readily accessible oceanic extensional
Transfer faults were first recognized and described in the domains, such as Iceland, transfer faults are lacking and
Viking Graben, North Sea (Gibbs, 1984). Since then, the dominant type of interaction are relay ramps (Acocella
various authors have reported the occurrence of transfer et al., 2000).
faults at various extensional settings, such as the Atlantic To investigate whether the selective occurrence of relay
margin of Brazil (Milani and Davison, 1988), the Atlantic ramps and transfer zones is controlled by the differential
margin of Newark (Schlische, 1992), the Basin and Range extension along the rift, we have performed analogue
(Duenbendorfer and Black, 1992; Martin et al., 1993), the experiments using sand (brittle crust analogue) and silicone
Atlantic margin of Galicia (Boillot et al., 1995), the Atlantic (ductile crust analogue). The experiments simulate crustal
margin of Namibia (Clemson et al., 1997), the Atlantic blocks undergoing differential extension and show that
margin of Norway (Dorè et al., 1997; Tsikalas et al., 2001), transfer faults, relevant at the rift scale, can form only with a
the Suez Rift (McClay and Khalil, 1998), the Atlantic percentage of differential extension DeO21G3%. Below
margin of W Africa (Watts and Stewart, 1998) and the this threshold, relay ramps, relevant at the rift scale, occur.
Tyrrhenian (Acocella et al., 1999a) and Japanese (van der This value roughly corresponds in nature to the threshold of
 V. Acocella et al. / Journal of Structural Geology 27 (2005) 397–408 399

Table 1
Model and nature ratios applicable to the present study

Parameter Model value Nature value Model/nature ratio

4
Length 0.01 m 10 m L*w10K6
Density 1.2–1.4 kg/m3 2.4–2.8 kg/m3 r*w0.5
Gravity 9.8 m/s2 9.8 m/s2 g*w1
Stress (sZrgL) s*w5!10K7
Viscosity (mZs/3) 104–105 Pa s 1020–1021 Pa s m*w10K16
Strain rate 5!10K6 sK1 10K15 sK1 3*w5!109
Time (tZ1/3) t*w2!10K10

extension between narrow continental rifts and oceanic Silicone putty with Newtonian behaviour has been used
ridges, where extension is lower and relay ramps are at the base of the sand-pack to simulate the plastically
dominant, and wide rifts, passive margins and back-arc deforming crust. In order to reproduce different amounts of
basins, where transfer faults are often more numerous. extension across the model, we use two adjacent layers of
silicone, with different viscosities of 7.9!104 Pa s and
4.5!105 Pa s (Fig. 2b); the silicone has a density
w1310 kg/m3.
2. Experimental procedure
The following relation applies to Newtonian ductile
materials (Benes and Davy, 1996):
2.1. Scaling and materials
s1* K s3* Z m * 3 * (1)
Analogue experiments were constructed to simulate the
interaction between adjacent crustal blocks experiencing where m* and 3* are the viscosity and the strain rate ratios
different extension rates. Scaled models should be geome- between model and nature, respectively. Where
trically, kinematically and dynamically similar to natural s1* K s3* w5 !10K7 , the m* and 3* ratios have to be scaled
examples (Ramberg, 1981, and references therein). In the accordingly (Table 1). Considering the viscosities of
experiments, the length ratio L between model and nature is silicone and the mean viscosity of the lower crust (1020–
L*Z10K6 (1 cm in our experiments corresponds to 10 km 1021 Pa s; Ranalli, 1995), m*w10K16 and, as a result (from
in nature), the density ratio between rocks and common Eq. (1)), 3*w5!109. For a mean extensional strain rate
experimental materials is r*w0.5 and the gravity ratio 3nw10K15 sK1, commonly found in nature, to have
between model and nature is g*Z1. The corresponding 3*w5!109, requires a strain rate 3mw5!10K6 sK1 in
stress ratio between model and nature is s*Zr*g*z*w5! the experiments. As 3*Z1/t*, 1 s in our experiments
10K7 (Table 1). Cohesion c has the dimensions of stress; corresponds to 2!1010 s (w634 years) in nature (Table 1).
assuming a Mohr–Coulomb criterion and natural cohesion
cw107 Pa, a material with cw5 Pa is required to simulate 2.2. Set-up
the brittle crust: for this purpose, dry quartz sand, with
cw0 Pa, is used; the dry sand has a density w1400 kg/m3. The experimental set-up is characterized by a 1–2-cm-
thick sand layer, a 1–2-cm-thick silicone layer and a basal
plate (Fig. 2). In all the experiments, the silicone layer
consists of two adjacent silicone portions with different
viscosity (Fig. 2a). The silicone layer is confined on three
sides and has a free boundary along the fourth side.
Extension within the model is obtained by means of a
moderate tilt (4–68) of the basal plate towards the free
boundary (Fig. 2c). The tilt angle has been chosen in order
to maintain a consistence between the mean extension rates
of the model and nature (Table 1). Such a tilt develops in
fact a tangential (parallel to the dip of the tilted plate)
component of the gravity force. The gravity force at the base
of an experiment characterized by 1.5 cm of sand and
1.5 cm of silicone is:

s Z ðrgzÞsand C ðrgzÞsilicone Z 398:37 N

Fig. 2. Sketch of the experimental apparatus. (a) Map view; (b) frontal For a basal tilt of 48, the tangential stress at the base of the
section view; (c) lateral section view. experiment is:
400 V. Acocella et al. / Journal of Structural Geology 27 (2005) 397–408

Table 2
Main imposed and observed features in the experiments. TbZsand thickness; TdZsilicone thickness; LZlength of model; WZwidth of model; m1Zviscosity
of silicone in one plate; m2Zviscosity of silicone in the other plate; aZangle of tilting of the basal rigid plate; DeZpercentage of differential extension defining
the experimental threshold between relay ramps and transfer faults

Experiment Tb (cm) Td (cm) L (cm) W (cm) m1 (Pa s) m2 (Pa s) a (8) De

5 4
TIR 2 1 2 18 28 4.5!10 7.9!10 4 21
TIR 3 1 2 20 40 4.5!105 7.9!104 6 18
TIR 4 1.5 1.5 20 40 4.5!105 2.7!105 4 O10
TIR 5 1.5 1.5 25 40 4.5!105 7.9!104 4 23
TIR 6 1 2 20 50 4.5!105 7.9!104 4 20
TIR 7 1.5 2 20 40 4.5!105 7.9!104 4 24

st Z 398:37cos48 Z 27:8 N the experiment. As the sand and silicone layers thin, st
decreases: st at the end of the experiment can be estimated
This value represents the tangential component of the at 50% of the initial st. This process is responsible for a
gravity force responsible for the flow of silicone along the decrease in the stretching rate with time, resulting in
slope at the beginning of the experiment (Fig. 2c). significant deformation of the model at the earlier stages and
The flow of silicone induces the thinning and extension moderate deformation at later stages.
of the overlying sand pack. As the silicone has different Variable extension rates are also found in natural
viscosities, it undergoes different flow velocities and extensional settings and, to a first approximation, assuming
therefore stretching. Since the amount of extension within a variable extension rate should not limit the applicability of
the sand is proportional to the amount of stretching of the the experiments. Despite the variable extension rate, the
underlying silicone, this set-up simulates the interaction of total duration (w5 h) and extension (up to w70%) of the
adjacent blocks of upper crust with differential extension. experiments correspond to a realistic duration of rifting
At the end of each experiment, sand was added and (w1.1!107 years) and amount of stretching in nature.
levelled onto the surface of the model, to preserve The experiments do not take into account any control of
topography. The model was then saturated with water and pre-existing brittle structures on the development of the
cut. Six experiments were performed in order to test: the interactions, even though these may play an important role.
role of the ratio between the thickness of the brittle and The purpose of the experiments is in fact the study of the
ductile materials in the deformation pattern; the silicone type of interaction due to differential extension in the
viscosity contrast; the slope angle; the lateral dimensions of simplest conditions.
the models (Table 2). Finally, the types of interaction on which this work is
This apparatus permits us to examine the interaction focused are those characterizing the continuity and shape of
between structures undergoing differential extension with- the rift zones; therefore, only those interacting structures
out the control of a rigid basal velocity discontinuity, as that appear relevant at the rift scale, both in nature and the
commonly used in previous experiments of interacting experiments, are considered.
extensional structures (Courtillot et al., 1974; Elmohandes,
1981; Serra and Nelson, 1988; Naylor et al., 1994; Mauduit
and Dauteuil, 1996; Acocella et al., 1999b). The lack of a
3. Experimental results
velocity discontinuity has the important advantage of
avoiding the related modifications of the deformation
The evolution of the experiments is here summarized by
pattern at surface.
model TIR 7 (TbZ1.5 cm; TdZ2 cm, where Tb and Td are
the sand and the silicone thickness, respectively), with
2.3. Limits and assumptions of the experiments silicone viscosities of 7.9!104 (left plate in Fig. 3a) and
4.5!105 Pa s (right plate in Fig. 3a). At tZ0 0 (minutes) the
These experiments study the effect of differential experiment is undeformed (Fig. 3a).
extension within a brittle crust; to achieve this, two silicone The tilt of the rigid base induces the flowing of silicone
putties with different viscosities are used. The use of and extension in the two plates. At tZ60 0 several
silicone with different viscosity does not necessarily depressions form. The plate with more viscous silicone
correspond to the simulation of different types of ductile has four regularly spaced depressions bordered by normal
crust and therefore does not have a specific counterpart in faults, whereas the plate with less viscous silicone has six
nature. Silicone viscosity is thus varied to simulate a regularly spaced wider depressions bordered by normal
differential extension, not to simulate different types of faults (Fig. 3b). The lateral termination of these graben-like
ductile crust. features along the contact between the adjacent plates is
The tangential component of the gravity force st marked by relay ramps or accommodation zones. These
responsible for the flow of silicone is not constant during consist of a broad deformed area, delimited by normal
 V. Acocella et al. / Journal of Structural Geology 27 (2005) 397–408 401

Fig. 3. Evolution of experiment TIR 7. Map views of: (a) undeformed stage; (b) experiment at tZ60 0 ; (c) experiment at 180 0 ; (d) experiment at 300 0 . Enlarged
section views along: (e) the plate with higher viscosity silicone; (f) the plate with lower viscosity silicone. Arrows in sections indicate the direction of
extension; dashed rectangles show examples of horst and graben.
402 V. Acocella et al. / Journal of Structural Geology 27 (2005) 397–408

At tZ300 0 the experiment does not show significant

differences compared with the previous stage, even though
the depressions are more accentuated and the strike-slip
faults better developed (Fig. 3d).
The section view of the more viscous plate at the end of
experiment is shown in Fig. 3e. Its thinning has been
achieved through the development of a set of grabens and
horst-like structures. The areas of maximum thinning in the
brittle part correspond to the areas of rise of the underlying
silicone putty.
The section view of the less viscous plate at the end of
experiment is shown in Fig. 3f. The horst–graben configur-
ation is here more pronounced, with local elision of the
Fig. 4. Time variation of the mean percentage of extension for each plate. brittle overburden and rise of the silicone. The more severe
thinning of this plate is related to the lower viscosity of the
faults, slightly oblique to the extension direction and silicone.
displaying arcuate shapes. The mean percentages of extension measured for each
At tZ180 0 the depressions have become wider and plate during the experiment are given by:
deeper (Fig. 3c). The interaction between the normal faults
em Z ðLm K Li Þ=Li
bordering these grabens now occurs with different mod-
alities. Far from the free boundary, the interaction still where Li is the initial length of the plate and Lm is the
occurs through relay ramps (Fig. 3c). Near the free incremental length of the plate after a given time interval.
boundary, the normal faults bordering the grabens are These percentages represent average values for each plate:
interrupted by a set of left-lateral faults subparallel to the at 60 0 (15 and 32%), 180 0 (32 and 60%) and 300 0 (37 and
extension direction (Fig. 3c). Therefore, the interaction 67%) they show an overall decay in the amount of extension
between extensional structures here occurs by means of with time (Fig. 4).
transfer faults. The displacement vectors derived from the extension of

Fig. 5. Vectors of displacement of the nodes of the reference grid of experiment TIR 7 at the final stage of deformation (tZ300 0 ). The initial position of the
nodes coincides with the origin of the arrows. e1 and e2 are the percentages of extension related to each couple of nodes (origin of the dotted arrows) at the two
sides of the plates contact. The difference (e1Ke2) between the percentages of extension for each couple of nodes at the sides of the contact gives the percentage
of differential extension De. The inset highlights the variations of De (also represented as shades of grey) along the plate contact.
 V. Acocella et al. / Journal of Structural Geology 27 (2005) 397–408 403

contact gives the local percentage of differential extension

De (inset in Fig. 5).
The shades of grey in Fig. 5 show that the presence of
transfer faults (grey lines) connecting the two extending
plates is limited to differential extension DeO24%; below
this threshold, the extending plates are connected by relay
ramps. The differential extension of 24% represents there-
fore, in this experiment, the threshold between two types of
interaction between extensional structures, characterized by
relay ramps and transfer faults.
The remaining experiments, with the exception of TIR 4,
showed an overall deformation pattern similar to TIR 7,
despite their different model attributes (Table 2). The
similar deformation pattern was produced during the
Fig. 6. Diagram showing the mean spacing of the depressions on the two
plates for each experiment. The insets above report the mean spacing
development of relay ramps far from the free boundary
relative to all the experiments. and transfer faults near to the boundary. The values of
differential extension associated with the presence of
the reference grid at tZ300 0 are shown in Fig. 5. In general, transfer faults are similar to TIR 7, giving a mean threshold
the vectors are longer on the low-viscosity plate and of 21G3% (Table 2). These data show that the presence of
increase towards the free boundary. This implies that the transfer faults in all the experiments is restricted, at any time
differential extension at the contact between the two plates during the evolution of an experiment, to a mean differential
increases towards the free boundary as well. extension O21%; at lower values, relay ramps occur.
Fig. 5 also shows the percentages of extension (e1 and e2) The depressions in both plates in all the experiments are
related to each couple of nodes (origin of the dotted arrows) regularly spaced. Their mean spacing, related to each plate
at the border between the plate contact; these percentages and experiment, is shown in Fig. 6. The depressions on the
are given, for each node, by: lower viscosity plate have a slightly larger mean spacing
(SZ6.7) with regard to those on the higher viscosity plate
e Z ðLf K Li Þ=Li (SZ4.8).
The experimental transfer faults are usually arranged in
where Li is the initial length of the model and Lf is the subparallel segments. Their evolution is shown through
incremental length of the node plus the initial length of the different stages of experiment TIR 3 (Fig. 7); transfer faults
model Li. The difference (e1Ke2) between the percentages usually grow in length through the linkage of en-échelon
of extension for each couple of nodes at the sides of the segments. Their propagation is mainly away from the free

Fig. 7. Evolution of a transfer fault in experiment TIR 3.

404 V. Acocella et al. / Journal of Structural Geology 27 (2005) 397–408

Table 3
Estimates of b (and related references) for rift zones on Earth, Venus and Mars. The references describing the type of interaction (r.r.Zrelay ramps, t.f.Z
transfer faults) are given in the introduction

Rift type Reference Stretching b Interaction type

Narrow rifts
Rio Grande (Golombek et al., 1983) 1.08 r.r.
Rhine graben (Illies, 1979; Villemin et al., 1986) 1.13G0.03 r.r.
East African Rift System (Ebinger, 1989a; Prodehl et al., 1997) 1.15 r.r.
Baikal (Agar and Klitgord, 1995) 1.16 r.r.

Oceanic ridges
Iceland (Forslund and Gudmundsson, 1991; !1.06 r.r.
Dauteuil et al., 2001)

Extra-terrestrial narrow rifts

Valles Marineris (Mars) (Mege and Masson, 1996) 1.06 r.r.
Beta Regio (Venus) (Foster and Nimmo, 1996) 1.1G0.1 r.r.
Wide rifts
Mojave Desert (Martin et al., 1993) 1.5 t.f.Cr.r.
Basin and Range (Wernicke, 1985) O1.5 t.f.Cr.r.
North Sea (Latin and White, 1990) 2.5 t.f.Cr.r.

Passive margins
Brazil (Milani and Davison, 1988) 1.39 t.f.Cr.r.
Newark (Schlische, 1992) 1.45 t.f.Cr.r.
Suez (Angelier, 1985) 1.45 t.f.Cr.r.
NE Atlantic (Tsikalas et al., 2001) 2.2 t.f.Cr.r.
Galicia (Boillot et al., 1995) O2.75 t.f.Cr.r.
W Africa (Watts and Stewart, 1998) O3 t.f.Cr.r.

Back-arc basins
Tyrrhenian margin (Faccenna et al., 1997) O1.45 t.f.Cr.r.
Japanese margin (Jolivet et al., 1994) 1.8 t.f.Cr.r.

boundary, because the percentage of differential extension is The rate of extension shows an overall decrease with time
greatest at the free boundary and increases with time at all (Fig. 4), consistently with the decrease of st.
points along the transfer zone. The different viscosities of silicone at the base of the
TIR 4 was the only experiment where transfer faults did models are responsible for the differential extension. The
not develop and was characterized by a lower viscosity experiments are all consistent with a differential extension
contrast between the two silicone layers (viscosities of associated with graben-like structures; these are, on the less
1.7!105 and 4.5!105 Pa s). As a consequence, their viscous plate, more numerous, wider and deeper. The
maximum differential extension was w10% and, consist- grabens show moderate variations in their spacing (Fig. 6),
ently with the above results, the interaction between controlled by the interplay between the variations in Tb, Td
extensional structures in the two plates was characterized and the viscosity of silicone. In particular, the spacing S
only by relay ramps. between instabilities responsible for the initiation of
stretching of competent materials with thickness Tb follows
the relationship Sw4Tb (Ricard and Froidevaux, 1986).
4. Discussion and conclusions Also, the instabilities developed during extension lead to the
thinning of the sand overburden and the consequent rise of
4.1. Interpretation of the experiments silicone; the lower the viscosity and the higher the thickness
of silicone, the easier it will rise (Brun, 1999). A higher
The tilt of the basal plate induces the downward flow of amount of risen silicone results in the higher spacing of the
the silicone, causing extension of the model. At the very depressions (Fig. 3f). Similar processes have been used to
beginning, the low amount of extension within a restricted study different modalities of continental extension, assum-
area is responsible for a deformation pattern interpretable as ing therefore a general significance (Brun, 1999, and
similar to the one of narrow rifts. At later stages, the higher references therein).
amount of extension within a wider area is responsible for a Since the silicone nearer to the free boundary moves
deformation pattern more similar to the one of wide rifts. further (Fig. 5), the differential extension De (DeZe1Ke2)
 V. Acocella et al. / Journal of Structural Geology 27 (2005) 397–408 405

increases towards the free boundary. As a result, two types known about a rift zone is its overall stretching factor b,
of interaction develop due to the gradient of differential where eZ(bK1)!100, rather than the differential exten-
extension. sion between its portions.
The type of interaction is a function of the percentage of The b estimates for various extensional settings and the
differential extension between the two plates De. When related references are shown in Table 3; the references
De!21G3%, the interaction occurs through relay ramps or describing the type of interaction observed in each setting
accommodation zones. Their evolution can be characterized are listed in the introduction. Table 3 shows that the narrow
by the following stages (Fig. 3): (a) lateral propagation of rifts (EARS, Baikal Rift, Rhine Graben, Rio Grande Rift)
the faults; (b) linkage with the adjacent structure. Their are characterized by a stretching factor b!1.16. In contrast,
overall evolution is therefore similar to that of the natural the wide rifts (Basin and Range, North Sea), back-arc basins
prototypes (e.g. Acocella et al., 2000, and references (Tyrrhenian Sea, Japan Sea) and passive margins at various
therein). When DeO21G3%, the interaction occurs stages of evolution (Suez Rift, Atlantic margins of Brazil, W
through transfer faults. The experiments display the growth Africa, Newark, Galicia and Norway) are characterized by
of the transfer faults, due to the linkage of smaller en- higher stretching factors, where bO1.39.
échelon strike-slip segments (Fig. 7): because of their These data suggest that at narrow rifts, characterized by a
geometries and kinematics, these segments can be inter- limited amount of stretching (b!1.16), significant (relevant
preted as Reidel systems R of the growing transfer fault. at the rift scale) transfer faults cannot usually form. In fact,
The experiments suggest that relay ramps, usually as the experiments suggest that transfer faults occur for
characterized by a minor component of strike-slip, accom- DeO21%, a differential extension within the narrow rift
modate minor differential displacements between two requires, to form transfer faults, stretching values higher
adjacent extensional structures. Conversely, larger differ- (bO1.21) than those measured. As a result, at narrow rifts
ential displacements can be only accommodated by relay ramps are the commonly observed type of interaction
predominant strike-slip systems parallel to the extension between extensional structures (Illies, 1975; Cordell, 1978;
direction, such as transfer faults. Sherman, 1978; Morley, 1988; Ebinger, 1989a,b; Brun et
The very moderate scatter (G3%) of the experimental al., 1991; Hutchinson et al., 1992; Mack and Seager, 1995)
threshold suggests that geometric parameters (such as the and significant transfer faults are lacking. The formation of
ratio between the thickness of the brittle and ductile transfer faults for De!21% may indeed be possible,
materials, the slope angle and the lateral dimensions of provided there is a presence of pre-existing structures,
the models) do not significantly affect the overall modalities subparallel to the extension direction, along the boundary
of deformation, the types of interaction and their threshold between extending crustal portions. In this case, their
value. The 3% scatter is moderate also when compared with reactivation may form transfer faults. Evidence for
the variations in the spacing (approximately G20%; Fig. 6) reactivated transfer faults is found in the Apennines of
of the depressions; this suggests that the development of the central Italy, where pre-existing NE–SW structures are
type of interaction is independent from the overall reactivated as transtensive under NW–SE regional exten-
configuration (frequency or spacing) of the extensional sion, with De!21% (Acocella and Funiciello, 2004). Also,
structures. Moreover, the fact that the two types of transfer-like features in the EARS have been interpreted as
interaction occur above the contact between the silicone due to the reactivation of pre-existing structures (Rosendahl,
layers confirms that the latter does not influence the type of 1987).
deformation, depending entirely on the amount of differen- Conversely, on wide rifts, back-arc basins and passive
tial extension. margins, where bO1.39, the kinematic conditions required
to develop transfer faults can be fully met. Where the
4.2. Comparison with nature percentage of differential extension remains at De!21%,
relay ramps continue to form. Nevertheless, because of the
A qualitative comparison between our experiments and larger extension involved, a non-uniform extension may
nature shows that the analogue faults share close geometric locally allow DeO21%, developing transfer faults; this
and kinematic similarities with those observed in exten- could be more easily achieved in those settings character-
sional settings Also, the overall deformation pattern is ized by the highest b (Table 3). As a result, both transfer
consistent with the ones observed at various rifts, such as the faults (Gibbs, 1984; Milani and Davison, 1988; Schlische,
Basin and Range, the Aegean Sea or the Tyrrhenian area 1992; Martin et al., 1993; Boillot et al., 1995; Clemson et
(Duenbendorfer and Black, 1992; Gawthorpe and Hurst, al., 1997; Dorè et al., 1997; McClay and Khalil, 1998; Watts
1993; Martin et al., 1993; Acocella and Funiciello, 2004). and Stewart, 1998; Acocella et al., 1999a; van der Werff,
A quantitative comparison is limited by the insufficient 2000; Tsikalas et al., 2001) and relay ramps (Larsen, 1988;
knowledge of the percentages of differential extension Gawthorpe and Hurst, 1993; Anders and Schlische, 1994;
between crustal portions in rift zones. This constitutes a Moustafa, 1996; Ferrill et al., 1999; Peacock et al., 2000b)
significant limitation for the complete and rigorous are commonly observed as types of interaction between
application of the results to nature. In fact, what is usually extensional structures where bO1.21 (Table 3).
406 V. Acocella et al. / Journal of Structural Geology 27 (2005) 397–408

Fig. 8. Types of interaction within rift zones on Earth, Venus and Mars as a function of the stretching factor b. Relay ramps are widespread and independent
from b. Transfer faults are observed only where bO1.39. The experimental threshold (eZ21% or bZ1.21) coincides with the separation between domains
where relay ramps (continental narrow rifts, oceanic ridges and extra-terrestrial narrow rifts) and relay ramps and transfer faults (wide rifts, passive margins,
back-arc basins) have been observed.

These considerations show an overall consistency 1.16) and both of transfer faults and relay ramps within rift
between the experimental data and the continental exten- zones characterized by higher stretching factors (bO1.39).
sional domains. As far as oceanic extensional domains are In the latter case, it is proposed that the amount of
concerned, the best-studied ridge is possibly the Icelandic differential stretching De between adjacent extensional
Ridge. Calculations of its crustal dilation suggest a structures will locally determine the occurrence of transfer
maximum bZ1.06 (Forslund and Gudmundsson, 1991; faults (DeO21%) or relay ramps (De!21%).
Dauteuil et al., 2001) and therefore transfer faults should be
inhibited. This is in agreement with the fact that transfer
faults are not observed within the ridge of Iceland and that
the dominant type of interaction are relay ramps (Acocella Acknowledgements
et al., 2000).
Rift zones are also found on other terrestrial planets, C. Faccenna is acknowledged for helping in the set-up of
characterized by a rigid lithosphere. Among these, the best- the experiments and a critical read of the manuscript.
known cases are Beta Regio (Venus) and Valles Marineris N. D’Agostino kindly provided Fig. 5. Suggestions from
(Mars), both consisting of narrow rifts, with overall J.P. Brun, A. Nicol and A.E. Clifton helped to significantly
deformation patterns similar to the rifts on Earth (Frey, improve the work. Work partly financed with GNV (Campi
1979; Foster and Nimmo, 1996; Anderson et al., 2001). Flegrei Project) Funds.
These planets exhibit gravity forces and crustal thickness
that are different to those on Earth. Even though the
performed experiments have not been specifically built to
take into account these variations, it is nevertheless References
interesting to consider that these rift zones are likely
Acocella, V., Funiciello, R., 2004. Transverse structures and volcanic
associated with a maximum bZ1.1 (Foster and Nimmo, activity along the Tyrrhenian margin of central Italy. Proceedings of the
1996; Mege and Masson, 1996) and lack of transfer fault- 32nd International Geological Congress, Florence, Italy, p. 106.
like structures. Therefore, their overall geometry and Acocella, V., Salvini, F., Funiciello, R., Faccenna, C., 1999a. The role of
kinematics may be consistent with the experimental results transfer structures on volcanic activity at Campi Flegrei (Southern
Italy). Journal of Volcanology and Geothermal Research 91, 123–139.
and the Earth analogues.
Acocella, V., Faccenna, C., Funiciello, R., Rossetti, F., 1999b. Sand-box
The comparison between the experimental data and the modelling of basement controlled transfer zones in extensional
considered continental and oceanic rifts on Earth and domains. Terra Nova 11, 149–156.
selected examples form terrestrial planets suggests a Acocella, V., Gudmundsson, A., Funiciello, R., 2000. Interaction and
consistency in the type of interaction between extensional linkage of extensional fractures: examples from the rift zone of Iceland.
Journal of Structural Geology 22, 1233–1246.
structures within rift zones (Fig. 8). This consistency Agar, S.M., Klitgord, K.D., 1995. Rift flank segmentation and propagation:
consists of the widespread presence of relay ramps within a neotectonic example from Lake Baikal. Journal of the Geological
rift zones characterized by moderate stretching factors (b! Society London 152, 849–860.
 V. Acocella et al. / Journal of Structural Geology 27 (2005) 397–408 407

Aldrich, M.J., 1986. Tectonics of the Jemez Lineament in the Jemez Frey, H., 1979. Martian Canyons and African Rifts: structural comparison
Mountains and Rio Grande Rift. Journal of Geophysical Research 91, and implications. Icarus 37, 142–155.
1753–1762. Gawthorpe, R.L., Hurst, J.M., 1993. Transfer zones in extensional basins:
Anders, M.H., Schlische, R.W., 1994. Overlapping faults, intrabasin highs their structural style and influence on drainage development and
and the growth of normal faults. The Journal of Geology 102, 165–180. stratigraphy. Journal of the Geological Society, London 150, 1132–
Anderson, R.C., Dohm, J.M., Golombek, M.P., Haldemann, A.F.C., 1137.
Franklin, B.J., Tanaka, K.L., Lias, J., Peer, B., 2001. Primary centers Gibbs, A.D., 1984. Structural evolution of extensional basin margins.
and secondary concentrations of tectonic activity through time in the Journal of the Geological Society, London 141, 609–620.
western hemisphere of Mars. Journal of Geophysical Research 106, Gibbs, A.D., 1990. Linked faults in basin formation. Journal of Structural
20563–20585. Geology 12, 795–803.
Angelier, J., 1985. Extension and rifting: the Zeit region, Gulf of Suez. Golombek, M.P., McGill, G.E., Brown, L., 1983. Tectonic and geologic
Journal of Structural Geology 7, 605–612. evolution of the Espanola basin, Rio Grande rift: structure, rate of
Benes, V., Davy, P., 1996. Modes of continental lithospheric extension: extension and relation to the state of stress in the Western United States.
experimental verification of strain localization processes. Tectonophy- Tectonophysics 94, 483–507.
sics 254, 69–87. Huggins, P., Watterson, J., Walsh, J.J., Childs, C., 1995. Relay zone
Boillot, G., Beslier, M.O., Krawczyk, C.M., Rappin, D., Reston, T.J., 1995. geometry and displacement transfer between normal faults recorded in
The formation of passive margins: constraints from the crustal structure coal-mine plans. Journal of Structural Geology 17, 1741–1755.
and segmentation of the deep Galicia margin, Spain, in: Scrutton, R.A., Hutchinson, D.R., Golmshtok, A.J., Zonenshain, L.P., Moore, T.C.,
Stoker, M.S., Shimmield, G.B., Tudhope, A.W. (Eds.), The Tectonics, Scholz, C.A., Klitgord, K.D., 1992. Depositional and tectonic frame-
Sedimentation and Palaeoceanography of the North Atlantic Region work of the rift basins of Lake Baikal from multichannel seismic data.
Geological Society Special Publication, 90, pp. 71–91. Geology 20, 589–592.
Brun, J.P., 1999. Narrow rifts versus wide rifts: inferences for the Illies, J.H., 1975. Recent and paleo-intraplate tectonics in stable Europe and
mechanics of rifting from laboratory experiments. Philosophical the Rhine graben Rift System. Tectonophysics 29, 251–264.
Transactions of the Royal Society London 257, 695–712. Illies, J.H., 1979. Holocene movements and state of stress in the
Brun, J.P., Wenzel, F., Ecors-Dekorp Team, 1991. Crustal-scale structure Rhinegraben Rift System. Tectonophysics 52, 349–359.
of the southern Rhinegraben from ECORS-DEKORP seismic reflection Jolivet, L., Tamaki, K., Fournier, M., 1994. Japan Sea, opening history and
data. Geology 19, 758–762. mechanism: a synthesis. Journal of Geophysical Research 99, 22237–
Clemson, J., Cartwright, J., Booth, J., 1997. Structural segmentation and the
22259.
influence of basement structure on the Namibian passive margin.
Kornsawan, A., Morley, C.K., 2002. The origin and evolution of complex
Journal of the Geological Society London 154, 477–482.
transfer zones (graben shifts) in conjugate fault systems around the
Cordell, L., 1978. Regional geophysical setting of the Rio Grande Rift.
Funan Field, Pattani Basin, Gulf of Thailand. Journal of Structural
Geological Society of America Bulletin 89, 1073–1090.
Geology 24, 435–449.
Courtillot, V., Tapponier, P., Varet, J., 1974. Surface features associated
Koukouvelas, I.K., Asimakopoulos, M., Doutsos, T.T., 1999. Fractal
with transform faults: a comparison between observed examples and an
characteristics of active normal faults: an example of the eastern Gulf of
experimental model. Tectonophysics 24, 317–329.
Corinth, Greece. Tectonophysics 308, 263–274.
Dauteuil, O., Angelier, J., Bergerat, F., Verrier, S., Villemin, T., 2001.
Larsen, P.H., 1988. Relay structures in a Lower Permian basement-
deformation partitioning inside a fissure swarm of the northern
involved extension system, Early Greenland. Journal of Structural
Icelandic rift. Journal of Structural Geology 23, 1359–1372.
Geology 10, 3–8.
Dawers, N.H., Anders, M.H., 1995. Displacement–length scaling and fault
Latin, D., White, N., 1990. Generating melt during lithospheric extension:
linkage. Journal of Structural Geology 17, 607–614.
Dorè, A.G., Lundin, E.R., Fichler, C., Olesen, O., 1997. Patterns of pure shear vs. simple shear. Geology 18, 327–331.
basement structure and reactivation along the NE Atlantic margin. Macdonald, K.C., Fox, P.J., 1983. Overlapping spreading centres: new
Journal of the Geological Society, London 154, 85–92. accretion geometry on the East Pacific Rise. Nature 302, 55–58.
Duenbendorfer, E.M., Black, R.A., 1992. Kinematic role of transverse Mack, G.H., Seager, W.R., 1995. Transfer zones in the Southern Rio
structures in continental extension: an example from the Las Vegas Grande Rift. Journal of the Geological Society, London 152, 551–560.
Valley shear zone, Nevada. Geology 20, 1107–1110. Martin, M.W., Glazner, A.F., Walker, J.D., Schermer, E.R., 1993. Evidence
Ebinger, C.J., 1989a. Tectonic development of the western branch of the for right lateral transfer faulting accommodating en echelon Miocene
East African Rift System. Geological Society of America Bulletin 101, extension Mojave Desert, California. Geology 21, 355–358.
885–903. Mauduit, T., Dauteuil, O., 1996. Small-scale models of oceanic transform
Ebinger, C.J., 1989b. Geometric and kinematic development of border zones. Journal of Geophysical Research 101, 20195–20209.
faults and accommodation zones, Kivu–Rusizi Rift, Africa. Tectonics 8, McClay, K., Khalil, S., 1998. Extensional hard linkages, eastern Gulf of
117–133. Suez, Egypt. Geology 26, 563–566.
Ebinger, C.J., Deino, A.L., Drake, R.E., Tesha, A.L., 1989. Chronology of Mege, D., Masson, P., 1996. Amounts of crustal stretching in Valles
volcanism and rift basin propagation: Rungwe Volcanic Province, East Marineris, Mars. Planetary and Space Science 44, 749–761.
Africa. Journal of Geophysical Research 94, 15785–15803. Milani, E.J., Davison, I., 1988. Basement control and transfer tectonics in
Elmohandes, S.E., 1981. The central European graben system: rifting the Reconcavo–Tucano–Jatoba rift, Northeast Brazil. Tectonophysics
imitated by clay modelling. Tectonophysics 73, 69–78. 154, 41–70.
Faccenna, C., Mattei, M., Funiciello, R., Jolivet, L., 1997. Styles of back- Morley, C.K., 1988. Variable extension in Lake Tanganyika. Tectonics 7,
arc extension in the central Mediterranean. Terra Nova 9, 126–130. 785–801.
Ferrill, D.A., Stamatakos, J.A., Sims, D., 1999. Normal fault corrugation: Morley, C.K., Nelson, R.A., Patton, T.L., Munn, S.G., 1990. Transfer zones
implications for growth and seismicity of active normal faults. Journal in the East African Rift System and their relevance to hydrocarbon
of Structural Geology 21, 1027–1038. exploration in rifts. American Association of Petroleum Geology 74,
Forslund, T., Gudmundsson, A., 1991. Crustal spreading due to dikes and 1234–1253.
faults in southwest Iceland. Journal of Structural Geology 13, 443–457. Moustafa, A.R., 1996. Internal structure and deformation of an accommo-
Foster, A., Nimmo, F., 1996. Comparison between the rift systems of East dation zone in the northern part of the Suez rift. Journal of Structural
Africa, Earth and Beta Regio, Venus. Earth and Planetary Science Geology 18, 93–107.
Letters 143, 183–195. Naylor, M.A., Larroque, J.M., Gauthier, B.D.M., 1994. Understanding
408 V. Acocella et al. / Journal of Structural Geology 27 (2005) 397–408

extensional tectonics: insights from sand-box models, in: Roure, F., Schlische, R.W., 1992. Structural and stratigraphic development of the
Ellouz, N., Shein, V.S., Skvortsov, I. (Eds.), Geodynamic Evolution of Newark extensional basin, eastern North America: evidence for the
Sedimentary Basins, International Symposium, Moscow, pp. 69–83. growth of the basin and its bounding structures. Geological Society of
Nelson, R.A., Patton, T.L., Morley, C.K., 1992. Rift–segment interaction America Bulletin 104, 1246–1263.
and its relation to hydrocarbon exploration in continental rift systems. Serra, S., Nelson, R.A., 1988. Clay modeling of rift asymmetry and
American Association of Petroleum Geology Bulletin 76, 1153–1169. associated structures. Tectonophysics 153, 307–312.
Peacock, D.C.P., 2003. Scaling of transfer zones in the British Isles. Journal Sherman, S.I., 1978. Faults of the Baikal rift zone. Tectonophysics 45, 31–
of Structural Geology 25, 1561–1567. 39.
Peacock, D.C.P., Sanderson, D.J., 1991. Displacement, segment linkage Tsikalas, F., Faleide, J.I., Eldholm, O., 2001. Lateral variations in tectono-
and relay ramps in normal fault zones. Journal of Structural Geology 13, magmatic style along the Lofoten–Vesteralen volcanic margin off
721–733. Norway. Marine and Petroleum Geology 18, 807–832.
Peacock, D.C.P., Knipe, R.J., Sanderson, D.J., 2000a. Glossary of normal Villemin, T., Alvarez, F., Angelier, J., 1986. The Rhine graben: extension,
faults. Journal of Structural Geology 22, 291–305.
subsidence and shoulder uplift. Tectonophysics 128, 47–59.
Peacock, D.C.P., Price, S.P., Whitham, A.G., Pickles, C.S., 2000b. The
Walsh, J.J., Watterson, J., 1991. Geometric and kinematic coherence and
world’s biggest relay ramp: Hold with Hope, NE Greenland. Journal of
scale effects in normal fault systems, in: Roberts, A.M., Yelding, G.,
Structural Geology 22, 843–850.
Freeman, B. (Eds.), The Geometry of Normal Faults Geological Society
Pollard, D.D., Aydin, A., 1988. Progress in understanding jointing over the
of London Special Publication, 56, pp. 193–206.
past century. Geological Society of America Bulletin 100, 1181–1204.
Watts, A.B., Stewart, J., 1998. Gravity anomalies and segmentation of the
Prodehl, C., Fuchs, K., Mechie, J., 1997. Seismic refraction studies of the
Afro–Arabian rift system—a brief review. Tectonophysics 278, 1–13. continental margin offshore West Africa. Earth and Planetary Science
Ramberg, H., 1981. Gravity, Deformation and the Earth’s Crust in Theory, Letters 156, 239–252.
Experiments and Geological Applications. Academic Press, London. van der Werff, W., 2000. Backarc deformation along the eastern Japan Sea
Ranalli, G., 1995. Rheology of the Earth. Chapman & Hall, London. margin, offshore northern Honshu. Journal of Asian Earth Sciences 18,
Ricard, Y., Froidevaux, C., 1986. Stretching instabilities and lithospheric 71–95.
boudinage. Journal of Geophysical Research 91, 8314–8324. Wernicke, B., 1985. Uniform-sense normal simple shear of the continental
Rosendahl, B.R., 1987. Architecture of continental rifts with special lithosphere. Canadian Journal of Earth Sciences 22, 108–125.
reference to East Africa. Annual Review of Earth and Planetary Science
15, 445–503.