Chapter

Salt Pillows and Salt Anticlines

4
4.1 Introduction and more geometrically complex if it formed above alloch-
thonous salt. The swell declines laterally, either to thinner salt
Salt-cored folds are among the simplest types of salt-tectonic
or to welded salt. The margin of a salt pillow is commonly
structures. These folds are abundant, easily recognized, and
gradational and subjective to map.
present in all salt-tectonic basins. Salt pillows and salt anti-
Salt pillows and anticlines can form in three ways. First,
clines are upwellings of salt having concordant overburden;
they can form by halokinesis, meaning that they grow purely
that is, they are overlain by strata that are parallel to the upper
by gravity in the absence of lateral tectonic forces. Second, they
salt contact (Figure 4.1). In map view, structural domes having
can form as contractional folds. Finally, they can form in the
a length-to-width ratio of less than 2 are termed salt pillows
cores of normal-fault rollovers.
(Trusheim 1960). Folds having a ratio of 2 or more are called
salt anticlines (DeGolyer 1925). In a single profile, salt anti-
clines and salt pillows are indistinguishable. In the simplest
form of either, the overburden is a gently curved arch having a
4.2 Formation of Salt-Cored Folds by
smooth crest. The fold limbs typically dip gently but can be Halokinesis
isoclinal, having vertical or even overturned limbs. The base of
the salt swell is typically simple if it rests on autochthonous salt
4.2.1 Mechanics
In the absence of lateral tectonic forces, salt pillows and salt
anticlines form by halokinesis (Figure 4.2(a)). Their growth is
driven by differential loading (Section 3.8.3.1, Figure 3.26).
The differential load is applied by lateral variations in density
or thickness in the overburden. Salt flows from areas of higher
gravitational load toward areas of lower load, where the salt
swell forms.
Growth purely by gravitational loading means that salt
pillows or anticlines cannot rise by halokinesis under a uni-
formly thick overburden even if the overburden is denser than
Salt pillow
the salt (Figure 4.3). In fact, any halokinetic salt pillow or
anticline having an overburden of uniform thickness would
deflate.
In order for salt to rise, the gravitational driving forces
must overcome the strength and weight of the roof over
the crest of the structure. Halokinetic pillows thus grow
most efficiently when the roof is thin and weak; growth
tends to slow down as the overburden becomes thicker and
stronger. Late-stage growth of halokinetic pillows and anti-
clines may also be slowed by decreasing salt supply from the
thinning source layer. However, if salt is imported from out
of the plane of section, these salt-cored folds can continue
to amplify even after the adjoining source layer appears
Salt anticline welded.
Limitations placed on halokinetic inflation of salt-cored
folds owing to the weight of the roof may be overcome if folds
Figure 4.1. Salt pillows are equant in plan view, whereas salt anticlines are form by flank deflation (Figure 4.2(a)). The crest of the fold
elongated. Both are cored by salt and have concordant overburden.

62
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 4.2 Formation of Salt-Cored Folds by Halokinesis

a HalokinesisResidual salt pillow Onlap and thinning Amplified salt pillow

Long
synkinematic
interval
Inflation

Deflation
Salt
Original thickness of salt Negligible prekinematic interval

b Contraction Onlap Truncation Erosional truncation of fold crest Salt sheet

Postkinematic interval
Short synkinematic interval

Contractional Thrust
kink band
Thick prekinematic interval
Detachment fold

c Extension
Depocenter Growth wedges

Weld
Extensional pillow Extensional rollover
Figure 4.2. Features of salt pillows and salt anticlines formed by (a) halokinesis, (b) contraction, and (c) extension.

remains at its original elevation, and the flanks subside by synchronous, so one flank may subside while the other
expulsion of salt into other nearby salt structures. The roof is remains static.
not lifted, so less work against gravity is needed to form
the fold. 4.2.3 Crestal Structures
The most common structures above the crest of halokinetic
4.2.2 Diagnostic Features folds are bending-related normal faults (Figures 4.4 and
Halokinetic salt pillows and salt anticlines typically form grad- 4.6(b)). They form a keystone graben by outer-arc extension
ually, unlike folds formed by shortening (Section 4.3). Thus, in anticlinal crests. Near the surface, opening-mode fractures
halokinetic folds have a long growth interval, during which perpendicular to bedding can form. Fault throw can diminish
synkinematic sediments thin over the crest of the fold upward if faulting accompanies sedimentation. However, the
(Figure 4.4). A long growth history is the chief diagnostic lower parts of faults decrease in throw downward to a neutral
feature of halokinetic folds. Because halokinetic folds form surface subparallel to the folded strata. If the neutral surface is
most readily below a thin roof, the prekinematic interval above well within the folded strata, then thrust faults can deform the
the salt core is typically thin (Figure 4.2). Care is needed when fold cores below the neutral surface. Commonly, however, a
interpreting thinning of strata over pillow crests, because thin- neutral surface is not reached within the folded overburden, so
ning may also result from differential compaction. only outer-arc stretching is evident, and this occurs by tangen-
A second feature of halokinetic folds that can help distin- tial longitudinal strain, which is concentrated in fold hinge
guish them from contractional folds is that growth on oppos- zones. Crestal grabens thus do not necessarily signal regional
ing limbs of halokinetic folds need not be the same age extension. The map-view pattern of crestal normal faults
(Figure 4.5). Salt withdrawal on the two flanks need not be reflects the shape of the underlying salt structure: radial

63
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 Salt Pillows and Salt Anticlines

patterns are typical of salt pillows, and more elongate networks

are characteristic of salt anticlines (Figure 4.7).
4.3 Formation of Salt-Cored Folds
A less common type of crestal structure that does not by Shortening
require regional stresses is delamination extension (Figures
4.6(c) and 4.8). The extensional detachment follows weak over-
4.3.1 Diagnostic Features
pressured muds or evaporites interlayered with the overburden Salt anticlines in mountain belts were recognized as a mani-
within the anticline, as in the North Sea, where thin Triassic festation of regional contraction very early in the history of
salts, such as Röt and Muschelkalk, provide detachments for salt tectonics (Stille 1910, 1925). Pillows, by contrast, were
delamination above the main Permian Zechstein salt (Owen
and Taylor 1983; Coward and Stewart 1995; Hooper and More
1995). As a fold limb steepens, gravitational shear stress West East
increases until the dip slope fails.

A B
3
Water hw No growth
ho
Overburden

Two-way time (s)

ho
hs
Salt
4
PB Early, short-

Growth
PA Decreasing wavelength
pressure folds

s = salt density o = overburden density

w= water density g = gravitational acceleration
5
Pressure at PA = ( o
· g · ho ) + ( s ·g · hs )
Pressure at PB = ( o ·g · ho ) + ( w·g · hw )
Because s > w and hs = hw , PA > PB
Figure 4.3. Fluid statics indicates that when overlain by an overburden having Oligocene–Recent Albian
Fault
uniform thickness and density greater than salt, a salt pillow or salt anticline Upper Cretaceous–Eocene Aptian salt
must diminish in amplitude. The equations show static pressures in salt at two
points, PA and PB, having the same elevation. The difference in pressure head 0 3 km
reduces to the density difference between salt in the anticline core and an
equal thickness of water in the syncline core. That density difference is about 2:1 Figure 4.5. Diachronous sedimentation is evidence of halokinetic growth. The
and would be even greater for subaerial folds. The static pressure within the salt west side of the structure began subsiding during deposition of the pale-green
core is much greater than that below the syncline, so salt flows out of the unit, whereas the east side does not show growth until after the orange unit
anticline core and inflates below the adjoining syncline. had accumulated. Kwanza basin (Angola); after Hudec and Jackson (2011).

0 Figure 4.4. Halokinetic salt anticlines typically

show evidence of long, slow growth. In this section
all overburden units thin over the crest of the salt
pillow, indicating continuous growth for more than
100 Myr (Late Jurassic to Eocene). The keystone
graben over the fold crest is interpreted as having
1 resulted from outer-arc bending. East Texas basin
Two-way time (s)

(USA); after Inderweisen (1983).

Salt

3
0 3 km

64
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 4.3 Formation of Salt-Cored Folds by Shortening

Figure 4.6. Second-order structures in the crests

a Indentation extension b Keystone extension of salt pillows and salt anticlines.

Local discordance
in top of salt Neutral surface
at top of salt

c Delamination extension d Thrusted crest

Salt

Fishtail thrust
e Breakthrust f

g Unconformity

historically regarded as halokinetic because their equant plan- preexisting diapirs to expel salt sheets (Chapter 6). Thus, salt
forms seemed unlikely to result from shortening. sheets may be found emplaced at the level of the synkinematic
In contrast to halokinetic structures, which grow under section or, in many cases, lying on the crestal unconformity
continuous gravity forces over long periods of time, contrac- (Figure 4.2(b)).
tional salt-cored folds form during shortening events (Section
11.2.1.1). These events tend to be relatively brief compared
with the overall history of a basin. Thus, the most diagnostic
4.3.2 Contractional Fold Geometries
feature of contractional folds is a long-lived prekinematic Most contractional salt pillows and salt anticlines deform by
phase, followed by a short synkinematic period (the shortening buckling, which is a form of layer-parallel shortening under
event), followed by a postkinematic phase of quiescence. regional compression. As the buckling overburden rucks up,
Although these synkinematic sediments are typically deposited anticlinal cores fill with salt or other weak mobile rocks. These
during a short time, they may be thick if folds have high relief detachment folds are also known as lift-off folds or pop-up
and sedimentation is rapid (Figures 4.2(b) and 4.9). folds. Buckle folds are abundant, so their fold mechanics
Other features, too, can be useful to identify contractional has been exhaustively studied. A near-infinite variety of fold
folds (Figure 4.10). (1) Crestal thrusts or contractional kink geometries can form, depending on precursor structures and
bands indicate that growth of the underlying fold involved mechanical stratigraphy. We summarize those most common
shortening (Figure 4.6(d)–(f)). (2) Contractional folds build in salt tectonics, which are restricted to unmetamorphosed
relief quickly, so they are commonly capped by erosional sedimentary rocks.
unconformities (Figures 4.6(g) and 4.8). Unconformities
may, of course, form above any type of salt-cored fold, but 4.3.2.1 Shortening of a Prekinematic, Originally Flat-Lying Layer
they tend to be more common and more deeply discordant If a sedimentary layer is loaded by compressive stresses parallel
above those formed in shortening. (3) Synkinematic sedi- to bedding, the layer is initially stable, though it may shorten
ments thinning onto contractional folds should be the same laterally by compaction. As stresses increase, a layer containing
age on both sides of the fold crest. (4) In many salt basins the some mechanical contrasts becomes unstable and buckles by
shortening event causes folding but may also cause internal strain in two ideal ways.

65
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 Salt Pillows and Salt Anticlines

a phone book being flexed. Strain builds by slip between strata,

a Hawkins pillow, East Texas basin
D
D U
concentrating in the fold limbs and increasing with limb dip.

46
U The second ideal fold model is neutral-surface folding, also

00
0 1 km known as tangential longitudinal strain (Ramsay 1967) and
orthogonal flexure (Twiss and Moores 1992). Whereas
00 00 U D flexural-slip folding is typical of layered sedimentary rocks,
46 44
neutral-surface folding is typical of massive rocks. Both mech-
D
U anisms produce parallel fold geometry, but the internal strains
40 in profile are radically different. Neutral-surface folding con-
00 centrates strain in the fold hinge rather than on the fold limbs.
00
41

U Bedding extends on the outer arc of the fold (in the crest of an
D anticline or base of a syncline), typically by a fan of normal
faults. In contrast, bedding shortens on the inner arc of a fold
4600

(in the cores of anticlines or synclines) and is typically

U expressed in sediments by reverse faults. In buckle folds
D detaching on salt, this zone of shortening is rarely present
because the inner arc is in the salt, where shortening is accom-
0
U 420 modated by salt flow.

0
460
D Most natural salt anticlines and salt pillows have curved
N geometries in cross section. However, where there is a strong
U U anisotropy imposed by interlayered beds of contrasting
D D D strength, more angular dip-domain folds form (Groshong
1999). Dip-domain folds have a kinked profile where each kink
00

00
46

U 4600
separates a panel of more or less uniform dip (Figure 4.11).
48

D
U Typically the stronger layers are sandstone, chert, or carbonate,
whereas the weaker layers are shale, marl, and especially
b Quitman anticline, East Texas basin 5900 evaporites.
0

-
Despite the prevalence of moderate anisotropy and
90
-5

rounded folds in nature, kinematic models of growth folding

are generally based on dip-domain folds because they are the
95
0 easiest to analyze. Two ideal kinematic models explain the
-5 growth of dip-domain anticlines without coeval sedimentation
00
(Poblet and McClay 1996). In the limb-rotation model,
0 -59 straight fold limbs rotate and steepen but maintain constant
80 N
-5 0 length (Figure 4.12(a)). Fold hinges and axial surfaces are
0 0
0 -5 80 -60 fixed. In the kink-band-migration model (Suppe et al. 1992),
80
-5 0 fold limbs lengthen but remain constant in dip (Figure 4.12
70
00

-5 (b)). Particles move toward the anticline through synclinal

90 -59

axial surfaces, which actively migrate outward. Hybrid forms

0

0
00

80

10
0

0 would combine limb rotation and kink-band migration so that

-58

-5

0 -6
57
-5

0 - 0 1 km limbs steepen and lengthen as the fold amplifies.

-580 00
0
-6 Strata commonly thin over the crest of salt pillows and
-5900
900
-6000 -5 anticlines because of two accompanying processes: sedimenta-
-6100
tion and differential compaction. Sedimentation effects are most
easily analyzed in dip-domain growth folds, using the end-
Figure 4.7. Map patterns of bending-related faults reflect the shape of the member models of limb rotation and kink-band migration
underlying salt structure. Faults above Hawkins pillow are radial, whereas those with variable balance between the rates of uplift and sedimen-
above Quitman anticline define an elongate pattern. U, upthrown; D, tation (Poblet and McClay 1996; Shaw et al. 2005; Figures 4.13,
downthrown. Hawkins map from Wendlandt et al. (1946); Quitman map from Galloway
et al. (1983). and 4.14). Differential compaction can be distinguished from
depositional thinning in two ways. First, onlaps and trunca-
tions cannot form by differential compaction, so their concen-
The first ideal fold model is flexural-slip folding, which is tration in pillow crests signifies synkinematic growth strata.
also known as flexural-shear folding or flexural-flow folding. Second, axial surfaces in compactional drapes form exten-
During flexure, bedding provides innumerable slip surfaces for sional kink bands (Figure 4.15(b)) (Shaw et al. 2005). In
simple shear, so that beds slip past each other like the pages of contrast, axial surfaces in growth sequences form contractional

66
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 4.3 Formation of Salt-Cored Folds by Shortening

South-southwest North-northeast Figure 4.8. Tilting during folding may cause

0 beds to slide down a detachment along one flank
of the structure, a process known as delamination
extension. Stratigraphic relationships indicate that
Erosional this anticline underwent a complicated growth
unconformity history, with two phases of shortening indicated by
erosional unconformities, separated by a period of
Two-way time (s)

Erosional Diachronous halokinetic growth, as indicated by diachronous

unconformity growth
1 flank subsidence. After Owen and Taylor (1983).

Delamination
2 fault

Tertiary Jurassic Middle Triassic Permian - Zechstein salt

Cretaceous Upper Triassic Lower Triassic Subsalt Fault

0 4 km

South-southeast North-northwest South-southwest North-northeast

0
Seaward Landward

Synkinematic Sigsbee 4
sediments salt canopy
Isopachous
Cretaceous–
Early Cenozoic
6
Kink band

Depth (km)
8

10

12

14

Possible depositional
edge of salt 0 15 km
Louann Salt
Neogene Jurassic–Paleogene V.E. × 2.75
Basement

Figure 4.9. Contractional salt anticlines have a long-lived prekinematic phase and a brief synkinematic episode. Perdido fold belt (northern Gulf of Mexico);
from Hudec and Jackson (2011), courtesy of CGG.

67
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 Salt Pillows and Salt Anticlines

kink bands. These diagnostic kink bands form only where

Erosional unconformity the sediments are strongly anisotropic, as in contrasting
multilayers.
Weld During buckling, the dominant wavelength is affected by
the thickness of folded layers, the degree of anisotropy, the
viscosity contrast between the layers, whether the viscosity
is linear or follows a power law, whether the folding is infini-
tesimal or finite, the size of the fold, and the rate of aggrada-
tion. Thus, there is no universal quantitative relationship
between wavelength and layer thickness. However, other
factors being equal, fold wavelength increases with layer thick-
ness. This relation can produce complex, polyharmonic
Thrust fault geometries when a unit is folded in several stages, or when
shortening continues while the layer thickens (Figures 4.16,
4.5, and 4.17).

4.3.2.2 Shortening of Precursor Structures

Pillows formed by halokinesis respond variably to later
shortening. Pillows originally elongated normal to the
shortening direction tighten and lengthen. Conversely, pillows
Salt
originally elongated parallel to the shortening direction
shorten in that direction and widen in a perpendicular direc-
Figure 4.10. Geometric criteria can be used to distinguish contractional tion. Regardless of their orientation, pillows being shortened
from halokinetic salt anticlines. The deeply buried fold here has a thrust fault
in the roof, which is crosscut by an erosional unconformity. A salt sheet was
tend to rise and their flanks tend to steepen. As the pillows
emplaced across this unconformity, and it is now welded on this plane of buckle, their crests can shift laterally if they comprise
section. After Veritel (2008).

a Tight, narrow Broad box

syncline anticline
b c

Narrow
box al l
fold a c tion Extensiona
t r n d kink band
Conink ba
k
Miter fold

Conjugate contractional kink bands

d Multilayered
overburden
Source
layer

0 1 cm

Figure 4.11. Mechanically anisotropic multilayers kink during shortening parallel to the layers. (a) Physical model of a modeling-clay multilayer lubricated by
graphite interlayers (after Price and Cosgrove 1990). (b, c) contractional and extensional kink bands (after Price and Cosgrove 1990). (d) Narrow, sinusoidal anticlines and
broad synclines (black multilayer) just above the source layer pass upward (red multilayer) into kink bands and broad, flat-crested box anticlines and tight, narrow
synclines as the anisotropy dominates (centrifuge model by Hemin Koyi comprising silicone putty [pale] and modeling clay [dark]; Hudec and Jackson 2011).
68
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 4.3 Formation of Salt-Cored Folds by Shortening

Figure 4.12. Angular anticlines rise in two

a Constant limb length, b Variable limb length, idealized ways, shown here without synkinematic
variable limb dip constant limb dip sedimentation: (a) limb dips steepen but have
constant length, or (b) fold limbs lengthen but
have constant dip. After Poblet and McClay (1996).
Salt

Fold grows by Fold grows by axial

limb rotation surface migration

Passive axial surface Material moves through

Limited-activity active axial surfaces
axial surfaces
Active axial surface

Constant limb length,

low sedimentation rate nonparallel strata, in ways that depend on the fold mechanism
and prefolding geometry (Ramsay 1967). Where pillows are
a laterally offset from one another, the resulting buckle folds are
also offset. An array of precursor pillows having various shapes
and sizes creates an irregular template for any subsequent
Salt shortening. The resulting array of structures has been termed
a pillow fold belt (Jackson et al. 2011; Dooley et al. 2013).
Bed dips decrease upsection The original halokinetic pattern of pillows and the super-
Forelimb onlap, posed folds interfere to form fold patterns different from those
Backlimb onlap then offlap
in classic fold belts. For example, Figure 4.18 is an erosional
section through interference folds in the Great Kavir of north-
ern Iran. The folds probably detach on the same Miocene
evaporites that are exposed a few kilometers away in diapirs.
X-shaped saddles form at quadruple junctions where two
domes and two basins meet. Four domes form an anticlinal
ring around a central structural basin, as is typical of dome-
Variable limb length, and-basin interference. In this part of the Alborz foreland the
low sedimentation rate regional shortening structures trend roughly east–northeast.
b The orthogonal trend, which is more open, is likely to be that
of salt pillows that predate shortening.
Large faults can be folded during later buckling. For
example, normal growth faults in a rift basin can be inverted
as thrusts and became folded during subsequent continental
Backlimb and forelimb offlap collision (Figure 4.19; Section 11.3.3). Alternatively, major
Fold scarp
Synkinematic

thrusts can form early during shortening and be later folded.

These folded thrusts ramp up from the detachment of thin salt
strata

through the fold core, and then flatten obliquely across the
crest (Figure 4.20).

Prekinematic strata 4.3.3 Crestal Structures

Figure 4.13. Sediments form offlapping patterns when they accumulate Contractional folding involves bending of beds, so keystone
more slowly than an anticline rises. The details of the pattern depend on normal faults are possible just as in halokinetic structures
whether the anticline is rising by (a) steepening limbs of constant length, or (b) (Figures 4.6(b) and 4.21), although perhaps less common
lengthening limbs of constant dip. After Poblet et al. (1997).

69
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 Salt Pillows and Salt Anticlines

Constant limb length, Variable limb length, Figure 4.14. Sediments thin over the crest of a
a high sedimentation rate b high sedimentation rate fold when they accumulate faster than the
anticline rises. The details of the pattern depend on
whether the anticline is rising by (a) steepening
limbs of constant length, or (b) lengthening limbs
Salt of constant dip. After Poblet et al. (1997).
Growth
Early onlap No onlap triangle

Growth axial Active axial

Dips fan outward surface surface

Bed dips Crestal Bed dips parallel to Anticline

decrease upsection syncline prekinematic strata widens upward

Synkinematic
strata
Minor growth triangle Major growth triangle Prekinematic strata
with constant limb dip

Postkinematic drape fold Figure 4.16.

Extensional
a λ1 Multiwavelength growth
folds form as they
kink band Axial surface
dips away from crest amplify and sediments
Postkinematic

thicken during
shortening. (a, c)
Dominant wavelength,
b λ2
λ1, forms proportional to
thickness of overburden
(blue) relative to salt
(red). (b) A larger
wavelength, λ2, forms
Synkinematic growth fold after the overburden
thickens and shortening
Contractional Axial surface resumes. (d) A larger
kink band dips toward crest wavelength, λ2, forms if
Synkinematic

the overburden thickens

as shortening continues.

c λ1

Figure 4.15. The dip direction of kink bands helps to distinguish between a
postkinematic drape fold and a synkinematic growth fold. After Shaw et al. (2005). λ2
d
owing to compressional horizontal stresses. Delamination
extension (Figure 4.6(c)) is also possible, and it is perhaps
more likely in contractional folding because of the greater
uplift and rotation possible in shortening.
With enough shortening, the various types of thrust geom-
etries are diagnostic. Small thrusts offsetting the crest of a anticlines are breakthrusts (Willis, 1893; Fischer et al. 1992;
pillow or anticline are rare because this shortening is coun- Rowan et al. 1999) (Figure 4.6(e)). The thrusts do not pass into
tered by extension in a fold hinge zone (Figure 4.6(d)). Thrust higher detachments, and their displacement decreases upward
faults that splay upward and outward from the core of (Figure 4.22). Between breakthrusts the anticlinal roof rises as

70
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 4.3 Formation of Salt-Cored Folds by Shortening

a pop-up. Breakthrusts form while shortening continues after follows a weak layer, such as shale (Harrison 1995; Letouzey
strata have already buckled. Steep seismic reflectors in con- et al. 1995; Sans and Vergés 1995).
tractional kink bands in fold limbs can resemble thrusts if Erosion is particularly likely in contractional folding
seismic processing is inadequate (Camerlo and Benson 2006). because shortening can produce large uplift, exposing crestal
Fishtail thrust faults in anticlines abruptly reverse their ver- areas to erosion unless sedimentation is rapid (Figure 4.6(g)).
gence in zigzag fashion (Figure 4.6(f)) where the thrust flat Furthermore, many contractional provinces are subaerial,
making erosion even more likely. Deep erosional thinning
weakens roof strata and enhances differential loading of the
West East salt, facilitating upwelling and diapiric breakout of salt at the
6 fold crest (Trusheim 1960; Coward and Stewart 1995).
St. Malo Long
Late growth
of west flank
anticline synkinematic
phase 4.3.4 Salt Flow during Shortening
8 During moderate shortening, synclines tend to remain at

Depth (km)
Early, short- much the same elevation while anticlines rise, which thickens
wavelength folds the source layer. Salt flows into the low-pressure cores of
10 anticlines, which are encased by overburden that is much
higher in pressure than at equivalent depth in the synclines
(Figure 4.23(a)). Thus, the salt does not amplify the fold but
12 merely passively fills the fold core as the strong overburden
buckles upward.
Plio–Pleistocene Cretaceous When buckling intensifies, salt is expelled downward from
Miocene Late Jurassic the anticlinal cores, which further thickens the source layer
Oligocene Salt
0 10 km (Figure 4.23). As salt is expelled, the fold limbs converge. Fold
Paleocene–Eocene Basement
V.E. × 3 limbs that are isoclinal can eventually meet to form a weld
(Figure 4.24). Alternatively, if the fold limbs rotate beyond
Figure 4.17. Early contractional folds have shorter wavelengths than folds parallel, they trap salt in the fold core (Figure 4.23(b)). Within
formed later when sediments were thicker. A long synkinematic phase suggests
that the later folding may have been halokinetic. St. Malo anticline (deepwater the core the salt pressure rises to equal that of the overburden
Gulf of Mexico); after Hudec and Jackson (2011), courtesy of CGG. and is much higher than in deeper salt below the weld.

Figure 4.18. Interference structures in the

Miocene Upper Red Formation in the Great Kavir
(Iran). After Hudec and Jackson (2011).
34.75° N
34.60° N
54.50° E
54.30° E

Pr tre

0 8 km Anticline
eb nd

g
uc

lin Syncline
ck nd
kli

u
n

B tre
g

71
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 Salt Pillows and Salt Anticlines

Present day Figure 4.19. Cretaceous extensional growth

North-northeast South-southwest faults were inverted as thrusts during Neogene
Onteniente-1 (proj) shortening in the Betic Cordillera (Spain); after De
Sierra Mariola Ruig (1992).
Sierra Solana anticline -0.667° -0.467°
anticline

38.833°
0
Depth (km)

Sec
2

tion
Salt
3

38.667°
Alcoy
4
0 4 km VE × 1

Restored to Paleogene

Basement

Neogene Senonian Lower Cretaceous Lower-Middle Jurassic

Paleogene Cenomanian Upper Jurassic Triassic

North South Oligocene

Weissenstein anticline
Folded Topography Oxfordian to Kimmeridgian lime-
thrust stone topped by Eocene karst
Elevation
(m) Lower Oxfordian Effingen
Formation (calcareous shale)
1000
Lower Oxfordian
Birmenstorf limestone
Bathonian to Callovian
argillaceous limestone
500
Bajocian to Bathonian
oolite (Hauptrogenstein)
Lower Bajocian
argillaceous limestone
0
Aalenian Opalinum shale

Lias
Salt
Upper Triassic (Keuper)
V.E. × 1
Middle Triassic carbonate
Extensional fault (Rhine graben age?) (Hauptmuschelkalk)

0 500 1000 m Faults Middle Triassic evaporite

Basal Triassic sandstone

(basement)
Figure 4.20. This southward-verging anticline is cut by folded thrusts, created during Jura shortening, and by younger extensional faults, partly reactivated by
sinistral strike-slip. After Meier (1977).

72
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 4.3 Formation of Salt-Cored Folds by Shortening

2° tilt
Overburden pressure high in
anticline core
258,000 yr

4.0 Overburden
Two-way time (s)

Salt

Salt pressure low in anticline core

458,000 yr
4.5

Salt

0 1 km

Figure 4.21. Late-stage shortening arched this diapir roof. Despite the overall
contraction of the fold, outer-arc extension in the fold crest created a keystone
graben. Lower Congo basin (Gabon); courtesy of Direction Générale des
Hydrocarbures, Gabon.
Anticlines widen upward
Synclines widen downward
Figure 4.23. Low pressure in the core of an anticlinal box fold at an early
stage (258,000 yr) of buckling shows that salt flows passively into a fold core in
the overburden. After the fold core pinches shut (458,000 yr), salt within the
core is pressurized. The overburden comprises alternating beds of viscous and
plastic layers. Numerical model by Dan Schultz-Ela (Hudec and Jackson 2011).

Figure 4.22. Break thrusts cut the angular hinge of a box fold detaching on Paleogene salt. The East Qiuliake anticline is exposed in the Kezilenuer River, Kuqa
foreland basin (China). Person in river bed provides scale. Photograph by Martin Jackson.
73
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 Salt Pillows and Salt Anticlines

The thickness of salt strongly affects how the overburden

buckles. If the salt is much thinner than the overburden,
4.4 Formation of Salt-Cored Folds
folding is limited (Figure 4.25(a)). The detachment folds have by Extension
low amplitude because there is not enough salt to fill large Most salt pillows and anticlines form by halokinesis or buck-
anticlinal cores. Furthermore, the proximity of the rigid sub- ling, but some form by extension (Figure 4.2(c)). Extensional
salt precludes rounded synclines. Thin salt thus promotes low- rollovers form between separating blocks of overburden trans-
amplitude, symmetric box-fold anticlines having isoclinal lating along a low-angle décollement. As the hanging wall
limbs. Where salt is thicker, folds are more rounded and open
even though the mechanical properties of the overburden and
salt are unchanged (Figure 4.25(b)–(c)). Compared with folds
over thin salt, strain is more diffuse and fold amplitude is
a
Moving endwall Fixed endwall
greater.
2° tilt 293,000 yr
Symmetric box fold Low
amplitude

1 0.3-km-thick salt Closed core

b 250,000 yr

2
Salt
Depth (km)

1-km-thick salt
526,000 yr

3
c Asymmetric rounded fold High amplitude

Salt Open
4 Deformed core
VE × 1
passive grid

0 1 km
V.E. × 1.5
4-km-thick salt
Figure 4.24. Continued shortening of a salt-cored fold squeezed the isoclinal
limbs together, forcing most of the salt back down into the main salt body. Figure 4.25. Thicker salt favors more open, higher amplitude, less symmetric
A small remnant of salt is still present at the top of the weld. Northern Gulf of folds. The overburden is isotropic. Numerical model by Dan Schultz-Ela (Hudec and
Mexico; courtesy of CGG. Jackson 2011).

Figure 4.26. A salt anticline is bounded by

extensional rollovers and salt diapirs on both sides.
Offshore Brazil; after Schlumberger and TGS-NOPEC
(2000).

74
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007
 4.4 Formation of Salt-Cored Folds by Extension

Figure 4.27. Salt pillows can form by extension,

a Bend as shown in this example from offshore southern
Gabon (Astrid Marin area). (a) Seismic profile
showing that the eastern limb of the salt pillow is
3
part of an extensional rollover; the other limb
subsided by expulsion and welding of salt. (b) Dip
map showing how normal faults within this
rollover crosscut the domal salt pillow, whose
upper contact is contoured in yellow. Seismic data
courtesy of Direction Générale des Hydrocarbures, Gabon.
Two-way time (s)

Salt

0 2000 m Survey 0 2000 m

boundary
b

0 2000 m

slides down a listric normal fault, strata bend toward the fault. field results from arching of the overburden and is influenced
A fan of synkinematic wedges accumulates during bending by the planform of the salt swell. The other stress field results
and thickens toward the master fault. As this extensional roll- from the master extensional fault and is controlled by the
over sags, it expels underlying salt, which flows away from the shape of the fault surface. The crestal faults in Figure 4.27(a)
fault and thickens the source layer there. The rollover forms are clearly localized over a salt pillow, but their linear
one flank of a salt pillow or anticline. This association with a map pattern is controlled by the master extensional fault,
rollover is diagnostic of this type of structure. Other flanks of not by the subcircular salt pillow (Figure 4.27(b)). The
the structure may form by salt expulsion or by an opposing regional stress field related to the master extensional fault
rollover (Figure 4.26; Section 10.2.2.2). dominates the local stress field caused by arching over the
If crestal faults form above the salt pillow or anticline salt pillow, as demonstrated by physical modeling (Withjack
their pattern is influenced by two stress fields. One stress and Scheiner 1982).

75
Downloaded from https:/www.cambridge.org/core. Columbia University Libraries, on 22 Mar 2017 at 18:19:42, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms.
https://doi.org/10.1017/9781139003988.007