Chapter 13

Structure and structural history

A. J. B A R B E R & M. J. C R O W

The present structure of Sumatra is dominated by the effects of form part of the Nicobar Fan and are 2 km thick. The turbidites
the current subduction system in which the Indian Plate is were derived from the Himalayas following their uplift during
being subducted northeastwards beneath the island at a rate of the Miocene, and formed the eastern branch of the Bengal Fan,
c. 7 cm a-~. The structure of Sumatra was described by van before sediment supply was cut off by the collision of the northern
Bemmelen (1949) and in terms of plate tectonics by Hamilton end of the Ninety-East Ridge with the subduction trench in Pleis-
(1979). The main structural elements of Sumatra and its surround- tocene times. On the ocean floor the sedimentary cover decreases
ing region are defined with respect to the Sumatran subduction in thickness southwards, until at the southern end of Sumatra, the
system (Fig. 13.1). thickness of the fan sediments is reduced to less than 1 km
(Fig. 13.2, inset). Sediments of the Nicobar Fan are covered by
(1) Forearc region, which includes the subduction trench, part
a thin veneer of Recent pelagic sediments.
of the Sunda Trench extending from Myanmar to eastern
Seismic reflection profiles obtained by the Scripps Institution of
Indonesia, the developing accretionary complex, composed Oceanography (SIO) around Nias in the 1970s and 1980s as a con-
of ocean floor materials scraped off the Indian Plate, the tribution to the Sumatra Transect, part of the SEATAR (Studies in
forearc ridge which rises above sea level to form the forearc
East Asian Tectonics and Resources) Program (CCOP-IOC 1981),
islands, and the forearc basins which lie between the ridge, show that Indian Ocean lithosphere, and its covering of sediments,
and the volcanic arc on the mainland of Sumatra (Fig. 13.2). are being subducted in the Sunda Trench northeastwards beneath
(2) Barisan Mountains and the Sumatran Fault System. The Sumatra (Fig. 13.3). More recently very similar seismic sections
Barisan Mountains are composed of an uplifted basement of
have been obtained to the south of Enggano by the R / V Sonne
Upper Palaeozoic and Mesozoic sedimentary and volcanic as part of the GINCO (Geoscientific Investigations along the
rocks, variously metamorphosed, deformed and intruded
active Convergence zone between the eastern Eurasian and
by granites, overlain by Cenozoic sediments and volcanics,
Indo-Australian plates off Indonesia) Project (Kopp et al. 2001).
including the products of the volcanoes related to the The subduction trench lies about 250 km to the SW of the main-
present subduction system, which form the currently active land of Sumatra and 100 km to the SW of the outer arc islands
volcanic arc. The Sumatran Fault system is a complex of (Fig. 13.2). At the northern end of Sumatra the subduction
dextral strike-slip faults running the whole length of the trench is 4000 m deep, but the trench increases gradually in
island through the centre of the Barisan Mountains from depth southeastwards, until at the southern end of the island it is
NW to SE, with zones of compression and extension,
more than 6000 m deep (Fig. 13.2, inset). A compilation of
forming areas of uplift and pull apart basins which form
echo-sounding measurements from the floor of the trench, and
grabens along the line of the fault system. Movement along seismic reflection and refraction determinations of the depth of
this transcurrent fault system is attributed to the oblique sub- the underlying oceanic basement shows that this increase in
duction of the Indian Plate beneath Sumatra, which is carrying depth is due entirely to a decrease in the amount of sediment on
the west coast of Sumatra and the whole of the forearc region the ocean floor (Moore et al. 1982) (Fig. 13.2, inset). The SIO
northwestwards as a 'sliver plate' (Curray 1989). seismic reflection profiles show sub-horizontally bedded Nicobar
(3) Backarc region, extending northeastwards from the Barisan Fan sediments on the floor of the trench overlain by a thin
Mountains, across the Malacca Strait to the east coast of the
wedge of more recent sediment at the foot of the inner slope.
Malay Peninsula, occupied by Tertiary sedimentary basins, The Indian Ocean floor slopes gently northwestwards at 2 ~
formed by Palaeogene rifting and subsidence and in filled
towards the trench and as the trench is approached the overlying
by Neogene to present day sedimentation. The sediments sediments and the ocean floor are broken by normal faults down-
are affected by folding and faulting and contain coal and the throwing towards the trench and parallel to the trench axis. At the
major oil and gas resources of Sumatra. base of the inner slope of the trench the sediments on the Indian
Ocean Plate are seen in seismic sections to have been uplifted
along thrust surfaces and imbricated to form an accretionary
complex (Fig. 13.3a). The trenchward outer slope and normal
The Sunda forearc faulting in the ocean floor are attributed to a downward flexure
and a complementary bulge on the incoming plate, resulting
Subduction trench and accretionary complex from loading by the overlying accretionary complex.
The inner slope is made up of a series of ridges and troughs par-
To the west of Sumatra and the outer arc islands, the floor of the allel to the trench axis which rise steeply from the floor of the
!ndian Ocean increases in depth from 4000 m at the northern trench, and then flatten out in the outer arc ridge (Fig. 13.3a).
end of the island to over 5000 m in the south (Fig. 13.2). Two Karig et al. (1980, fig. 4) interpret fans of recent sediment on
linear north-south submarine volcanic structures, the Ninety the floor of the trench as formed by material slumping down the
East Ridge and the Investigator Ridge, considered to be based of the steep lower face of the accretionary complex. These fans
on oceanic transform faults, rise several kilometres above the impede the flow of sediments along the trench axis.
general level of the ocean floor (see Fig. 1.2). The basaltic crust Seismic profiling of the trench shows that the trench sediments
of the Indian Ocean, which is here of Cretaceous to Eocene age and the underlying turbidites are uplifted along thrust faults at the
(Sclater & Fisher 1974; Liu et al. 1983) (Fig. 13.2), is overlain toe of the accretionary complex. The ridges on the face of the
first by Cretaceous-Eocene pelagic sediments and then by accretionary complex are formed by successive anticlinal folds
Miocene turbidites. At the northern end of Sumatra the turbidites of ocean floor sediments, broken by faults and converted into

175
176 CHAPTER 13

I I I I I I
96~ 98 ~ 100 ~ 102 ~ 104 ~ 106 ~
6~_
9~,,wB~nda Aceh A Active Volcano

Sumatran Fault System

FOREARC BACKARC
.4 ~
Deformation Front of the 4~_
Sumatran Subduction
&-40:Medan Complex
Fig.13.4(a)
--,'.',LakeTob~
[~~lue ~L~~ ~. 0 100 200 300 400 500km

- '
2 ~_

2~ % %
Banyak Is .
Nias~ iiiiiiiiiiiiiiiii iiiiiiiiiii
~iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii:i:..ii~
:::::::::::::::::::::::::::::::::::::::::::::::::

;:OiPekanbal
_0 o ,~ 9...-o ~ 0~
l i n e s of sections in :!!!iii!i!i!i!iiii!iiii~
:.:.:.;.:.:.:
F i g s 1 3 . 6 ( a ) & (b).
TigalJiJ/oh
lines of seismic
p r o f i l e s in :!:i:i:i:i:i:i 9 "'i:?:i:?:?:i:i:?:?:?:?:?:i:?:::
F i g s 1 3 . 3 ( a ) & (c)

2~ 2~-
Paga~
INDIAN
PLATE ~i:i:i:i:i:!:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:iOiPalembang
, Sipor~

~5 Bengkul
_4 ~
tiiiiiiiiii iiiiiiiiiliiiiiiiii!iiii
Quaternary-Recent volcanics

Tertiary Sediments and volcanics % Enggano

604
I Pre-Tertiary Basement
96~ 98 ~ 100 ~ ~ 102 ~ 104 ~ 106 ~
I I I \ I I I
Fig. 13.1. Structural map of Sumatra.

thrust slices which are underthrust and uplifted by the formation of of 19 km beneath the outer arc ridge, and at 21 km beneath the
later thrust slices as the incoming plate passes down the subduc- forearc basin, 80 km landwards of the deformation front (Kopp
tion zone (Stevens & Moore 1985) (Fig. 13.3). As is the case in et al. 2001; Schldter et al. 2002). Seismic profiles obtained by
other accretionary complexes (e.g. Barbados, the Makran and the Shell Company to the south of Java show steps in the basement
the Nankai Trough etc.) the steep (c. 35 ~ dip of the faults which suggest that the decollement at the base of the sediments
seen near the surface flatten out at depth into bedding-parallel sometimes extends down into the basement, and that slices of
decollement surfaces in the pelagic sediments a short distance oceanic crust have been uplifted into the base of the accretionary
above the oceanic basement (Moore & Curray 1980, Fig. 7). complex (Hamilton 1979). Karig et al. (1980) reached the same
On the SIO profiles the oceanic basement of the Indian Plate conclusion for the accretionary complex off Sumatra, as melanges
can be traced landwards beneath the accretionary complex for a on the outer arc ridge in the island of Nias contain blocks of
distance of 25 km (Moore & Curray 1980). In the R / V S o n n e pro- serpentinite, pillow basalt and pelagic sediments derived from
files to the south of Enggano the surface of the downgoing slab the oceanic basement.
dips at 3 ~ at the deformation front, increasing to 5 ~ beneath the Troughs on the face of the accretionary complex become
outer arc ridge. The depth to the surface of the subducting plate broader towards the upper, flatter part of the slope. Seismic pro-
was determined along the strike by seismic refraction at a depth files show that these troughs are slope basins filled by sediments
 STRUCTURE AND STRUCTURAL HISTORY 177

I I I I Islands and F o r e a r c B a s i n s
100 ~

Sumatran Forearc 1
o
.
--. A
Nicobar SimeulueNiasSiberut
tsA / ' k / ~ A /'~ ~c,
Enggano
r~ sea level

1
N 2
km 3
4
5
,, | ~ 'N | :} :N
6
BANDAACEH 7
8 I I ' ~
o 1ooo 2000 3000 4000km
Sunda Trench
3
4
5
6
km
7
MEDAN
8

0 1000 2000 3000 4000km

Paleocenei
BOLGA

Batu
b Islands

PURA

% BENGKULU
Paleocene

Cretaceous

0 100 200 300 400 500km

Fig, 13.2. Structural map of the Sumatran Forearc based on Hamilton (1979), with transform faults, magnetic anomalies and age of oceanic crust (double lines at 45 Ma
mark an extinct spreading ridge) in the Indian Ocean from Sclater & Fisher (1974) and Liu et al. (1983); structures in the forearc from Izart et al. (1994), Matson & Moore
(1992) and Diament et al. (1992); structures in the Nias Basin from Matson & Moore (1992), normal faults with ticks, and monoclinal flexures with triangles, indicating
the downthrown sides. Onland extensions of the forearc basins are shown in white. The inset shows topographic and bathymetric profiles parallel to the arc system through
the forearc islands, the forearc basins and the Sunda subduction trench after Moore et al. (1982).
178 CHAPTER 13

iii .~ 9

LU
r +tlllil ++ Z
~-~o

~cEi > o -~
<:1311
~E
Q.)

g--
rr" 9

C,,3
'L
<E
Q)

+_,

/(( /
0 O'Boo
rr
,,<
ILl .~
9 ._. ~2,
=9 ~-'.5 .~
rr
0 d3
L.I_
0e ,~ .-=..~=
Z
,< 03,-
X O 04 03 ,~-
rr O
I-- O. eLU!I-~eN~-O~I s p u o o e s .~.- b
,<
:E E
0 9~ Z ""
:D 0
>,
Seconds two-way-time ".~
ILl c- C~ CO ~ L~ r.D ~_~
"-r O
i
' i , ; /,'I ~// .....
I I I I ~~ "~ B
I.-- I/ll,~lllr+,,+, llh .....
bA
O0 0
E
CO 0
< ~i'11t/1'".,,
0 B
fr-
O

O0 t-
kt~i,t,,I i 1. h i,'lt, 'l' (~)) (~) .~ ~ 7,
00 >
O . ~,,-;i~,,,,~ "~C::~'r~
Z
-r
0 c O
0 o
I-- Z 0 '-,.-
t-
tel
00rn
~.~
0 UJ r '~ o

L.L.I rr o (3
.~2
r~ I-- O & =
o
~g
<1: c-
0 "~=
$Tr,',/a]/
Z e-
E :D e- ;_~
"...~ ' / t,, ,I
03 0

o 6 o
/
~-- 04
I
03 ~
I I
v E I & CO
mm q~,dec] JeleAA
 STRUCTURE AND STRUCTURAL HISTORY 179

(Moore et al. 1980a; Karig et al. 1980). While more recent sedi- accretionary complex from the SIO seismic reflection profiles
ments in these slope basins are sub-horizontal, older sediments (Fig. 13.3), completed a series of traverses across the outer arc
are tilted landwards, and deeper in the basins are increasingly ridge where it is exposed onland Nias, in collaboration with the
folded and more highly deformed and disrupted by thrusts, Indonesian National Institute of Geology and Mining (Moore
suggesting that the accretionary complex is under compression et al. 1980a; Moore & Karig 1980) as part of the Sumatran Trans-
and that imbricate thrusts in the accretionary complex are continu- ect of the Studies in East Asian Tectonics a n d Resources
ally re-activated to deform the sediments in the basins (Stevens & (SEATAR) Programme (CCOP-IOC 1981).
Moore 1985) (Fig. 13.3). Karig et al. (1980) suggest that the The melange deposits, described as the Oyo Complex, were
greater part of the sediments in the slope basins are derived found to occur as linear belts several hundred metres wide and
locally by slumping of soft sediment from the face of active several kilometres long, parallel to the N W - S E trend of the
fault scarps, rather than from erosion on the island of Nias island (Fig. 13.4). The melange alternates with belts of bedded
higher up the slope. sediment, described as the Nias Beds. The older sediments
In the seismic sections to the south of Enggano Kopp et al. within the bedded succession are turbidites, which coarsen and
(2001) make a distinction between the active accretionary thicken upwards. The oldest part of the succession lacks calcar-
complex with low seismic velocities, indicating that it is formed eous microfossils, interpreted as due to deposition below the
of unconsolidated sediments, and an older accretionary complex CCD (carbonate compensation depth). Both the age and depth of
forming the outer arc ridge. The older complex, while still com- deposition of the younger units were determined by their con-
posed predominantly of sediments, is more highly consolidated tained microfossils. It was found that Lower Miocene sediments
and has higher velocities. They suggest that the older complex were deposited at bathyal-abyssal depths > 2 0 0 0 m , Upper
is of Palaeogene age and acted as a backstop to the younger Miocene at depths of 2 0 0 0 - 5 0 0 m , while Pliocene deposits
Neogene to Quaternary complex (Kopp & Kukowski 2003). were accumulated on the continental shelf at < 5 0 0 m, and
Pleistocene deposits were formed near sea level in a reef environ-
ment. The bedded units were folded and faulted contempora-
neously with their deposition, with the older units being more
Outer arc islands highly deformed than the younger units.
From their study of the offshore seismic data Karig et al. (1979)
The accretionary complex rises steeply from the floor of the trench developed a model to account for the evolution of the accretionary
to form an outer arc ridge, c. 120kin wide (Fig. 13.3a) which complex, the development of the forearc ridge and the geology
appears above sea level in a chain of islands to the west of and structure of Nias (Fig. 13.5a, b). The Oyo m61ange was inter-
Sumatra. In the north the ridge rises 5.5 km from the floor of the preted as a trench-fill deposit composed of fragments of ocean
trench to the island of Simeulue, and in the south for c. 6.5 km crust and turbidites that had slumped down the inner trench
in Enggano (Fig. 13.2). slope and were accreted into the base of the accretionary
During the 1980s the geologists of the Indonesian Geological complex (Moore & Karig 1980). The chaotic and sheared nature
Research and Development Centre (GRDC) mapped most of the of the m61ange was considered to be due to the dynamic tectonic
outer arc islands using aerial photographic interpretation and environment within the accretionary complex, in which the
field traverses. The resulting maps were subsequently modified original thrust surfaces were continually reactivated and new
in the 1990s by interpretation of SAR (synthetic aperture radar) thrust planes developed, disrupting the oceanic basement and
imagery, supplemented by additional field checking. These geolo- breaking it up into blocks. The oceanic basement material was
gical maps were published by GRDC at the scale of 1:250 000 continually uplifted into the accretionary complex along the
(Simeulue and the Banyak Islands--Endharto & Sukido 1991 developing thrusts. The age of the m61ange was not determined
(Fig. 13.4); Nias--Djamal et al. 1994 (Fig. 13.4); Batu directly, but the youngest blocks incorporated in the melange
Islands--Nas & Supandjono 1994; Pagai and Sipora--Budhitrisna appeared to be the Eocene limestones, so that the m41ange was
& Andi Mangga 1990; Siberut--Andi Mangga et al. 1994b; considered to be of Eocene age. No stratigraphic contacts were
Enggano--Amin et al. 1994a) (Fig. 13.4). seen between the Oyo Complex m61ange and the bedded units,
Most of the islands show restricted outcrops of melange, with but the m41ange was considered to be the oldest unit, forming a
blocks of serpentinite, gabbro, basalt, chert, calcilutite and rare basement to the overlying bedded Nias Beds. The two belts
limestones with large foraminifers, Nummulites, Discocyclina of sediments on Nias were found to be broadly synformal but
and Pellatispira of Eocene age (Douville 1912; Budhitrisna & complicated by faults and small scale folds.
Andi Mangga 1990), and granitic and metamorphic rocks, amphi- Although no depositional contacts between the basement and
bolites, schists, phyllites and slates, together with abundant the sediments were seen, Moore et al. (1980a) and Moore &
greywacke, sandstone, shales and claystone, in a sheared scaly Karig (1980) suggested that on the southwestern margins of the
clay matrix, in addition to the chaotic melange there are also basins the sediments were deposited unconformably on the under-
more extensive oucrops of bedded units composed of sandstones, lying m~lange, as near these contacts the m~lange is highly
siltstones and clays, often tuffaceous, peats and coals, the latter sheared, while the Nias Beds are only fractured. On the other
indicating mangrove swamps, marls and limestones with abundant hand the northeastern boundaries of the basins were found to be
benthonic and planktonic microfossils, indicating abyssal to tectonic, with the Nias Beds sheared and mixed with the
sublittoral environments of deposition. Microfossils show that m61ange along the contact. Along strike the melange was observed
the sediments range from Late Oligocene-Early Miocene to to be in contact with different units of the Nias Beds. These con-
Pliocene in age. These older units are generally folded, faulted tacts were interpreted as high angle reverse faults. Near the contact
and thrust and are overlain unconformably by reef limestones the Nias Beds are folded into tight asymmetric synclines on
and associated reef debris of Plio-Pleistocene age. The islands N E - S W axes with NE dipping axial planes and SW vergence.
are surrounded by modem mangrove swamps and coral reefs. The fold axes plunge at low angles either to the NE or SW, and
In many areas, particularly on the northeastern coasts of the cannot be traced for more than a short distance along strike.
islands, drowned mangroves indicate recent subsidence, and on In some examples the hinges of the folds have been sheared out
southwest facing coasts raised reefs indicate recent uplift. along small scale reverse faults. Moore & Karig (1980) report
that the older strata are more highly deformed than the younger
The Karig model (Figs 13.5a, b and 13.6a). The most intensively units.
studied of the outer arc islands is the island of Nias. Karig et al. Moore et al. (1980a) and Moore & Karig (1980) interpret their
(1980), having made a detailed study of the trench and the observations of the geology and structure of Nias in terms of
180 CHAPTER 13

I 1
95~ 96o00 , B A N Y A K ISLAN~DS 97o15, 2~

reefs (a) SIMEULUE

SIBOGO
TUANGKU

reefs

oooTp,~,~ \ 97o15 '

! 2o00'N ~
I

Fault
- Lineament
Sibau ~ Thrust
Gabbro ~ Anticline
Group "-.X--- Syncline
(40Ma)

Quaternary reef and alluvium ~efs 2~ '-

.2030 `

i Melange (blocks, often ophiolitic,

in clay - may be diapiric)
Dihit Formation ) Late Miocene-Pliocene
Layabaung Formation )
Sibogo Formation ) Middle Miocene
Sigulai Formation )
Lasikin Member (Early Miocene)
0 5 10 15 20km
} ~ reef,~
96o00'E 96015 '
I I

I I
1 ~"X_~ Gomo r,,.

II" ~
97o15'E 97o30 '
~ b a s i n
(b) NIAS

B I~30'N 1"30'

CD

07,s, "',/
GUNUNG-
SITOLI 1 ~Sub- klrf I~'1-k /
1o15 '

Fault
- - Lineament
Thrust
"~ Anticline
Syncline
-- 1~ l~ '-

Mud volcanoes ~ ~ B
Alluvium (3 CZ)d~ o B Fig. 13.4. Geological and structural maps of
9
i Melange (blocks in clay % Simuelue, the Banyak Islands and Nias.
Simeulue and the Banyak Islands are after
~ GunungsitoliFormationiPlio-Pleistocene) B Endharto & Sukido ( 1994); the location of
Gomo Formation 'Nias Beds' (Early Miocene-Early Pliocene) the ophiolite complex on Bangkaru and the
Lelematua Formation age of the Sibau Gabbro are from Kallagher
Conglomerate Member (Late Oligocene-Early Miocene) (1990). The geology of Nias is based on
Ophiolitic basement (B) Djamal et al. (1994) but the maps are after
Samuel (1994) and Samuel & Harbury
(1996). The inset map of Nias showing the
0 10 20kin sedimentary sub-basins, faults and
I reefs TELUKDALAM lineaments and the interpretative strain
97o15'E 97030 ' 97045 , reefs ellipse, comes from Samuel & Harbury
I I I (1996).
 STRUCTURE AND STRUCTURAL HISTORY 181

MODELS FOR THE EVOLUTION OF THE SUMATRAN

ACCRETIONARY COMPLEX AT NIAS ISLAND

a. EARLY MIOCENE (Karig et al. 1980 Model)

Melange formed Slope basins filled with sediment from Sumatra
Ocean floor with thin by slumping into continually reactivated by thrusts on landward
pelagic sediments trench and tectonic side with slumped sediment; earlier sediments
deformation in toe compressed and deformed
of accretionary complex
J

Trench ~ ~ ~
flexure
9~ ~ l u m p ~ / ~ - ~ \ \ \ ~ ",. . . . . . Rifted Sumatran
~ ~ Palaeogene Continental basement
INDIAN PLATE

b. LATE MIOCENE TO RECENT (Karig et al. 1980 Model)

Uplift of accretionary complex (melange)due to accretion of thick sediments
Slope basins with shallowing upwards sequence - from below CCD to
Pleistocene reefs and recent exposure.
Thick sediments Slope basins overthrust on landward side
of the Nicobar Fan Forearc Basin
on oceanic crust NIAS Pleistocene subsidence

flexure ~ ~ - ~ \ 1 - ~r 4 ~ . \ ~ '" / - ~~ \ \ ~/ - ~~- - -~- - ~ -

INDIAN P L A T E ~ ~ h m ~ ~ . ~ U~lift flexure

c. LATE MIOCENE TO RECENT (Samuel et al. 1997 Model)

Accretion of Nicobar Fan sediments caused uplift and extension
of the accretion complex forming rift basins in the old accretionary
complex filled with sediment from the Sumatran Shelf until the
Toe of Accretionary Pleistocene subsidence of the forearc basin.
Complex advances Melange formed by shale diapirism with mud volcanoes.
over incoming plate< NIAS Pleistocene subsidence

flexure

Fig. 13.5. Comparison of models for

INDIAN P L A T E ~ o m p l e xComplex evolution of the Sumatran Forearc base on
studies on Nias Island (a) & (b) by Karig
et al. (1980) and (e) by Samuel et al. (1997).

the model derived from the study of the seismic sections Moore & Karig (1980) predict that if the SW margins of the
(Karig et al. 1979) (Fig. 13.5a, b). They suggest that the bedded basins were exposed they would show the original unconformable
units were deposited on top of the accretionary complex in two relationships between the melange and the Nias Beds. With
slope basins developed on the lower trench slope. At this stage, continual accretion the contacts and the layering in the overlying
in the Early Miocene, the oceanic plate had only a thin sedi- sediments were rotated to give their present steep angles of dip.
mentary cover, so that the sediments were deposited directly on On the other hand the NE margins of the basins are steep
the oceanic basement. In the Late Miocene, when there was a reverse faults along which the basement has been uplifted, com-
greater thickness of Nicobar Fan sediments on the oceanic plate, pressing and folding the bedded sediments in the intervening sedi-
the slope basins were uplifted as new material was accreted to mentary basins. The reverse faults were continually reactivated
the base of the slope. At the initiation of a slope basin, near the during the deposition of the sediments, so that older units are
base of the slope, the sediments were deposited below the CCD, more highly deformed (Moore & Karig 1980).
but as uplift continued sediments were deposited at progressively
shallower depths, until the youngest deposits on Nias are uplifted The Samuel model (Figs 13.5c and 13.6b). In the 1980s and early
coral reefs resting on the older slope basin sediments (Moore et al. 1990s University of London Group for Geological Research in
1980a). Southeast Asia, in collaboration with the Indonesian Research
182 C H A P T E R 13

CROSS SECTIONS OF THE SUMATRAN FOREARC Volcanic

SW Sunda ACCRETIONARY COMPLEX NE

Trench Inner Trench Slope Forearc Ridge Forearc Basin coast
km
with slope basins NIAS Musala line . ~ ~ Arc km2
~ __ _ '~'~[~'~.+'+: ~-.-. .... ~ ~ . . " - ~ ~ . . , _ ..,_~'..,.. ~-"~,.-Y N-"- -" I o

8~ - - ~ - ~ _ _ _ . ~ ~ - - - - - ~ ~ earlier accretionary
lOJ ~ complexes 0 10 2.0 30 40 50kin

(a) Karig Model

SW NE
Volcanic Arc
Forearc ridge and
Sunda Sumatran Fault Zone
Indian Ocean Trench Accretionary wedge NIAS Forearc Basin km
km
0 ~ ~ ~ : " ' ~ ' . ' T " 7 ' . ~ .. ~ . .. ,-" .-" ,," .," ,-" ,," ")' . .F . . .~ ~ 20

t ~
~ ~
\ ~
x ~
\ /%qe-"'~/
/ "y'7 ~.;,<....---..~-.----~
\ N.
., ,.., ,. ,. -, ,. ,, ,. -, ,..,i,, ,~,.-,.,o
". ", "- "- % "- ", ", ", ,~ "- % ", ",/ % "4 I/,, ",,
~ ~ ~ ~ ~ ~ Crust of . ,,"..". , ..'-". , .-"," . ..-"z . ...",". . .,"..' . ..."," .,."," ....."," ~-,"'..~'.'~"_-.",.[/~.q
-"
.-" .-" I .."
v' ," ," z
I' "
," I,." .,"
'

1 8 ~ ~ ~ ~ ~ ~ ~ ~ unknown nature ............ . , . , . , 4~.~..,'.'~".',.'i.,

% % ",. "., ",, ",, ", '., ",. ",, % ",. % ",,,' I,. % t,, "., "..,
~ "-, "., "., ",. % ",, '-, "., ",,, "-, "., "., ",. ",C ",, % "',, % "-,

"% % % % % % % ~. "- % % "- X% %1% %'

"~ ~.,/%~.%/%/%]%.~%.~....~ ~. ~. ~- .., ~. ,a ,,, ,,

9,,,..,, ,.. ,., ,,, .,, % *,. ,., ~',,~ ~.~ ",, % ",.
-,~,~ ,., ,. .,- .,, ..e ,,, ,," Jt I.," ,-' ,,."

0 10 20 30 40 50km 44,, "-" ." ". x. ,:,r<.~.'.".,,'~,r

I i I ! I I
O+o"~,'.,?.,:.f~,,.:~.~
i,
i~,~?
~ ,,,% ",, "=~ "., "q ",. ",. "..
01/i,,
, ~ "~,~."x~,"... "
u O,~"L",",'-.~".","

(b) Samuel Model

Fig. 13.6. Interpretative cross-sections of the Sumatran Forearc from the Indian Ocean through Nias to the volcanic arc on the mainland of Sumatra. (a) Karig model. In
this model the sedimcntary basins on Nias are considered to have developed as as slope basins on the inner trench slope and to be overthrust on their northeastern sides by
slices of accreted oceanic basement (after Karig et al. 1979). (b) Samuel model. In this model the sedimentary basins on Nias are considered to have originated as half
graben due extension of the forearc; thrusts occurred subsequently due to inversion of the bounding normal ['aults (after Samuel & Harbury 1996) N.B.The vertical scale in
(b) has been increased so that it is the same as in (a) for easier comparison of the two models.

and Development Centre for Oil and Gas Technology (LEMIGAS), igneous minerals have been replaced by low-grade metamorphic
studied several of the forearc islands. Reconnaissance visits were minerals, including pumpelleyite. A high positive Bouguer
made to Simeulue and the Banyak Islands (Situmorang et al. 1987; gravity anomaly over the gabbro and elsewhere in Simeulue
Kallagher 1989; Harbury & Kallagher 1991) (Fig. 13.4) and the suggest that the island is underlain by a substantial slab of
Batu Islands and Siberut (Barber et al. 1992). oceanic basement (Milsom et al. 1990). Samples of gabbro from
As part of this study the island of of Nias was remapped in detail Simeulue were analysed by Kallagher (1989, 1990) and showed
by Samuel (1994) and Samuel et al. (1995, 1997) with tight an enriched MORB-type geochemistry and gave K - A t ages
stratigraphic control provided by microfossils. The oldest unit of 35.4 4-3.6Ma and 40.1 _+ 2.7Ma (Late Eocene-Early
which can distinguished on Nias, as well as on the other forearc Oligocene). Kallagher (1989, 1990) did not consider that these
islands, is a basement unit, the Bangkaru Ophiolite Complex, ages represented the age of the original ocean floor since the pre-
named from one of the Banyak islands (Fig. 13.4). This consists sence of pumpelleyite indicates that the gabbro had been affected
of ocean floor material occurring either as a coherent unit, or as by a low-grade metamorphic event. She suggests that K - A r age
blocks in the melange. Rock types include serpentinised peri- indicates the time of metamorphism during subduction/accretion
dotites, gabbros, diorites and plagiogranites, dolerites, basalts, (Kallagher 1989, 1990).
generally showing pillows, pillow breccias and hyaloclastites, Scattered outcrops of coherent Bangkuru Ophiolite Complex
garnet amphibolite (reported by Moore & Karig 1980), hornblende were mapped by Samuel et al. (1995) on the SW coast of Nias,
gneiss and hornblende schist, palagonite tufts, foraminiferal and at one location on the north coast ('B' in Fig. 13.4b). On
limestones, banded cherts, ochres, greywackes and quartz and Sigata coarse gabbro, cut by gabbro pegmatite and black dolerite
barroisite schists (Samuel et al. 1997, Table 1). dykes and veined by epidote, crops out on the southern shore of
In the islands mapped or visited by the University of London the island below the lighthouse. The lack of a positive gravity
Group coherent units outcrop on Bangkuru in the Banyak anomaly over this island (Milsom et al. 1990) suggests that the
Islands, Simuelue, Nias and Sigata and Barogang in the Batu gabbro could be a large block in the m61ange. On Barogang
Islands. On Bangkuru serpentinite and gabbro outcrop in the diorite containing fine-grained dark xenoliths crops out on the
hanging wall of a reverse fault with a strong shear fabric parallel northern side of the island, and the same rocks, highly brecciated,
to the fault. On Simuelue the Sibau Gabbro Group crops out on form a small offshore islet (Barber et al. 1992).
the NE coast towards the southern end of the island (Fig. 13.4). Much more commonly, components of the Bangkaru Ophiolite
It is composed of coarse and fine metagabbro in which the Complex occur throughout the forearc islands as clasts in the
 STRUCTURE AND STRUCTURALHISTORY 183

mdlange, ranging in size from centimetres to more than 30 m The origin of the mdlange. As part of his study Samuel (1994) and
(Samuel 1994). The ultramafic and basic rocks, serpentinites, Samuel et al. (1997, Table 2) made a systematic study of the
gabbros, dolerites, pillow basalts and hyaloclastites, occurring as mdlange and its relationships to the bedded units. They found,
blocks in the m61ange, are compatible with an origin as part of that in addition to the ophiolitic components, at least 50% and
an ocean floor assemblage and show the effects of low grade commonly 90% of the clasts in the mdlange were derived from
ocean floor metamorphism, such as occurs in the region of a the Oligocene and Lower Miocene units, while some outcrops
spreading ridge. In central Nias blocks of serpentinite in the also include clasts of the Middle Miocene to Pleistocene units.
m61ange with foliation and linear structures show that they had Very commonly the clasts showed the same sedimentary and
been subject to ductile deformation under high temperature structural features as seen in adjacent bedded units. It was found
conditions in a mid-ocean ridge environment, possibly in a trans- that the mud matrix has the same mineral composition, contains
form fault zone, prior to serpentinization. Hornblende gneiss the same microfossils and also shows the same thermal history,
and schist blocks may have a similar origin. However, some with the same range of vitrinite values, as mudstones in the
metasedimentary rocks, analogous to the metagreywackes of Oligocene to Lower Miocene succession. The scaly foliation
Moore & Karig (1980) and Kallagher (1989), are reported to which pervades the matrix is commonly vertical, but may be
contain prehnite, and an unusual rock composed of quartz, folded, wrapping around the clasts, and is parallel to the margins
pyrite and an amphibole identified as barroisite (Samuel 1994; of the m61ange outcrops. Contacts between the m61ange and the
Samuel et al. 1997), suggests that ocean-floor sediments had bedded units are always intrusive, with the matrix penetrating
been subducted. along the bedding planes and fractures in the bedded units.
The pelagic limestones, bedded cherts and ochres found as M61ange is found cutting bedded units of all ages from Oligocene
blocks in the m61ange are also compatible with an ocean floor to Recent. It appears that the major period of m61ange formation
origin. A sample of pelagic chert from central Nias was found to occurred during the Pliocene, but m61ange formation on Nias
contain foraminifers of Campanian (Late Cretaceous) age. While continues to the present day, as indicated by the eruption of mud
samples of bedded red chert yielded radiolaria of Mid-Eocene volcanoes extruding blocks and a grey mud slurry identical to
age (Samuel 1994; Samuel et al. 1997). These ocean floor sedi- the clay matrix of the m61ange (Figs 13.4 and 13.5c).
ments are compatible with the age of the ocean floor that As a result of this study Samuel (1994) and Samuel et al. (1995,
has been subducted beneath Sumatra, as deduced from Indian 1997) concluded that the mdlange was the product of shale diapir-
Ocean spreading history indicated by magnetic anomaly patterns ism and not due to the tectonic disruption of trench fill sediments;
(Sclater & Fisher 1975) (Fig. 13.2). it, therefore, does not constitute the basement upon which the
Samuel (1994) and Samuel et al. (1997) subdivided the Nias bedded units were unconformably deposited, as was proposed by
Beds of Moore et al. (1980a) into six units which could be corre- Karig et al. (1979). The evidence suggests that Oligocene and
lated across the island. The oldest sediments (Oyo Formation) Lower Miocene deep-marine muds near the base of the bedded
are thick-bedded, massive, micaceous sandstones. The unit is succession were periodically mobilized to intrude the Bangkaru
highly disrupted, so that coherent successions are rare, but this Ophiolite Complex and the overlying bedded sediments, incorpor-
lithology commonly occurs as blocks in the m61ange. Early ating blocks of these units into the m61ange matrix.
Oligocene to earliest Miocene fossils were obtained from this Mdlange mapped on the other outer arc islands also contains
unit, but Samuel (1994) considers that the Early Oligocene ophiolitic and sedimentary clasts in a scaly c l a y - m u d matrix
fossils were reworked, like the limestone clast with N u m m u l i t e s similar to those recorded on Nias. Circular outcrops of m61ange
of Eocene age found in a conglomerate (Douville 1912). Samuel and the active mud volcanoes mapped on Simeulue, Siberut,
(1994) and Samuel et al. (1997) conclude that the Oyo Formation Sipora and Pagai (Endharto & Sukido 1994; Andi Mangga et al.
is of Mid-Oligocene-earliest Miocene age and was deposited as 1994b; Budhitrisna & Andi Mangga 1990) suggest that the diapiric
turbidites in a deep marine setting below the CCD, as suggested mechanism is responsible for occurrences of m~lange in all the
by Moore et al. (1980a). The Oyo Formation is overlain conform- outer arc islands.
ably by the Gawo Formation of Early to Late Miocene age
which has similar characters, but is thinner bedded and finer Structural evolution of the Forearc ridge. From mapping and strati-
grained. Samuel (1994) and Samuel et al. (1997) were unable to graphical study of the bedded sediments of Nias, Samuel (1994)
confirm the conclusion of Moore et al. (1980a) that the 'Nias and Samuel et al. (1997) recognized three sedimentary sub-
Beds' show a coarsening upwards succession. basins, including a basin in the NW of the island, the Lahewa
The Gawo Formation is overlain by the sandstones and mud- sub-basin, in an area that had not been visited by Moore et al.
stones of the Olodano and Lahomie formations of Early to (1980a) (Fig. 13.4, inset). It was found that the earliest structural
Mid-Miocene age, which also include coral-algal limestone features in the sediments were syn-depositional extensional faults.
units formed as carbonate build-ups in a shallow marine environ- Samuel et al. (1995) therefore suggest that although the basins
ment. The progressive uplift with a shallowing upward sequence, on Nias were formed on top of the accretionary complex, they
from lower to upper bathyal and then sublittoral identified by developed during a phase of extension and are bounded to the
Moore et al. (1980a), was confirmed by the study of the benthic NE by major extensional faults (Fig. 13.5r The increase in
foraminifera (Samuel et al. 1997). The transition from the deep thickness of the bedded succession towards the northeastern
water facies of the Gowa Formation to the shallow-water facies margins of the basins indicates that-these margins were formed
of the Olodano Formation is highly diachronous, occurring as active growth faults during the deposition of the sediments.
in the Early Miocene in the east, but not until late in the Mid- Thrusts, sometimes bringing slices of the Bangkaru Ophiolite
Miocene in central Nias. Complex over bedded sediments, and folds are superimposed on
The conglomerates, sandstones and mudstones of the following earlier extensional features, indicating that the basins were sub-
Middle Miocene-Early Pliocene Lahomie formation indicate sub- sequently compressed. Localized inversion in the western part of
sidence, with carbonate build-ups of the Olodano Formation Nias took place in Early Miocene times and was followed by the
covered by a blanket of mudstones, indicating that they had infilling of the basins, indicated by the upward shallowing of
been drowned. Subsidence was followed by uplift and erosion the depositional environments. Subsidence was renewed in
during the Pliocene as the Late Pliocene to Recent Tetehosi (sili- M i d - E a r l y Pliocene times, but was followed again by widespread
ciclastic) and Gunungsitoli (reef limestone) Formations rest inversion with deformation during the Pliocene. The major bound-
unconformably on the older units. Very recent uplift is confirmed ing faults to the sedimentary basins have been reactivated
by C 14 dating of 800 year old raised reefs of the Gunungsitoli as thrusts during inversion. The alternations of extension and
Formation (Vita-Finzi & Situmorang 1989). subsidence, compression and uplift in the sedimentary basins are
184 CHAPTER 13

attributed to variations in the rates of convergence at the subduc- Diament et al. (1992) carried out a seismic survey of the zone of
tion zone, and the effects of transpression and transtension along disturbance. Their profiles show an uplifted block with a complex
transcurrent faults due to the oblique subduction. Matson & pattern of horsts and grabens, bounded on both sides by normal
Moore (1992) proposed another model for the development of faults, with downthrown forearc ridge sediments on one side,
the Sumatra forearc in which the accretion of massive quantities and downthrown forearc basin sediments on the other. This struc-
of sediment from the Nicobar Fan in the late Mid-Miocene led ture was interpreted as a positive flower structure, and together
to the depression of the incoming Indian Ocean Plate during the with the straight trace of the fault over several hundred kilometres,
Mid-Miocene to Early Pliocene, causing the subsidence of led Diament et al. (1992) to suggest that the zone of disturbance
the forearc ridge and its sedimentary basins, as recorded in the was a major transcurrent fault, the Mentawai Fault, named after
stratigraphic sequence (Samuel 1994). the outer arc archipelago (Fig. 13.2). Diament et al. (1992) went
Clasts in the m~lange on Nias indicate that the island is under- on to suggest that the Sumatran forearc was dissected into
lain by upper mantle, oceanic crust and pelagic sediments derived several narrow fault slivers along strike-slip faults, parallel to
from the Indian Ocean floor and built up into the accretionary the main Sumatran Fault on the mainland. They suggest that
complex. The only components of the m61ange which do not fit these fault slivers are being displaced differentially northwards
this model are garnet amphibolite and barroisite schist clasts in response to the oblique subduction of the Indian Ocean Plate.
reported by Moore et al. (1980a), Samuel (1994, 1997) and They further suggest that in Nias the Mentawai Fault passed by
Samuel et al. (1997). A greater variety of clasts is reported fom way of the Batee Fault into the main strand of the Sumatran
the islands of Pagai and Sipora (Budhitrisna & Andi Mangga Fault in northern Sumatra. North of Simeulue Izart et al. (1994)
1990). In addition to ophiolitic rocks and garnet amphibolite, found several faults in the accretionary complex to the west of
clasts include garnetiferous mica schist, and granodiorite with the trace of the Mentawai Fault and suggests that the fault
biotite, and granitic gneiss with orthoclase and muscovite, breaks up into a horsetail pattern of subsidiary faults at its northern
suggesting that continental basement rocks underlie the eastern end. Further north, opposite Banda Aceh, they found that the
parts of some of the outer arc islands. main trace of the fault was replaced by an easterly directed
Samuel et al. (1995) found that the Upper Palaeogene and thrust fault, the West Andaman Fault (Fig. 13.2).
Neogene stratigraphic sequences and lithologies in the Banyak By careful mapping and age determinations in eastern Nias,
and Batu islands, which lie within the forearc basin, and from bore- Samuel & Harbury (1996) found that the western limb of
holes in the forearc basin itself, resemble the stratigraphy and the monoclinal flexure consists of 5 km of easterly dipping
lithology of the same units on Nias. As will be discussed below, Oligo-Miocene sediments. Seismic sections to the east of the
there is evidence that the forearc basin, which separates the outer flexure on eastern Nias, and in the offshore area, show that
arc islands from the mainland of Sumatra, has developed relatively about 3 km of the Oligo-Miocene sediments seen to the west of
recently. Samuel et al. (1995) suggest that prior to the Pleistocene, the flexure are absent, and that to the east Upper Miocene sedi-
sedimentation was continuous across the present forearc basin to ments rest unconformably on the forearc basement in the Mola
the outer arc islands. The common occurrence of well-rounded basement high. They, therefore, concluded that the flexure
quartzose and metamorphic clasts in the Oligocene and Lower passed at depth into a major extensional fault, the boundary
Miocene sandstones and conglomerates on Nias, indicate that the fault of a complex half-graben, rather than a thrust. Thrust features
greater part of these sediments were derived from a mature conti- seen in the rocks at the surface are attributed to the effects of Late
nental provenance. Samuel et al. (1995) suggest that sediments Pliocene inversion.
were eroded from basement uplifts in the forearc region or were Samuel & Harbury (1996) also studied the fold traces and
transported across the site of the present forearc basin from lineaments seen in the SAR (synthetic aperture radar) imagery
the mainland of Sumatra, to be deposited in extensional basins of Nias. Major anticlinal and synclinal traces and the dominant
developed on top of the accretionary complex at the continental N N W - S S E lineaments run sub-parallel to the length of the
margin. Some conglomerates, however, contain locally derived island (Fig. 13.4). The dominant lineaments are faults, bounding
ophiolite clasts, and coarse breccias, composed of large angular the sedimentary basins, which Samuel & Harbury (1996) consider
ophiolite and siltstone clasts, are interpreted as rock falls from have been reactivated as thrusts during later inversion (see
active fault scarps, indicating that slices of the oceanic basement Fig. 13.4b, inset). Contrary to the suggestion of Diament et al.
within the accretionary complex were being uplifted and eroded (1992) evidence of strike-slip movement has not been seen in
along the basin margins during sedimentation, as suggested also outcrop in faults with this N N W - S S E orientation. However, com-
in the model of Moore et al. (1980a). plimentary N W - S E , and approximately north-south faults
cutting across the strike of the beds do show strike-slip features
Monoclinal flexure and the M e n t a w a i Fault. In the eastern part of in outcrop and are interpreted as conjugate shears (Fig. 13.4b,
Nias Moore & Karig (1980) mapped steeply dipping or overturned inset). Samuel & Harbury (1996) also recognize E N E - S W E
Nias Beds with westerly dipping shears and reverse faults in a zone lineaments, which they interpret as extensional faults, indicating
3 km wide along the eastern boundary of the easterly sedimentary that the island, and presumably the forearc as a whole, has been
basin. To the east this zone of steep dips is followed by Upper extended parallel to its length since the Pliocene. They consider
Pliocene and younger sediments with a low easterly dip. Pliocene that it is very unlikely that the Batee Fault passes into the
sediments rest on the older rocks with an angular unconformity, Mentawai Fault (cf. Diament et al. 1992), it is more likely that
but further east in the forearc basin this angular discordance disap- it is represented by one of the north-south shears.
pears. Karig et al. (1979) and Moore & Karig (1980) interpreted Similar patterns of fold traces and fault lineaments are seen on
this structure as a large 'homocline' or monoclinal flexure, the 1:250 000 Geological Maps of all the other forearc islands,
between the deformed rocks of the forearc ridge and the flat- from Enggano in the south to Simeulue in the north (Amin et al.
lying sediments of the forearc basin. The downward displacement 1994a; Budhitrisna & Andi Mangga 1990; Andi Mangga et al.
of the forearc basin sediments across the flexure was estimated at 1994b; Nas & Supandjono 1994; Djamal et al. 1994; Endharto
3 km. They suggested that the flexure was the surface expression & Sukido 1994).
of a SSW-dipping back thrust at depth, on which the accretionary
complex had been thrust over the forearc basement, which had
acted as a back stop during the development of the complex. Forearc basins
This flexure zone can be recognised in seismic reflection profiles
(Fig. 13.3a & c), and can be traced southwards as a belt of struc- Between the forearc islands and the mainland of Sumatra are a
tural disturbance to the east of the forearc islands, as far as Siberut. series of forearc basins (Fig. 13.2). At the present time the sites
 STRUCTURE AND STRUCTURAL HISTORY 185

of the basins are depressions, with the sea floor lying at depths of Oligocene dolomitic limestones, calcareous mudstones and
up to 3000 m opposite north and south Sumatra, but rising oppo- pyritic shales with steep dips up to 50 ~. These rocks are poorly
site central Sumatra, where basin sediments and forearc basement imaged in seismic profiles, but can be traced from the Sumatran
are exposed in the islands of the Banyak and Batu groups, and mainland westwards beneath the continental shelf as far as the
in islands offshore Sibolga. The area of uplift coincides with a shelf edge. The sediments are at least 2 km thick, show variable
marked bend in the subduction trench, the 'Nias elbow' of dips, are cut by faults, and occupy a trough to the northwest of
Milsom (Chapter 2) (Fig. 13.1). It is probable that this area of the Banyak Islands extending for 100 km parallel to the arc
uplift is due to the subduction of the Investigator Ridge and poss- (Beaudry & Moore 1985).
ibly the thermal and topographic perturbations caused by the In the Bengkulu Basin to the south a borehole penetrated a
extinct Wharton spreading ridge (Liu et al. 1983) as it passed sequence of ?Upper Eocene to Oligocene volcaniclastic sand-
down the subduction zone beneath this region (Malod & Kemal stones interbedded with claystones which are correlated with the
1996) (Fig. 13.2). Lahat Formation volcaniclastics exposed onland in southern
Sumatra (Hall et al. 1993). Seismic profiles indicate that these
Nature of the j~rearc basement. Hamilton (1979) in his review of sediments occupy faulted half graben, up to 6 km deep, trending
the tectonics of the Indonesian region suggested that subduction north-northeastwards and cut by NW-trending transfer faults.
complexes were developed within oceanic crust and that forearc The trend of the graben has led to the suggestion that they may
areas are underlain by segments of remnant oceanic crust attached be the continuation of similar graben of the same age to the east
to the margin of the continent. Seismic refraction studies in the of the Barisan Mountains in the Sumatran backarc area, displaced
forearc to the east of Nias showed that the forearc basement had by c. 100 km along the Sumatran Fault (Howles 1986; Hall et al.
seismic velocities between those of oceanic and continental 1993; Yulihanto et al. 1995). This correlation will be discussed in
crust, which were compatible either with continental or thickened the section on the backarc area.
oceanic crust (Kieckhefer et al. 1980). The occurrence of clasts of The Palaeogene rocks are overlain with major unconformity by
garnetiferous mica schist, garnet amphibolite, granodiorites and Lower Miocene and younger rocks (Fig. 13.3c), which on the
granitic gneisses in m61ange on the outer arc islands of Pagai mainland of Sumatra to the NW rest directly on the Palaeozoic
and Sipora (Budhitrisna & Andi Mangga 1990) suggest that con- and Mesozoic basement and to the west rest on the accretionary
tinental crust may extend as far as the outer arc ridge. Recent complex. In the Late Oligocene (29 Ma) the whole of the
seismic refraction studies during the cruise of the R / V Sonne in forearc area was exposed to subaerial erosion, probably with a
the forearc basin to the east of Enggano show that the basement landscape of significant relief, which supplied coarse sediment
in this area is of continental type (Kopp et al. 2001). to the extensional basins which were developing on the accretion-
Karig et al. (1979) concluded from the tectonic history of ary complex which formed Nias to the west (Samuel & Harbury
Sumatra, that the forearc basins were underlain by Pre-Miocene 1996). In the Early Miocene the forearc region underwent a
accretionary complexes which formed the continental margin marine transgression. In the shelf area sediments immediately
against which the present complex was accreted (Fig. 13.6a). above the unconformity are littoral sands, followed by Lower
They proposed that the original, pre-present subduction phase, Miocene siltstones with shallow water foraminifera (Beaudry &
continental margin coincided approximately with the monoclinal Moore 1985). In the Early (?), M i d - L a t e Miocene carbonates
flexure along the eastern side of the outer arc islands. The occur- were developed in the shelf area (Rose 1983).
rence of ophiolitic material in the Banyak islands within the The Barisan Mountains, to the east on the mainland of Sumatra,
forearc suggested that this margin was irregular, with oceanic were uplifted and eroded in the Late Miocene, supplying large
embayments (Karig et al. 1979). From the account of the Pre- quantities of terrigenous sediment to the forearc region (see
Tertiary geological development of Sumatra given in this Chapter 7). At the same time the forearc region itself underwent
volume (Chapters 4 & 14) the forearc basement is the western major subsidence. Prograding shallow-water clastic sediments
extension of the Bentaro-Saling Volcanic arc and the associated overwhelmed the carbonate banks and, as sediment supply
Woyla Accretionary Complex, intruded locally by Late Cretac- exceeded the rate of subsidence, built out to form a continental
eous and Tertiary granitoids and overlain by Palaeogene sediments shelf and a continental slope towards the west. Further west, in
and volcanics. the deeper part of the Nias Basin, deep water turbidites of Late
Miocene age buried earlier Upper Miocene shallow-water carbon-
Depositional history of the forearc basins. The forearc basins are ate mounds, which had been constructed directly above the uncon-
from north to south: the Aceh Basin; the Meulaboh (or Simeulue) formity in the early phase of subsidence. This pattern of
Basin; the Nias (West Sumatra, Sibolga, or Singkel) Basin; and sedimentation, with the progradation of the shelf and the depo-
the Mentawai and Enggano (Bengkulu) Basin (Fig. 13.2). At the sition of pelagic turbidites in the deep basins, has continued
present day the greatest depth of the three northern basins through Late Miocene and Pliocene times to the present day.
decreases from north to south: Aceh Basin, 2710 m; Meulaboh The same broad sequence of events affected all the forearc
Basin, 1150 m and the Nias Basin, 610 m; and increases again basins from the Aceh Basin in the north (Izart et al. 1994) to the
to the south: > 1000 m in the Mentawai Basin and > 2 0 0 0 m in Bengkulu Basin in the south (Hall et al. 1993).
the Enggano Basin to the south (Fig. 13.2). The basins are asym- In the Banyak islands, between the Simuelue and Nias basins,
metrical, for example in the Nias Basin the Sumatra continental Middle-Upper Miocene turbidites deposited in deep water are
shelf offshore the mainland of Sumatra deepens westwards to a overlain directly by Pleistocene to Recent reefs, indicating that a
shelf edge at c. 200 m, and drops down a continental slope into once continuous forearc basin has been separated into two
a deep-water basin, up to 610 m deep, further west. Sediment basins by recent uplift, localised in this area. This uplift has
cores obtained from the floor of the basin are turbidites (Karig been attributed to the passage of the Investigator Ridge and an
et al. 1979). The basin is cut off on its western side by a steep extinct Indian Ocean spreading ridge beneath the forearc
slope rising to Nias, coinciding with the monoclinal flexure and (McCann & Habermann 1989; Malod & Kemal 1996; Fauzi
the Mentawai Fault. et al. 1996). However, Matson & Moore (1992), from their
Seismic reflection surveys across the Meulaboh and Nias forearc study of the Banyak to Pini section of the forearc, discount the
basins calibrated by boreholes (Karig et al. 1979; Beaudry & possibility that subduction of the Investigator Ridge was respon-
Moore 1981, 1985; Matson & Moore 1992; Izart et al. 1994) sible for the uplift and subsidence in the forearc region. They
show seismic sequences ranging in age from Palaeogene to the suggest that due to the oblique subduction of the Indian Plate,
present day (Fig. 13.3c). The oldest dated rocks found in uplift and subsidence caused by the subduction of the ridge
exploratory oil company boreholes are Upper Eocene and Lower would be expected to progress southwards along the arc with
186 CHAPTER 13

time, whereas throughout this segment of the arc, uplift and fluviatile and lacustrine sediments in the more central parts. This
subsidence are synchronous (Matson & Moore 1992). pattern of sedimentation probably extended into the Sumatran
Matson & Moore (1992) following the earlier work of Beaudry forearc where it is poorly imaged in seismic sections and only
& Moore (1981, 1985) made a detailed study of SIO and industry rarely penetrated by oil company boreholes (e.g. Beaudry &
seismic profiles and used stratigraphic controls provided by oil Moore 1985). At this time subduction still continued along the
company boreholes to determine the sedimentation history and western margin of Sundaland represented by Palaeogene plutons
structural evolution of the Nias Basin. They found that the basin and volcanic rocks which outcrop along the west coast of
consists of two sub-basins separated by a broad area of elevation, Sumatra, while India was moving northwards towards its collision
the Singkel Basin to the northwest, limited to the west by the with the southern margin of Asia.
Batee Fault and to the east by the Singkel Fault, and the Pini In the Oligocene the forearc area was uplifted and exposed to
Basin to the south, limited by a fault to the west and a fault or subaerial erosion, supplying sediment to basins developed on the
monoclinal flexure to the east. They suggest that the location of accretionary complex to the west (Samuel et al. 1997). Karig
the basins is controlled by irregularities in the forearc basement et al. (1979) suggest that this period of uplift was related to
which, following Karig et al. (1980), is considered to represent compression of the forearc due to an increase in the rate of move-
the position of the original continental margin. These sub-basins ment of the Indian Ocean Plate. Marine transgression, with the
were found to have different subsidence histories. The rapid renewal of sedimentation, in the Late Oligocene and Early
2 km subsidence of the Singkel sub-basin in the Lower Miocene Miocene was due to the general world-wide rise in sea level
is attributed to a 25 km northward movement between the trans- which occurred at this time (Haq et al. 1987). An increase in the
current Batee and Singkel faults and the movement of the basin subduction rate from 5 to 6.5 cm a - l between 5 and 10 Ma may
down the inclined surface of the subducting plate (Matson & have been responsible for the Pliocene uplift and unconformity
Moore 1992, Fig. 13). They attribute the even greater subsidence seen in Nias and the other islands (Karig et al. 1979).
of the Meulaboh Basin which moved 100 km northwards along the
Batee Fault to the same mechanism. The accretionary wedge. In the Palaeogene the Indian Ocean floor
Along the Mentawai Fault on the western side of the Nias Basin that was being subducted or accreted into the accretionary
Miocene-Pliocene basin sediments are seen in seismic profiles to complex consisted of Cretaceous-Eocene basaltic crust with a
dip steeply eastwards into the basin, forming the monoclinal thin veneer of pelagic ocean floor sediments. Accretionary com-
flexure (Fig. 13.3c). Upper Pliocene to Pleistocene deposits rest plexes that are formed largely of basaltic ocean crust, are able
unconformably on the tilted rocks showing that the uplift of to maintain a steep inner trench slope as is the case with the accre-
Nias to form the flexure occurred in Late Pliocene times and tionary complex to the south of Java at the present day. In the mid-
that the present form of the forearc basin developed only recently. Miocene the Indian Ocean floor received a vast influx of terrige-
The dislocation represented by the flexure occurred approximately nous sediment derived fi'om the collision zone of the Indian
along the original contact between the Sumatran continental continent with the southern margin of Asia and the uplift and
margin and the accretionary complex (Karig et al. 1980; erosion of the Himalayas. This influx continued through the
Diament et al. 1992). As already described this flexure is attributed Pliocene until the Nicobar Branch of the Bengal Fan was cut off
by Samuel & Harbury (1996) to the reactivation of a basin- from sediment supply by the collision of the Ninety-East Ridge
bounding normal fault as a thrust, due to later compression. with the Sunda Trench adjacent to the Andaman Islands.
The Bengkulu Basin to the south has been studied by After the mid-Miocene the great thickness of terrigenous sedi-
Mulhadiono & Sukendar Asikin (1989), Hall et al. (1993) and ments which was scraped off the Indian Plate altered the dynamics
Yulihanto et al. (1995). They found that the Bengkulu Basin has of the accretionary complex. When the accretionary complex
a similar sedimentation history to the forearc basins to the north. is composed largely of incompetent sedimentary materials
Mulhadiono & Sukendar Asikin (1989) suggest that the graben the inner slope of the trench will have a much lower angle of
in the Bengkulu Basin developed as pull-apart basins on strike- slope than one composed of basaltic ocean crust. The surface of
slip faults driven by the oblique subduction. Hall et al. (1993) the accretionary complex will adopt a wedge-shaped cross-
and Yulihanto et al. (1995) suggest that the Bengkulu Basin origi- sectional profile with a critical taper, the angle between the
nated in the Palaeogene as a series of north-south extensional topographic surface and the inclination of the downgoing plate.
graben similar to those developed in eastern Sumatran at the The critical taper will depend on factors such as the frictional
same time, and that during the Early Miocene transgression the resistance at the base of the wedge, the strength of the material
Bengkulu forearc basin was connected with the South Sumatra composing the wedge, and pore fluid pressures which will also
Basin to the east, across the present site of the Barisan Mountains. influence the strength (Davis et al. 1983). The Sumatran accre-
They suggest that the basins later developed as pull-aparts by tionary wedge is composed of relatively weak materials with a
reactivation of earlier N W - S E transfer faults, related to the topographic slope of the order of 4 ~ and a 5 ~ inclination of the
Palaeogene extensional graben. downgoing plate, giving a critical taper of 9". The critical taper
represents an equilibrium condition. As thrust slices of sedimen-
tary material are compressed into the toe of the wedge by the
movement of the incoming plate the angle of the topographic
Tectonic evolution o f the f o r e a r c region slope will be increased and the wedge will adjust to re-establish
the critical taper by moving forward across the incoming plate.
Cretaceous to Oligocene history. At the end of the Cretaceous the This process provides a mechanism for the continual extension
area of the present Sumatran forearc formed the southwestern of the upper parts of the accretionary wedge and accounts for the
margin of the Sundaland continent. It was composed of the formation of the half graben developed on the surface of the wedge
Bentaro-Saling Arc and associated accretionary ocean crust as mapped by Samuel (1994) on Nias (Figs 13.5c and 13.6b).
which had been amalgamated with the continent in the mid- Matson & Moore (1992) have also pointed out that the continual
Cretaceous, and in the Late Cretaceous was the site of a magmatic increase in the volume of material incorporated into the wedge
arc related to subduction of the Indian Plate beneath Sundaland will exert a downward pressure on the incoming plate increasing
(Barber 2000). in the Palaeogene the forearc area, in common the angle of inclination which will counteract the increase of the
with the rest of Sumatra, and SE Asia as a whole, was subject to topographic slope and retard the forward movement of the toe of
regional extension. In Sumatra, extension led to the formation the wedge across the incoming plate. This in turn caused the
of graben structures which were occupied by lakes, with the uplift of the eastern part of the wedge. This mechanism would
deposition of screes and alluvial fans around their margins and account for the continual uplift recorded in the Lower to Middle
 STRUCTURE AND STRUCTURAL HISTORY 187

Miocene in the stratigraphic sequences in the basins on Nias, until Evidently the whole of the forearc is being deformed and has
in the Pliocene the upper part of the wedge and its overlying changed its shape by contraction normal to the trench and exten-
sediments emerged above sea level. They also point out that sion parallel to the trend of the arc.
with time this mechanism will result in the westward movement Prawirodirdjo et al. (1997) and Bock et al. (2003) have demon-
of the load exerted by the wedge, causing a dislocation during strated from GPS measurements of the displacement of 60 sites
the Late Pliocene along the Mentawai Fault between the uplifted on mainland Sumatra and on the outer arc islands, that at the
forearc ridge and subsiding sedimentary basins in the western present time the forearc to the south of the Batu islands is
part of the forearc basin (Matson & Moore 1992). coupled to the Indian Plate and is moving parallel to the con-
The processes and effects proposed by Matson & Moore (1992) vergence direction, but at a slightly slower rate than the incoming
would have operated with increasing intensity while sedimentary plate (44 mm a -1 compared with 75 mm a - i ) , while to the north
material was added to the surface of the Nicobar Fan and the thick- the forearc has an important component of northward movement
ness of sediment on the Indian Plate was continually increased. parallel to the Sumatran Fault at a much slower rate (Fig. 13.7).
When the sediment supply was cut off by the collision of the The change in the rate and direction of movement occurs at the
Ninety-East Ridge with the Sunda Trench, the accretionary point where the Investigator Ridge is entering the subduction
wedge was able to reach an equilibrium, and Upper Pliocene system, and it is suggested that this is due to the incorporation
to Pleistocene sediments were deposited unconformably on the of water rich sediments of the Nicobar Fan into the accretion
eroded surface of the older rocks. Later minor uplift and sub- system to the north of this point. The effects of subduction
sidence can be attributed to continual adjustments to the shape of the Indian Plate extend for a few tens of kilometers to the NE
of the accretionary wedge and to the fluctuations in sea level of the Sumatran Fault, but the greater part of eastern Sumatra
during the Pleistocene. Continual recent uplift has been documen- belongs to the Sunda Plate, which extends through Borneo
ted on Simeulue and Nias with five raised intertidal platforms to to western Sulawesi and is moving southeastwards at
the south of Sinabang on Nias. Dating of molluscs from reef 6 + 3 mm a-1 relative to the remainder of the Eurasian Plate
terraces on these islands gave ages ranging from c. 6000 to (Bock et al. 2003) (Fig. 13.7).
< 3 0 0 m m a -~ BP, with rates of uplift between 0.3 and
1.0 mm a - l (Vita-Finzi & Situmorang 1989). Drowned mangroves
on the eastern side of Siberut indicate that uplift is not uniform and
The Barisan Mountains
the islands may be tilted, and in Nias the east is uplifted while the
NW is drowned. Karig et al. (1979) propose that this is due to the
The Barisan Mountains extend for 1700 kin, from Banda Aceh in
displacement of the crest of the outer arc ridge towards the west
the north to Banda Lampung in the south, along the whole length
with the westward growth of the accretionary complex.
of the island of Sumatra, parallel and close to the west coast. Over
much of their length the mountains reach 1000 and 2000 m above
Effects o f transcurrent faulting. An important influence on the
sea level, locally rising above 3000 m in Aceh (Gunung (Mount)
tectonic evolution of the Sumatran forearc is the obliquity of con-
Leuser, 3381 m) and to the west of Lake Toba, and isolated vol-
vergence and subduction of the Indian Ocean Plate beneath
canoes rise above the general surface in Gunung Kerinci (3805)
Sumatra. In models of oblique subduction the strain in the over-
lying plate is considered to be partitioned between compression
normal to the subduction trench, which is taken up by inversion
of the sedimentary basins during the Pliocene, with N N W - S S E I \ I / I I

thrusting and folding, seen in all the outer arc islands, and trans-
,~o ~ 1t 102
~~ ~0 104~ 106"
2
lation parallel to the trench along transcurrent strike-slip faults
(Fitch 1972; Platt 1993; McCaffrey 1996). In the Sumatran sub-
duction system the major component of translation is the Sumatran
Fault which separates the forearc region from the Eurasian Plate
as a separate Burma sliver plate (Curray 1989).
There is a major difference in the amount of displacement along
the Sumatran Fault System from north to south. To the north of ; o
Sumatra the displacement is represented by extension, indicated
by the development of oceanic crust in the Andaman Sea, differ- t & 7~ SUNDAPE/ATEO
ential displacement being taken up along a series of closely
spaced transform faults with a total displacement of about Siberu
460 km, the westernmost of which passes southeastwards into
the Sumatran Fault System (Curray et al. 1979). On the other
hand displacement of the fault system in the Sunda Strait at the
southern end of Sumatra is less than 100 km (Huchon & Le
Pichon 1984; Harjono et al. 1991). Some of this discrepancy
may be accounted for by transcurrent movement along the
Mentawai Fault (Diamant et al. 1992), and some may be taken
up along splays of the Sumatran System, such as the Batee Fault
which extends into the forearc region from northern Sumatra INDIAN PLATE Enggano
(Fig. 13.2). Minor strike-slip faults, like those described by
Matson & Moore (1992) in the Singkel Basin, may be distributed
throughout the Ibrearc and the occurrence of transcurrent faults Fig. 13.7. Movementsin the forearcand withinSumatrarelativeto the estimated
within the submerged part of the accretionary complex is Sunda Shelf reference frame from GPS measurements 1991-2001 (after Bock
unknown. However, it is probable that the bulk of the differential et aI. 2003). The bold arrow shows the Australia/Eurasiamovement vector for
movement, must be taken up along the large numbers of minor the Indian Plate the finer arrows show the directions and amounts of movement
transcurrent faults, which form conjugate sets marked by linea- measured at specific locations.The lengths of the arrows are propbrtionateto the
ments seen in all the forearc islands, and by small scale rate of movement in mm a i. Ellipses of 95% confidencelimits have been
extensional faults which bisect the obtuse angle of the conjugate omitted, but are in general much larger for measurement to the NE of the
shears (e.g. Nias--Samuel & Harbury 1996) (Fig. 13.4b). Sumatran Fault System, than to the SW.
188 CHAPTER 13

in the centre, and Gunung Denpo (3159) in the south. The moun- characterised by a Cathaysian Flora (Hutchison 1994). In
tain range is broadest in the north, 100 km wide, occupying almost the Devonian Indochina separated from Gondwana (northern
the whole width of the island, narrowing to 50 km in the south. Australia) with the development of Palaeotethys, and amalga-
In the north the mountain range is formed of Pre-Tertiary rocks mated with South China in the Late Carboniferous (Metcalfe
of Carboniferous to Cretaceous age forming the basement of 1996). In the Early Permian Sibumasu (SIkkim, BUrma,
Sumatra, which are overlain by Tertiary sedimentary and volcanic MAlaya, S..._U_Umatra),distinguished by glacial sediments, separated
rocks which thicken into the basins in the forearc and backarc from Gondwana, and in the Late Permian or Early Triassic joined
areas, forming low ground of less than 100 m, to the SW and Indochina and East Malaya along the Bentong-Raub Suture and
NE. Locally Tertiary rocks occupy intramontane basins within its northern extension into Thailand and China (Metcalfe 2000).
the mountain range. Towards the southeastern part of the island Also probably in the Early Triassic the West Sumatra Block
the basement rocks are increasingly covered by Tertiary to joined the previously amalgamated blocks along the Medial
recent sediments and volcanics, with the older rocks being Sumatra Tectonic Zone, by strike slip faulting (Barber & Crow
exposed only in scattered inliers. At intervals along the chain base- 2003). The final component of the Pre-Tertiary basement of
ment rocks are overlain by Late Pleistocene to Recent volcanic Sumatra is the Woyla Nappe, which originated in Tethys as an
piles, some of which are active volcanoes (Fig. 13.1). oceanic island arc, and together with an accretionary complex
composed of imbricated oceanic crust, was thrust over the
western margin of Sundaland in the mid-Cretaceous (Barber
2O00).
Pre-Tertiary rocks in Sumatra
The Bentong-Raub Suture and the Bentong-Billiton Accretionary
Mapping the Pre-Tertiary units. Although Pre-Tertiary rocks in Complex (Figs 13.8 and 13.9). Metcalfe (2000) has given a full
Sumatra form mountainous terrain, they are in general poorly account and discussed the significance of the Bentong-Raub
exposed because of dense tropical rain forest and deep weathering. Suture Zone in Peninsular Malaysia. The zone bisects the Malay
In addition, apart from areas immediately adjacent to the roads Peninsula from north to south, for a distance of over 400 kin,
or the larger rivers, much of the area is difficult of access. where its trace is marked by outcrops of serpentinite, ribbon-
During the mapping of northern Sumatra by the Indonesian chert, schist and melange. The suture zone is considered to mark
Directorate of Mineral Resources (DMR) and of southern the site of the destruction of Palaeotethys, due to the collision
Sumatra by the Geological Research and Development Centre between the Indochina and Sibumasu blocks. There has been a
(GRDC) in collaboration with the the British Geological Survey long-standing controversy concerning the southward extension
(BGS), use was made of aerial photographs, Landsat and SAR of the suture into Sumatra. Metcalfe (1996, Figs 1 & 10) illustrates
(synthetic aperture radar) imagery to identify major geological four distinct paths which have been suggested in the literature and
structures. Major lineaments, mainly fault traces and wherever proposes another of his own.
possible bedding or foliation traces were plotted to outline fold Hutchison (1994) recognized that the suture was the eastern
structures. These features are plotted on the geological map margin of a much broader zone of deformation which he termed
sheets published by the Indonesian Geological Research and the Palaeotethys Suture Zone. His interpretation was confirmed
Development Centre, together with lithological and structural by Metcalfe (2000) who found that radiolarian cherts of the
data recorded during the fieldwork programme. Massive lime- Semanggol Formation 120 km to the west of the Bentong-Raub
stones are commonly well exposed, but form karstic terrain diffi- Suture consisted of two components, a steeply dipping sequence,
cult of access. Outcrops, together with float, of other rock types Lower to Upper Permian in age, repeated either by isoclinal
occur commonly in river gorges and along rivers and stream folds or by thrusts, and a unit of cherts, rhythmites and conglom-
networks and in coastal exposures. Artificial exposures occur in erates of Middle to Upper Triassic age which is only gently folded.
occasional quarries and in new road cuttings during road building The time of collision between Sibumasu and Indochina is marked
programmes. Due to the high rate of tropical weathering these by the unconformity between the Permian and Triassic within the
roadside exposures commonly last only for a few years. During Semanggol Formation.
the survey several areas that were found to be particularly well- Metcalfe (2000) suggested that the Permian, Devonian and
exposed were subject to more detailed structural examination. Carboniferous cherts, identified in the western part of the Malay
Maps illustrating the distribution of all the Pre-Tertiary units in Peninsula were deposited on the floor of Palaeotethys and were
Sumatra are given in Chapter 4 of this volume, together with their subsequently incorporated into an accretionary complex. The
definition, sedimentary features, palaeontology, stratigraphy and zone of collision between Indochina and Sibumasu in Malaya is
the palaeogeographic interpretation, while the overall tectonic marked by a broad accretionary complex, rather than by a
evolution of Sumatra is discussed in Chapter 14. This chapter narrow suture. This re-interpretation means that it is no longer
will concentrate on the structure and tectonic relationships of necessary to search in southern Sumatra for a discrete suture
these units. marking the collision between East Malaya and Sibumasu.
In this account it is proposed that the accretionary complex
Crustal blocks in Sundaland. The Pre-Tertiary units of Sumatra recognised in the Malay Peninsula extends southeastwards into
form part of Sundaland, the southeastern extension of the Eurasian the islands of Bangka and Billiton. It is therefore termed the
tectonic plate. Sundaland is considered to have been formed by Bentong-Billiton Accretionary Complex (Fig. 13.8).
crustal blocks which were rifted from the northern margin of An account of the lithology, stratigraphy and structure of the
the Gondwana continent, were separated during the Mid-Late Carboniferous, Permian and Triassic rocks on Bangka to the SE
Palaeozoic, and amalgamated to form Sundaland in the Late of the mainland of Sumatra is given by Ko (1986). Although
Palaeozoic and Early Mesozoic. The definition of the structural fossils are scarce, the oldest unit, the Pemali Group is considered
blocks is based on the work of Pulunggono & Cameron (1984), to be of Carboniferous and Permian age, and indeed Permian
Hutchison (1994), Metcalfe (1996, 2000) and Barber & Crow fossils have been found (De Roever 1951). The bulk of the
(2003). island is made up of slates and schists showing isoclinal folding
The pattern of crustal blocks which make up Sumatra and and a steeply dipping N W - S E foliation, imbricated with basalts,
adjacent parts of Sundaland is illustrated in Figure 13.8 and the andesites, bedded cherts, distal turbiditic sandstones, some-
stratigraphic units included within blocks forming Sumatra are times graded, mudstones, black pyritic shales and limestones.
shown in Figure 13.9. The Indochina Block, forms the core of Barber & Crow (Chapter 4) have interpreted the Pemali Group
Sundaland and with its southern extension into East Malaya is as oceanic material formed on the floor of Palaeotethys and
 STRUCTURE AND STRUCTURAL HISTORY 189

!....

SIBUMASU NDOCHINA
A(
~:: B LO C K

EAST MALAYA:
MEDA N
:~:~ B L O C K ) ~
"%%~io
Situtup
2~ Klippen

-% o,
,% ~

\ , %~ "% ~..oo

~, %o~,. ~ ~.. ~176176176

BENGKULU
;T SUM,
% % o
BLOCK

LAMPUNG
~

100 200 300 400 500km

% % on % % ,oo % % "%

" " :iii:!i:-iiiiiiiI L / JAVa

96 ~ 98 ~ 100 ~ 1020 1040 % 106 ~
I I i I I I

Fig. 13.8. Crustal blocks that comprise the pre-Tertiary basement of Sumatra, based on Hutchison (1994), Metcalfe (2000), Barber & Crow (2003). Reverse arrows
indicate dextral transcurrent movement on the Sumatran Fault System.

incorporated into an accretionary complex related to the Late been encountered in oil company boreholes (De Coster 1974;
Permian-Early Triassic collision between Sibumasu and East Eubank & Makki 1981). However, at Toboali on the southern
Malaya. tip of Bangka, Ko (1986) describes 'pebbly mudstones', similar
Permian fossils found on Bangka and Billiton include fusulinids to those described from other areas of Sibumasu Block to the
and a poorly preserved flora of Cathaysian affinity, indicating that NW (Cameron et al. 1980; Stauffer & Lee 1987; Mitchell et al.
the islands are related to the East Malaya Block (Van Overeem 1970). The Sibumasu Block is therefore considered to extend
1960). In eastern Sumatra and the offshore islands between the southwards into southern Bangka.
Malay Peninsula and Bangka the Bentong-Billiton Accretionary On Bangka the Pemali Group is locally intruded and hornfelsed
Complex is largely covered by Tertiary and Quaternary deposits, by Late Permian-Triassic granites (see Chapter 5), constraining
although rock types which may belong to the complex have the age of formation of the accretionary complex and the age of
190 CHAPTER 13

WEST SUMATRA BLOCK EAST SUMATRA BENTONG-BILLITON

WOYLA GROUP (Schiefer Barisan, MSTZ (SIBUMASU) ACCRETIONARY
Vorbarisan, Kluet and BLOCK COMPLEX
Kuantan Units)

Granitic Intrusions
Bintan F o r m a t i o n

Woyla Group Granitic Intrusions Granitic Intrusions

(oceanic and volcanic
arc assemblages)

(3_
o Situtup & T u h u r -~ ~) Kaloi, Batumilmil o
Limestone blocks Formations = E rr and Kualu ~ c Tempilang
in melange t- O Formations ~ ~ Formation
~
~F ~~ ~ ~~:~~~"~~"~'"~ ~ ~ ~~ II
,~ - -

N ~Z
:4 :g,:N::~ N # Stutup S ungkang <~ Kaloi and
Palepat and Menqkaranr E'E" :3 Batumilmil
Format ons- - Formations
(tropical Jambi flora) -~
o _~-c ~~
Pangururan Bryozoan Bec
O B o h o r o k (tilloids) Pemali Group
,~ ~ Kluet and Kuantan ,, :~o a n d Alas
Formations = ~ "6 C9 Formations F i g . 13.9. T h e s t r a t i g r a p h i c s e q u e n c e s a n d
(tropical f a u n a in ~-o -5 (temperate fauna p h a s e s o f g r a n i t i c i n t r u s i o n that c h a r a c t e r i z e
limestones) ~= ~= o~ z~ in l i m e s t o n e s )
the crustal b l o c k s w h i c h m a k e up the
s
we-Tertiary basement of Sumatra. MSTZ,
Medial Sumatra Tectonic Zone.

the collision between East Malaya and Sibumasu to Late to end- lithologies occur in the Mentulu Formation in the Tigapuluh
Permian. Outcrops of the Pemali Group form east-west bands Hills in central Sumatra, and as has already been mentioned,
across the island and alternate with outcrops of undeformed sand- pebbly mudstones also crop out at Toboali at the southern tip
stones and mudstones of the Triassic Tempilang Formation, which of the island of Bangka. This lithological association, with the
is folded into broad open folds. Because of their difference in presence of pebbly mudstones, is regarded as characteristic of
degree of deformation the Tempilang Formation is considered the Sibumasu Block which therefore occupies the whole of the
to have been deposited unconformably on the Pemali Group, but eastern part of Sumatra (Figs 13.8 & 13.9).
in places later deformation has thrust rocks of the Pemali Group Although there is no direct evidence for the age of the Bohorok
over the Tempilang Formation (Ko 1986). Formation, support for the correlation with the Lower Permian
of southern Thailand is given by an outcrop of decalcified lime-
stone, the Pangururan Bryozoan Bed, on the western shore of
Lake Toba (Aldiss et al. 1983) (Figs 13.10 & 13.11). This
East S u m a t r a ( S i b u m a s u ) B l o c k (Figs 13.8 a n d 13.9) limestone is associated with slates and sandstones which were
attributed to the Kluet Formation, presumably because it contains
Tapanuli Group. During the D M R / B G S Northern Sumatra no pebbly mudstones, but its location NW of the Medial Sumatra
mapping project Pre-Tertiary rocks in northern Sumatra were Tectonic Zone of this account, suggests that it should be more cor-
assigned to the Tapanuli (Carboniferous-Permian), Peusangan rectly attributed to the Bohorok Formation. This limestone con-
(Permo-Triassic) and Woyla (Jurassic-Cretaceous) groups tains fenestellid bryozoans deformed in the slaty cleavage and
(Cameron et al. 1980). The Tapanuli Group was further divided forming ideal strain markers (Ramsay 1967). The fenestellids
into three units, the Bohorok, Alas, and Kluet formations, outcrop- and the other fossils indicate a Late Carboniferous or Early
ping from NE to SW, in that order (Fig. 13.10). Of these units Permian age and the bed has been correlated with the Lower
only the Alas can be confidently ascribed palaeontologically to Permian Bryozoan Bed of Peninsular Thailand (Mitchell et al.
the Carboniferous, but the Bohorok and Kluet formations were 1970; Cameron et al. 1980). This outcrop is critical to
also considered to be of Carboniferous or Early Permian age determining the age of the Tapanuli Group and also the age of
because they are associated with the Alas Formation in the field, its deformation and metamorphism.
and a proposed stratigraphic correlation with similar rocks in A brief account of the structure of the Bohorok Formation is
western Malaya and southern Thailand (Cameron et al. 1980). given in the explanatory notes which accompany each of the
The Bohorok Formation, with a type locality in the Bohorok GRDC 1:250 000 geological map sheets (Bennett et al. 1981c;
River 60 km to the west of Medan, is characterized by the occur- Cameron et al. 1982a; Clarke et al. 1982a, b). Unfortunately,
rence of 'pebbly mudstones', interpreted as glacigenic deposits, very few structural observations are recorded on the map sheets,
together with massive sandstones, sometimes conglomeratic, and but it is reported that general strike of bedding throughout the
intervening shales interpreted as turbidites. Similar lithologies outcrop of the Bohorok and Mentulu formations is N W - S E ,
in northern Sumatra, but without the pebbly mudstones, were parallel to the trend of the Barisan Mountains and of Sumatra as
mapped as the Kluet Formation. It is possible that outcrops of a whole (Sumatran trend), and that the rocks are folded with
Kluet Formation which are shown on the quadrangle sheets steep and often vertical dips. Massive sandstones show little evi-
lying to the NE of the Medial Sumatra Tectonic Zone should dence of penetrative deformation, with only irregular jointing
more properly be assigned to the Bohorok Formation. The and quartz veining, although fracture cleavage is sometimes
pebbly mudstones of the Bohorok Formation have been correlated developed, but the intervening shales are generally tightly to
with the Lower Permian pebbly mudstones of Phuket in southern isoclinally folded and converted to slates with an axial plane
Thailand (Cameron et al. 1980; Mitchell et al. 1970). Similar slaty cleavage. Crenulation cleavages, kink bands and shears are
 STRUCTURE AND STRUCTURAL HISTORY 191

I
97~ SIBUMASU (EAST SUMATRA) BLOCK
.? Peusangan Group
(Permo-Triassic)
LHOKSEUMAWE
"~ ~Bo~J Tapanuli Group-Bohorok Formation
,:{~,1 (Carboniferous-Early Permian)
MEDIAL SUMATRA TECTONIC ZONE
L i m e s t o e'-.. -, (MSTZ)
Alas Formation (limestones etc.)
~ . ~ "'" "'" GON~ % ~ AmphiboliteFacies
Metamorphic Rocks
BO ~"~ >10ppm tin in stream sediment samples

;erbajadi
Granite

~ Bohorok C/ " ~ M E DAN

\
KUTACANE
\ ~R Toba Tufts
TAPAKTUAN Kualu (K)
,ilmil

X Antiform
LAUBALENG
X Synform :~ Toba Tuffs

Graniticintrusions
SIDIKALAN(

WOYLA NAPPE Pangururan

Bryozoan B
Woyla Group
P-~ (Jurassic-Cretaceous)
KLUET (WEST SUMATRA) BLOCK

Kluet Formation
Amphibolite Facies
Metamorphic Rocks
\,Granite//
264+6Ma'
0 50 100km \ /~
SIBOLGAq
I I
97~ 98~ Q ~
i I /
Fig. 13.10. Outcrops of pre-Tertiary units in northern Sumatra showing the distribution of formations in the Carboniferous to lower Permian Tapanuli Group and the
Permo-Triassic Peusangan Group (after Stephenson & Aspden 1982, with modifications from the present study). Near Kutacane the Medial Sumatra Tectonic Zone is
coincident with the outcrop of the Alas Formation and is distinguished by the juxtaposition of unmetamorphosed sediments and high-grade metamorphic rocks,
syntectonic granitoid intrusions and a tin anomaly. Further north the MSTZ is traced through Takengon following outcrops of phyllite, schist and gneiss, recognised in the
primary mapping, but not incorporated in the compilations. Turbiditic sediments, without pebbly mudstones to the NE of the MSTZ which were originally mapped as
Kluet Formation lie on the Sibumasu Block, and are here assigned to the Bohorok Formation. Pre-Tertiary rocks are covered by Tertiary and Quaternary sediments and
volcanics in areas left blank.
192 CHAPTER 13

reported indicating that the rocks have been subjected to multiple grade rocks (shown as Kluet Formation) by a marginal zone of
deformation (Bennett et al. 1981c; Cameron et al. 1982a; Clarke gneisses and schists. These metamorphic rocks were interpreted
et al. 1982a, b). as an aureole forming a carapace brought up t?om depth together
In a more detailed structural study of the Bohorok Formation with the batholith (Bennett et al. 1981c; Cameron et al. 1983), but
on the Pematangsiantar Sheet, in the area to the south and SE of is here suggested to form part of the Medial Sumatra Tectonic
the Hatapang Granite, Clarke et al. (1982a) (Fig. 13.11) recog- Zone (Fig. 13.10). Cameron et al. (1982a) describe the sequence
nized two stages of folding, the earlier on N W - S E axes with of rocks seen in metamorphic aureoles around granitoid intrusions
SW-dipping axial planes, and the later with axial planes inclined in the Medan area as: fine-grained hornfels ~ coarse m u s c o v i t e -
at a shallow angle to the west. The intersection of the two axial biotite hornfels with segregations of epidote, chlorite, hornblende
surfaces defines a lineation plunging at a shallow angle (c. 15 <') and tourmaline with quartzofeldspathic rims -+ schistose hornfels
to the NW. In the Pakanbaru Quadrangle to the south, dip with flattened segregations, andalusite and cordierite ---> biotite-
measurements show a wide range of orientations, but Clarke muscovite schists, sometimes garnetiferous--+ banded sillima-
et al. (1982b) report that slate units show widespread tight to iso- n i t e - b i o t i t e - m u s c o v i t e gneiss with feldspar porphyroblasts--+
clinal folding on axes which vary from e a s t - w e s t to N W - S E , coarse migmatitic gneiss with quartz-feldspar lit-par-lit layers,
with axial planes which are vertical or dip steeply to the SW. pods and ptygmatic veins. Flattened clasts in the hornfelsed sand-
SAR imagery of the same area shows bedding plane traces with stones show that the rocks had been deformed and converted to
complex folding and fold axial plane traces trending N E - S W slates before they were thermally metamorphosed and before the
and N W - S E , the latter direction becomes dominant towards emplacement of the igneous intrusions.
the SW where the Bohorok Formation is in contact with the To the SE in the Tigapuluh Hills in central Sumatra Pre-Tertiary
Tanjungpuah Member of the Kuantan Formation (Fig. 13.12). rocks of the Tigapuluh Group outcrop as an inlier among Tertiary
The Bohorok Formation is intruded by large and small igneous sediments (Fig. 13.1) 9 The group is composed of the Mentulu,
bodies with the development of hornfelses, schists and gneisses Pangabuhan and Gangsal formations (Simandjuntak et al. 1991;
in the adjacent country rocks. One of the largest is the Serbajadi Suwarna et al. 1991). The Mentulu Formation contains pebbly
Batholith to the north of Medan on the Takengon and Langsa mudstones, similar to those of the Bohorok Formation; the other
sheets. The intrusion is separated from the surrounding slate formations are turbiditic sandstones and shales, with the Gangsal

,,v- ill " ~ i I

~. ~ T~Tebingtinggi"~99~
To Prapat 1 Pleistocene Toba Tufts
Tertiary sedimentsand volcanics

2~ Tt N Middle-LateTriassic Kualu Formation

Sibaganding and Pangunjungan
Te Limestone Members
\ Oo
.....,
I-o Tebingtinggi

LAKE TOBA
Tt
_4q1
Te

Tt i
.....

Tt \

Tt '
2~ ~'~~O451E __ 0~ Li.... iorlplLInge
Beddingstrikeanddip
LAKE TOBA Ve~oalbeds Tt
~1~nal Invertedbeds
SAMOSIR ~ CI.... g......... ddip ~d an
ha (Tuffaceous k v..... ,e,.... ,e Te
sed'men's ,ntrusions,
Zt Carboniferous-Permian ~ ~ T(~*'~ Halqbia
,ozoan Bed Tapanuli Group , Rant
Haria ~ Bohorok Formation ~,~~
TtPintu (pebbly mudstones)
4e ~ Undifferentiated ~ 0 5 10 15 20km
99015'

Fig. 13.11. The geology of the area between Lake Toba and Rantauprapat showing the relationship between the Carboniferous-Permian Tapanuli Gl"oupand the Triassic
Kualu Formation based on the GRDC Pematansiantar (Clarke et al. 1982a) and Sidikalang (Aldiss et al. ]983) Quadrangle sheets. While the Tapanuli Group is isoclinally
folded with slaty cleavage and shows the effects of multiple deformation the Kualu Formation shows one set of upright fold and argillaceous units are not cleaved.
Although all the contacts are faulted the Kualu Formation must have an unconformable relationship to the Tapanuli Group. The inset map shows the location of the
Pangururan Bryozoan Bed (PBB) on the western shore of Lake Toba.
 STRUCTURE AND STRUCTURAL HISTORY 193

I
Pasirpangarayan.
lOkm
90km Igneous intrusions
I

Kuantan Formation
Pakanbaru (West Sumatra Block)
disilan 70kin .
Tanjungpuah Member
(Medial Sumatra Tectonic
Zone)
Bohorok Formation
(Sibumasu Block)
70
Strike and dip of bedding
75 Strike and dip of cleavage

~"~20 Plunge of lineation

- 0~ ..~ Photodip 0~
2,, _ ~I Hot spring

.._TJ

Ban

F~angkalan-kota-baru
9 Siasam

Bukit
Tinggi
50km,
9 Muaraketua
100~ . . . . 101~
0o00'

Fig. 13.12. Structure across the Medial Sumatra Tectonic Zone (MSTZ) from GRDC Pakanbaru Quadrangle Sheet (Clarke et al. 1982b), central Sumatra, with the
addition of bedding traces from SAR imagery. Irregular refolded folds in the Bohorok and Kluet Formations trending approximately east-west, contrast with isoclinal
folds trending NW-SE within the MSTZ which incorporates the Tanjunpuah Member of the Kuantan Formation. N.B. Granitic rocks within the MSTZ show a gneissose
foliation parallel to the trend of the zone. The identification of units on the map has been modified in the light of the interpretation SAR imagery. Pre-Tertiary basement
rocks are overlain by Tertiary sediments in the areas left blank.

Formation being finer-grained. SAR imagery indicates that the et al. 1991; Suwarna et al. 1991). The possibility of the influence
Gangsal Formation is also more deformed than the other units, of strong Tertiary deformation in this area on the orientation of
with a strong N W - S E trend. the cleavage in the basement rocks has not been clarified.
Structures in the Tigapuluh Group are similar to those reported Pelitic rocks of the Tigapuluh Group are altered to cordierite
from the Bohorok Formation, with folded bedding and steeply hornfels, biotite schist and gneiss in metamorphic aureoles
dipping cleavage in pelitic units, indicating tight to isoclinal around granitoids of Jurassic age (Schwartz et al. 1987). Defor-
upright folds (Simandjuntak et al. 1991; Suwarna et al. 1991). mation, with folding and the development of cleavage preceded
Although the contact between stratigraphic units and measure- the intrusion of the granitoids, as the thermal metamorphism
ments of the orientation of the bedding within the group have affects rocks which were already cleaved.
the N W - S E Sumatran trend, cleavage measurements made in
the field and shown on the Rengat and Muarabungo map sheets The Permo-Triassic Peusangan Group in the Sibumasu Block. In
show a wide scatter, but trend predominantly e a s t - w e s t and dip northern Sumatra the Permo-Triassic Peusangan Group is
either to north or south. The reasons for the discrepancy represented mainly by isolated limestone outcrops in the northern
between the orientation of the cleavage and the bedding is not part of the Sibumasu Block (Fig. 13.10). Each of the isolated
clear, and requires further study. Two phases of folding are limestone outcrops has been given a separate formation name
reported, the first e a s t - w e s t and the second N W - S E , but no (for a detailed account of these formations see Chapter 4). The
examples of refolded folds were identified in the field, although limestones are generally massive and recrystallized, and unlike
crenulation cleavage, indicating that the rocks were affected by the slates and sandstones of the Bohorok Formation, do not gener-
a second phase of deformation, was recorded (Simandjuntak ally show folding or penetrative deformation. For this reason the
194 CHAPTER 13

Fig. 13.13. Map of the outcrops of the Bohorok,Alas and Kluet tbrmationsbetween Kutacaneand Laubaleng,based on the GRDC Medan Quadrangle Sheet (Cameron
et al. 1982a), with the additionof beddingtraces (dashed lines) from SAR imagery;solid lines are faults. Open foldingin the Bohorokand Kluet lormationscontrasts with
tighter foldingin the Alas Formation,which lies within the Medial Sumatran Tectonic Zone. In areas left blank the the we-Tertiary basement is covered by Tertiary and
Quaternary sediments and volcanics,includingthe alluviumin the Kutacane Graben.

surveyors considered that the undeformed Permo-Triassic rocks which show open folds (Bennett et al. 1981c). To the south of
rest unconformably on the deformed Bohorok Formation, although Medan at Parapat on Lake Toba bedded limestones and shales
no stratigraphic contacts have been described (Cameron et al. of the Kualu Formation are moderately to tightly folded about
1980). On this basis it was suggested that the major phase of sub-horizontal, N W - S E axes with steep NE-dipping axial
deformation seen in the Tapanuli Group occurred in the Early to planes. The intensity of the tblding increases towards the west
Mid-Permian, before the deposition of the Peusangan Group and the steep western limbs of the folds may be overturned,
(Cameron et al. 1980). giving a westerly vergence. However, cleavage is not developed
Many of the limestone outcrops are recrystallized and in the argillaceous interbeds in these outcrops, although highly
apparently unlbssiliferous, but a few have yielded Permian and deformed slates of the Bohorok Formation occur only a short
Triassic fossils. M i d - L a t e Permian fossils have been obtained distance away across a fault contact (Clarke et al. 1982a; Aldiss
from the Situtup Formation to the NW of Takengon, and the et al. 1983) suggesting that the relationships between the Kualu
Kaloi and Batumilmil formations to the NW and west of Medan and the Bohorok formations are unconformable (Fig. 13.11).
(Fig. 13.10). The Situtup Formation has yielded Mid-Permian In the Pematangsiantar Quadrangle to the east, thin-bedded
fusulinids with a Cathaysian affinity, indicating that this area limestones and cherts of the Pangunjungan Member of the
forms part of the West Sumatra Block. M i d - L a t e Triassic Kualu Formation show tight disharmonic folds which have been
fossils have been obtained from the Situtup, Kaloi and Batumilmil attributed to slumping (Clarke et al. 1982a).
formations (Fontaine & Gafoer 1989) but the relationships It has been argued by Barber & Crow (Chapter 4) that the obser-
between the Permian and Triassic components of these outcrops, vations concerning the structural and stratigraphic relationships
whether conformable or unconformable, have not been estab- of the Tapanuli and Peusangan groups in northern Sumatra have
lished. Other outcrops of Triassic rocks belonging to the Kualu been misinterpreted. It is commonly observed in slate-grade
Formation occur on the eastern and western shores of Lake metamorphic terranes that massive limestones behave as compe-
Toba and 35 km south of Medan (Fig. 13.11). tent materials, while incompetent argillaceous materials are
In the outcrop of the Kaloi Formation massive undeformed deformed around them. This may well be the case in northern
limestones are associated with Triassic limestones and shales Sumatra, where the argillaceous sediments of the Tapanuli
 STRUCTURE AND STRUCTURALHISTORY 195

Group are highly deformed, while the massive limestones of the Quadrangle near Kutapanjang, the rocks are schists and gneisses
Peusangan Group are unaffected. In this respect it is notable that that enclose garnetiferous granites with gneissose or foliated
massive limestones of the Alas Formation, considered to form margins (Cameron et al. 1983). The granitoids contain narrow
part of the Tapanuli Group are also undeformed, with the preser- zones of mylonite and cataclasite showing horizontal slickensides
vation of fossils and delicate sedimentary structures, while the and are separated from slate-grade country rocks by a meta-
argillaceous rocks around them have been altered to slates and morphic envelope. The syntectonic granitoids are considered
schists. to have been intruded into active sub-vertical shear zones and
Evidence for the age of deformation in the Sibumasu Block of to be responsible for the amphibolite-facies metamorphism of
northern Sumatra is found around Lake Toba, where on the the adjacent rocks (Cameron et al. 1983). The MSTZ is also
western shore of the lake, the Pangururan Bryozoan Bed associated with elevated values of tin in stream-sediment
(Figs 13.8 & 13.9), a decalcified argillaceous limestone of Late samples > 10 ppm (Stephenson et al. 1982) (Fig. 13.10). The tin
Carboniferous to Early Permian age, is interbedded with slates was derived either directly from the Pre-Tertiary basement and
and sandstones of the Kluet Formation (?Bohorok Formation in the associated granites (e.g. the Kais Complex, Fig. 13.10) or
this account) and has been deformed to the same extent. On the indirectly from Tertiary sediments.
other hand slaty cleavage is not developed in the folded argillac- The southern segment of the Medial Sumatra Tectonic Zone is
eous beds of the Middle to Upper Triassic Kualu Formation generally very poorly exposed, but crops out in a N W - S E -
on the eastern shore of the lake. Determining the age of the trending belt as the Pawan and Tanjungpuah members attributed
Pangururan Bryozoan Bed is critical for defining the age of defor- to the Kuantan Formation, between Pakanbaru and Lubuksikaping
mation more precisely. However, taking the evidence available, (Clarke et al. 1982b; Rock et al. 1983) (Figs 13.12 & 13.14).
it is considered that the major deformation in northern Sumatra These rock units are composed of intensely folded muscovite,
occurred within units classified in the Peusangan Group between tremolite, chlorite, carbonate and quartz schists. Locally coarse-
the Permian and Triassic. Since no fossils representing the latest to fine-grained banded marbles are interbedded with fine-grained
Permian or earliest Triassic have been found anywhere in northern chlorite schists derived from basic volcanics or tufts (Rock et al.
Sumatra, it is therefore most probable that the major phase of 1983).
deformation occurred during the Late Permian and Early Triassic. Large scale folds on N W - S E axial traces have been mapped in
This is the age of the deformation seen in Peninsular Malaya, the Tanjunpuah Member in the Pakanbaru Quadrangle using aerial
where it is regarded as marking the collision of the Sibumasu photographs (Clarke et al. 1982b). On SAR imagery, irregularly
and East Malaya terranes (Metcalfe 2000). oriented folds in the Bohorok Formation to the NE, and the
Kuantan Formation to the SW, pass into tight to isoclinal folds
with N W - S E axial traces in the Pawan and Tanjungpuah out-
crops. The Pawan and Tanjungpuah units lie within a zone of
The M e d i a l S u m a t r a T e c t o n i c Z o n e (Fig. 13.8) intense deformation identified here as the MSTZ (Fig. 13.12).
In the field Clarke et al. (1982b) recognized three phases
The Medial Sumatra Tectonic Zone (MSTZ) forms a zone of of folding in the Pawan Member: the earliest on moderate to
highly deformed rocks, extending for the whole length of steeply plunging axes, the second forming tight to isoclinal folds
Sumatra, separating the Sibumasu Block from the West Sumatra on sub-horizontal N W - S E axes and the third as narrow zones of
Block. The MSTZ is separated into three segments by major brittle-style refolding and kinking on NW-dipping axial planes,
faults. The northern segment of the zone abuts against the with the development of quartz tension gashes.
Samalanga Fault and must pass beneath Tertiary sediments into The metamorphic rocks are intruded by granites with a lensoid
the Andaman Sea to the west of the fault. The central segment shape, elongated parallel to the schistosity in the adjacent slates.
has been displaced southwards by c. 50 km along the Lokop- The granites also contain a steeply dipping N W - S E internal
Kutacane Fault and near Sibolga the southern segment has been foliation (e.g. the Pulaugadang Granite in Figs 13.12 & 13.14)
displaced southeastwards for a distance of 150 km along the that is defined by oriented mica flakes, encloses aligned felspar
Sumatran Fault Zone (Fig. 13.8). megacrysts, and also affects cross-cutting microgranitic veins.
The northern segment of the MSTZ is marked by a zone of Adjacent to the granitic intrusions the metasediments are
phyllitic, schistose and gneissose rocks which were identified on converted to schists with a steep schistosity. The schistosity
the 1:100000 field maps and described in the initial reports related to the second phase of folding has been dated by the
prepared during the DMR/BGS Northern Sumatra Survey, K - A t method as Early Jurassic (Clarke et al. 1982b).
but are not represented separately on the 1:250 000 Quadrangle Southeastwards, the MSTZ may be represented by the intensely
sheets. These include the Uneuen Unit on the Lhokseumawe deformed Gangsal Formation on the southwestern side of the
and Takengon Quadrangle Sheets, the Toweren Member of the Tigapuluh Hills, but it is not exposed further to the SE, and its
Peusangan Group along Lake Tawar, and amphibolite facies trace can only be inferred from schistose and basic lithologies
schists, gneisses and marbles which were interpreted as forming encountered in oil company boreholes put down through Tertiary
the aureole of the Serbajadi Granite (Keats et al. 1981; Cameron sediments (De Coster 1974; Eubank & Makki 1981) (Fig. 13.14).
et al. 1983) (Fig. 13.10). A calcareous bed in the Uneuen Unit The zone identified here as the MSTZ has long been recognized
yielded Triassic fossils (Cameron et al. 1978, appendix), a as an important tectonic boundary in Sumatra. Van Bemmelen
feature seen in the MSTZ further south. (1949) following the survey by Von Steiger (1922) drew the
The central segment of the MSTZ corresponds with the outcrop boundary between his Tectonic Zones I and III along this line,
of the Alas Formation (Figs 13.8 & 13.13) characterized by and his lensoid Zone II incorporates the Pawan and Tanjungpuah
massive limestones, which locally contain a Early Carboniferous Members. Lithologies described within Zone II include quartzites,
(Vis6an) fauna (Metcalfe 1983; Fontaine & Gafoer 1989), but phyllites, shales, diabase-schists, limestone, radiolarian chert,
also includes sandstones and shales, identified as turbidites, conglomerates with granitic boulders and mylonitized breccia,
similar to those of the Bohorok Formation. The temperate and the zone is characterized by tin mineralization (see Figs
Vis~an fauna identifies the Alas Formation as part of the Sibumasu 13.8 & 13.12). In their tectonic synthesis of Sumatra Pulunggono
Block which has been disrupted and incorporated into the MSTZ. & Cameron (1984, Fig. 1) draw a line separating the Bohorok and
Apart from these limestones, the Alas Formation is schistose and Kuantan formations through central Sumatra, from the outcrop
metamorphosed in the greenschist to amphibolite facies. Locally of the Alas Formation in the north to Palembang in the south.
the limestones have been altered to coarse graphite-phlogopite They identified a lens of material along this line, including the
marbles. In the Rikit Gaib (Fig. 13.10) area in the Takengon Pawan and Tanjungpuah members, Triassic rocks of the Kualu
196 CHAPTER 13

Fig. 13.14. The Medial Sumatra Tectonic Zone in central Sumatra distinguished by highly deformed metamorphic rocks, syntectonic granitic intrusions and a tin
anomaly, separating the East and West Sumatra blocks.

Formation, and the Mutus Assemblage, previously described along the shear zone. Multiple phases of deformation recorded
from southern Sumatra by Eubank & Makki (1981). Eubank & within the rocks of the shear zone indicate that movements
Makki (1981) defined the Mutus Assemblage from oil company occurred at different periods of time. The occurrence of relatively
boreholes which had penetrated the Pre-Tertiary basement. The undeformed Triassic rocks of the Kualu and Tuhur formations
material included 'radiolarian chert, meta-argillite, red-mauve adjacent to the shear zone suggest that initial movements occurred
shale, thinly-bedded limestone and deep water rhythmite before the Mid-Triassic. Granites, foliated together with the
sequences'. Basalt from one borehole, and the association with associated metasediments, were emplaced syntectonically. From
deep-water sediments, led them to speculate that the assemblage regional correlations these granites are considered to be of
might include ophiolitic material; the chert and rhythmites were Late Triassic to Early Jurassic age. If the ages of the syntectonic
con-elated with the Triassic Kualu Formation of northern Sumatra. granitic rocks are confirmed it suggests that movements along
The absence of records of large bodies of ophiolite along the the MSTZ continued throughout these periods. Further age deter-
MSTZ suggest that the zone does not mark a major suture zone minations on the granites would constrain the period of movement
representing the collision of continents and the subduction of more precisely. Records of mylonitization, cataclasis, brecciation
oceans, although tremolite schist in the Pawan and Tanjungpuah and surfaces with slickensides within the MSTZ (Cameron et al.
may represent slivers of deformed and metamorphosed serpenti- 1982a) suggest that strike-slip movements have occurred con-
nite, and chlorite schists may represent basic igneous rocks. tinually within the zone. The latest movements along the MSTZ
We interpret the MSTZ as having been initiated as a transcurrent are probably associated with the movement of the Sumatran
fault along which the West Sumatra Block was emplaced against Fault Zone.
the Sibumasu Block. The subsequent uplift of high-grade meta-
morphic rocks, the emplacement of syntectonic granitic magmas
and the flux of tin mineralizing fluids indicate that the MSTZ is
a major lineament on a crustal scale. West Sumatra Block
Some of the material incorporated in the zone are slivers of the
adjacent Bohorok, Alas, Kluet and Kuantan formations, and their Kluet Unit (Fig. 13.8). The Kluet tectonic unit is coincident with
original structural patterns have been truncated and drawn out into the outcrop of the Kluet Formation, occupying the western part
conformity with the N W - S E trend of the shear zone (Figs 13.12 & of the Barisan Mountains between Sibolga and Tapaktuan to the
13.14). Local amphibolite-facies metamorphic rocks juxtaposed southwest of the Medial Sumatra Tectonic Zone, and is overlain
with unmetamorphosed sediments, including fossiliferous lime- to the NW and SE by the Woyla Nappe (Fig. 13.8). The Kluet
stones, indicate that material has been uplifted from deeper in Unit is considered to form the northern part of the West Sumatra
the basement and that material has subsided during movements Block which has become separated from the remainder of the
 STRUCTURE AND STRUCTURAL H|STORY 197

block to the south by movements along the Sumatran Fault complex have given Late Triassic and Early Jurassic ages. Roof
System. pendants of Kluet Formation in the granite, when they are not
Lithologically the Kluet Formation is composed of alternating hornfelsed, show isoclinal folding and cleavage, with graded
quartz-wackes, siltstones and shales, with some limestones. Over bedding indicating that some beds are inverted (Aspden et al.
most of the outcrop the argillaceous rocks have been converted 1982b). It is therefore presumed that the bulk of the intrusion is
to slates, but amphibolite-grade metamorphic rocks occur on post-tectonic. A detailed study of the relationships between the
the western side of the outcrop to the east of Tapaktuan structures and dated intrusive phases of the Sibolga Complex
(Fig. 13.10). Cameron et al. (1982b) describe a southwestwards may elucidate the history of deformation in the Kluet Formation.
change from predominantly slates to pelitic schists and phyllites
in the Simpali area and further southwest, in the type locality of Kuantan Unit (Fig. 13.8). The Kuantan Unit in central Sumatra is
the Kreung (River) Kluet to amphibolite facies pelitic schists, coincident with the outcrop of the Kuantan Formation. It is
calc-schists and quartzo-feldspathic gneisses with small concor- limited to the NE by the Medial Sumatra Tectonic Zone and to
dant gneissose granitoid bodies. Barber (2000) has suggested the SW across the Takung Fault by the Vorbarisan Unit (Tobler
that these higher-grade metamorphic rocks mark the footprint 1910, 1917). The Kuantan Formation is composed of turbiditic
of the overlying Woyla Nappe. These rocks have not yet been sandstones and shales with some limestones and resembles litho-
dated to test this hypothesis. logically the Kluet Formation of northern Sumatra, and indeed
Structurally the sandstones and slates of the Kluet Formation the boundary between the two formations was defined arbitrarily
are folded on both the large and small scale. In the outcrop to at a break in outcrop along 99~ longitude. Massive limestones
the south of Sidikalang, large scale folds of the bedding, with syn- within the Kuantan Formation with a Carboniferous (Vis6an)
lbrmal and antiformal axial traces 20 km apart, have been traced fauna suggest a correlation with the Alas Formation of northern
on aerial photographs (Aldiss et al. 1983) (Fig. 13.15). The fold Sumatra. However the Kuantan fauna is of tropical type, while
axial traces trend W N W - E S E and the folds plunge to the ESE. the Alas fauna is of temperate type, indicating that the two units
Variations in the general strike of bedding and cleavage from cannot be correlated directly (Fontaine & Gafoer 1989). The
N W - S E to N E - S W recorded in some areas of the Kluet Kuantan Block is considered to be related to the Indochina or
outcrop are presumed to be due to the effects of the later lbld East Malaya Block, and was emplaced in its present position
phases (Aspden et al. 1982b; Aldiss et al. 1983). In the field by transcurrent movements along the Medial Sumatra Tectonic
pelitic rocks are often tightly to isoclinally folded, and an axial Zone (Hutchison 1994; Barber & Crow 2003).
plane slaty cleavage is developed. Again the cleavage generally Structures in the turbidites of the Kuantan Formation are the
strikes N W - S E or W N W - E S E and dips vertically or steeply to same as those reported from the Kluet Formation, with steep
the SW. dips of both bedding and cleavage, indicating that the rocks are
In the same area a detailed structural study was made of the affected by tight or isoclinal folding. Local variations in strike,
minor structures in the Kluet Formation exposed in a road cut from dominantly N W - S E to east-west, suggest that the rocks
4 km to the south of Sidikalang (Clarke & Bagdja 1979; Aldiss were affected by more than one phase of folding (Aspden et al.
et al. 1983) (Fig. 13.15). The rocks are thick, coarse, massive sand- 1982b; Rock et al. 1983). In the Pakanbaru Quadrangle Clarke
stones interbedded with finer-grained, laminated sandstones and et al. (1982b) report refolding of the cleavage and bedding on
siltstones. The fine-grained sandstones frequently show grading sub-horizontal N W - S E fold axes on near vertical axial planes.
and occasionally small-scale current bedding. Three fold phases Aspden et al. (1982b) suggested that in the Padangsidempuan
were recognized (Fl, F2 and F3). F1 folds are tight to isoclinal, and Sibolga Quadrangle the east-west phase of folding was the
the shape depending on lithology. The axial planes are upright earlier. This interpretation is confirmed by the SAR imagery
to vertical and strike W N W - E S E , but may be horizontal locally which shows tight folds with east-west axial traces refolded by
where the rocks are refolded. Grading in sandstone beds in asym- folds with N W - S E axial plane traces (Figs 13.12 & 13.16).
metrical folds indicate that the limbs of folds are overturned Around igneous intrusions the slates are converted to hornfels,
towards the NE (Aspden et al. 1982b). Slaty axial plane cleavage schist and gneiss.
(S0 is developed in pelitic bands and refracted through graded Further to the south the Terantam Formation in the Duabelas
sandstone beds. Fracture cleavage is developed in semi-pelitic Mountains, 120 km along strike to the SE, the Tarap Formation
bands, and quartz tension gashes occur normal to bedding in of the Garba Mountains in South Sumatra, and the Gunungkasih
thicker and more massive psammites. The fold axes, and a Complex of Lampung have all been correlated with the Kuantan
bedding-cleavage intersection lineation (L1), plunge to the SE. Formation, and although they have not been studied in detail,
It is probable that this is the same phase of folding is seen on a from the descriptions these occurrences are very similar in
large scale to the south on aerial photographs (Fig. 13.15). lithology, structure and metamorphism (Simandjuntak et al.
F2 folds are confined to narrow bands 100-200 m wide. These 1991; Gafoer et al. 1994; Amin et al. 1994b; Andi Mangga
folds refold the bedding, the earlier Fl folds, which may become et al. 1994a). In the Gunungkasih Complex the schistosity
recumbent, and the slaty cleavage developed during F1, on near strikes N W - S E , but is folded about east-west axes and refolded
vertical axial planes. A crenulation cleavage ($2) is developed in by NW-SE-trending upright folds and then by variably oriented
the slates, forming a prominent sub-horizontal crenulation linea- kink-band folds (Barber 2000).
tion (Lz) trending N W - S E . F3 folds are small-scale chevron The Gunungkasih Complex is intruded by gabbros and granites
folds, 1 cm to 10 m in amplitude with fold axes striking from of the Sulan Pluton that have given K - A r ages of 151 • 4 Ma and
west to SW and a lineation plunging NW at shallow to moderate 113 ___ 3 Ma respectively (McCourt et al. 1996). In the same area
angles. Directions of overturning and the predominant SW dip granites and basaltic dykes have been deformed by strike-slip
of the axial planes of both the first and second phase folds, show movements to form banded gneiss. Diorites from this gneiss
a predominant northeasterly vergence. A similar sequence of complex gave a K - A r age of 89 +_ 3 Ma (McCourt et al. 1996).
folds is seen in the road section between Pakkat and Barus some In southern Sumatra deformation had occurred by the Late
25 km to the south of Sidikalang (Fig. 13.15). Jurassic, but continued along shear zones into the mid-Cretaceous
The Kluet Formation is intruded and extensively hornfelsed by (Barber 2000).
the plutons of the Sibolga Granite Complex. A wide range of ages
has been obtained from the main body and satellite intrusions Vorbarisan Unit (Fig. 13.8). The 'Vorbarisan' tectonic unit was
by different dating methods. The oldest is a R b - S r whole-rock proposed by Tobler (1910, 1917, 1919) for the area occupied
age of 264 • 6 Ma (Mid-Permian) from the main outcrop north by Permo-Triassic rocks between the Takung and Musi faults.
of Sibolga, but other widely distributed granite phases in the To the NW the Vorbarisan Unit is transected by the Sumatran
198 CHAPTER 13

Fig. 13.15. Detail of the GRDC Sidikalang Quadrangle sheet showing the outcrop of the Kluet Formation to the SW of Lake Toba (Aldiss et al. 1983). The highly
irregular outcrop pattern is due to the infilling of valleys in a mountainous terrain by tufts from the 70 000 years Bp eruption of the Toba volcano. Bedding strikes
and dips were collected in the field. Traces of bedding are from airphoto and Landsat imagery. Solid lines are faults and dashed lines are possible faults and/or
pholo-lineaments plotted from the imagery.
 STRUCTURE AND STRUCTURAL HISTORY 199

Fig. 13.16. Detail of the GRDC LubuksikapingQuadrangle Sheet (Rock et al. 1983) showingthe outcrop of the KuantanFormation,with the additionof bedding traces
(dashed lines) from SAR imagery. Solid lines are faults. The bedding traces show folds on east-west axial traces refoldedby north-south or NW-SE folds. In the areas
left blank pre-Tertiary rocks are overlain by Tertiary sediments and volcanics.

Fault system of which the Takung and Musi faults appear to be these beds before the rocks were folded. Similar small duplex
splays. Further to the NW the extension of the Vorbarisan Block structures have been observed in the Palepat Formation in the
lies beneath the Woyla Nappe. Batang Tantan area. Throughout the Silungkang, Palepat and
Permian stratigraphic units in this area have been described Mengkarang formations finer grained units within folded beds
under the names of the Silungkang, Palepat and Mengkarang do not show cleavage.
formations. The Silungkang and Palepat formations are composed The relatively undeformed nature of these Permian units in
of lavas and tufts, interbedded with shales, siltstones, sandstones contrast to the Kuantan Formation to the NE and the slates and
and crystalline limestones, in the Palepat Formation passing up phyllites of the 'Schiefer Barisan' to the SW, together with the
into an upper Limestone Member. The massive lavas and the lime- volcanics and the Cathaysian 'Jambi Flora' in the Mengkarang
stones in these units are faulted and jointed and the thinner bedded Formation, led Zwierzycki (1935) to suggest that they formed an
units are sometimes strongly folded (Suwarna e t al. 1994). overthrust 'Jambi Nappe'. Because of the affinities of the
However, coal bands and plant fossils in the Mengkarang For- Permian lavas and the Cathaysian flora to those of East Malaya
mation are well-preserved, showing that the rocks have not been Zwierzycki (1935) suggested that this nappe was overthrust
metamorphosed. In the Batang Tembesi south of Muarabungo, from the northeast over a distance of 350 km. Van Bemmelen
sandstones of the Mengkarang Formation at Pulau Bayer are (1949) considered that this amount of movement was too great
folded into an anticline on an east-west axis with an overturned to have occurred during a single phase of movement (in the
northern limb. In the limb of the fold thin-bedded sandstone Cretaceous Varanginian Stage according to Zwierzycki 1930a)
layers are imbricated along small scale thrusts, indicating that and suggested in his 'undation hypothesis' that the nappe had
thrust movements directed towards the west had occurred in been gravitationally moved westwards by successive uplifts
200 CHAPTER 13

cored by plutonic belts from the Triassic to the Cretaceous. The are folded by open asymmetrical folds overturned to the NNE,
nappe hypothesis has been discounted by more recent authors with a spaced axial plane cleavage cutting across the earlier
(Katili 1970; Cameron & Pulonggono 1984; Hutchison 1994; cleavage, and an intersection lineation plunging at c. 30 ~ to the
Barber & Crow 2003). Barber & Crow (2003), following WNW. Sandstone bands up to 40 cm thick are fractured at right
Hutchison (1994), have suggested that the Vorbarisan Unit and angles to the bedding and the fractures are filled with quartz;
the Kuantan Unit constitute a West Sumatra Block separated similarly limestones are cut by calcite veins.
from Cathaysia and emplaced in Sumatra along a major transcur- The Peneta Formation covers the same age range as the Rawas
rent fault, represented by the Medial Sumatran Tectonic Zone, Formation and the description is very similar to that of the non-
to the SW of the Sibumasu Block. volcanic, finer-grained parts of the Rawas Formation (Suwarna
et al. 1994). It consists of slates, shales, siltstones, sandstones
Schiefer Barisan Unit (Fig. 13.8). The Schiefer Barisan tectonic unit and meta-limestones. The siltstones are strongly folded with a
lies to the SW of the Vorbarisan Unit from which it is separated by slaty cleavage striking N W - S E , emphasized by new mica growth.
the Musi Fault; it is overlain further to the SW by the Woyla The Asai, Peneta and Rawas formations are considered to
Nappe. The Permian Barisan, the Triassic Tuhur and the Jurassic have been deposited in shallow-water environments, passing into
and Lower Cretaceous Asai, Rawas and Peneta formations out- deeper water in a foreland or forearc basin on the SW margin of
cropping in the unit, as the name of the implies, are characterized Sundaland (Pulunggono & Cameron 1984; Barber 2000). Defor-
by penetrative cleavage and foliation, in contrast to the Permian mation, with the development of folds and slaty cleavage occurred
and Triassic units of the Vorbarisan Unit. in the mid-Cretaceous, later than the youngest sediments but
The Barisan Formation shown cropping out to the south of earlier than the intrusion of Late Cretaceous granites. Deformation
Solok is composed of phyllite, slate, sandstones, limestones and of the other units of the Schiefer Barisan, the Tuhur and Barisan
cherts (Rosidi et al. 1976). Lower Permian fusulinids in massive formations, presumably occurred at the same time. The defor-
limestones in the eastern part of the outcrop shown as Barisan mation is attributed to the collision and overthrusting of the
Formation on the geological map indicate that these rocks, Woyla volcanic island arc and its associated accretionary
although more highly deformed, should be correlated with the complex over the margin of Sundaland in mid-Cretaceous times.
Silungkang and Mengkarang formations of the Vorbarisan Block The deformed low-grade metamorphic rocks of the Schiefer
(Fontaine & Gafoer 1989). The foliation in the phyllites and Barisan mark the footprint of the Woyla Nappe. Folds overturned
slates strikes N N W - S S E and limestone lenses are elongated in towards the NE and SW-dipping cleavage indicate that the nappe
the same direction. Later deformation is indicated by kinking of was emplaced from the SW and overthrust towards the NE. The
the slaty cleavage and small scale shear zones (Rosidi et al. 1976). absence of penetrative cleavage in the Permian rocks forming
The Triassic Tuhur Formation outcrops in the area from Lake the Vorbarisan Unit suggests that they were never covered by
Singkarak to Dibawah and Diatas lakes (Silitonga & Kastowo the nappe. The Takung and Musi faults which separate the
1975; Rosidi et al. 1976). Lithologies include argillaceous Kuantan, Vorbarisan and Schieferbarisan units may be older
sediments with brown cherts, thin turbiditic sandstones and thin structures, but have been reactivated in the Neogene, during
limestones which resemble those described from the Kualu movements along the Sumatran Fault.
Formation of northern Sumatra. Silitonga & Kastowo (1976)
distinguished a Slate and Shale Member and a Limestone
Member. The Slate and Shale Member, in which the slaty cleavage
strikes N W - S E , occupies the greater part of the outcrop. Woyla N a p p e (Fig. 13.8)
The Limestone Member includes limestone conglomerates that
contain blocks of Lower to Middle Permian fusulinid limestone The Woyla Group crops out discontinuously in the Barisan
(Silitonga & Kastowo 1975). Similar conglomerates were Mountains along the west coast of Sumatra from Banda Aceh in
described to the north of the equator by Turner (1983) from the the north, through Natal and Padang in central Sumatra, to the
Cubadak Formation and his Limau Manis Formation, in an area Gumai and Garba Mountains and Bandar Lampung in the south.
which was mapped by Rock et al. (1983) as part of the Silungkang The further extent of the Woyla Group in southern Sumatra can
Formation. The Cubadak and Limau Manis formations contain be traced in oil company boreholes beneath the Tertiary sedimen-
Halobia and ammonites indicating a M i d - L a t e Triassic age. tary cover (Barber & Crow 2003). A comprehensive review of the
The limestone conglomerates indicate that Permian limestones Woyla Group in Sumatra, based on the work of Bennett et al.
were uplifted during the formation of a horst and graben structure (198 i a, b) and Cameron et al. (1982, 1983) in northern Sumatra,
and were subjected to erosion between the Mid-Permian and the Rock et al. (1983), Wajzer et al. (1991) and McCarthy et al.
Mid-Triassic (Barber & Crow 2003). (2001) in central Sumatra and Gafoer et al. (1992c, 1994) in
The Middle Jurassic to Lower Cretaceous Asai, Rawas and southern Sumatra, has been given by Barber (2000). Barber's
Peneta formations outcrop in the foothills of the Barisan (2000) review covered the areas of outcrop, the lithologies and
Mountains to the southwest of Bangko and Sarolungan (Suwarna structures, the environments of deposition of the stratigraphic
et al. 1994). The Asai Formation is composed of greywacke, units or formation of the volcanic units, palaeontological and
meta-sandstone, siltstone, slate, phyllite and limestone. The isotopic evidence of their age, and presented a tectonic synthesis
slates are blue-grey or reddish in colour with a strong penetrative of their origin and emplacement as the Woyla Nappe on the
cleavage striking N W - S E and dipping steeply. Sandstones are cut southwestern margin of Sundaland (Fig. 13.8).
by quartz veins and limestones by calcite veins. Fossils indicate Cameron et al. (1980) distinguished two lithological assem-
that the Asai Formation is of Middle Jurassic age and it appears blages in the Woyla Group of northern Sumatra, an oceanic
to be the oldest of the three formations (Fontaine & Gafoer 1989). assemblage and a volcanic arc assemblage (Fig. 13.17) shown
The Rawas Formation consists of basalt lava flows with on the 'Simplified Geological Map of Northern Sumatra'
intrusive dolerite dykes, associated with conglomerates, grey- (Stephenson & Aspden 1982). This distinction was extended to
wacke sandstones and siltstones, described as a turbidite sequence cover other outcrops of the Woyla Group throughout Sumatra
(Suwarna et al. 1994). Extensive outcrops of grey slate with thin (Barber 2000).
sandstone and siltstone bands and limestone lenses of the Rawas The oceanic assemblage, which generally lies to the NE of the
Formation are exposed in the Rawas River. Slaty cleavage arc assemblage, consists of serpentinites, gabbros, mafic to inter-
strikes N W - S E and dips at 40 ~ to the SW. The bedding lamination mediate volcanic rocks, commonly basaltic and showing pillow
in the siltstones is parallel to the cleavage, but the cleavage is axial structures, hyaloclastites, volcaniclastic sandstones, red radiolar-
planar to small isoclinal folds in limestone lenses. The slates ian cherts, red and purple manganiferous shales, sometimes with
 STRUCTURE AND STRUCTURAL HISTORY 201

%
I I
96~
o
Cretaceous granites
BANDA ACEH WOYLA GROUP
Lho'nga ::.,,:,.,\. Oceanic assemblage
Lamn( s Arcassemblage
volcanics/limestones
i~]lll~/~ .~ Geumpang~'---~---,,~ ~ Meukuk Gneiss
Bentaro
Volcanics
Lam Minet m Undifferentiated
-5ON Peridotite/serpentinite 5~

-"Batholith-".,-'. euem
Tel
CALANG " ~ , . - ~ tke

Penarum
"~ " ~ ~ i " . . . . %>%.

r :undifferentiated

0 ~
4~ 0 4~
4" ~abahrot
-~- Faults
'~ Thrusts
SFZ Sumatran Fault System
KL Kla Line Tapaktuan
GL Geumpang Line Volcanics
Meukuek
0 50 100km
.....t It -7--:: :::::==========================
. . . . . . . . .

TAPAKTUAN
96~ 97~
I I
Fig. 13.17. Outcrops of the Jurassic-Cretaceous Woyla Group in Aceh, northern Sumatra, modified from Stephenson & Aspden (1982).

manganese nodules, and rare limestones. These rock units form crust ranging in age from Triassic to mid-Cretaceous was being
lens-shaped outcrops, usually separated by steep faults that show subducted or imbricated to form the accretionary complex. A
slickenside evidence of thrust, and sometimes strike-slip, move- K - A r isotopic age of 105 + 3 Ma was reported by Koning &
ment (Wajzer et al. 1991). Interlayered melange units, composed Aulia (I985) from a tuff at Indarung.
of angular fragments of the other units in a clay or serpentinous The volcanic arc assemblage which lies along the west coast
matrix, occur within the oceanic assemblage (Fig. 13.18). of northern Sumatra in Aceh is described as the Bentaro Volcanic
The oceanic assemblage is interpreted as an ocean-floor Formation and consists of basaltic to andesitic volcanics, which
sequence composed of serpentinized mantle peridotite, gabbroic are not pillowed, and volcaniclastic sandstones (Bennett et al.
and basaltic oceanic crust, with overlying oceanic sediments, 1981a) (Fig. 13.17). The volcanics are associated with massive
imbricated at a subduction zone to form an accretionary complex. to bedded limestones with a variety of formation names, of
Blocks of massive limestone, sometimes occurring in the mel- which the Teunom Limestone Formation is typical (Bennett
anges, have been interpreted as derived from the carbonate et al. 1981b). The assemblage is interpreted as representing an
cappings of sea-mounts. Triassic foraminifers from a massive oceanic island arc with carbonate fringing reefs and its sedimen-
limestone block in melange in Natal (Wajzer et al. 1991), Mid- tary apron (Cameron et al. 1980; Barber 2000). A similar assem-
Jurassic radiolarian fossils from cherts at Indarung near Padang blage of rock types crops out in the Gumai Mountains inland from
(McCarthy et al. 2001), Late Jurassic to Early Cretaceous stroma- Bengkulu in southern Sumatra, where they are identified as the
toporoids, corals and foraminifers also at lndarung (Yancey & Alif Saling Volcanic Formation, the Lingsing Formation, composed
1977) and in Aceh (Cameron et al. 1983) indicate that oceanic of volcanics with interbedded sediments, and the Sepingtiang
202 C H A P T E R 13

r -d
..t r./]

Z
~b
E
z~ o r
I 0 "~=

0 .~ "~
"'E ;~ ~ o
,.r
a
0 E~ ,'~ ~ -- 2~ ".:
"" c~ ~

%
O--

r-
,i/
Z / &
0
Z

Z~
eh ~

<

~S
tm ~ 9
o
.1
-1

_~-=
Z
0 0

0 "~

Z
0
~a
,-'~
I ..............
~ 0
 STRUCTURE AND STRUCTURALHISTORY 203

Limestone Formation (Gafoer et al. 1992c). There are no isotopic deformed, although mylonitized limestones in which the foliation
ages from any of the volcanic rocks, but the limestones have was affected by kink-band folds were observed near Banda Aceh
yielded Late Jurassic to Early Cretaceous corals, stromatoporoids (Barber 2000). Bedded limestones, however, are commonly folded
and forams (Bennett et al. 1981a, b) and an Albian (mid- on a large scale, as seen in the quarries in the Lho'nga Formation
Cretaceous) foraminifer was obtained from the Sepingtiang also near Banda Aceh (Bennett et al. 198 la) (Fig. 13.17). Two fold
Formation, indicating that the oceanic and arc assemblages were phases were seen, F~ isoclinal folds with vertical to steeply dipping
contemporaneous. axial planes with the intersection of bedding and cleavage
A detailed study of the imbricated oceanic assemblage in a plunging at a low angle to the SE. The cleavage is folded into
2 0 k m road and river section at Natal on the west coast of small crenulations. Elsewhere in the quarry the first fold phase
Sumatra near the equator was carried out by Wajzer et al. is represented by tight folds with cylindrical cores and recumbent
(1991; see also Chapter 4) (Fig. 13.18, inset). Lithologies axial planes on W N W - E S E axes and are cut by a second phase
include serpentinite, pillow basalt, bedded chert, volcaniclastic folds on N E - S W axes. Where limestones are interbedded with
sandstone, shale and melange. They found that lithological units, shales they show pinch-and-swell structures, boudinage and
1 - 2 km in width, trending N W - S E , were separated by steeply calcite-filled tension gashes normal to the bedding.
dipping or vertical faults. Also the units are disrupted internally There are variations in the structural trends in the scattered
by faults every few metres. Some fault surfaces show slickensides outcrops of rock units correlated with the Woyla Group in
indicating normal or reversed movements, while others show sub- central and southern Sumatra. In the Siguntur Formation, cropping
horizontal slickensides indicating strike-slip movement. There is out to the south of Padang, the general strike of the bedding and
no apparent pattern in the sequence of lithological units and slaty cleavage is east-west (Rosidi et al. 1976). In the Gumai
some units are repeated several times in the section (Wajzer Mountains east of Bengkulu the contact between the Saling and
et al. 1991). The grade of metamorphism varies along the Lingsing formations trends east-west, and while the rocks
section with some units being metamorphic schists and slates of are reported to be highly deformed and folded, the strike of
prehnite-pumpellyite greenschist facies, while others are unmeta- the bedding and cleavage trends north-south (Gafoer et al.
morphosed. There is no apparent pattern in the state of meta- 1992c). A massive limestone, the Sepintiang Formation, rests dis-
morphism, with greenschist facies schists juxtaposed against cordantly across the contact of the Saling and Linsing formations.
unmetamorphosed units. The limestone evidently represents a fragment of a fringing reef
In Natal where foliation and slaty cleavage is developed the emplaced tectonically over the other formations. In the Gumai
general trend of the strike is N W - S E with steep but variable Mountains, to the SW of Baturaja, outcrops of volcanics, cherts
dips. Some finer-grained units, including the clay matrix of and mrlanges are associated with metamorphosed Palaeozoic
melanges units, show isoclinal folding (Ft) with axial plane rocks and bounded by NW-SE-trending thrust faults (Gafoer
cleavage. The highest grade unit, the Si Gala Gala Schist is a et al. 1994; Barber 2000). Foliation in the scaly matrix of the
quartz-muscovite-chlorite schist in which the schistosity ($1) m~langes also trends N W - S E ; blocks enclosed in the mrlange
contains a rodding lineation (Lj). In some units the foliation, are elongated in the same direction and cut by tension fractures
schistosity or cleavage and the earlier formed folds are refolded normal to their long axes. In the Garba Volcanic Formation two
by open to close folds (F2) on N N W - S S E or N W - S E axes, lbld phases are distinguished; the first, with east-west axes, is
with the development of crenulation cleavage and crenulation refolded by later folds on N W - S E axes (Gafoer et al. 1994).
and intersection lineations. Again there is no pattern in the Radiolarian and foraminiferal fossils found in the oceanic
amount of deformation along the section, units with simple defor- assemblage of the Woyla Group show that the ocean floor of
mation being juxtaposed against those with multiple deformation. which it formed a part existed from Triassic to Early Cretaceous
Similar distributions of lithological units, and variations in times. Mid-Cretaceous foraminifers from the Sepingtiang
metamorphism and deformation are reported from the oceanic Formation and Late Jurassic to Early Cretaceous fossils that
assemblage of the Woyla Group in Aceh (Bennett et al. 1981a, commonly occur in limestone members of the volcanic arc assem-
b; Cameron et al. 1982b, 1983) with the finer-grained sedimentary blage, together with the K - A r age of 105 + 3 Ma from a tuff at
and volcaniclastic rocks converted to slates and phyllites. Higher- Indarung (Koning & Aulia 1985) show that the volcanic arc was
grade metamorphic rocks of the Meukek Gneiss Complex occur in constructed on the ocean crust in the Late Jurassic and remained
a fault-bounded block in the Woyla Group to the north of active until mid-Cretaceous times. Granites were intruded into
Tapaktuan, including a garnet-biotite amphibolite with garnets both the oceanic and the volcanic arc assemblages of the Woyla
up to 8 cm in diameter (Cameron et al. 1982b). An area of high- Group after they were accreted to the western margin of
grade metamorphic rocks between strands of the Sumatran Fault Sumatra. These include the Sikuleh Batholith in Aceh, the
described as 'undifferentiated', which includes coarse banded Manunggul and Kanaikan batholiths in Natal and the Garba
marbles, hornblende schists and mylonitized biotite-garnet- Pluton in the Garba Mountains. The Manunggal Batholith has
staurolite schist (Cameron et al. 1983) has probably been included yielded a Late Cretaceous age of K - A r age of 87.0 Ma (Kanao
in the Woyla Group erroneously and may belong to the Kluet et al. 1971, reported in Rock et al. 1983). The Woyla volcanic
Block or possibly the continuation of the Medial Tectonic Zone. arc and its associated oceanic crust were evidently accreted to
In Aceh the general trend of the lithological units, separated by and thrust over the western margin of Sumatra in early Late
faults and thrusts, and of the strike of the schistosity, foliation Cretaceous times (Barber 2000).
and cleavage, is N W - S E , with moderate to steep dips. Isoclinal
folds with an axial-plane cleavage can be seen wherever the
bedding lamination can be distinguished, with a bedding-
cleavage intersection lineation plunging to the SE. The slaty clea- The Sumatran Fault Zone
vage is sometimes refolded by more open folds on subhorizontal
axes and NE-dipping axial planes, with the development of sec- The Barisan Mountain Range is split along its length by the
ondary cleavages and lineations. Large-scale upright folds with N W - S E Sumatran dextral transcurrent fault system (Fig. 13.19),
a 7 km amplitude and subsidiary folds on the scale of 1 - 2 km a transform fault linking the Andaman Sea spreading centre in
are reported from the Tapaktuan Quadrangle (Cameron et al. the north to an area of spreading in the Sunda Strait in the
1982b). south. From the Andaman Sea to the Sunda Strait the Sumatran
The massive volcanic units of the Bentaro Formation and the Fault is c. 1900 km long, and cuts through all the rock units in
limestones of the Tuenom Formation and its equivalents are Sumatra, including Recent volcanic tufts and alluvial sediments.
faulted, fractured and jointed, but are not obviously internally The overall shape of the fault is a lazy S, the segment to the
204 CHAPTER 13

Fig. 13.19. A simplifiedmap of the Sumatran Fault System in its tectonic setting, showing the locationof figures illustratingdetailed sections of the fault systemdiscussed
in this account. Inset map shows the distribution of areas of subsidence, forming grabens, and areas of uplift along the trace of the Sumatran Fault, which Holder et al.
(1994) attribute to the formation of subsiduary splays with strike-slip movement, due to transpression across the fault, with the principal compressive stress (o-I) oriented
ENE-WSW.

north of the equator being concave to the SW, while the segment concerning the time of initiation and the amount of displacement
to the south is concave to the NE. The fault is currently active along the fault. In some parts of its length the active fault trace
along much of its length, as indicated by frequent historic and is a single discontinuous strand, with mainly right step-overs,
recent earthquake shocks and measured rates of differential but in other areas it bifurcates and splits into a number of
m o v e m e n t across the fault using GPS measurements. Splays of strands that may rejoin to isolate fault blocks, some of which
the main fault extend into the forearc and also into the backarc have subsided to form lakes, or have been partially or completely
region. It is probable that prominent Pre-Neogene faults mapped filled by Quaternary lacustrine and fluvial sediments.
in the backarc area have been reactivated in association
with more recent movements along the main fault trace. This Age of the Sumatran Fault System. The time of the initiation of the
has not always been appreciated and may have led to confusion fault and the amount of m o v e m e n t along the fault have been
 STRUCTURE AND STRUCTURAL HISTORY 205

matters of continual speculation. Since the Sumatran Fault is a displaced by 150 km between Natal and Lake Toba by dextral
transform fault, clearly related to the Andaman Sea spreading movement along the Sumatran Fault (Fig. 13.8). McCarthy &
system, it is most reasonable to suppose that it was initiated in Elders (1997) also suggested a displacement of 150kin in
the Mid-Miocene (c. 13 Ma) together with the present phase of central Sumatra from a possible correlation between the mid-
opening of the Andaman Sea (Curray et al. 1978; Curray 1989; Jurassic Siguntur Formation on the SW side, with the contempora-
McCarthy & Elders 1997). The trace of the fault is commonly neous Asai Formation on the NE side of the fault.
seen cutting Quaternary sediments and volcanics, but sometimes In central Sumatra Posavec et al. (1973) identified east-west-
mylonites are exposed in outcrops along the line of the fault, trending aeromagnetic anomalies that cut across the Sumatran
indicating that the fault had an earlier history of movement Fault. These anomalies correspond with volcanic centres and are
(McCarthy 1997). Also Wajzer et al. (1991) report N W - S E attributed to dioritic intrusions at depth. They found that a series
strike-slip faults in the Woyla Group in the Natal area of North increasingly deeply eroded volcanic edifices extend to the NW
Sumatra which are cut by the Late Cretaceous Manunggal of the Maningjau Centre as far as Padang, and suggest that this
Batholith and Madingding Diorite, and Barber (2000) reports indicates that the crust to the SW of the fault has moved northwest-
N W - S E foliated syntectonic granitic and basaltic intrusions ward with respect to the volcanic centre, indicating dextral displa-
dated by the K - A r method at 89 + 3 Ma in (mid-Late Cretac- cement for a distance of 90 km (Fig. 13.25, inset). On the other
eous) in the Sekampung River near Bandarlampung in southern hand volcanic edifices are displaced southeastwards by 35 km
Sumatra, suggesting that strike-slip movements occurred along on the northeastern side of the fault, giving a total relative displa-
the same trend as the Sumatran Fault during the Late Mesozoic. cement of 125 km. This movement must have occurred during the
Pulunggono et al. (1992) have interpreted a series of W N W - Quaternary.
ESE lineaments recognized in SAR imagery in the Tertiary Beaudry & Moore (1985), from their study of the distribution
sediments of the backarc area as the traces of successive strike- of facies in the West Aceh and West Sumatra forearc basins
slip faults in the basement which were developed in the Sumatran suggested that these two basins were contiguous in Mid-
continental margin during the Mesozoic. They suggest that the Miocene times. Since that time the West Aceh Basin has been
most northerly of these lineaments to the south of Palembang displaced by some 65 km along the Batee Fault, a splay of the
is of Triassic age and that the lineaments become progressively Sumatran Fault (Fig. 13.20). Further evidence for movement of
younger towards the SSW. this order of magnitude was presented by Kallagher (1989) from
her study of the West Aceh Basin. Here, fine-grained clastic and
Displacement along the Sumatran Fault System. During the primary calcareous sediments of Lower Miocene age are juxtaposed
mapping of Sumatra it was appreciated that the Barisan Mountains across the Batee Fault against coarse volcaniclastic deposits of
were bisected by a series of discontinuous rift valleys, a 'longi- the same age. These deposits must have been separated by tens
tudinal valley', which extended all the way from Aceh in the of kilometres in Early Miocene times, before they were juxtaposed
north to Semangka Bay in the south (van Es 1919). Van Bemmelen by movements along the fault. Malod & Kemal (1996) suggest that
(1949) interpreted this longitudinal rift system as due to the domal the northern part of the Sumatra Forearc constitutes an indepen-
uplift of the Barisans with extension and the collapse of a central dent Aceh Plate, bounded by the Mentawai, West Andaman,
'keystone'. Durham (1940) was the first to recognize the strike- Batee and the northern segment of the Sumatran Fault.
slip nature of the fault in its central section and subsequently In central Sumatra, Hahn & Weber (1981b) proposed 42 km
this was recognized for other segments of the fault. The first of dextral displacement from the correlation of coarse- and fine-
description of the nature of the Sumatran Fault Zone in modern grained facies in the Permo-Triassic Air Mabara and Sopan
terms was given by Katili & Hehuwat (1967), who also presented granites across the Lubuksikaping Fault (Rock et al. /983)
evidence from the displacement of buildings and other structures (Fig. 13.24). Katili & Hehuwat (1967), Posavec et al. (1973)
for the amounts of strike-slip movement along segments of and Sieh & Natawidjaja (2000) have reported dextral offsets
the fault during earthquakes, and over a longer term from stream of 2 0 - 3 5 km from stream courses which cross the trace of the
displacements. Sumatran Fault. Again these movements must have occurred
The overall amount of movement along the fault can be deduced during the Quaternary.
from its tectonic setting as a transform fault between the Andaman
Sea spreading centre and the zone of extension in the Sunda Strait. Current movements along the Sumatran Fault System. In the dis-
Segments of continental crust on either side of the Andaman cussion of the structural evolution of the forearc above, it was
spreading ridge are now separated by c. 460 kin, with Indian pointed out that many authors have proposed that during the
Ocean side moving northwards with respect to the rest of SE oblique subduction of the Indian Ocean Plate beneath Sumatra
Asia. To the south this movement is tranformed into the West the convergence between the two plates is partitioned between
Andaman Fault, or into strands of the Sumatran Fault System. thrusting normal to the subduction trench and shearing parallel
At the southern end of the Sumatran Fault System the amount of to the arc (Fitch 1972; Hamilton 1979; Jarrard 1986; Curray
extension in the Sunda Strait between Java and Sumatra has 1989; McCaffrey 1996). Shearing parallel to the arc is taken
been estimated as 100 km since the Miocene, this extension up along the Sumatran Fault System, including the Batee and
being taken up by movements along the fault zone (Huchon & Mentawai faults.
Le Pichon 1984; Harjono et al. 1991; Malod & Kemal 1996; The obliquity of convergence increases northwestwards along
Sieh & Natawidjaja 2000). the arc, from zero opposite Java, where convergence is normal
Many suggestions for the amount of movement along the to the trench, to 45 ~ opposite north Sumatra. Increasing obliquity
Sumatran Fault Zone in the Barisan Mountains have been pro- is matched by an increase in the rate of movement along the
posed from the displacement of units which have been correlated Sumatran Fault System, confirmed by measurements of actual
across the fault. Page et al. (1979) found a displacement of at movement along the fault by the displacement of recent sediments,
least 200 km in northern Sumatra from lithium values from from the differential movement of trigonometrical survey points
stream sediment samples, which are commonly > 6 0 ppm on the over the last 100 years, and from repeated GPS surveys over the
NE and < 3 0 ppm on the SW side of the fault. However, this past few years (cf. Prawirodirdjo et al. 2000). Slip rates are
could be because the basement to the NE is largely composed of calculated as c. 6 mm a-z at the southern end of the fault near
the Tapanuli Group of continental derivation, while to the SW, the Sunda Strait (Bellier et al. 1991, 1999), < 1 0 m m a -1 near
basement is the Woyla Group composed of rocks of volcanic- 5~ (Bellier et al. 1991), c. 1 0 m m a - l at the equator and
arc and ocean-floor origin. In north central Sumatra the Medial 28 mm a - l near 2.2~ (Sieh et al. 1994). At the northern end of
Sumatra Tectonic Zone, identified during the present study, is the fault, the rate of opening of the Andaman Sea is calculated
206 CHAPTER 13

Fig. 13.20. Dextral displacement of the Meulaboh and West Sumatra Forearc basins, the reef complex (brick pattern) and the shoreline along strands of the Sumatran
Fault System since the Mid-Miocene (from Beaudry & Moore 1985).

as 40.4 mm a-1, averaging 37.2 mm a-J since the Mid-Miocene along the fault in the 1964 earthquake. However, Plio-Pleistocene
(Curray 1989). sediments are highly deformed against the fault, showing that it
As has been discussed in the account of the forearc above, was certainly more active in the recent past. A depression filled
this differential movement implies that the forearc area is being with recent alluvium and volcanic products, the Banda Aceh
extended, the extension being accommodated within a forearc Embayment, lies between the Seulimeum and Aceh faults
sliver by movements along the Batee and Mentawai faults and (Fig. 13.22). Towards the south rocks of the Woyla Group
along minor strike-slip and extension faults within the forearc forming the basement, rise from beneath the younger sediments
islands and the accretionary complex. of the embayment to form a horst block. Genrich et al. (2000)
It has been demonstrated by GPS measurements that at the estimate a rate of displacement across the two strands of the
present time the forearc to the south of the Batu Islands is Sumatran Fault at 15 mm a -~, with the Banda Aceh Embayment
moving in the same direction as the Indian Plate, but at a slower being extruded towards the NW at a rate of 5 + 2 mm a -1
rate, while the forearc to the north has a component of movement Pre-Tertiary and Tertiary rocks cropping out on the southwes-
northwards, parallel to the Sumatran Fault (Prawirodirdjo et al. tern side of the Aceh Fault are affected by a discontinuous series
1997) (Fig. 13.7). of thrusts, the Geumpang Line (Fig. 13.22). The thrusts are
Sieh & Natawidjaja (2000) have recently prepared detailed steep against the fault, but flatten towards the SW, becoming
maps of the active traces of the Sumatran Fault System, based horizontal in Gle Cuplet (Bennett et al. 1981a). The thrusts
on their geomorphic expression, using 1:50 000 topographic bring together Tertiary sediments and different units of the
maps, 1:100 000 stereoscopic aerial photographs, the 1:250 000 Woyla Group, which are also thrust across the Late Cretaceous
geological maps and SAR imagery at the same scale. They Sikuleh Batholith. Serpentinites and serpentinous m~lange,
divided the fault into 19 segments, named after major rivers or presumably derived from the oceanic assemblage of the Woyla
bays within the segment, and show the relationship of the fault Group, sometimes outcrop along the thrusts. M~lange near
traces to active volcanoes (Fig. 13.21). Rumah Baru contains blocks of Early-Mid-Miocene fossiliferous
limestone (Bennett et al. 1981a; Cameron et al. 1983), while Ni
The Sumatran Fault System in Aceh. At the northern end of the and Cr anomalies in Plio-Pleistocene sediments show that the
Sumatran Fault System in Aceh the fault bifurcates into two serpentinites had been uplifted and exposed to erosion during
strands, the Seulimeum and Aceh faults (Fig. 13.22). Geomorphic the Neogene. Cameron et al. (1983) attribute the development of
features show that the Seulimeum Fault has been active recently, the thrusts to transpression due to the northwards movement
as it cuts through Plio-Pleistocene sediments and volcanic pro- of the forearc sliver into the constraining bend formed by the
ducts of the active Seulawai Again volcano; hot springs occur SW concavity at the northern end of the Sumatran Fault System.
at its southern end. The fault transects and displaces the axial To the SE of Rumah Baru the Sumatran Fault bifurcates to
traces of east-west trending folds affecting Pliocene deposits form the Anu Batee and Blangkejeren faults, and again further
with the SW side of the fault having been moved northwestward. south to form the K l a - A l a s Fault (Cameron et al. !983)
Bennett et al. (1981a) suggest that this displacement amounts (Fig. 13.22). The outcrop of the Blangkejeren Fault is marked
to 5 km, but Sieh & Natawidjaja (2000) suggest a movement by a zone of gouge; breccia, phyllonite and m~lange. Thick
of 2 0 k m , corresponding to stream offsets along the fault conglomerates in the Peutu Formation adjacent to the fault indi-
further south. cate that the fault was active during the early Mid-Miocene. The
The Aceh Fault does not show any geomorphic effects of recent A n u - B a t e e Fault extends southwards into the offshore region,
movement and is considered to be currently inactive, although where it influenced the development of the Sumatran forearc
Soetadi & Soekarman (1964) reported that a school and other basins (Beaudry & Moore 1981). From the landward part of the
buildings were displaced by up to 0.5 m in a N W - S E direction fault Sieh & Natawidjaja (2000) report that several of the larger
 STRUCTURE AND STRUCTURAL HISTORY 207

Fig. 13.21. Active traces of the Sumatran Fault System identifiedby their geomorphicexpression, fault segments and estimated rates of dextral movement, the location of
active volcanoes, lakes and extensional graben (from Sieh & Natawidjaja 2000).

river channels appear to be displaced across the fault by distances 12 + 5 m m a-1 from the offset on three stream courses and the
up to 1 0 k m , while smaller stream courses are unaffected, estimated age of the streams, calibrated against m o v e m e n t on
suggesting that there have been no m o v e m e n t s along this fault the main Sumatran Fault where it cuts through the c. 70 000 year
for the past tens of thousands of years. On the other hand Bellier Toba Tufts, and therefore the time of initiation of the stream
& S~brier (1995) have calculated a present rate of m o v e m e n t of courses is known.
208 CHAPTER 13

Fig. 13.22. Thrust structures related to the

northern end of the SumatranFault Systemin
Aceh. C-P, Carboniferous-Permian
Tapanuli Group; P-T, Permo-Triassic
Peusangan Group; ,l-K, Jurassic-Cretaceous
Woyla Group; Tom; TertiaryOligo-Miocene
sediments.

The east-west-trending Kla Line (thrust) (Fig. 13.22), between to 1000-2000 m on each side of the graben, which forms a long
the Kla-Alas and Blangkejeren faults, which brings the Permo- narrow depression (75 km long and 9 km wide at its widest
Carboniferous Kluet Formation to rest on the Jurassic-Cretaceous part), with a floor at 180-200 m, occupied by Quaternary to
Woyla Group, is attributed to Late Cretaceous tectonism Recent alluvium.
(Cameron et al. 1983). Near Takengon the east-west Takengon The emplacement of the Kembar Volcano and the subsidence
Line, a southward-directed thrust, bringing Permo-Triassic of the Kutacane Graben are attributed to transtension within a
Peusangan Group over the Woyla Group and Oligocene sediments releasing bend on the concave side of the complex of faults
was formed prior to the deposition of the Peutu Formation which which forms the the Sumatran Fault System in this area. The
is unaffected by the thrust (Cameron et al. 1983). The outcrop of bounding faults cut the products of the Kembar Volcano and
the thrust forms a marked topographic feature where Peusangan displace alluvium at the northern and southern ends of the
limestones rest on soft Tertiary sediments. At its western end graben, indicating that recent movement has occurred along
the Takengon Line links with the dextral strike-slip Geureuggang the faults (Cameron et al. 1982a). To the south of the graben
Fault which extends to the north coast (Fig. 13.22). the Toru Fault cuts and displaces the 73 000 year Toba Tufts,
Movements along the Geureuggang Fault, the Takengon Line giving an average rate of movement since their eruption of
and the formation of the east-west folds in Pliocene sedi- 27 mm a - j (Sieh & Natawidjaja 2000); GPS measurements
ments are due to north-south compression. Curray (I989, Fig. 1) indicate that the current rate of movement along the Toru Fault
suggested that a southward-directed subduction system had is 26 __ 2 mm a - i (Genrich et al. 2000).
developed in the Andaman Sea off the north coast of Sumatra
(Fig. 13.22) which could account for the compression. This postu-
lated subduction system has also been invoked to account for the The Equatorial b~[urcation. Between l~30'N and the equator the
volcanoes lying to the east of the general trend of the volcanic arc Sumatran Fault System splits into two branches, which enclose a
in northern Sumatra (see Chapter 7). But these volcanoes are lens of structurally complex geology (Fig. 13.24). This structure,
much more likely to be related to the eastward-dipping Sunda s- formed by the Barumun and Angkola fault segments, is termed
ubduction system (see the contours on the Indian Plate in Seih the Equatorial Bifurcation by Sieh & Natawidjaja (2000)
& Natawidjaya 2000) and there is no other evidence for southward (13.21). Rock et al. (1983) suggest that the E N E - W S W trend of
subduction, so that there is no obvious cause for the north-south the lithological units within the fault block, compared with the
compression. general N W - S E trend of the rock units outside it, indicate that
the lens has been rotated c. 30 ~ in an anticlockwise direction by
movements along the bounding faults. In the Barumun segment,
Kembar Volcano and the Kutacane Graben. To the SE of Aceh the movement along the Lubuksikaping Fault, which is concave
K l a - A l a s and Blangkejeren faults define the SW margin of a towards the SW, has formed the Rau Graben in a releasing bend
faulted block, with the Lokop-Kutacane Fault on its NE at its southern end. The floor of the Rau Graben lies at 300 m
margin, into which the active Kembar Volcanic Centre has been with the mountains on either side rising to heights of 600-1700 m.
emplaced (Fig. 13.23). Further south the Lokop-Kutacane Fault Based on their mapping programme in central Sumatra Hahn &
passes into the Toru Fault which forms the NE margin of the Weber (1981b) suggested that the Sopan Granite on the
Kutacane Graben. The Eastern and Central Barisan ranges rise eastern side of the Lubuksikaping Fault, in which coarse- and
 STRUCTURE AND STRUCTURAL HISTORY 209

Fig. 13.23. Kembar Volcano and the

Kutacane Graben, North Sumatra, a volcanic
centre and a graben filled by Quaternary
alluvium in a releasing bend of the Sumatran
fault System (detail from GRDC map of
Medan--Cameron et al. 1982a).

fine-grained facies can be recognized, could be matched by the either side rise to 1 0 0 0 - 1 7 0 0 m. The volcanic centre of Sorik
Air Mabara Granite on the western side of the fault, indicating a Merapi has been intruded to the south of the graben, near the
right-lateral displacement of c. 42 kin (Fig. 13.24). McCarthy & sharp bend between the Gadis and P u n g k u t - B a r i l a s faults
Elders (1997) visited these localities and urge caution in accepting (Fig. 13.24). Katili & Hehuwat (1967) found right-lateral offsets
this correlation as these granite bodies are petrologically hetero- of 2 0 0 - 1 2 0 0 m on tributaries of the Angkola River at the northern
geneous. Sieh & Natawidjaya (2000) recognised 20 km of right- end of the Panyabungan Graben, and many streams on the north-
lateral offset on the channel of the Barumun River, but consider eastern slopes of Sorik Merapi also show dextral offsets. The
that this segment of the fault is relatively inactive at present. present rate of sli~ along this segment of the fault is estimated
The faults in the Angkola segment bounding the lens to the SW at 23 + 4 mm a - (Genrich e t al. 2000). The Gadis Fault was
are the Gadis and Pungkut-Barilas faults. These faults are the site of an earthquake in 1892 in which a right-lateral dis-
concave towards the NE and the Panyabungan Graben has been placement of 2 m was recorded trigonometrically (Mfiller 1895).
formed in a releasing bend against the Gadis Fault. The floor of The original survey data have been recalculated and have
the Panyabungan Graben lies at 200 m, while the mountains on shown that the amount of dextral displacement was actually
210 CHAPTER 13

Fig. 13.24. Fault block bounded by

segments of the Sumatran Transcurrent Fault
System based on GRDC Geological Map of
Lubuksikaping (Rock et al. 1983),
interpreted as an extensional stepover and in
terms of the strain ellipsoid. Reversed arrows
indicate strike slip faults; toothed lines are
reversed faults; and lines with blocks are
extensional faults. The correlation of the Air
Mabara and Sopan granites indicating dextral
transcurrent movement of 42 km on the
Lubuksikaping Fault is taken from Hahn &
Weber (1981b). If the correlation of the
granites is correct the fault-bounded block
has rotated counter-clockwise by some 40~.

4.5 • 0.6 m (Prawirodirdjo et al. 2000). As well as topographic to a kilometre. Major historical earthquakes have occurred along
expression, the outcrop of the Barilas-Pungkut Fault is marked this segment of the fault; the 1926 Pandangpanjang earthquake
by a 20 m wide fault zone with a fault gouge, composed of to the north of the lake, and the 1822 and 1943 earthquakes near
sulphide-rich clays and silicified breccia with gypsum (Rock Solok to the south. In the Pandangpanjang earthquake, buildings
et al. 1983). in the town were displaced up to 6 0 c m towards the NW by
In Figure 13.24 the Equatorial Bifurcation is interpreted in terms dextral fault movement (Katili & Hehuwat 1967). Genrich et al.
of a strain ellipse in which the Panyabungan and Rau grabens (2000) calculate that the current rate of dextral displacement is
occupy the extensional segments. In this figure, apart from the 23 ___ 5 mm a-~ along this segment of the Sumatran Fault.
bounding faults, for which there is good evidence of dextral Bellier & S6brier (1994) used SPOT (Satellite Propatoire d'Ob-
strike-slip, movement on the other faults is inferred from their servation de la Terre) imagery to distinguish between active
orientation with respect to the strain ellipse. The Equatorial (young) and inactive (old) fault traces in their study of the fault
Bifurcation is also interpreted as an extensional right-stepping system. They suggest that the lake formed within an extensional
step-over, developing complementary pull-apart grabens. right step-over which developed as a graben bounded by faults
to the NE and the SW. These faults have been superseded by a
Lake Singkarak (Fig. 13.25). Lake Singkarak, in West Sumatra to active major through-going fault which passes through the
the north of Padang, occupies a depression flanked by escarpments centre of the lake, displacing the northwest bank right laterally
which rise 400 m above the lake surface. The escarpments mark for a distance of 2500 m (Fig. 13.25).
the outcrop of two opposing oblique normal faults, forming a Holder et al. (1994) made a study of the lineament pattern in
pull-apart graben structure within the Sumatran Fault System Sumatra to the south of the equator from SAR (synthetic aperature
(Fig. 13.25). Tjia & Posavec (1972) report that fault traces are radar) imagery. They found that the Sumatran Fault was marked
seen to displace lahars from recent volcanic eruptions, lake by series of V-shaped graben between the main fault trace and
terraces and valley alluvium, and to offset stream courses for up W N W - E S E splays at intervals of 5 0 - 1 0 0 kin; the apex of the
 STRUCTURE AND STRUCTURAL HISTORY 211

Fig. 13.25. The pull-apart basin of Lake

Singkarak, based on GRDC maps of Padang
(Kastowo & Leo 1973) and Solok (Silitonga
& Kastowo 1975), with modificationsto the
faults from Bellier & S6brier (1994) and
detail of normal fault scarps from Sieh &
Natawidjaja (2000). The inset shows the
progressive displacementof volcaniccentres
from the Kerinci Centre towards the NW and
SE by movements along the Sumatran Fault
System (SFS) according to the hypothesisof
Posavec et al. (1973).

Vs being towards the north (Fig. 13.19, inset). Segments along the is closely associated with volcanic activity and with many hot
fault between the graben are areas of recent uplift, with perched springs. Ranau Lake at the northwestern end of the depression,
river terraces and the erosion of the Tertiary and Quaternary sedi- occupies the caldera of a volcano that erupted in a releasing
ments. Holder et al. (1994) suggest sinistral strike-slip movements right stepover between the two fault segments. The unusual rec-
along the splays, together with dextral movements along the main tangular walls of the caldera represent the bounding strike-slip
fault, induced subsidence between the main fault and the splay and and normal faults of the pull-apart basin into which the volcano
uplift along the ENE sides of the splays, as crustal blocks moved was emplaced (Bellier & S6brier 1994). A resurgent volcanic
along the fault. They suggest that these movements were due to dome has developed on the southeastern margin of the caldera.
oblique compression across the fault zone (Fig. 13.19, inset). The southwestern component of the step-over is a presently inac-
tive fault strand (North Semangka Fault) that extends from the
southern bank of the lake along the southern side of the Liwa
Lake Ranau and the Semanka Depression (Fig. 13.26). A 150 km long depression. The currently active Ranau-Suwoh Fault cuts
depression, filled with the products of Quaternary volcanic through the northeastern part of the lake replacing the stepover
products and alluvium, extends from Lake Ranau to Semangka and the pull-apart basin and offsetting the caldera rim by
Bay in southern Sumatra (Fig. 13.26). The depression is 2300 i 100m (Bellier & S6brier 1994). Bellier & S6brier
bounded by the Ranau-Suwoh and Semangka fault segments at (1994) estimate a rate of displacement of 6 _+ 4 mm a -~. This
the southern end of the Sumatran Fault System. The fault zone segment of the fault was the site of the 1933 and 1994 earthquakes.
212 CHAPTER 13

Fig. 13.26. The pattern of faults between Lake Ranau and Semangka Bay at the southern end of the Sumatran Fault System, from S6brier et al. (1991 ) based on SPOT,
Landsat and aerial photographic interpretation. Inset shows the relationship between Java, Sumatra and the Sunda deformation front in solid lines, compared to their
relationships at 13 Ma (from Huchon & Le Pichon 1984) in dashed lines. The cross-hatched area indicates the area of extension and cruslal thinning in the forearc, the
shaded area indicates the zone of extension in the Sunda Strait, opened up as western Sumatra moved c, lO0 km northwestwards along the Sumatran Fault.

Fifty kilometres to the SE of the lake, Quaternary alluvium fills forearc basin is hardly developed. A well-developed accretionary
the Suwoh Graben, occupying a releasing bend in the R a n a u - complex and forearc basin is again developed further north
Suwoh Fault. A small group of calderas, one of which erupted opposite Sumatra. The curvature of the subduction trench
in 1933, occurs on the southwestern side of this graben, at the towards the NW means that subduction becomes increasingly
northern end of the Semangka Fault. The Semangka Fault with a oblique in this direction. Seismic profiles show that these vari-
significant dip-slip component downthrowing to the NE, defines ations in the development of the accretionary prism and the
the southwestern side of Semangka Bay, and a complementary forearc basin are not due to a change in the attitude of the subduct-
fault defines its NE margin. A subsidiary fault, the Banding ing plate which has a constant rate of movement (c. 7 cm a - ~) and
Fault, limits a triangular depression filled with alluvium at the a constant angle of subduction (c. 7 ~, Kopp et al. 2002).
head of the bay (Fig. 13.26). Malod et al. (1995) used existing Sea Beam data and the results
of a new echo-sounding survey to compile a bathymetric and tec-
The Sunda Strait (Fig. 13.27). The Sunda Strait lies within the tonic map of the area of the Sunda Strait. The accretionary
zone of transition in which normal subduction of the Indian complex is represented by a series of small parallel basins, anticli-
Ocean Plate beneath Java is replaced by oblique subduction nal ridges and large scarps, some of the latter show the character-
beneath Sumatra. Opposite Java there is a well-developed accre- istics of reverse faults, with a N W - S E trend, culminating in a
tionary complex and forearc basin, while opposite the Sunda series of rift basins and SW- and NE-facing escarpments identified
Strait the deformation front of the accretionary complex is as the Ujung Kulong Fault Zone (Fig. 13.27). To the north of
deflected northeastwards for a distance of 40 km, the topographic the accretionary complex the central part of the Sunda Strait
expression of the accretionary complex is much reduced, and the is occupied by a closed n o r t h - s o u t h depression 1800 m deep.
 STRUCTURE AND STRUCTURAL HISTORY 213

Fig. 13.27. Extensional fault system in the

Sunda Strait between Sumatra and Java
(from Malod et al. 1995). Bathymetry in
metres; toothed lines are normal faults;
arrows indicate strike slip faults; triangles are
active volcanoes.

The depression is bounded to the south by the accretionary The relationship between the Sumatran Fault System and the
complex and to the north by the northern shore of Semangka Quaternary volcanic arc. Quaternary volcanic centres and currently
Bay, defined by the Sunda segment of the Sumatran Fault active volcanoes show a close relationship to the trace of the
system. The tectonic map of Malod et al. (1995) shows the Sumatran fault system. Posavec et al. (1973) claimed that this
depression bounded to east and west by north-south escarpments relationship is seen particularly in central and southern Sumatra
representing faults downthrowing into the depression. Within the between Lake Toba and Semangka Bay. They remark that when
bounding faults the depression is cut by fault scarps trending plotted on a small-scale map the volcanic centres lie at intervals
N W - S E and downthrowing either to the NE or to the SW of 7 5 - 1 0 0 km along the fault trace 'like a string of pearls'.
(Fig. 13.27). Seismic reflection and refraction data obtained by However, from their mapping of active fault traces and of volcanic
Lelgemann et al. (2000) confirmed the general structural pattern centres Sieh & Natawidjaja (2000) show that this relationship is
identified by Malod et al. (1995) and show substantial crustal not as close as has been supposed (Fig. 13.21). Plotting the distri-
thinning with the development of a horst and graben structure bution of volcanic centres relative to the line of the fault they
within the strait. Major north-south graben structures occur demonstrate that the centres switch back and forth across the
both to the east and west of the central depression. The graben fault along its length, with centres occurring up to 20 km from
contain up to 6 km of Neogene and Quaternary sediment. the fault on its SW side near the equator, (Talakmau, Maninjau)
Malod et al. (1995) interpret the Sunda Strait as a north-south and up to 50 km on the northeastern side in Aceh and in the
extensional pull-apart basin, bounded to the north by the Sunda Sunda Strait (Kapal, Krakatoa). Page et al. (1979) suggested that
segment of the Sumatran Fault System, and to the south by the the eastward displacement of volcanic centres from Lake Toba
Ujung Kulon Fault Zone. Extension evidently continues at the northwards into Aceh is due to a fracture in the downgoing plate
present day, as a north-south zone of earthquake epicentres along the Investigator Fracture Zone which has been subducted
extends through the strait, paralleled by a line of volcanoes extend- in this area. They suggest that presence of the fracture is respon-
ing northwards from Krakatoa into southern Sumatra. sible for the extent and the intensity of the explosive eruption of
The opening of the Sunda Strait is interpreted as the result Toba, and that to the north of Toba the subducted plate is
of oblique subduction that has thinned and extended the crust passing into the mantle at a lower angle, so that the depth at
above the down-going plate, resulting in the concave form of which magmas are generated (c. 100 kin) is displaced towards
the deformation front and the poor development of the accretion- the east (Page et al. 1979).
ary complex. Huchon & Le Pichon (1984) suggested that forearc As has already been reported, Posavec et al. (1973) found in the
material to the west of the strait, including the accretionary area of their study in central Sumatra that active volcanic centres
complex, has been translated c. 100 km northwestwards along are grouped around east-west aeromagnetic anomalies that
the Sumatran Fault System since the Miocene (Fig. 13.26, intersect the fault zone, and which they suggest are due to grano-
inset). Sieh & Natawidjaja (2000) have recently confirmed this dioritic/dioritic intrusions, representing an underlying magma
estimate using more rigorous calculations. chamber. These east-west zones of volcanic activity at a high
214 CHAPTER 13

angle to the Sumatran Fault trace may be due to north-south North Sumatra and N W A c e h basins
extension related to the northward movement of the forearc
sliver plate. Again, as noted above, Posavec et al. (1973) found The North Sumatra Tertiary sedimentary basin and its westward
that the present volcanic centres have given rise to trails of extension in NW Aceh occupy the northeastern part of Sumatra
earlier volcanic edifices which extend towards the NW on the between Banda Aceh and Medan, extending northwards into the
southwestern side of the fault, and to the SE on the northeastern Andaman Sea (Fig. 13.29). Knowledge of the geology and
side, and are increasingly eroded with distance from the volcanic structure of these basins is largely due to work of the companies
centre. This displacement of the volcanic centres with time is holding concessions in the area, including Inpex, Mobil (now
attributed to dextral movement along the Sumatran Fault during ExxonMobil), Asamera (now ConocoPhillips) and Pertamina
the past few million years (Fig. 13.25, inset). (the Indonesian National Petroleum Company).
In detail, as has been pointed out in the preceding account The Tertiary sediments rest unconformably on low-grade meta-
of local areas along the fault, volcanic centres are often located sediments of Carboniferous-Permian age intruded by granites
in stepovers and in releasing bends where they are associated which are exposed in the Barisan Mountains to the south
with normal faulting and the formation of pull-apart sedimentary and west of the basins. Outcrops of Tertiary sediment also occur
basins (Bellier & S~brier 1995) (e.g. Kembar Volcano-- within the Barisans as fault-bounded basins or tilted caps to
Fig. 13.23, Sorik Merapi--Fig. 13.24, Ranau--Fig. 13.26). horst blocks. In the Malacca Strait towards the NE, Tertiary
The apparent close relationship between the trace of the fault sediments thin out over the Malacca Shelf, and further east the
zone and the distribution of the volcanic centres has led to the Pre-Tertiary basement rises above sea level in Peninsular
suggestion that there is a genetic relationship between faulting Malaya. Tertiary sediments also thin out to the SE towards the
and volcanicity (e.g. Saint Blanquat et al. 1998). The suggestion Asahan Arch, a basement high that separates the North from the
is that the generation of magmas in the upper mantle and their Central Sumatra Basin (Fig. 13.28). There is no clear boundary
intrusion into the upper crust has formed a weak zone of ductile to the basin towards the north where the basins pass into Thai
material extending from the upper surface of the downgoing territorial waters as the Mergui Basin (Polachan & Racey 1994).
plate to the surface, along which the shear component of strain The Tertiary sediments are covered extensively by Pleistocene
partitioning has been focused. In the upper crust fractures to Recent alluvium, swamp deposits and the products of
related to the fault zone provide channels for the passage of Quaternary volcanism, including volcanic edifices, and to the
magmas to the surface to construct volcanic edifices, indeed south of Medan, by the Toba Tufts (Fig. 13.29).
earthquake hypocentres extend vertically below the fault zone The earliest sediments in NW Aceh, and extending westwards
for 100-135 kin, down to the surface of the downgoing plate as across the Barisan Mountains into the West Aceh Basin, are
defined by the Wadati-Benioff Zone below (Seamans 1993). conglomerates, sandstones, siltstones and shales with interbedded
However, as noted above this relationship is not as close as limestones (Meucampli and Agam formations) of Eocene to Early
first appears, and elsewhere in the world, in other regions of Oligocene age (Bennett et al. 198 la). Some of the conglomerates
oblique subduction and strike slip faulting, volcanism does not contain volcanic clasts suggesting that volcanicity occurred in
always coincide with the active fault zone (Sieh & Natawidjaja this region at that time. Apart from the active volcanoes, it is
2000). Sieh & Natawidjaja point out that the active volcanic evident that at this time the Barisan Mountains did not form a
centres are much younger than the initiation of the active fault topographic feature, and that sedimentation in fluvial, coastal
traces, hundreds of thousands of years as opposed to millions. and restricted marine environments was continuous from the
They concede that the location of the fault zone may have North Sumatra Basin into the West Aceh Basin. In the North
been controlled by earlier Neogene volcanism, but conclude Sumatra Basin the earliest Tertiary sediments are marine platform
that the relationship seen in Sumatra at the present time between carbonates, of presumed Eocene age (Tampur Formation), exten-
active faults and modern volcanoes is not cogenetic but sively exposed in a karstic plateau to the west of Langsa, which
coincidental. rest unconformably on the eroded surface of the Pre-Tertiary
basement.
From a detailed study of seismic sections Situmorang &
Tertiary basins in the backarc area Yulihanto (1985) reconstructed sub-surface horizons and ident-
ified fault patterns in the Pertamina Block, between Pangkalan
The backarc area of Sumatra, to the east of the Barisan Mountains Brandan and Medan. They found that fault traces in the Pre-
and the currently active volcanic arc, is a relatively low-lying area Tertiary basement, which they identified as strike-slip faults,
declining in relief into the Malacca Straits, crossed by meandering have a predominantly north-south orientation. In the Late
rivers and passing into mangrove swamps towards the straits. Palaeogene to Early Miocene the platform broke-up in a 'rift
Beneath the present alluvial and swamp deposits this area is phase' with the formation of extensional pull-apart basins and
underlain by Tertiary sediments which rest unconformably on the development of horsts (highs) and graben or half-graben
the Pre-Tertiary basement and occupy a series of sedimentary (deeps). This horst and graben structure now forms the underlying
basins. The basins hold major reserves of oil and gas and locally structure of the basin (Figs 13.30 & 13.31). In the Lho Sukon
coal, and have been intensively studied by geophysical methods (Pase) Deep this basement now lies at depths of more than
and by drilling by companies that hold concessions for the 3000 m (McArthur & Helm 1982).
exploration and exploitation of oil and gas. The results of these Scree deposits formed marginal to the horsts extended out
studies have been reported mainly in the Proceedings of the into the grabens as alluvial fans, with coal swamps passing into
Annual Conventions of the Indonesian Petroleum Association lacustrine, estuarine and shallow marine deposits (Bruksah
(IPA). Figures modified from these Proceedings have been used Formation). In the Late Oligocene to Early Miocene, the basin
with the written permision of the IPA to illustrate the following entered a 'sag phase', with marine conditions extending through-
account. out the basin (Fig. 13.31). The separate graben coalesced into a
The backarc region is divided by the Asahan and Tigapuluh regionally extensive basin, with more rapid subsidence to the
arches into the North, Central, with its associated Ombilin west of a hinge line (Rayeu Hinge) at the margin of the Malacca
Basin, and South Sumatra Basins (Fig. 13.28). The lithologies Shelf (Figs 13.30 & 13.31). Subsidence outpaced sedimentation,
and sedimentary history of these basins has been described submerging the horsts, including the Arun and the Lho Sukon
earlier in this volume by De Smet & Barber (Chapter 7), the highs, and the western part of the Malacca Shelf, on which carbon-
environments for oil and gas by Clure (Chapter 10) and the coal ate build-ups developed (Peutu Formation). These build-ups
deposits by Thomas (Chapter 11). host important gas fields (McArthur & Helm 1982). Continued
 STRUCTURE AND STRUCTURAL HISTORY 215

Fig. 13.28. The geographical and tectonic

setting of the Sumatran backarc basins. The
volcanic arc follows approximately the trace
of the Sumatran Fault, based on Davies
(1984, Fig. l).

subsidence, coinciding with a global rise in sea level, resulted Fold structures. The Tertiary sediments are folded (Figs 13.29 &
in maximum marine transgression during the Mid-Miocene 13.32). Fold structures can sometimes be recognized by outcrop
(Collins et al. 1995). The reefs were submerged, source areas patterns, bedding traces on aerial photographs and in outcrop by
became restricted, and fine grained sediments (Baong Formation) the dip of the bedding, especially on the margins of the Barisan
were deposited throughout the basin. Carbonate reefs were buried Mountains and in temporary roadcuts, but many folds have been
beneath fine-grained sediments forming effective traps for oil and recognized only in seismic sections during the exploration for
gas. At this stage the basin extended westwards over much of the oil and gas.
area which now forms the Barisan Mountains. From the Mid- In the NW Aceh Basin between Banda Aceh and Lhokseumawe
Miocene to the present time the rate of subsidence has decreased the fold trends are approximately east-west, parallel to the north
and the basin has undergone a regressive phase. This coincided coast. This is surprising as the underlying basement structures
with the progressive uplift of the Barisan Mountains, together trend north-south (Fig. 13.30). It has been suggested that the
with erosion and the eastward spread of fluvial deposits (Keuta- east-west orientation of the folds is due to the incipient develop-
pang, Seureula and Julu Rayeu formations), followed by the emer- ment of a southward-dipping subduction system in the southern
gence of the southern part of the basin, with continued uplift of the Andaman Sea, offshore northern Sumatra (Bennett et al. 1981a;
Barisan Mountains and the growth of the volcanic arc, while to the Curray et al. 1979), but there is no evidence of such a system in
north beneath the Andaman Sea the basin is still submerged and the structural syntheses prepared by the hydrocarbon industry
deposition continues. It is estimated that the original thickness (Nur'aini et al. 1999) (Fig. 13.30).
of the sediments in the central part of the basin reached over The east-west folds affecting Plio-Pleistocene sediments
5 km (Kingston 1988). are open, symmetric to slightly asymmetric, mainly synclinal
Studies of the surface lineaments, representing fault structures folds, arranged en echelon. The corresponding anticlines are
in the northern part of the North Sumatra Basin using SAR (syn- absent, or represented by interference and accommodation stuc-
thetic aperture radar) imagery showed that N W - S E (Sumatran) tures, especially in argillaceous units. The folding occurred after
and N E - S W (antithetic) trends are dominant throughout the the Early Pleistocene as the present volcanic edifices have been
basin, with subordinate W N W - E S E and E N E - W S W trends constructed on rocks which had already been folded. An earlier
(Sosromihardjo 1988). Surprisingly the north-south trend that east-west phase of folding is recognized in the Takengon
dominates the subsurface horst and graben structure of the basin Quadrangle to the south where folded early Tertiary rocks are
is not represented in the surface lineaments, which must reflect overlain unconformably by Late Oligocene sediments (Cameron
more recent stress systems. et al. 1983).
216 CHAPTER 13

Fig. 13.29. Structuralmap of the North Sumatra Basin and the distributionof Tertiary and Quaternarysedimentsin northernSumatra. The locationof the cross-sectionin
Figure 13.32 is indicated.

In the North Sumatra Basin, in the area to the south of & Marinoadi 1977) and mudstones cropping out in the cores of
Lhokseumawe, fold traces swing round into a N N W - S S E the anticlines are often vertical, crushed, sheared and slickensided.
direction, parallel to the margin of the Barisan Mountains The anticlines are commonly associated with mud volcanoes and
(Fig. 13.29). The swing in strike is attributed to dextral strike- oil, gas and warm water seepages. These features are attributed to
slip movement on the Lhokseumawe Fault (Bennett et al. mud-diapirism in which the rapidly deposited, water-saturated
198 lc). Further south this fault is seen to have a major downthrow mudstone, buried beneath the sandstones, became overpressured,
to the east and joins the Lokop-Kutacane Fault to mark the producing a density inversion that has caused the activated
margin of the Barisan uplift. To the east of the fault zone mudstones to rise diapirically towards the surface. This process
the Simpang Kanan Monocline forms the western margin of the is considered to have commenced in the Pliocene, but continues
structural North Sumatra Basin. The Tampur Formation of (?) to the present day (Bennett et al. 1981 c). Kinking and bifurcation
Eocene age, the oldest of the Tertiary units, forms a plateau on of the anticlinal fold traces seen in the area to the west of Aru Bay
the flat limb of the monocline and is seen in aerial photographs (Fig. 13.29) has suggested that the locations of the anticlines are
to be intensely fractured and jointed, with a karst topography. controlled by dextral movement along strike-slip faults in the
The monocline is considered to be the surface expression of a basement.
major normal fault at depth with a 3 km downthrow to the east Fold structures in the younger Tertiary units to the NE, become
(Bennett et al. 1981c). The vertical limb is composed of mud- difficult to recognize on aerial photographs as they form low-
stones of the Bampo Formation (Upper Oligocene-Lower amplitude domes and basins, resembling interference structures,
Miocene), which are sheared and slickensided and cut by and are covered by alluvial deposits.
west-dipping reversed faults. Locally the Bampo mudstones are In a detailed study of seismic data from the Pertamina Block to
altered to dark slates containing deformed septarian nodules. the north of Aru Bay Situmorang & Yulihanto (1985) examined
Tight, extremely elongated anticlines and broad synclines the orientation of faults, fractures and fold axes at different horizons
occur in a belt to the east of the monocline in which the cores of in the Tertiary sequence and demonstrated that at the level of
the anticlines are formed of mudstones of the Baong Formation the basement the structure is extensional, with dominant north-
(Middle-Upper Miocene) and the cores of the synclines of south-trending normal faults, while above the base of the Baong
sandstones of the Keutapang and Seureula formations (Upper Formation the dominant structures are compressional, with fold
Miocene-Pliocene) (Fig. 13.32). From field studies and in axes and strike-slip faults trending N W - S E (Sumatran Trend).
seismic sections it can be seen that the Baong Formation is In a further study of the same area, Ryacudu et al. (1992) plotted
excessively thickened over the crests of the folds (Mulhadiono contours of three horizons, the Belumai Formation, the Middle
 STRUCTURE AND STRUCTURAL HISTORY 217

Fig. 13.30. Basementstruclure in the North

Sumatra Basin showingthe highs and
depressions which have controlledTertiary
sedimentation,based on Nur'aini et al.
(1999) with modificationsafter Collinset al.
(1995). The whitedashed line marks the limit
of thick Tertiary sediments on the Malacca
Shelf, and the bold dashed line marks the
position of the Early MioceneRayeu Hinge;
subsidence was more rapid to the west of this
hinge (after Kingston 1988). The location of
the cross-sectionin Figure 13.31 is indicated.

Baong Sandstones and the Lower Keutapang Sandstone, from Barisans the area of sedimentation was continuous with the
seismic reflection profiles. From this study they compiled a West Sumatra Basin on the west coast of Sumatra. Exposure of
S W - N E cross-section that illustrated the structure and the struc- the Tertiary sediments is poor, except in the Barisan foothills
tural evolution of the area (Fig. 13.32). At the base of the section and to the south around the Tigapuluh Hills, where outcrops
the Pre-Tertiary basement is poorly imaged in the seismic data, occur in river and more transient road sections. Over the greater
but is overlain by the Tampur Formation which has been encoun- part of the basin the Tertiary sediments are covered by Recent allu-
tered in several boreholes. At this level extensional normal vium and swamp deposits. However, the Central Sumatra Basin is
faulting is dominant, but the SW end of the section is cut by a a major oil province and has been intensively investigated during
dextral strike-slip fault parallel to the N W - S E 'Sumatran' trend. oil exploration by seismic reflection profiling and by boreholes by
Further to the NE are several complementary NE-SW-trending P.T. Caltex, Pertamina and P.T. Stanvac, so that the subsurface
'antithetic' sinistral strike-slip faults. Following the deposition structure (Fig. 13.34) and the sedimentation history of the basin
of the Belumai Formation these strike-slip faults were re-activated are very well known (Fig. 13.35).
and inverted in a compressional tectonic regime and in the upper The Central Sumatra Basin is separated from the North Sumatra
part of the section have developed as positive flower structures, Basin to the NW by a basement ridge, the Asahan Arch, and less
with reverse rather than normal sense of movement, and form sharply from the South Sumatra Basin to the SE by the Tigapuluh
fold structures which increase in amplitude upwards through the Arch (Fig. 13.28). Pre-Tertiary rocks have been penetrated in
section. At the SW end of the section the Keutapang Sandstone many boreholes during oil exploration, as the fractured basement
is exposed at the surface in the core of an anticline. Thickening has locally proved to be productive. It has therefore been possible
in the Upper Baong Formation seen in this fold indicates that to reconstruct the nature of the basement to some extent (Eubank
the structure developed by the diapiric flow of shales into the & Makki 1981). The Pre-Tertiary basement is composed of a
anticlinal core. The cross section is interpreted as showing that series of terranes with a N W - S E structural grain (see Chapter 4).
prior to the Mid-Miocene the structure of the area was developed In the NE beneath the Malacca Straits, boreholes encountered a
in a transtensional tectonic regime, while after the Mid-Miocene 'quartzite terrain', followed to the SW by a zone of radiolarian
the tectonic regime was transpressional. cherts, mauve-shales, thin limestone and sandstones and shales
(rhythmites) which has been termed the Mutus Assemblage and
correlated with the Triassic Kualu Formation, which crops out
near Medan to the north. Further to the SW is a zone of greywacke
Central Sumatra Basin sandstones and mudstones correlated with the Carboniferous-
Permian Tapanuli Group of northern Sumatra. These rocks also
The Central Sumatra Basin with a width of nearly 300 km from crop out to the SE in the Tigapuluh Hills. Forming the southwes-
the Malacca Straits in the NE, to the foothills of the Barisan tern margin of the Tigapuluh Hills and extending to the NW is a
Mountains in the SW, occupies the greater part of Sumatra from zone of highly deformed schists termed have the Medial
2~ to l~ (Fig. 13.33). Faulted outliers of Tertiary deposits, Sumatra Tectonic Zone (MSTZ). In the Barisan foothills along
such as the Ombilin Basin, suggest that before the uplift of the the southwestern margin of the basin Tertiary sediments are
218 CHAPTER 13

Fig. 13.31. Diagrammatic cross-section to

illustrate the tectonostratigraphic evolution
of the North Sumatra Basin modified from
Kingston (1988) and Collins et al. (1995).
The location of the section is shown on
Figure 13.30.

seen to rest u n c o n f o r m a b l y on the C a r b o n i f e r o u s K u a n t a n alternating with t r o u g h s or d e e p s w h i c h are filled with Tertiary

F o r m a t i o n , the P e r m i a n P a l e p a t and M e n g k a r a n g f o r m a t i o n s a n d s e d i m e n t , to a d e p t h o f o v e r 5 k m in the B a r u m a n B a s i n to the
the Jurassic a n d C r e t a c e o u s R a w a s , P e n e t a a n d Asai f o r m a t i o n s . n o r t h o f R a n t a u p r a p a t (Fig. 13.34). T h e ridges a n d t r o u g h s are
T h e t o p o g r a p h y o f the b a s e m e n t is h i g h l y irregular, with ridges, c o n s i d e r e d to be c o n t r o l l e d by o l d e r N N W - S S E l i n e a m e n t s in
h i g h s or ' u p l i f t s ' w h e r e the b a s e m e n t a p p r o a c h e s the surface, the b a s e m e n t , r e p r e s e n t i n g the M S T Z and the m a r g i n s o f Triassic

Fig. 13.32. Diagrammatic cross-section of the structure in the Simpang area to the north of Aru Bay, modified from Ryacudu et al. (1992, fig. 18) based on the
interpretation of seismic profiles. Normal or transtensional strike-slip faults in the lower part of the succession were inverted as transpressional faults during and after the
deposition of the Middle Miocene Baong Formation. Fold structures are developed over positive flower structures related to dextral or sinistral strike-slip faults. The
amplitude of the anticline at the SW end of the section has been increased by the diapiric flowage of shale into its core. The line of section is indicated on Figure 13.29.
 STRUCTURE AND STRUCTURAL HISTORY 219

Fig. 13.33. The Central SumatraBasin,based on GRDC maps, with additionof subsurface structure from Heidrick& Aulia (1993). Tertiarysedimentsare only exposedat
the surface in the southwesternpart of the basin in the foothillsof the Barisan Mountainsand also around the TigapuluhHills to the south. ElsewhereTertiarysediments
are covered by Recent alluviumand swamp deposits.

graben, but also seen in the Malay Peninsula, and younger regional subsidence, so that sedimentation became more wide-
N W - S E or 'Sumatra' trend lineaments seen in the Barisans to spread, extending from the graben across the adjacent horsts.
the west. The troughs occur as two groups, a western group The sediments are sands and marine shales of the Menggala and
along the front of the Barisan Mountains, including the Baruman Bangko formations. In the Early Miocene deltaic sediments
Basin in the north, separated by the Kubu High from the Balam derived from the Sunda Shelf in the region of the Asahah Arch
and Kiri troughs to the south. A series of highs, including the in the NNE extended southwards into the basin, with some input
Dumai High, the Rokan Uplift and the Minas and Kampar highs from the Malay Penisula to the east (Sihapas Group). Delta front
separate the western troughs from the Bengkalis Trough towards sand deposits interfinger with marine shales (Telisa Formation)
the Malacca Straits. In the western part of the basin, ridges and towards the south. As the deltas advanced southwards marine
troughs trend in a N W - S E direction, but in the east the structure deposits were gradually replaced by terrestrial sediments and
is dominated by the north-south Bengkalis Trough and its coal seams were developed on the delta tops. Subsidence was
extensions to the south in the Genako and Bukit Susah troughs not uniform throughout the basin, with greater subsidence in the
(Wain & Jackson 1995). The ridges and troughs were formed troughs. Subsidence and rapid sedimentation was greatest in
as horsts and graben by extension in the earliest phase in the the north, so that the greatest thickness of sediments is found
structural development of the Central Sumatra Basin. in the Barumen Basin (>5000 m) and the sediments thin out
The sedimentation history of the Central Sumatra Basin as over the Kampur High to the south (Fig. 13.34). With continuing
illustrated by Wongsosantiko (1976) (Fig. 13.35) is similar to subsidence, but a decrease in sediment supply, a major marine
that of the North Sumatra Basin. The earliest sediments are brec- transgression occurred in the Mid-Miocene, so that marine depos-
cias, conglomerates and sandstones interbedded with shales and its of the Telisa Formation were deposited across the delta surface.
coal seams, which were eroded from the ridges and deposited in At the time of maximum trangression marine sedimentation
subsided troughs or half-graben. The evironments of deposition extended westwards across the present site of the Barisan
are interpreted as scree, alluvial fan, fluvial and lacustrine with Mountains to reach the Ombilin Basin (Fig. 13.33), well beyond
rare marine incursions (Pematang Formation). Although the age the bounds of the Central Sumatra Basin. In the Ombilin Basin
of the earliest sediments is poorly constrained they are considered Mid-Miocene sediments include a carbonate reef (Ombilin
to be of Late Eocene to Oligocene age. Again, as in the North Formation), indicating that at that time the mountains did not
Sumatra Basin, the rift phase was followed by a sag phase with form a topographic feature. Uplift and erosion of the Barisan
220 CHAPTER 13

Fig. 13.34. Basement slructure in the Central Sumatra Basin showing highs and depressions which controlled Tertiary sedimentation, simplified after Heidrick & Aulia
(1993, Fig. 3). The lines of sections (a) and (b) in Figure 13.35 are indicated.

Mountains late in the Mid-Miocene provided a source of sedi- The troughs are bounded by steep normal listric growth faults
ments which advanced across the basin from the west, depositing on their western or southwestern margins with hinges to the east
a regressive sequence of grey sandstones, siltstones and shales up or NE broken by small normal faults. Rollover folds were devel-
to 1.5 km thick (Petani Formation) through the Pliocene and Early oped against the major bounding faults. The troughs show
Pleistocene. These deposits are overlain, above an unconformity, dog-leg bends at the accommodation zones which are associated
by Pleistocene to Recent alluvial and swamp deposits of the with N E - S W oblique faults, and the troughs terminate at faults
Minas Formation (Fig. 13.35b). with the same orientation.
The Dumai and Pakanbaru Quadrangle sheets (Cameron et al. Soeryowibowo et al. (1999) have made a structural study of
1982d; Clarke et al. 1982b) show the stratigraphic sequence Tapung Half-Graben in the southern part of the Kiri Trough
exposed in the foothills of the Barisan Mountains with local and (Fig. 13.34). This graben is 25 km long, 8 km wide, trends
restricted outcrops of the Pematang Formation adjacent to base- N N W - S S E , and lies immediately to the SW of the Minas Field.
ment horsts, with more extensive outcrops of the Sihapas Group The graben is bounded on its SW side by a series of three
and the Telisa Formation forming broad N W - S E anticlines and arcuate listric normal faults which are considered to detach at a
syclines faulted into the Pre-Tertiary basement. Away from the depth of less than 6.5 km. The graben has a syn-rift section of
mountain front broad anticlines with a N W - S E trend, cored by 1500 m, the extension factor (/3-value) varies along fault frag-
the Sihapas Group and Telisa Formation, including the folds ments between 5 and 12% with a maximum extension of 2 kin.
marking the site of the prolific Minas oilfield (Fig. 13.33), occur It is suggested that the graben developed as the result of extension
among extensive Quaternary sands, gravels and swamp deposits. on north-south faults in the underlying pre-Tertiary basement.
The anticlines occur above highs in the underlying Pre-Tertiary Comparing the pattern of faulting in the Tapung Graben with
basement or mark the inversion of the sediments deposited in the sandbox model studies of normal and oblique graben for-
the troughs ('Sunda Folds', Eubank & Makki 1981). mation by McClay & White (/995), Soeryowibowo et al. (1999)
conclude that the extension did not occur in an east-west direc-
Balam Trough. The structure of the Balam and the associated tion, normal to the basement structures, as had been previously
Rangau, Kiri and Aman troughs, on the western side of the assumed, but obliquely in a N E - S W direction. They point out
Central Sumatra Basin (Central Deep on Fig. 13.34) has been that the experiments show that only normal faults are developed
studied by Williams et al. (1985) and Yarmanto et al. (1995) during oblique extension, and that neither strike-slip nor
who describe them as a series of en echelon graben, with inter- oblique-slip faults are involved. Plio-Pleistocene compression,
vening complex basement highs or accommodation zones. also in a N E - S W direction, inverted the Tapung Graben. On the
 STRUCTURE AND STRUCTURAL HISTORY 221

Fig. 13.35. (a) Diagrammatic cast-west cross-section across the western part of the Central Sumatra Basin showing troughs and highs and sediment provenance (after
Williams & Eubank 1995); (b) Diagrammatic north-south cross-section to illustrate the tcctonostratigraphic development of the Central Sumatra Basin (modified
from Wongsosantiko 1976, fig. 3). The lines of section are shown on Figure 13.34.
222 CHAPTER 13

Fig. 13.36. The Bengkalis Graben. (a) Outline of the graben and controlling faults from Moulds (1989) with the addition of fault traces from Heidrick & Aulia (1993). (b)
Model for the formation of the Bengkalis Graben due to extension on N W - S E and N E - S W basement fractures and the collapse of rhomboid blocks from Moulds (t989,
fig. 5). (e) Cross-section showing the Bengkalis Graben as a half-graben, based on a seismic profile in the southern part of the graben, after Heidrick & Aulia (1993).
Length of section is c. 60 km, vertical scale is not given. (d) Cross-section showing the Bengkalis graben as a half-graben with normal faults re-activated as thrust faults at
the NNE end of the section, based on a seismic profile from the northern part of the graben after Santy (2001). Length of section is c. 30 kin. The locations of sections (b)
and (c) are shown on (a). Circled 'A' and 'T' against vertical faults indicate 'away' and 'towards' on dextral strike-slip faults.
 STRUCTURE AND STRUCTURAL HISTORY 223

other hand, in their study of the Amin Trough and related graben, producing normal faulting and the formation of the half graben
Williams et al. (1995) suggest that dextral strike-slip fault structure. Whether there was a component of transtension is not
movements along the main boundary faults caused compression always possible to determine. The faults were inactive thoughout
at the dog-leg bends, with complementary sinistral strike-slip the deposition of the Sihapas Group and the Telisa Formation.
on the N E - S W faults, during both the Middle Miocene and Deformation with strike-slip faulting and N N E - S S W com-
Plio-Pleistocene detbrmation events. pression, causing, the reactivation and inversion of the normal
fault and the formation of the fold structure in the Tandun Field,
Bengkalis Trough. Heidrick & Aulia (1993) made an intensive occurred during the deposition of the Petani Formation in the
study of the sub-surface structure of the 'Coastal Plains Block' Late Miocene to Plio-Pleistocene. The western troughs in the
covering the area to the south of Bengkalis island, including the Central Sumatran Basins show a similar sequence of events, as
Bengkalis Trough, on behalf of P.T. Caltex Pacific Indonesia. has been diagrammatically illustrated by Yarmanto et al. (1995)
The 265 km long Bengkalis Trough originated as a series of exten- (Fig. 13.38).
sional half grabens on north-south normal faults (Fig. 13.36).
Seismic sections show steep normal faults at the surface passing
into listric faults, and an inferred flat-lying decollement surface Ombilin Basin
in the basement at a depth ofc. 6 km (Fig. 13.36c). At the northern
end the major bounding fault is on the SW side of the trough, A group of en echelon intramontane basins within the Barisan
while in the south it is on the NE side (Fig. 13.36c, d). During Mountains, faulted into Pre-Tertiary basement rocks, lie to the
the Plio-Pleistocene one of the normal faults, the Padang Fault, west of the Central Sumata Basin. From north to south these are
at the northern end of the trough, was re-activated as a reverse the Mandian, Kampar Kanan, Payakumbuh and Ombilin basins
fault in a phase of N E - S W compression (Fig. 13.36d).The base- (Fig. 13.33). The best studied of these is the Ombilin Basin in
ment structure of the trough has been modelled by Moulds West Sumatra, some 15 km to the SW of the Barisan Mountain
(1989) as due to the subsidence of the basement as rhomboidal Front, and about 10 km to the NE of the active strand of the
blocks between north-south- and NNE-SSW-trending faults as Sumatran Fault at Solok. The basin has been described by
the result of regional extension (Fig. 13.36b). Koesoemadinata & Matasak (1981), Koning & Aulia (1985),
Heidrick & Aulia (1993) recognized a complex history of struc- Whateley & Jordan (1989), Situmorang et al. (1991), De Smet
tural development with two intersecting dominant structural (1991) and Howells (1997a). The Tertiary rocks are preserved in
trends, north-south and N N E - S S W , which controlled the struc- a synclinal basin divided into two sub-basins, the Talawi and
tural development of the Central Sumatra Basin and were continu- Sinamar sub-basins, by the north-south Tanjung-Ampolo Fault
ally reactivated throughout its history. These structures behaved (Fig. 13.39). The basin is surrounded by Pre-Tertiary rocks
as dextral wrench faults, normal faults or reverse faults, depending of the Carboniferous Kuantan Formation to the NE and the
on the orientation of the stress system at different stages in the Permo-Triassic Silungkang and Tuhur formations to the SW. To
structural evolution of the basin. Heidrick & Aulia (1993) calcu- the NW the Tertiary sediments are overlain by volcanic products
late nearly 9 km of extension across the Bengkalis Trough and a of the Quaternary Malintang and Merapi volcanoes.
minimum of 43 km total dextral strike-slip displacement across The average topographic height of the basin is c. 400 m with
north-south faults. The earliest phase of deformation was rifting some peaks in the southern part of the basin reaching over
on north-south or N N E - S S W normal faults and reactivated 1000 m. Much of the basin is easily accessible and the Tertiary
W N W - E S E basement fractures during Eocene to Oligocene sediments are well exposed in mountainous terrain, with many
time. A second phase of deformation with N N E - S S W transten- river and road sections, so that conventional geological outcrop
sional wrenching in the Early Miocene was associated with mapping is possible. In addition there are also several large
the regional sag phase and re-activated the north-south faults as open-cast coal mines in which the small-scale structures may be
dextral wrench faults, and causing counter-clockwise kinking. examined in detail. The basin has also been investigated in the
in the period from the Mid-Miocene to the present N N E - S S W search for oil and gas, so that the subsurface structure has been
compression has reactivated the N N W - S S E wrench faults as explored by seismic sections and boreholes.
WSW-directed thrust faults (Fig. 13.36d). As presently exposed the basin is elongated in a N W - S E direc-
tion, the longer axis being c. 64 km, with a width of c. 25 km and
Pungut and Tandon Fields. The complex interaction between folds a present depth of c. 4600 m (Williams & Eubank 1995). The basin
and faults in the structural development of anticlinal structures is considered to have originated as a half-graben in the Late
which form traps for oil fields is illustrated by the Pungut and Eocene or Early Oligocene, during the same phase of extension
Tandon fields 65 km to the NNW of Pekanbaru (Mertosono that formed the troughs in the Central Sumatran Basin.
1975; Eubank & Makki 1981) (Fig. 13.37). A N N W - S S E anticli- Particular attention has been paid to the Ombilin Basin, as it is
nal and synclinal fold pair are transected and apparently displaced considered to be a well-exposed analogue for the early stages in
for some 3 km by a major dextral strike-slip fault. The Pungut development of the basins of the Central Sumatra Basin and the
Field to the north is bounded to the east by a north-south other basins in the Sumatran backarc area that can only be
segment of the strike-slip fault. The oilfield occupies a narrow studied by seismic methods and from borehole data. The Takung
anticlinal structure developed over an upfaulted sliver of the base- Fault that bounds the northeastern margin of the basin is con-
m e n t (Fig. 13.37). The Tandun Field to the south occupies an sidered to be the major bounding fault to the half-graben, as the
anticlinal fold to the east of the strike-slip fault, which here sediments thicken towards the fault, but the original normal
trends N N W - S S E . The strike-slip fault follows the trace of fault has now been partially inverted as a thrust. The hinge zone
a normal fault which bounded the western margin of a half to the SW is also broken by faults, but to the NW around the
graben, filled with a thick sequence of the Upper Oligocene Tungkar High, and on the SW side of the basin near Kolok,
Pematang Formation (Fig. 13.37). The change in the orientation the unconformity between Tertiary and Pre-Tertiary rocks is
is significant, as this segment of the fault has been reactivated as well-exposed (Fig. 13.39). Although the unconformity is well-
a reverse fault. The oilfield occupies the anticlinal structure devel- exposed where the Tertiary rocks rest on the Tungkar Granite,
oped by the inversion of the thick sediments forming the graben its position is difficult to define precisely in the field, as weathered
fill, uplifted along the reverse fault. This is an example of the granite passes into arkosic sandstone without a distinct break.
'Sunda Folds' as described by Eubank & Makki (1981).
The sequence of events which can deduced from these relation- Sedimentation histo 9 Against Pre-Tertiary units the oldest
ships is that the earliest stage was a period of east-west extension, deposits (?Late Eocene-Early Oligocene) are marginal screes
224 CHAPTER 13

Fig. 13.37. Structuralmap from Mertosono ( 1975, figs 7, 8) and line drawings from seismic sections of the Pungut and Tandun oilfields. Central SumatraBasin (Eubank &
Makki 1981; Williams et al. 1995). 'U', upthrown sides; D, downthrown sides of faults.

and alluvial fans, passing out into braided stream sandstones algal reefs. Fine sandstones with fragments of coal and amber
and lacustrine sediments deposited in an anoxic environment probably represent beach sands. Howells (1997a) suggests that
in the central part of the basin (Sangkarewang Formation). The this marine incursion came from the backarc area to the east.
Sangkarewang Formation is equivalent to the Pematang Formation Because the sediments in the lower part of the sequence were
of the Central Sumatra Basin, and is estimated to be some 3000 m deposited in a terrestrial environment it has proved difficult to
thick (Williams & Eubank 1995). It is followed by the (?) date them precisely, although fish occur in the Sangkarewang
Oligocene Sawahlunto Formation, composed of sandstones, silt- Formation and palynomorphs have been recovered from the
stones, mudstones and coals, deposited in meandering river and Sawahlunto and Sawahtambang formations, these have not
flood plain environments, 172 m thick in the Sinamar No.1 well proved to be age-diagnostic, although a Late Eocene to Oligocene
(Fig. 13.40). Coal seams up to 10 m thick are worked in open- age is inferred (Bartram & Nugrahaningsih 1990; Humphreys
cast pits and underground mines in the Talawi area. The et al. 1991). This general age is confirmed by the marine fauna
Sawahlunto Formation is overlain by the (?) Upper Oligocene in the overlying Ombilin Formation, which includes foraminifers
Sawahtambang Formation, with thick coarse, quartz-rich fluvial of Early Miocene age, giving an upper age limit for the older for-
sandstones deposited from braided streams, with overbank and mations (Silitonga & Kastowo 1975; Koesoemadinata & Matasak
flood plain silts and coals, 1365 m thick in Sinamar No.l well. 1981; Howells 1997a).
Outcrops of massive sandstones form cliffs and plateaux to the
west of Sawahlunto. The Sawahtambang Formation is equivalent Origin o f tile Ombilin Basin. Although there is general agreement
to the Sihapas Formation and marks the continued subsidence of that the Ombilin Basin developed as a half-graben, there is no
the basin and the renewed influx of sediment due to the uplift of agreement concerning the relative importance and timing of exten-
the source areas. The increase in volcanic clasts upwards in the sion, strike-slip faulting and compression in the development of
section indicates that volcanicity had commenced in the source the basin. Koning & Aulia (1985) and Situmorang et al. (1991),
area, which lay to the SW of the basin in the present forearc impressed by its close proximity to the Sumatran Fault System,
area (Howells 1997a). Towards the top of the Sawahtambang suggested that the basin had originated as a pull-apart basin in a
Formation fine green sandstones are less quartz-rich and contain dextral transcurrent fault regime. Although the basin has a major
glauconite as well as volcanic clasts (Howells 1997a), indicating controlling fault on its NE margin, this is a normal fault which
a marine incursion into the Ombilin area. The Lower Miocene has been inverted as a thrust. Strike-slip movement at some
Ombilin Formation, which overlies the Sawahtambang Formation stage is indicated by mismatch between clasts in conglomerates
conformably, is entirely marine, and consists of fine sandstones, and the adjacent basement lithologies along this margin
siltstones and claystone, often carbonaceous, with local lime- (Howells 1997a). There is, however, no complementary major
stones, 50 m to 100 m thick, which include lenticular coral and fault on the southwestern side of the basin. It is not known
 STRUCTURE AND STRUCTURAL HISTORY 225

Fig. 13.38. Diagrammatic

representation of the structural
development of the Central
Sumatra Basin, modified from
Yarmanto et al. (1995, fig. 3).
White circles with crosses and
dots indicate 'away' and
'towards', respectively, on
dextral strike-slip faults.

whether the northern margin of the basin is fault-bounded, but the downwards into a horizontal decollement surface, marked by a
sediments increase in thickness until they are covered by the thin band of comminuted coal. On the downthrown side of the
volcanic products of Malintang Volcano. At the southern end of fault a wedge of sandstone thickens towards the fault trace,
the basin the basement emerges from beneath the basin fill, with which has a total throw of 1.75 m. The structure is covered by a
no evidence of a major basin-bounding fault. The Ombilin Basin 2 m thick bed of unfaulted sandstone. A much larger version of
does not show the characteristic rhomboidal shape or pattern of a listric normal fault, with a throw of 4 - 5 m, is seen in the same
faulting seen in strike-slip pull-apart basins. road section. Other examples are seen in the open-cast coal pit,
but are continually being removed during the excavation of
Extensional structures. Detailed structures in the stratigraphic units
the coal. Extensional faulting also occurs much higher in the
in the Ombilin Basin can be studied in numerous river sections, succession, as a listric fault with a rollover anticline, broken by
roadcuts, quarries and in large open-cast coal pits. Evidence of small-scale normal faults forming a crestal graben, is seen in a
extensional faulting is ubiquitous. Normal faults are common in quarry in the Ombilin Formation, opposite the garage at Sijunjung
all stratigraphic units, in particular several spectacular outcrops on the Trans-Sumatra Highway.
of extensional listric growth faults have been described from the
Sawahlunto Formation. A fault in a road cut on the access road Compressional structures. As has already been mentioned, the
to the Parambahan open cast mine shows a NE-dipping curved Takung Fault which is the major bounding fault on the northeast-
surface marked with slickensides indicating normal move- ern side of the basin is interpreted as a reverse fault from mapping
ment (McCarthy 1997; Howells 1997a). The fault plane passes and in seismic section (Koning & Aulia 1985). Small-scale reverse
226 CHAPTER 13

Fig. 13.39. Geologicalmap of the

intramontane OmbilinBasin, Central
Sumatra, from Howells (1997a) based on a
compilationby de Smet (1991), from Musper
(1929), Koesmadinata& Matasak (1981)
and Koning& Aulia (1985).

faults are common in outcrops throughout the stratigraphic units and closures in the crests of anticlines, isolated by cross-cutting
in the Ombilin Basin. A clear example, dipping at 45', with N E - S W faults, which are potential oil-bearing structures. The
slickensides indicating up-dip movement towards the NW, is Sawahlunto and Sawahtambang formations are seen to thin
seen along the access road to the Parambahan Mine cutting a sand- towards the crest of the Palangki Anticline and the Ombilin
stone wedge associated with a listric normal fault, giving a clear Formation onlaps the flanks of this fold showing that this anticline
indication of relative age (McCarthy 1997). In the open-cast pit was a growth structure during the deposition of all the sedimentary
numerous reverse faults can be seen, varying in dip from steep units in the basin (Howells 1997a).
to fiat-lying. Fault planes often have a low angle of dip where Folding on the outcrop scale is seen in the Sangkarewang and
they pass through shales and become steeper when they cross Sawahlunto formations. Bedding in the laminated lacustrine
sandstone beds. One reverse fault was seen to be a reactivated shales of the Sangkarewang Formation often dip steeply or verti-
growth fault which has become a thrust related to the axial cally, with a strike parallel to the N W - S E trend of the basin.
plane of a monoclinal fold, passing into a hanging wall anticline Interbedded shales and sandstones of the Sangkarewang For-
where the thrust runs along a bedding plane (Howells 1997a). mation are folded in a complex fashion. Some of these folds
have been interpreted as due to sedimentary slumping, distin-
F o l d structures. Structurally the Ombilin Basin consists of two guished by being underlain and overlain by unfolded beds.
sub-basins. The Talawi sub-basin to the west trends N W - S E Howells (1997a) measured the orientation of slumps in the
and is relatively shallow, with outcrops of the Sangkarewang Malakutan River near Kolok and found the general trend of the
Formation around the margins in contact with the Pre-Tertiary fold axes was N W - S E , parallel to the basin margin, and the ver-
basement, and the Sawahlunto and Sawahtambang formations gence of the folds was to the NE. The inference is that the present
in the centre. This is separated from the NNW-SSE-trending basin margin is parallel to the margin of the basin during the depo-
Sinamar sub-basin to the east by the north-south Tanjung- sition of the sediments, and that the palaeoslope was northeast-
Ampolo Fault (Fig. 13.40). As may be seen from the cross- wards into the basin. The sediments of the Sangkarewang and
section (Fig 13.40), the Sinamar sub-basin is a composite syncline, Sawahlunto formations have certainly undergone a great deal of
with subordinate anticlines and synclines on axes trending gener- syn-sedimentary deformation, with the formation of unusually
ally N N W - S S E , parallel to the trend of the basin as a whole, and large-scale load casts (Moss & Howells 1996), injection of sand-
broken by a series of normal faults. Figure 13.40 shows contours in stone dykes into the interbedded shales and, presumably at a
seconds two-way-time on the top of the Sawahtambang Formation later stage, the injection of shales into lithified sandstone beds to
STRUCTURE AND STRUCTURAL H|STORY 227

Fig. 13.40. Structural map from Koning &

Aulia (1985, Fig. 7) and cross-section from
Williams & Eubank (1995, based on a
seismic profile from Koning & Aulia 1985)
of the Ombilin Basin, Central Sumatra. The
contours on the map are in ms two-way-time
on the top of the Early Miocene
Sawahtambang Formation, except for the
closure noted in the south, which is on the top
of the Oligocene Sawahlunto Formation. The
maximum depth of the basin is at least
4500 m.
228 CHAPTER 13

form flame structures, or in extreme cases, m61ange or 'broken Sawahlunto an outcrop of Triassic limestone is cut by a N W - S E
beds' (Howells 1997a). Evidently there was frequent earthquake fault with a 0.5 m breccia zone. On one side the breccia zone is
activity during the formation of the rift graben and the deposition planed off along a fault surface which shows horizontal grooving
of the graben fill. and slickensides indicating dextral strike-slip movement. Evi-
Folds, clearly of tectonic origin occur particularly in the thin- dently a phase of N E - S W extension was followed by a phase of
bedded lacustrine shales of the Sangkarewang Formation. N W - S E strike-slip faulting most probably related to movements
In road-cuts up to 5 m high between Atar and Sitankai, shales along the Sumatran Fault System.
and thin sandstones are folded on a large scale into chevron
Structural history. The structural development of the Ombilin
folds, with long limbs and tight angular hinges (45'~), overturned
Basin can be interpreted in terms of an initial phase of extension
towards the SW on NE-dipping axial planes. The beds show
during which the half-graben structure was formed. As indicated
slight thinning in the limbs and thickening in the hinges. Steeply
by the abundant listric normal growth faults, extension continued
dipping, alternating right-way-up and inverted beds along this
from the ?Late Eocene, during the deposition of the Sangkarewang
section, and in a similar 8 m high section along the road
Formation, until the Early Miocene, during the deposition of
between Talawi and Padang-Ganting, show that long-limbed
the Ombilin Formation. The evidence for the formation of the
folds with acute hinges are a common feature of the Sangkarewang
Palangki Anticline as a growth fold indicates that there was
Formation. In contrast to the Sangkarewang Formation, sediments
some differential subsidence of the basement within the basin.
in the overlying Sawahlunto, Sawahtambang and Ombilin for-
The extent to which extension was accompanied by a component
mations show gentle dips throughout the basin, although mono-
of transcurrent fault movement resulting in a pull-apart basin,
clinal folds are seen in open-cast coal pits in the Sawahlunto
as proposed by Koning & Aulia (1985) and Situmorang et al.
Formation, with one steeply inclined limb and with the hinge
(1991) is impossible to determine. A component of transtension
zone broken by thrust faults along the axial plane. The/bld axes
in the formation of the basin is probable, as normal extension
trend east-west and the folds are generally overturned towards
without some strike-slip component is exceedingly rare. As
the south (Howells 1997a).
in the basins in the backarc area, the extensional rift phase
was followed, during the deposition of the upper part of the
An unconformity?. The contrast between the steeply dipping and
Sawahtambang Formation and the Ombilin Formation, by a sag
highly folded Sangkarewang Formation and the relative lack of
phase due to thermal subsidence. Volcanic clasts indicate that
folding in the overlying units led de Smet (1991) to suggest that
an active volcanic arc lay to the west of the Ombilin Basin
there is an unconformity between the Sangkarewang and
showing that subduction of the Indian Plate was in progress at
Sawahlunto formations. Indeed de Smet (1991) and Howells
this time. From the vitrinite reflectance data and from projection
(1997a) report several sections where steeply dipping lacus-
of fold structures in the seismic section it is estimated that
trine beds of the Sangkarewang Formation are apparently
some 1800-2500 m of the Ombilin Formation has been eroded
overlain unconformably by horizontal or gently dipping, coarse,
following Plio-Pleistocene uplift of the Barisan Mountains.
Sawahlunto sandstones, although the actual contact is not
Vitrinite reflectance data indicate that the sedimentary units in
exposed. In his structural study of the Ombilin Basin, Lailey
the Talawi sub-basin did not subside to the same depth as those
(1989) pointed out that chevron lblds, with sharp angular hinges
in the Sinamar sub-basin, and that carbonaceous material in
and long limbs, seen in the laminated lacustrine shales of the
the sediments in this basin is more mature than sediments in the
Sangkarewang Formation, are characteristic of folds formed in
Central Sumatra Basin, due either to a greater depth of burial or
thin-bedded highly anisotropic rock units by compression parallel
to a higher heat flow.
to the layering. Chevron lblds are only formed where slip along the
Deposition in the Ombilin Basin was followed by compression,
bedding planes is possible during the formation of the folds
causing inversion of the basin with reversal of the movement
(Ramsay 1974). The geometry of the folds limits propagation of
on the listric normal faults, including the Takung Fault, the
the folds for any distance through the sequence, so that individual
major NE boundary fault which was inverted to form a thrust
folds die out both upwards and downwards. The folded package is
(Fig. 13.40 section). Compression also formed the major folds
bounded above and below by a decollement surface. Chevron
such as the Sinamar Anticline and Syncline, and intensified the
folding, commonly associated with monoclinal folds and box
Palangki Anticline. It was also responsible for the minor folding
folds with convergent axial planes, seen in the Sangkarewang
and thrusting seen in the Sangkarewang and Sawahlunto for-
and Sawahlunto tbrmations, is restricted to incompetent strati-
mations. The extent to which the compression was accompanied
graphic units with high anisotropy, and cannot be transmitted
by a component of transpression is again impossible to determine,
through the more competent and homogeneous sandstone units
but some component of transpression is probable.
of the Sawahlunto and Sawahtambang formations. The relation-
The final event in the structural development of the Ombilin
ships seen in the field, between steeply dipping shales and flat-
Basin was strike-slip faulting, dextral on a N W - S E trend and
lying sandstones above, do not therefore necessarily indicate
sinistral on the complementary east-west trend. Wherever the
an unconformity. Indeed, there is no evidence of a break at the
relative age of strike-slip faults, normal faults and folds could
boundary between the Sankarewang and the Sawahlunto for-
be determined, strike-slip faulting was always the youngest
mations in the vitrinite reflectance data from the Sinamar No.1
event. This major phase of strike-slip faulting is most probably
well, although other breaks in the succession were identified
related to movements on the Sumatran Fault Zone, which lies
(Koning & Aulia 1985).
only 10 km to the SW of the basin. As has been discussed in an
earlier section, movement on the SFZ commenced in the Middle
Strike-slip faulting. Small-scale strike-slip faults, with horizontal
Miocene, after the deposition of all the sediments now preser-
or sub-horizontal slickensides, are common throughout all the
ved in the Ombilin Basin. Uplift of the Barisan Mountains
units in the Ombilin Basin. Both Howells (1997a) and McCarthy
accompanied these movements. The Ombilin Basin then formed
(1997) made detailed studies of strike-slip faults from outcrops.
part of the source area for the Plio-Pleistocene sediments of the
They found that the faults are vertical or steeply dipping and fall
Central Sumatra Basin.
into two sets, one sinistral, trending east-west, and the other
dextral, trending N W - S E . Both sets of faults cut across dipping
beds and are therefore probably later than the folding. Where South Sumatra Basin
the relative age could be determined the east-west set is earlier
than the N W - S E set. Along the western side of the Ombilin The Central and South Sumatra basins have similar structural
Basin, on the road from the Trans-Sumatra Highway to and sedimentary histories and once probably formed a single
 STRUCTURE AND STRUCTURAL HISTORY 229

large basin, with a poorly defined division now marked by the the Muara Enim Deep (Benakat Gulley) and the Limau Graben,
exposed Pre-Tertiary basement in the Tigapuluh Hills and the occur to the SE, sometimes collectively refered to as the South
Duabelas Mountains (De Coster 1974) (Fig. 13.28). To the east Palembang Basin. Tertiary sediments reach a depth of 5 km in
the South Sumatra Basin is separated from the Sunda Basin the Benakat Gulley (Fig. 13.43). The basins are separated and sur-
in the Java Sea by the Lampung High, and its northward exten- rounded by upfaulted blocks where the Pre-Tertiary basement lies
sion in the islands of Bangka and Billiton; to the NE the basin at a relatively shallow depth, such as the Tigapuluh High in the
deposits thin out over the Sundaland basement in the Malacca north, the Musi and Kuang platforms in the south, and the Palem-
Straits; to the SW the basin is limited along the margins of bang, Tamiang and Lampung highs in the east (Fig. 13.42). From a
the Barisan Mountains by uplifted basement, exposed in the study of SAR (synthetic aperture radar) imagery and seismic data
Gumai and Garba mountains and the Gunungkasih Complex Pulunggono e t al. (1992) recognized lineaments with W N W -
(Fig. 13.41). Within the Barisans the basin deposits are covered ESE, N E - S W and north-south trends, which he considered rep-
by Pleistocene to Recent volcanoes and their volcanic products resent structures in the Pre-Tertiary basement which were re-acti-
(Kamal 2000). vated as normal faults during extension to form the highs, the
Internally the South Sumatra Basin is made up of the large basins and the troughs. Pulunggono e t al. (1992) suggest that the
Central Palembang Basin > 4 km deep, trending N W - S E with W N W - E S E lineaments, including the Lematang Fault, may
a northeastward extension into the N E - S W trending Jambi mark Mesozoic strike-slip faults in the basement, analogous to
Trough or (Jambi Sub-Basin) (Fig. 13.42). Two further basins, the present Sumatran Fault Zone, which were re-activated as

Fig. 13.41. Structure of the South Sumatra

Basin showing the distribution of folds and
faults, based on data from GRDC map
sheets, De Coster (1974), Pulunggono
(1986) Pulunggono et al. (1992) and Kamal
(1999). LBF, Lebak Fault; KF, Kikim Fault.
230 CHAPTER 13

Fig. 13.42. Basement structure of the South

Sumatra Basin based on an unpublished map
prepared by Pertamina/BElCIP (1985), with
additional data from GRDC maps, De Coster
(1974), Pulunggono (1986) Pulunggono et al.
(1992) and Kamal (1999) and showing
'deeps', 'troughs', grabens and 'highs', with
deplh to basement in seconds two-way-time
(TWT). The Lampung High which marks the
eastern boundary of the basin is about
100 km to the east of Palembang. The line of
the cross-section illustrated in Figure 13.43
is indicated.

normal faults during the Palaeogene. Unlike the North and Central the Pre-Tertiary basement. Conglomerate clasts include slate,
Sumatra basins, it has not yet been demonstrated that active phyllite, metasandstone, marble, basalt, andesite and vein
strike-slip faulting has played an important part in the develop- quartz derived from the underlying Tapanuli, Kuantan and
ment of the South Sumatra Basin, although Pulunggono et al, Woyla groups, and from intrusive granites. Towards the
(1992) report that the Lematang Fault is cut and displaced central parts of the basin the conglomerates pass into bedded
dextrally for 12 km by the north-south strike-slip Kikim Fault sandstones and siltstones with thin coals, and irregular carbonate
(Figs 13.41 & 13.42). layers and glauconitic and tuffaceous shales (De Coster 1974).
Environments of deposition are interpreted as scree, alluvial
S e d i m e n t a t i o n history. Apart from the greater importance of vol- fan, fluviatile and fresh to brackish water lacustrine. These
canic rocks, the sedimentary sequence in the South Sumatra deposits are followed by channel sandstones with silicified
Basin resembles those in the Central and North Sumatra wood, alternating with siltstones and carbonaceous shales, some-
basins (Fig. 13.43). The oldest deposits, the Lemat and Lahat times containing molluscs, with coal seams and tuffaceous units
formations (?Middle Eocene-Upper Oligocene), outcrop in the (Talangakar Formation, Upper Oligocene-Lower Miocene), laid
foothills of the Tigapuluh Hills and the Duabelas Mountains, down in a delta plain environment, from fluvial to lacustrine,
and are identified in boreholes and seismic sections along the lagoonal and shallow marine, becoming euxenic in the
margins of the troughs and graben throughout the basin. These troughs. In the troughs the Talangakar Formation follows
are volcanic and rift phase sediments, including breccias, conformably on the Lemat or Lahat Formation, but at the
conglomerates and 'granite wash', resting unconformably on basin margins becomes unconformable.
 STRUCTURE AND STRUCTURAL HISTORY 231

Fig. 13.43. Diagrammatic cross-section to

illustrate the tectonostratigraphic
development of the South Sumatra Basin
modified after Kingston (1988).

Differential subsidence, with reactivation of the marginal faults oldest Tertiary units in the basin, the Lemat and Lahat formations,
continued during the deposition of the Talangakar Formation, indicating later basement uplift. Fold structures, concentrated in
which marks a transgressive phase, and this is followed by the three broad anticlinal areas (anticlinoria), the Palembang, the
fully marine Baturaja and Gumai Formations (Lower-Middle Pendopo and Muaraenim anticlinoria, are best developed in the
Miocene), representing the period of maximum transgression. central part of the basin, where the Tertiary sediments are thickest
The Baturaja Formation is a thick platform carbonate unit, some- (De Coster 1974) (Fig. 13.41).
times including coral reefs, deposited on basement highs, passing The Palembang Anticlinorium extends southeastwards from
into bedded limestones and open marine shales in the intervening the Tigapuluh Hills to Palembang. It is made up of a series of
depressions. The area of deposition extended eastwards across N W - S E , elongated, narrow, periclinal, asymmetrical anticlines,
the Lampung High into the Sunda Basin. The Gumai Formation with intervening broader, basinal synclines. The more northerly
is composed of foraminiferal grey shales and siltstones, with inter- anticlines have steeper southern limbs, while the southern
calations of glauconitic and tuffaceous sandstone, which become folds have steeper northern limbs (Pulunggono 1986). In the
more important westwards towards the Barisans. At this stage Pendopo-Limau Anticlinorium SW of Palembang, the folds
the Barisan Mountains had ceased to exist and the area of sedi- have a more W N W - E S E orientation (Fig. 13.41), with limbs
mentation extended continuously from the backarc westwards dipping more steeply to the south; the fold axes are cut at frequent
into the forearc area. intervals by N E - S W normal faults. The anticline is considered to
Marine regression commenced with the deposition of the have formed as a drape over an uplifted basement block composed
Airbenakat and Muaraenim formations (Upper Miocene-Lower of Permian limestone and Cretaceous granite which outcrop in
Pliocene), which consist of sandstones and clays with coal beds the core (Gafoer et al. 1986) (Fig. 13.42). The Pendopo-Limau
and bands rich in molluscs and foraminifera. The overlying Anticline is limited to the south by the Lematang Fault, which
Kasai Fornaation (Pleistocene) rests with local unconformity on cuts the basement and has a throw of up to 1500 m to the south
the Muaraenim Formation and is composed of conglomerates, into the Benakat Gulley (Muara Enim Deep) (Pulunggono et al.
tuffaceous sandstones and tuffs with lignite and silicified 1992) (Fig. 13.42). The throw decreases eastwards and the fault
wood. The conglomerates contain clasts derived from the Pre- dies out into a monoclinal flexure.
Tertiary units and volcanic materials including pumice, marking The Muaraenim Anticline to the east of the Gumai Mountains in
the uplift of the Barisans and the eruption of active volcanoes. the southern part of the basin, is formed of a series of arcuate,
Sediment was also eroded from developing fold structures asymmetrical, periclinal folds with limbs which become steeper
within the basin and deposited locally. From the extrapolation and overturned towards the ENE, and are broken by thrusts
of the structure it is estimated that up to 1500 m of sediment (Pulunggono 1986) (Fig. 13.44). The folds are considered to be
has been removed from the crests of anticlinal folds (De Coster disharmonic, affecting Tertiary units above a detachment in the
1974). Gumai Formation (Fig. 13.44 section B - B ' ) . A gravitational
origin is suggested for these folds, formed by the slumping of
Structure. The structure of the South Sumatran Basin is dominated the Tertiary sediments towards the NE from the basement ridge
by outcrops of the Pre-Tertiary rocks in the Tigapuluh Hills, the which extends eastwards from the outcrop of Pre-Tertiary rocks
Duabelas Mountains in the north and along the Barisan front to in the Gumai Mountains (Pulunggono 1986; Holder et al. 1994)
the SW. The Pre-Tertiary rocks are fringed by outcrops of the (Fig. 13.44).
232 CHAPTER 13

Fig. 13.44. GeologicalMap and cross-sectionsof the Gumai Mountainsand the Muaraenim Anticlinoriumbased on GRDC 1:250000 Quadrangle Sheets of Bengkulu
(Gafoer et al. 1992c) and Lahat (Gafoeret al. 1986). Filled circles on the map are oil seeps, and open circles are gas seeps. The arcuate fold stuctures in the Muaraenim
Anticlinorium,shown in sectionB-B', are interpretedas due to gravitationalslidingfrom the upliftedGumai Ridgeon detachmentsurfaces withinthe GumaiFormation;
vertical lines in section B-B' are oil companyboreholes (after Pulunggono 1986).

Throughout the South Sumatra Basin anticlines are generally overlying Gumai Formation. Depocentres during the deposition
cored by outcrops of the older Tertiary units, the Talangakar, of the Talangankar Formation were situated in areas which later
Gumai and Airbenakat formations, while the synclines are cored became the sites of uplifted blocks (Pulunggono 1986). The
by the Plio-Pleistocene Kasai Formation. The folded rocks, rift phase was followed by thermal subsidence in a sag phase,
particularly in the southern part of the basin are covered uncon- which led to marine incursion and the deposition of fine grained
formably by Quaternary to Recent fluviatile and swamp deposits; marine sediments throughout the basin, with the formation of
the underlying structure being determined only from seismic data. carbonate reefs on the horst blocks. Continued subsidence led to
The deformation of the Tertiary sediments in the South Sumatra the drowning of the carbonate reefs and the deposition in deep
Basin evidently occurred in the latter part of the Pleistocene. water of the anoxic shales and marls of the Gumai Formation.
Study of microfossils and strontium isotopes in the Baturaja and
Structural history.The South Sumatra Basin was formed in the the Gumai formations from boreholes in the Muaraenim to
Late Eocene to Early Oligocene, at the same time as the North Baturaja area showed that the drowning of the carbonate platform
and Central Sumatra Basins, by extension of the Pre-Tertiary base- in a 'maximum flooding stage' was diachronous, and progressed
ment on pre-existing faults on W N W - E S E and N E - S W trends from west to east (Pannetier 1994). The shallowing upwards
and the subsidence of rift graben. North-south trends, dominant sequence in the Gumai Formation is correlated with a cooling
in the North Sumatra Basin and prominent in the Central event and a world-wide fall in sea level due to the formation of
Sumatra Basin, are less important in South Sumatra being rep- ice sheets (Pannetier 1994). Contrary to the earlier pattern, the
resented only by the Kikim Fault margining the Benakat Gulley greatest thickness of the Gumai Formation, occurred in the areas
and parts of the Lembak Fault terminating the Pendopo-Limau which are now depressions (Pulunggono 1986). Pulunggono
Anticlinorium (Pulunggono 1986), although the long straight (1986) attributes this tectonic inversion to the onset of com-
eastern coast of Sumatra, facing the Java Sea, appears to be con- pression in the South Sumatra Basin in the early Mid-Miocene,
trolled by a major north-south fault. The troughs were infilled due to the renewal of the subduction of the Indian Plate beneath
by erosion products derived locally from basement horsts in the west Sumatra.
rift phase. The marginal faults show their greatest amount of Marine deposition continued throughout the region in the
throw in the Talangakar Formation, and die out upwards into the Mid-Miocene, extending westwards into the forearc and eastwards
 STRUCTURE AND STRUCTURAL HISTORY 233

into the Sunda basins. As the subduction system became at about the same time throughout the whole SE Asian region.
established, volcanicity and uplift of the basement in the Barisan These basins are therefore the result of processes which affected
Mountains led to a marine regression, with the deposition of the whole of SE Asia. The precise age of formation of these
terrestrial sediments late in the Mid-Miocene, which gradually basins is difficult to determine, as the earliest deposits are
extended eastwards to cover the whole of the South Sumatra usually of terrestrial origin and do not contain age-diagnostic
Basin by the Pleistocene. Perhaps commencing in the Late Plio- fossils. However, it appears that during the Late Eocene to Early
cene, but completed during the Pleistocene, the basin became Oligocene extensional basins were formed across the area from
subject to N E - S W compression, reactivating basement faults, the Java Sea in the south, through Sumatra and the Malay Basin
uplifting basement blocks and generating folds on N W - S E axes and to Vietnam in the east.
in the overlying sediments. Variations in the vergence of the The regional extent of basin formation during the Palaeogene
fold structures, from NE to SW, are attributed to the movement has encouraged the search for a regional rather than a local expla-
of the developing folds in the Airbenakat and Muaraenim for- nation. The model proposed by Tapponnier et al. (1986) for the
mations away from the areas of basement uplift on decollement southeastward extrusion of SE Asia following the collision
surfaces within the underlying Gumai Formation (Pulunggono of the Indian continent with the southern margin of Eurasia,
1986). Locally, particularly on the Barisan Front along the commencing in the Eocene, seemed to provide such a solution.
western margin of the basin, both the basement and the overlying In this model SE Asia was extruded as a set of continental
sediments have been affected by N E - S W dextral strike-slip faults slivers separated by strike-slip faults, opening up pull apart
related to movements along the Sumatran Fault System. basins between the continental fragments as they moved away
differentially from the site of the collision.
Attempts have been made to interpret the basins in the Sumatran
Origin of basins in the Sumatran backarc basins backarc area as pull-apart basins formed during strike-slip move-
ments (e.g. Davies 1984; Daly et al. 1987, 1991). Davies (1984)
The review of the structural development of the linear belt of for example, attributed the formation of the Sumatran basins
Tertiary sedimentary basins in eastern Sumatra given above to pull-aparts between strike-slip faults which changed their orien-
shows that they were initiated as rift systems generated by tation in response to the rotation of a Sunda Plate, including
extension and thinning of the crust. The resulting high heat flow Sumatra, and variations in the direction and rates of subduction
was followed, after extension had ceased, by the development along the Sumatran margin. However, the consensus view is that
of sag basins due to thermal relaxation, enhanced by sediment the basins developed initially as extensional rifts, controlled
loading. In the literature these basins have generally been by the orientation of pre-existing lineaments in the Pre-Tertiary
described as 'backarc basins', as they occupy a backarc position basement. The orientation of these lineaments is different in the
relative to the active volcanic arc of Sumatra. The implication is three basins, being north-south in the North Sumatra Basin,
that these basins developed directly as the result of the activity N W - S E and N E - S W in the Central Sumatra Basin and N N W -
of the arc. For example Eubank & Makki (1981) have suggested SSE and N E - S W in the South Sumatra Basin. Strike-slip move-
that backarc extension was due to the establishment of convection ment in the backarc area is superimposed on these earlier trends
cells and diapirism in the mantle set up by the subduction of the and coincides with the uplift of the Barisan Mountains, the
Indian Plate. The implication is, that if extension had continued inversion of structures in the sedimentary basins and movements
the continental crust would have ruptured, with the generation along the Sumatran Fault System.
of oceanic crust, to form a backarc marginal basin, similar to Apart from specific problems with the application of the extru-
those associated with the subduction systems of the Western sion/strike-slip model to the basins in the Sumatran backarc,
Pacific. there are general problems in its application to SE Asia as a
One problem with this interpretation, evident from the foregoing whole. Many of the these basins, particularly on the Sunda
account, is that the formation of basins in Sumatra in the early Shelf in the South China Sea and the Java Sea are not correctly
Tertiary was not restricted to the backarc region. Outlying rem- oriented to have originated as pull-apart basins related to dextral
nants of Tertiary deposits within the Barisan Mountains indicate movements on strike-slip faults (Hall & Morley 2004). The
that these basins once extended across the site of the mountains impression given by the distribution of the basins is that there
to join with similar and contemporaneous forearc basins to the was overall expansion of the whole SE Asian area during
west. The term 'backarc basins' as applied to the Tertiary basins the Palaeogene. Hall & Morley (2004) point out that the area
of the Sumatran backarc area is therefore a misnomer. The between Sumatra, including the backarc area, and eastern
present position of these basins is due to the subsequent rise Borneo has a very high surface heat flow > 8 0 m W m -z, com-
of the mountains and the construction of the volcanic arc that pared with the average for continental crust of 40 mW m-2. The
separated the basins so that they now occupy forearc and highest surface heat flow, up to 1 8 0 m W m -2, occurs in the
backarc positions. Morley (2002b) has suggested that the for- Central Sumatra Basin, the Malacca Strait and the adjacent part
mation of basins across Sumatra was due to 'subduction roll- of the Malay Peninsula. Hall & Morley (2004) suggest that the
back' caused by the sinking of the incoming Indian Plate, high heat flow is due to high temperatures in the mantle, as indi-
drawing the whole subduction system forward, causing extension cated by low seismic velocities below the region, the result of
in both the forearc and backarc areas. long continued subduction. A contribution may also come from
Another problem with hypotheses that relate the formation of the upper crust, which in the SE Asian region contains a high pro-
the sedimentary basins in the Sumatran backarc area to their portion of radiogenic granites, insulated by a thick sedimentary
relationship to the present subduction system is that these basins pile including shales and coals. They go on to suggest that the
are not restricted to Sumatra, but form part of a network of exten- opening of the Tertiary sedimentary basins in SE Asia was due
sional rift basins, originating in the early Tertiary which formed to mantle/lower crustal flow.