By

Joep E.A. Storms

a, ,
,
Robert M. Hoogendoorn
a
,
Rien A.C. Dam
a
,
A.J.F. Hoitink
b, c
,
S.B. Kroonenberg
a

Abstract
The late-Holocene Mahakam delta, located along the tropical eastern shore of Kalimantan, Indonesia, is considered to be a textbook exa mple of a mixed tidefluvial dominated delta system. The
delta prograded about 60 km during the past 5000 years, which led to the development of a distinct network of distributary an d tidal channels. Wave action is low due the limited fetch in the
narrow strait of Makassar. Mahakam River discharge is about a quarter of the Mississippi River discharge and is characterized by absence of flood surges. Therefore, natural levees, crevasse
splays and avulsions are absent in the delta plain. For the past four decennia, both modern and ancient Mahakam delta deposit s have been studied in detail in order to better understand
subsurface Miocene and Tertiary Mahakam deposits, which host large volumes of hydrocarbons.
This study focuses on the dynamics and stratal patterns of delta plain, delta -front platform deposits and suspended sediments. Due to the predominance of semi -diurnal tides and the associated
flow reversals, depositional patterns are highly variable which has resulted in the formation of characteristic sand mud couplets. The distribution of the sandmud couplets found in this study
differs from previously proposed conceptual models. They are limited to the fluvial domain and form in the distributary chann els (lateral channel bar) or at the fluvial dominated delta-front platform,
which flanks the mouth bar deposits in offshore direction. The sand mud couplets which formed as delta-front platform and lateral channel bar deposits are similar and can only be identified
based on their
14
C age. The sand content decreases significantly towards the tidal dominated areas due to limitation in transport capacity. Tu rbidity measurements taken in front of the river mouth
also show rapid settlement of river plume sediments.
Some 22 new AMS
14
C dates show that late Holocene sea level history resembles the eustatic sea level curve giving a first approximation of the Late Holocene sea level history for East
Kalimantan. The dates suggest that the progradational delta system evolved under conditions of slowly rising sea level, which compares well to the eustatic sea level curve. In addition, calculated
averaged deposition rates of the sandmud couplets indicate that deposition is driven by the springneap tide cycles instead of the daily tidal cycle.
Keywords
Tidalites; Sea level; Radiocarbon dates;Subaqueous delta;Fluvial processes; Distributary channels
1. Introduction
The late-Holocene Mahakam delta, located along the tropical eastern shore of Kalimantan, Indonesia ( Fig. 1) prograded about 60 km during the past 5000 years which led to the development of a
distinct network of distributary and tidal channels, that has formed a lobate, fan shaped delta. The present day subaerial de lta covers 1800 km
2
of partly undeveloped tropical lowland
forest, Nypa swamps and a mangrove shoreline (Fig. 2). Based on the morphologic characteristics, the Mahakam delta is considered to be a textbook example of a mixed tide fluvial dominated
delta system ( Galloway, 1975). Yet, only few studies have examined the interaction between tidal and fluvial processes at the Mahakam delta, in contrast to other fluvial-tide dominated systems,
for example the GangesBrahmaputra ( Goodbred and Kuehl, 1999), McArthur river ( Jones et al., 2003) and the Seine ( Lesueur et al., 2003). The origin and distribution of fluvial deltaic facies in
relation to tidal processes remains controversial for the Mahakam delta, despite detailed studies of both modern and ancient Mahakam sediments ( Allen and Chambers, 1998, Allen et al.,
1977,Allen and Mercier, 1994, Gastaldo et al., 1995 and Gastaldo and Huc, 1992). The main focus of these studies has been the modern mouth bar, channel and delta -front slope that share
many similarities with the Miocene and Tertiary Mahakam delta deposits, which host large volumes of hydrocarbons ( Howes, 1997).
Fig. 1.
Location map for the Mahakam delta study area on the east coast of Kalimantan, Indonesia (after Roberts and Sydow, 2003).
Fig. 2.
Map of the study area showing the subaerial Mahakam delta plain with vegetation zones (after Allen and Chambers, 1998) and a shaded relief presentation of shelf and slope bathymetry (after Roberts and Sydow, 2003). Contours
are in metres.
The pioneering work by Allen et al. (1977) on the evolution of the modern Mahakam is still a standard reference work describing the individual depositional environments that comprise the delta
plain, delta-front platform, delta-front slope and prodelta environments (Fig. 2). A deep core (MISEDOR; 638 m) was drilled in 1984 by IFREMER in the southern part of the modern Mahakam
delta (Caratini and Tissot, 1988). Based on this core, Pleistocene and Holocene Mahakam facies are described using palynological studies. Four radiocarbon da tes establish the depths of the
Pleistocene base over which the early Holocene transgressive sequence and the late Holocene prograding sequence were deposite d (Caratini and Tissot, 1988 and Carbonel and Moyes, 1987).
In the 1990s, the focus of research shifted to the Pleistocene and Holocene offshore deposits as high -resolution shallow seismic data became available ( Roberts and Sydow, 1996, Roberts and
Sydow, 2003, Snedden et al., 1996 and Suiter, 1996). They reconstructed the delta development back to the Late Pleistocene in detail. Until then, very little was known about t he delta-front slope
and prodelta deposits of both the modern as well as the Pleistocene Mahakam depositional units. However, as no seismic and co re data were retrieved beyond the sub-tidal zone, the extent of
the distribution of deltaic facies in relation to tidal processes and fluvial dynamics remained unexplored.
Following the early work of Allen et al. (1977) this paper focuses on understanding the interaction between tidal and fluvial processes that led to the formation of the late -Holocene delta plain and
delta-front platform morphology and stratigraphy. Onshore shallow cores, suspended sediment measurements and new AMS
14
C dates provide new insight in the spatial distribution of delta plain
and delta-front platform deposits. We propose a depositional mechanism that explains the cross - and along-shore heterogeneity found in the delta plain and delta-front platform deposits. In
Late-Holocene evolution of the Mahakam delta, East Kalimantan,
Indonesia
14 Mei 2014
11:00
PETROLEUM Page 1
addition, the AMS
14
C dates give insights into spatial deposition rates and provide a first indication of the local sea level history.
2. Regional setting
2.1. Geologic setting
The modern Mahakam delta is located in the lower part of the Kutei Basin, which is situated on the east coast of Kalimantan, Indonesia, just south of the equator. The Kutei basin was initiated by
rifting during the middle Eocene and was subsequently filled by syn -rift clastic sedimentation as well as carbonates until the early Miocene ( Payenberg et al., 2003). The complete section up to
the lower Miocene was subsequently folded to form the Samarinda anticlinorium, which became the basement for continuous subse quent fluvio-deltaic deposition in the Kutei Basin (Tanean et
al., 1996). The Mahakam River incised the Samarinda anticlinorium during the middle Miocene and successive deltas have since prograded from a single point source near the present delta apex
(Allen and Chambers, 1998 and Payenberg et al., 2003). The long-term subsidence rate of the lower Kutei Basin is estimated at 0.20.5 m/ky (Roberts and Sydow, 1996 and Roberts and Sydow,
2003).
Three major deltaic complexes, separated by two marine transgressions, characterise deposition in the lower Kutei Basin since the lower Miocene. These complexes are dated as middle
Miocene, MiocenePliocene, and Holocene (Trevena et al., 2003). The MiocenePliocene deltaic sequences have been extensively described by Trevena et al. (2003) and consist of various
phases of delta development represented by heterogeneities in the deposits caused by different climate regimes and tectonic u plift of the catchment area. The most recent progradational phase
of the Mahakam delta started about 5000 yr BP ( Caratini and Tissot, 1988 and Comabaz and Matharel, 1978) and continues up to the present day.
2.2. River system characteristics
The present Mahakam River drains about 75,000 km
2
of the (upper) Kutei Basin as well as part of the uplifting central Kalimantan ranges. From available rainfall data and the s ize of the drainage
basin, a mean water discharge was evaluated by Allen and Chambers (1998) of the order of 3000 m
3
s
1
with large seasonal variations. Floods of up to 5000 m
3
s
1
may occur in the upper and
middle reaches of the catchment, which is separated from the river mouth of the drainage basin by a subsiding area characteri zed by a low relief alluvial plain and several large lakes, located
some 150 km upstream of the delta plain (Roberts and Sydow, 2003). The lakes create a buffer causing the dampening of the flood surges ( Allen and Chambers, 1998) and effectively level off
Mahakam River floods, resulting in a constant discharge for lower reaches of the Mahakam River and delta system.
The absence of peaks in river discharge has resulted in a delta plain that has no natural levees or crevasse splays. Therefor e, avulsions of distributary channels do not take place. Instead, as the
delta prograded, bifurcations of the fluvial distributaries occurred about every 10 km. The climatic conditions are tropical, with only a slight monsoon impact. Rainfall ranges from about 4000 to
5000 mm per year in the central highlands to 2000 to 3000 mm per year near the coast ( Roberts and Sydow, 2003) and has a high in January and a low in August. Temperature shows little
interannual variation and is on average 25.5 C. The average sediment discharge is estimated at 8 * 10
6
m
3
yr
1
(Allen et al., 1977), although this number is not based on actual field
measurements but on volumetric calculations.
2.3. Marine processes
The tides in the Strait of Makassar are semi -diurnal with a considerable diurnal inequality. Tidal amplitude ranges from less than 1 m in neap tides to about 3 m during s pring tides that occur with
a 15 day periodicity (Allen and Chambers, 1998). Although these tide amplitudes are not very large compared to macro tidal amplitudes, they are sufficient to generate stro ng tide-induced
currents which lead to reversing the flow direction of the Mahakam River as far upstream as Samarinda, located about 20 km up stream from the delta apex. The wave energy that affects the delta
is very low. Low wind speeds and the small fetch in the Makassar Strait result in a significant wave height of less than 0.6 m (Roberts and Sydow, 2003). Furthermore, the shallow submerged
delta-front platform, which extends to the 5 m isobath several kilometres offshore ( Fig. 2), dissipates wave energy so that only low-energy waves reach the coastline. Wave direction varies from
north to southeast. The Makassar Strait is subject to a southern throughflow that generally persists throughout the year ( Murray and Arief, 1988).
2.4. Delta morphology
The modern Mahakam delta (Fig. 2) forms a regular fan-shaped lobate system with a thickness around 50 to 70 m. It covers about 1800 km
2
of predominantly subaerial delta plain and about 2000
km
2
of delta-front platform and delta-front slope accumulations (Allen et al., 1977). During the past 5000 yr, the delta plain has prograded up to 60 km over older Holocene transgressive deltaic
deposits (Roberts and Sydow, 1996). The modern Mahakam has two active fluvial distributary systems directed northeast and southeast. The intervening areas con sist of tide-dominated areas,
with many high sinuosity channels which are mostly not connected to the fluvial system ( Fig. 2). This interdistributary zone occupies approximately 30% of the delta plain ( Allen et al., 1977).
The delta-front platform consists of a shallow (< 5 m water depth) intertidal to subtidal platform fringing the densely vegetated lower delta plain ( Fig. 2). The width of the delta-front platform varies
between 10 and 20 km and its outer limit is marked by a strongly inclined slope seaward of the 5 m isobath ( Allen et al., 1977). The delta-front platform contains subaqueous channels ( Lambert,
2003), which prolong both the tidal channels and fluvial distributaries. The delta -front platform environment is the area of continuous interplay between tidal and fluvial processes, which has led to
a complex depositional pattern ranging from fluvial related sandy bars to tide -related mud, which will be more extensively discussed later in this paper.
The delta-front slope is located beyond the 5 m isobath and marks a sharp increase in sea bottom slope. The outer limit of the delta -front slope is estimated at a water depth of around 70 m where
it passes into the prodelta. Deposition on the delta-front slope consists of suspended silt and clay ( Allen et al., 1977). The bathymetry of the shelf (Fig. 2) clearly shows rough bottom bathymetry
that is formed by individual and aggregated mound-like bioherms (Roberts and Sydow, 1996 and Wilson and Lokier, 2002). These bioherms (Halimeda calcareous alga) can be up to 20 to 30 m
thick and are located on a regional erosion surface that separates the latest Pleistocene deposits from the modern Mahakam pr odelta deposits (Fig. 3) (Roberts and Sydow, 2003). As sediment
from the Mahakam is preferentially transported in southerly direction due to a north to south directed longshore current ( Murray and Arief, 1988), the reefal build ups in the north appear to be
close to the prodelta (20 km), whereas in the south they are much further away ( 40 km) (Magnier et al., 1975).
Fig. 3.
A schematic representation of pre-Holocene and Holocene Mahakam stratigraphy (modified after Roberts and Sydow, 2003) with MISEDOR core (Carbonel and Moyes, 1987) location and main time and depth boundaries of
important Pleistocene and Holocene surfaces. In the offshore zone several carbonate bioherm structures are situated, which ar e remains of the late Pleistoceneearly Holocene transgressive phase.
The Mahakam delta plain is densely vegetated, with three vegetation zones indicating the relative elevation of the land to se a level (Fig. 2). A mixed-hardwood forest characterises the well -
drained supratidal upper delta plain and extends about 10 to 20 km from the landward limit of the delta. The upper delta plai n has a low flooding frequency and therefore a low deposition rate. The
lower delta plain is vegetated with Nypa palms and mangroves. The mangrove forest colonises newly formed intertidal flats and is very successful in trapping fine -grained sediment. As soon as
the intertidal flat reaches a higher elevation, the mangrove forest is replaced by the characteristic Nypa palms.
3. Data collection
In September 2003 a two-week field campaign was conducted in the Mahakam delta plain using a hand -augering device for onshore coring and a small speedboat for logistics. This enabled us to
disembark at any place along the river and core in densely vegetated areas. A total of 28 cores ( Fig. 4) were collected with a maximum depth of 8.5 m and a core diameter of 3 cm. Hand -augering
is a fast and easy-to-use method to recover a continuous subsurface sediment sample of 1 m length in unconsolidated sediments ranging in grain size from clay to fine sand, but is not capable of
penetrating into coarser sediments below the ground water table. This problem has at times limited the coring depth. The core s were described in the field, assessing grain size, organic content,
macrofossils and colour. 22 samples were collected for AMS
14
C dates (Table 1). To decrease the possibility of sampling reworked organic material, we collected predominantly soft fibres
from Nypa palms, which were easy to identify. In four cases we collected wood samples. A differential GPS system was used for determini ng the core position. This worked well and determined
the horizontal position of the coring-locations, within a 1 m radial resolution. The GPS measurements did not provide adequate vertical positioning. Therefore, it was assumed that the topographic
features in the delta plain are negligible and that a very slight sloping topography from apex to the delta -front prevails.
PETROLEUM Page 2
Fig. 4.
Map showing core locations and associated cross -sections, and the locations of the acoustic Doppler current meter measurements presented in Fig. 6 and Fig. 7, with the isobaths in the vicinity of these measurements.
Table 1.
Calibrated
14
C AMS dates in calendar years with interpreted facies types (DP = delta plain, TDDF = tidal dominated delta -front platform, FDDF = fluvial-dominated delta-front platform, LCB = lateral channel bar, MB = mouth bar)
Shipboard measurements were taken from the Geomarin research vessel. For current velocity observations the vessel was equippe d with a broad-banded 1.2 MHz Acoustic Doppler Current
Profiler (ADCP), manufactured by RDI. The ADCP transmitted 8 pings per ensemble along four beams, at an angle of 20 to the v ertical. The depth cell size was set at 0.25 m. Stratification of
water masses was investigated using a Seabird CTD (Conductivity Temperature Depth) probe, which sampled synchronously with th e ADCP observations. The CTD casts were taken using a
winch on the research vessel. Vertical profiles of salinity were obtained using the upcasts only, which covered the entire wa ter column except for the bottom 50 cm. A Seapoint Optical Backscatter
sensor (OBS) was attached to the CTD probe. For the OBS measurements, only the downcasts were used, as the CTD probe could cr eate some sediment clouds when it touched the seabed. For
calibration purposes, a Niskin bottle was attached parallel to the CTD to collect in situ water samples. The suspended sedime nt content (SSC) of the water samples was measured by vacuum
filtration of a fixed amount of water on pre-weighed polycarbonate filters with a pore size of 0.4 m. After filtration, the filters were cleaned with nano -pure water to remove salts, washed with
alcohol, and dried and weighed. The regression results of the SSC measurements taken from water samples against OBS voltage a re shown inFig. 5.
Fig. 5.
Plot of the OBS calibration. In situ water samples were taken at various sites in the southern river branch and adjacent rive r mouth of the Mahakam River.
Fig. 6 presents timedepth plots for longitudinal current velocity, salinity and suspended sediment concentration over a tidal cycle at an anchor station in the southernmost active river
distributary (Fig. 4). It shows that the location is close to the outer region up to which the salt intrusion reaches, with salinities remaining well below 10 PSU. The flood currents are
relatively weak because the river discharge is directed oppositely. During flood tide, suspended sediment concentrations grad ually decrease roughly from 100 to 50 mg/l, indicating that
the entering waters are relatively clear. To some extent, the opposite pattern occurs at the time of the ebb currents, which convey relatively turbid waters from upstream and cause
resuspension near the bed. The gradual decrease in suspended sediment concentration during flood tide in the main river distr ibutary indicates a negative downstream turbidity gradient,
which is in accordance with regional deposition of alluvial sediments on the delta plain.
PETROLEUM Page 3
Fig. 6.
Measurements of longitudinal flow velocity, salinity and suspended sediment concentration (SSC) at an anchor station about 20 km inland from the river mouth ( Fig. 4), indicating a negative downstream turbidity gradient.
Downstream of the ri ver anchor stati on, ri ver mouth processes were observed by taki ng CTD/OBS casts alternatel y at three l oca ti ons on the del ta-front pl atform, respecti vel y in 6, 14 and 30 m water depth ( Fi g.
4). The measurements l asted for a ti dal peri od about hal fway through the spri ng neap cycle (Fi g. 7). Just before sl ack water after fal l ing ti de, a water mass wi th a sal inity of about 25 PSU progressivel ymi grates
seaward, dumpi ng the majori ty of the sedi ments between the 6 and 14 m i sobaths on the channel i sed del ta -front pl atform. Peak l evel s of suspended sedi ment concentration amounted to 60, 40 and 25 mg/l at
the l ocati ons at 6, 14 and 30 m, respecti vel y. The observed rapi d settl ement of the sedi ments on the del ta -front pl atform i l l ustrates the day-to-day progradati on of the del ta due to suspended cl ays and silts. The
sal inity profile shows that the sedi ment pl ume remai ns i ntact wi thout mi xi ng wi th seawater, an i ndicator of rapi dsettl ement of ri ver sedi ments.
Fig. 7.
Quasi-simultaneous observations of the development of a river pulse in front of a distributary channel at the delta -front platform at three locations (6, 14 and 30 m water depth), showing rapid settlement of suspended sediment.
See Fig. 4 for location. SSC: suspended sediment concentration.
4. Sedi ments and strati graphy of the l ate-Hol ocene del ta devel opment
Whereas suspended sedi ments are predomi nantly composed of si lt and cl ay, bed l oad sedi ments are rel ativel y sandy, l eading to mouth bar formati on (Bates, 1953). At fal l i ng ti de, fl ow i s di rected seawards and
sedi ment i s transported towards the mouth bar and del ta -front pl atform. Deposi ti onof sand occurs i n a fan shaped manner, wi th decreasing grain size and quantity as fl ow competence decreases i n an offshore
di recti on (Fi g. 8A). The extent of the area of sand deposi ti on i s variable; i t i ncreases during the ebb phase of neap ti de and decreases duri n g the ebb phase of spri ng ti de. Measurements of suspended sedi ment
concentrati ons di rectl y in front of a mai n di stributary channel i ndicate that hi gh sedi ment l oads are restricted to the shal l ow part of the del ta-front pl atform (Fi g. 7). Li ttl e sedi ment i s transported beyond the 15 m
i sobath and the background rate of suspended sedi ment concentration i s l ess than 20 mg/l . Thi s i mpl ies that most of the sedi m ent transported by the Mahakam Ri ver i s deposi ted i n the del ta -front pl atform zone.
Fig. 8.
PETROLEUM Page 4
Fig. 8.
(A) Schematic representation of the dispersal of sand and clay under varying tidal conditions. During rising tide conditions (I) depositional processes lead to extensive clay deposition. As the tide falls (II and III), sand is dispersed by
the fluvial channel in a fan-like shape in the absence of a mouth bar (II) or in the presence of a mouth bar (III). (B) shows a schematic representation o f the facies distribution in the Mahakam delta.
Most mouth bars l ocated i n front of acti ve Mahakam di stributary channel s are i nterti dal and are aeriallyexposed duri ng l ow t i des. In these cases, the water fl ows around the mouth bar, thereby i ncreasing the
angl e of sedi ment di spersal over the del ta -front pl atform (Fi g. 8). Accordi ng to Al l en and Merci er (1994) thi s eventual lywi l l lead to a new bi furcati on point of the di stri butary channel . If, however, the mouth bar i s
not yet wel l devel oped, meani ng that i ts topographi c expression is small, then the fl ow wi l l be di rected both over and al ong the mouth bar, causi ng i t to mi grate seawards as wel l as upward. On the del ta pl ai n,
fl oodi ng water from the previ ous hi gh ti de cycl e i s drained from the i nterti dal l ower del ta plains during falling ti de.
Duri ng ri sing ti de, fl ow i s directed i nland i n both the di stri butary andti dal channels. The ri sing water l evel causes the de l ta pl ai n to be fl ooded as water i s forced i nto the extensi ve network of ti dal channel s and
creeks. The mangrove and Nypa vegetati on traps the cl ay parti cles whi ch are suspended i n the fl oodwater. On the del ta -front pl atform, cl ay deposi tion i s wi despread as fl ow competence i s l ow ( Fi g. 8A). Mud
deposi ti on i n the di stributary channel i s partly driven by the presence of a turbi di ty maxi mum i n the l ower reaches of the di stri butary channels ( Al l en and Chambers, 1998).
Sedi ments from both hi gh and l ow ti de condi ti ons accumulate wi th a vari able degree of erosi on or reworki ng as the del ta progr ades, whi ch l eads to a zonati on of fi ve faci es types across the del ta pl ai n anddel ta -
front pl atform (Fi g. 8B). The faci es types are cl assified according to thei r average grai n size as wel l as heterogenei ti es i n grain size and colour . The fi el d observati ons of the grai n si ze were confi rmed by anal ysis of
43 grai n si ze samples taken from core 12, usi ng a l aser diffraction spectrophotometer ( Konert and Vandenberghe, 1997).
4.1. Faci es
Lower delta plain facies (DP) consi sts of massive cl ays, deposited onshore i n the ti dal -domi nated i nterdistributary areas, whi ch are ri ch i n rootlets and have a variable degree of organi c content and macrofossils.
Sporadi cally there may be some sand or si l t i ncursions. The thi ckness of the l ower del ta pl ai n facies i s rel ated to the ti dal ampl i tude. Mangrove starts to col onize ti dal fl ats as soon as they become i nterti dal . Thi s
restri cts the thi ckness of the l ower del ta pl ai n facies to about 3.5 to 4 m.
Tidal-dominated delta-front platform facies (TDDF) consi sts of cl ayey to si l ty deposits that show di stinct colour l aminations at mi l limetre scale. Organic content i s l ow or abs ent and macrofossils are sparse, root
burrows may occur. Deposi tion typi callytakes pl ace offshore from the ti dal -domi nated i nterdistributary areas away from the ri ver mouth and hence thi s faci es l acks sandi nclusions. Neverthel ess, the va ri ability i n
fl ow regi me over the ti dal cycl es results i n mi llimetre-scale l aminati ons of cl ay drapes.
Fluvial-dominated delta-front platform facies (FDDF) consi st of di sti nct l aminated sand, silt and cl ay bands wi th thi cknesses from on average 1 to 5 mm, and someti mes up to a few centi metres. The boundari es
between l ami nae are sharp, however structural trends i n the thi ckness of the l ami nae as described i n other subti dal deposits ( Vi sser, 1980) were not observed. The fl uvi al -domi nated del ta-front pl atform facies
ori gi nates from fl ow reversal i n the vi ci nity of the ri ver mouth as sand i s deposited duri ng l ow ti de and cl ay i s deposited d uring hi gh ti de condi tions. Thi s facies type i s di rectly rel ated to the di stri butary channel
system as the fl ow capaci ty i n ti dal channels i s too l ow to transport sand ( Al l en et al ., 1977).
Lateral channel bar facies (LCB) consi st of sandmud coupl ets similar to fl uvi al -dominated del ta-front pl atform facies. Lateral channel bar facies form al ong the margi ns of a di stributary channel as the channel
thal weg sl owl y changes i n ti me. As such, l ateral channel bars are formed and eroded as a functi on of autocycl ic channel behav iour and therefore thei r age bears no di rect rel ati on to the depth of deposi ti on nor
wi th sea l evel ( Al l en et al ., 1977 and Al l en and Merci er, 1994). As a prograding di stributary channel erodes underl ying ol der del ta -front pl atform facies, l ateral channel bar facies are expected to be consi derably
younger than surroundi ng fl uvial -domi nated del ta-front pl atform facies. Cores that showed l ateral channel bar facies were l ocated wi thi n several metres from a di stri butary ch annel ( Fi g. 4, core 7, 10, 13 and 14).
Al though the di stri butary channel s are straight and do not mi grate, Al l en et al . (1977) poi nted out that the channel thal weg does meander. At the i nner bend of such thal weg meanders, subaqueous poi nt bar -like
features evol ve consi sting of sandmud coupl ets, and whi chwere cl assified as l ateral channel bars facies. The resembl ance to fl uvi al -dominated del ta-front pl atform facies l ead Gastaldoet al . (1995) to bel i eve
that al l sandmud couplets could be cl assified as l ateral channel bar deposits.
Mouth bar facies (MB) consi sts of massive, wel l sorted medi um grai ned sand (> 0.25 mm). It occurs di rectl y i n front of the di stri butary channel and therefore consists of much coarser sand and better sort ed than
the sands i n the surrounding fl uvial -domi nated del ta-front pl atform facies. The mouth bar can bui l d up to i nterti dal el evation and therefore may be subaeri ally exposed during l ow ti de. Si nce our cori ng techni que
cannot penetrate these coarse sand bodi es, we refer to Al l en and Merci er (1994) for a more detai l ed descri ption of the mouth bar sedi ments, thei r evol uti on and di stribution.
4.2. Faci es successi on
Core data are used to construct three faci es profi les ( Fi g. 9) whi ch gi ve i nsight i nto the spati al di stribution of the faci es successions. Profile A i s l ocated along a mai n northern di st ributary channel (Fi g. 4) and
records the progradati onal history of the del ta pl ai n and del ta -front pl atform. The faci es succession thi ckens seawards as accommodation space i ncreases. Mouth bar faci es are onl y present i n the western part of
the secti on. Duri ng the i ni ti al onset of the modern del ta progradati on, accommodation space was l ow and the number of di stri b utary channels was restricted. The mouth bars therefore were cl osel y spaced over a
rel ati vel y small and shallow area. As progradation continued the i ndi vidual mouth bars became l aterally separated by fl uvial andti dal domi nated del ta-front pl atform facies. These l ater faci es are therefore more
pronounced i n the eastern cores. The dates suggest that the fl uvi al -domi nated del ta-front pl atform facies ranges i n age between 3000 and 4300 yr BP ( Tabl e 1; Fi g. 9), but four dates are i nconsi stent. Core 8
(MD3;Tabl e 1) shows a rel ati vely ol d date from a wood sampl e that most l i kel y i s a reworked wood fragment. The date i n core 13 (MD12; Tabl e 1) appears to be too young when compared to dates of
nei ghbouri ng cores. The hi gh sand content i n thi s core, as wel l as i ts l ocation about 3 m from the di stri butary channel sugge sts that the deposi ts ori ginate from a l ateral channel bar rather that from del ta -front
pl atform deposi tion. The same pattern can be seen i n core 14, where two dates i ndi cate a modern age (MD 13 and MD 14; Tabl e 1).
Fig. 9.
Three cross profiles constructed using the core-data (Fig. 4 for location). The westeast Profile A shows the thickening of the facies in downstream direction. Profiles B and C are perpendicular to the main cha nnel showing a
decrease in sand content towards the tidal dominated area of the delta representing an increase in tidal domination. (See tex t for further explanation). The positions of calibrated AMS
14
C dates are shown using asterisks, the ages
are indicated below the core.
The two remai ni ng profi les of Fi g. 9 are si tuated perpendi cular to the mai n di stributary channel ( Fi g. 4). These cross profiles show the transi tion from a fl uvial -domi nated to a ti dal -dominated del ta-front pl atform
deposi ti onal envi ronment. Nei ther of the two profi l es shows mouth bar faci es, whi ch i s consistent wi th the di stri buti on of mo uth bar faci es shown i n profile A. The ti dal -domi nated del ta-front pl atform facies
shows a si gnificant i ncrease i n thi ckness towards the ti dal -dominated domai n. The absence of the ti dal -domi nated del ta-front pl atform facies i n core 2 (Profi le C) may be due to l ocal condi tions. Both profi les show
a decrease i n sand content towards the ti dal domi nated depositi onal envi ronment as i s expected based on the proposed sedi ment di spersal pathways shown i n Fi g. 8. The thi ckness of the del ta pl ai n deposi t has a
maxi mum of 4 m and does not i ncrease or decrease systemati cally. The AMS
14
C dates (Tabl e 1) i n profi l e B are consi stent wi th those of profi l e A except for the two dates retri eved from core 10 (MD7 and
MD8; Tabl e 1), whi ch are rel ati vel y young. Thi s core i s positioned along a wi de and l ow si nuous ti dal channel. In contrast to other ti dal channels i t has a narrow connection to the fl uvi al system. Channel
morphol ogy and the rel ati vel y young age i mpl y that thi s channel may i n fact be a former di stri butary channel, whi ch suggests that the encountered sandmud coupl ets are l ateral channel bar facies rather than a
fl uvi al-domi nated del ta-front pl atform faci es. Data from core 10 i ndi cate that thi s channel i s an ol dfl uvial di stributary, whi ch i s i n the process o f bei ng transformed i nto a ti dal channel . According to Al l en et al .
(1977) such transiti ons are not uncommon, especi al ly i nthe northern di stri butary system, si nce thi s di stributary system i s i n the p rocess of abandonment (Al l en and Chambers, 1998). Neverthel ess, recent satel lite
i mages show that hi gh concentrati ons of suspended sedi ment are di scharged from the northern di stri butarychannels during l ow ti de (Budhi man, 2004).
5.
14
C dates versus eustati c sea l evel hi story
No data exi st on l ate Hol ocene sea l evel change for eastern Kal i mantan or the Strai t of Makassar. Al though the new
14
C dates from thi s study have a l ow verti cal resolution, as the topographi c el evation could not
be accuratel y determi ned, the dates can sti l l be used to get a fi rst i mpression of l ate Hol ocene sea l evel . For each facies t ype except the l ateral channel bar facies, we know approxi matel y at what water depth
deposi ti on took pl ace. For exampl e, we assume that a di stri butary channel bi furcates i f the associated mouth bar grows hi gh e nough verti cal ly to reach the i nterti dal zone. Di stributary channel bi furcations take
pl ace every 10 to 20 km i n the del ta and occur i n both the northern and the southern di stri butary system. As soon as the mout h bar reaches the i nterti dal zone, whi ch has a di rect rel ation to mean sea l evel , i t
becomes vegetated and del ta pl ai n deposits wi ll cover the i nterti dal mouth bar. The depth i n the core at whi ch thi s transiti o noccurs has a di rect rel ation to mean sea l evel and can be dated. In the same manner,
del ta pl ai n anddel ta-front pl atform facies can be used as a proxy for sea l evel history. As l ateral channel bar deposition i s rel ated to channel d ynami cs rather than sea l evel , those dates could not be used.
Fi g. 10 shows a di agram of the
14
C dates versus depth. The carbon dates are compared to the eustati c sea l evel curve of Fai rbanks (1989), whi chgi ve a fi rst approximati onfor the l ate Hol ocene sea l evel hi story of
PETROLEUM Page 5
an otherwi se wel l studi ed del ta system. We excl uded the dates from cores that are i nterpreted to ori gi nate from l ateral chann el bar accreti on (MD1, MD2, MD7, MD8, MD12, MD13 and MD14; Tabl e 1) as wel l as
erroneous dates (MD3, MD10 and MD21; Tabl e 1). These l atter dates were ei ther based on reworked wood sampl es (MD3 and MD10; Tabl e 1) or were contami nated (MD21; Tabl e 1). Fi g. 10 suggests that sea
l evel has ri sen between 4500 cal yr BP and present. The data suggests a fai rl y good fi t wi th an exponenti al sea -level curve, where the fastest sea l evel ri se took pl ace between 4500 and 2500 cal yr BP. Thi s
suggests that sea l evel hi story i nEastern Kalimantan may not di ffer much from the eustati c sea l evel hi story ( Flemi ng et al ., 1998). The average subsi dence rate esti mated by Roberts and Sydow, 1996 and Roberts
and Sydow, 2003 between 0.2 and 0.5 mm yr
1
fi ts the data presented here, but i t shoul d be noted that the verti cal resoluti on of the depth measurements i s not suffi ci ent to refi ne the subsi dence esti mates.
Furthermore, there i s no i ndi cati onto suggest that there was a l ocal Hol ocene sea -l evel highstand as seen i n the nearby Sunda Shel f deposi ts ( Hanebuth and Stattegger, 2004).
Fig. 10.
Comparison between eustatic sea level ( Fairbanks, 1989) and AMS
14
C dates of various depositional environments in the Mahakam delta.
6. Deposi ti on rates
Two
14
C dates i n core 8 (MD4 and MD5; Tabl e 1), both representi ng fl uvi al domi nated del ta -front pl atform facies, were used to cal cul ate an average depositi on rate of 10.9 mm yr
1
. On average there are 2 to 5
sandy l ayers per verti cal centi metre of fl uvi al -dominated del ta-front pl atform facies whi ch are separated by mud l ayers. In general , sand beds are deposited duri ng l ow -tide condi tions (Fi g. 8A). Al though the
vari ability i n sandbed frequency and thi ckness i s l arge, i t i mpl i es that on average each year deposi ts rel ated to 2 to 5 i nd i vi dual ebb occurrences are preserved. The number of preserved sand beds per verti cal
uni t i s governed by the number of deposi ti on events and the number of erosi on events. The probabi l ity of sand deposition (bot h thi ckness and frequency) at a l ocation near the ri ver mouth i ncreases wi th the ti dal
ampl i tude. The hi ghest probabi lity for sand deposition at the fl uvi al domi nated del ta -front pl atform therefore occurs duri ng ebb stage of spri ng ti des, when ri ver fl ow vel oci ty i s hi ghest. It i s not cl ear i f san d i s
deposi ted at the del ta-front pl atform duri ng ebb stage wi th smal l er ti dal amplitudes. The rol e of erosi on cannot be esti mated but observati ons i ndic ate that preserved beds have sharp boundaries andshow no
gradi ng trends wi thi n the beds. Thi s i mpl ies that (some) erosi on takes place, removi ng ei ther the compl ete sand l ayer or just the top. Based on the above, we argue that the fl uvi al -dominated del ta-front pl atform
deposi ts consist predomi nantly of ebb stage deposi ts rel ated to spri ng ti des.
The deposi ti on rate for l ateral channel bar facies, calculated using
14
C dates i n cores 7 (MD 1 and 2; Tabl e 1) and 10 (MD 7 and MD 8; Tabl e 1), i s respecti vel y 9.7 and 10.4 mm yr
1
. Because the l ayer thi ckness for
the l ateral channel bar facies is comparable to the fl uvi al -dominated del ta-front pl atform facies, thi s i mplies that these sandmud coupl ets may al so represent neapspring ti de cycles, as suggested by Al l en et al .
(1977)and Gastaldo et al . (1995). However, we woul d expect that the rate of erosi on al ong a di stributarychannel woul d be hi gher and more vari able than at th e del ta-front pl atform.
7. Di scussi on
The data presented i n thi s paper gi ve a fi rst i mpression of the l ateral extent of the di fferent deposi ti onal envi ronments and thei r variability i n a mi xed fl uvial ti dal del ta system. The trends i n faci es succession are
consi stent through the constructed cross -secti ons as wel l as the new
14
C dates. There i s a seawards decrease i n sandcontent i n the uppermost 8 m, caused by the downward di p of the faci es successi ons. The
decrease i n sand content i n a cross-shore di recti on i s a result of decreasing fl ow competence. Wi thi n the fl uvi al domi nated del ta -front pl atform facies, we di d not noti ce a trend i n ei ther sand mud rati o or
thi ckness of the coupl ets, whi ch i s i ncontrast wi th the fi ndi ngs of Gastaldo et al . (1995) who noti ced an overal l i ncrease i n the sandmud rati o of the coupl ets i n a downstream di rection. Al l en and Chambers
(1998), however, reported a downstream fi ni ng trend i n some of the cores taken i n the Mahakam di stributaries. The ambi guous data ma y be expl ai ned by the fact that cross channel variability i n facies succession
i s i mportant (Fi g. 9; profi l e B and C) as wel l as making a di sti nction between the l ateral channel bar facies and fl uvial domi nated del ta -front pl atform facies. The l atter two faci es l ook i denti cal but are geneti cal ly
very di fferent and therefore shoul d not be compared i n terms of sand content. Our data suggest that the two faci es types cons i sti ng of sandmud coupl ets can onl y be di stinguished based on the radi ocarbon
dates.
At present, there are ei ght acti ve di stri butary channels enteri ng the Strai t of Makassar i n the south and four i n the north. For each new bi furcati on point, the di scharge i s di vided over the two newl y formed
di stri butary channels. A l arge number of coexi sti ng di stributary channels causes a rel ativel yl ow di scharge per di stri butary channel . Under i nvariable ti dal conditi ons, thi s i mplies that the bal ance between fl uvi al
and ti dal processes shifts slightly to ti dal -domination for each new bi furcation, so that depositi ongeneral ly becomes more cl ay ri ch. Assuming that both ri ver di scharge and ti dal character di d not change
si gni ficantl y during the past 5000 yr, one woul d expect a shi ft from a fl uvi al domi nated to a ti dal domi nated del ta system as progradation conti nued. Al so, duri ng the onset of progradati on, very l i ttl e
accommodati on space was present for ti dal i nfill and the di stributary mouth bars are therefore cl osel y spaced, possibly even i nterconnected. As progradation continued each bi furcation l ed to an i ncrease i n ti dal
accommodati on as the radi us of the del ta i ncreased.
Al l en et al . (1977) suggested that the ti dal processes have a dampeni ng effect on the fl uvi al dynami cs of the del ta, whi ch they use as an argumen t to expl ai n the l ack of typi cal fl uvial del taic features suchas l evees
and crevasses. As the i nfl uence of the ti dal processes i ncreased i n ti me whi l e the del ta prograded, the absence of any crevas ses or natural l evees suggests that the speci fi c non -flooding di scharge regi me of the
l ower Mahakam catchment area has l ed to thei r absence, rather than ti dal dampeni ng ( Bryant et al ., 1995). Thi s i s also suggested by the fact that there appears to be vi rtual l y no channel mi gration i n the del ta
pl ai n. Anal ysis of aerial photos from 50 years ago has shown no si gnificant changes i n channel l ocations ( Al len et al ., 1977).
The presence of l ateral channel bar facies does suggest ei ther channel mi gration or narrowi ng and i nfilling of the channel . T he fi rst has not been observed because both ti dal and di stributary channels appear to be
very stabl e. Neverthel ess, four cores (7, 10, 13 and 14), posi ti oned di rectl yal ong a channel, show l ateral channel bar facie s. A cl oser l ook at the core l ocati on map ( Fi g. 4) shows that these cores are posi ti oned
al ong a rel ati vel y narrow, meanderi ng channel whi ch appears to be ti dal i n ori gin, but whi ch sti ll has a narrow connection to the di stri butary channel s. As ti dal channel s are not capable of transporting sand ( Al len
et al ., 1977), the presence of sandmud coupl ets i ndi cates that some ti dal channels once were di stri butaries but l ater converted i nto ti dal channel s.
The vol ume of the l ate-Hol ocene Mahakam del ta i s esti mated to be 200 km
3
. Thi s number i s cal culated based on the sum of the aeri al extent of the del ta pl ai n (

1800 km
2
) and the del ta-front pl atform (

2000
km
2
), mul ti pl i ed by the average del ta thi ckness (50 m for the del ta pl ai n and 45 m for the del ta -front pl atform;Carati ni and Ti ssot, 1988 and Roberts and Sydow, 2003). Thi s vol ume of sedi ment has been deposi ted
duri ng the past 5000 yr. Based on the suspended sedi ment measurements i n the mai n southern di stri butary channel ( Fi g. 6) i t i s esti mated that the sedi ment l oad i s about 150 mg/l . Al l en and Chambers (1998)
esti mated the water di scharge to be i n the order of 3000 m
3
s
1
. Thi s woul d i mpl y that the vol ume of suspended sedi ment deposi ted over a 5000 yr peri od woul d be about 26 km
3
. Addi ng a rough esti mate for
bed l oad as 10% of the suspended sedi ment l oad ( Syvi tski et al ., 2003) bri ngs the total vol ume to 30 km
3
. The present-day suppl y of sedi ment i s cl early too l ow to have resul ted i n the formati on of the l ate -
Hol ocene Mahakam del ta compl ex. Al though the present -day water and sedi ment di scharges are onl y rough esti mates, i t woul d be unreal istic to i ncrease ei ther of them by a factor of si x to expl ai n the actual
vol ume of the present Mahakam del ta. The offset between del ta vol ume and present -day sedi ment l oad may i ndi cate that there have been di fferent hydraul i c conditions i n the past. Al though there have
conti nuously been l akes present upstream from the Mahakam del ta duri ng the course of the Hol ocene, thei r i nfl uence on the dow nstream Mahakam Ri ver may have vari ed i n ti me. In the absence of regul ati ng
l akes, ri ver fl oods, whi ch are presentl y absent i n the del ta, may have transported consi derable amounts of sedi ment to the de l ta. However, ri ver fl oods al so woul d have created l evees and crevasse splays, and no
evi dence has been found for thei r exi stence. Consi deri ng the l ocati on of the del ta at the Equator, i t i s unl i kel y that the cl i matic conditions have changed much duri ng the l ate Hol ocene, l eadi ng to changes i n
sedi ment di scharge. Addi ti onal
14
C dates may reveal changes i n progradation rate through ti me, whi ch may be rel ated to varyi ng water and sedi ment di scharge i n the past. The rel ati vel y di lute present-day ri ver
di scharge i s effecti ve at conveyi ng the l i mi ted sedi ments through the di stri butary channels wi th l i ttle l oss andtherefore th ere i s l i mi ted i nfi lling of the channel s. Thi s latter as wel l as the l ack of ri ver fl oods may
expl ai n the absence of channel mi grati on.
The typi cal del ta-front pl atform and slope of the Mahakam del ta can be i nterpreted as a subaqueous del ta whi chreveal s a compound -clinoform devel opment. Unfortunatel y, no data exi st on the progradati on
rate of the Mahakam del ta-front sl ope, but observati ons from other subaqueous del tas such as the Amazon del ta ( Ni ttrouer et al ., 1996), the GangesBrahmaputra del ta (Kuehl et al ., 1997), the Fl y ri ver (Harri s et
al ., 2004) and the Po del ta (Cattaneo et al ., 2003), suggest that the deposi ti onal mechanisms of the subaeri al del ta and the subaqueous del ta may be very di fferent. Subaqueous del tas are commonl y attri buted to
energeti c mari ne envi ronments ( Swenson et al ., 2005), whi ch are rel ated to energeti c waves and l arge ti dal ranges. Fi g. 11shows a comparison between the Mahakam del ta and other del tas wi th compound
cl i noform devel opment. Each subaqueous del ta i s di fferent i n size due to di fferences i n the ri ver di scharge and shel f bathyme try. Yet, the shape of the subaqueous del ta i s above al l formed by the l ocal
oceanographic regi me. Based on the morphol ogy of the subaqueous del ta, a di stinction can be made between the del ta -front sl ope and the prodel ta/shel f envi ronment ( Fi g. 11). Fi g. 11 shows that the shape of
the normal i zed subaqueous del tas may ei ther be smoothl y curved (Amazon, Yangtze and Red Ri ver del ta) or angul ar shaped (Mahak am, GangesBrahmaputra and Fly Ri ver del ta). The normal i zed profile of the
GangesBrahmaputra del ta front i s fairly simi lar to the normal i zed Mahakam del ta front profi le. Thi s fai rly qualitati ve analysis sho ws that the Mahakam del ta has a wel l devel oped subaqueous del ta. Based on
publ i shed ti de and wave data of these del tas ( Tabl e 2) i t cannot be determi ned whether waves or ti des are responsi bl e for the curved or the angul ar shape of the subaqueous del ta. Yet, the shal l ow depths of the
del ta-front pl atform of the Mahakam del ta may wel l be attri buted to the l ow wave energy i n the strai t of Makassar as tropical storm s are very rare. As a consequence, sedi ments deposi ted i n water depths
exceedi ng 5 m are not reworked.
Fig. 11.
A: Bathymetric profiles of subaqueous deltas (delta-front, prodelta, shelf) based on bathymetric maps from literature ( Harris et al., 2004, Hori et al., 2002, Hori et al., 2004, Michels et al., 1998, Nittrouer et al., 1995 and Tanabe et
al., 2003). B: bathymetric profiles of subaqueous deltas normalized between the shoreline and the base of the subaqueous delta (bottom set; black square in the top panel). Wave and tidal data of the subaqueous deltas are shown
in Table 2.
Table 2.
Wave and tidal data based on literature for the Red river delta ( Tanabe et al., 2003), Fly River delta (Harris et al., 2004), Yangtze delta (Hori et al., 2002), GangesBrahmaputra delta (Michels et al., 1998 and Mahboob-ul-Kabir et
al., 2002) and Amazon delta (Nittrouer et al., 1995 and Sternberg et al., 1996)
PETROLEUM Page 6
In terms of reservoir geometry, the sand-rich fluvial dominated delta-front platform, the mouth bar and the lateral channel bar facies can be interpreted as one continuous reservoir with a
maximum porosity in the mouth bar facies, which are flanked by lower porosity sand mud couplets. Because the mud couplets are not continuous ( Gastaldo et al., 1995), fluid flow is
possible in a vertical direction as well as in a horizontal direction. All sand -rich deposits are capped by impermeable delta plain deposits. Roberts and Sydow, 1996 and Roberts and
Sydow, 2003 showed the presence of continuous progradational (highstand to lowstand) pre -Holocene sediment wedges, which are in many respects similar to the present delta ( Allen
and Chambers, 1998). Incised channels that form during the falling sea level ( Roberts and Sydow, 2003) dissect the continuous character of the sand-rich deposits. But if the incised
valleys are sand-based, as is the case with the present channels ( Allen et al., 1977, Allen and Mercier, 1994, Gastaldo et al., 1995 and Gastaldo and Huc, 1992), they may enhance the
fluid flow properties of the reservoir.
8. Conclusions
Detailed analysis of new borehole data as well as new
14
C dates reveal the late-Holocene evolution of the Mahakam delta. Due to the predominance of semi-diurnal tides and the
associated flow reversals, depositional patterns are highly variable which has resulted in the formation of characteristic sandmud couplets. The distribution of the sandmud
couplets found in this study differs from previously proposed conceptual models. Based on facies analysis, a qualitative depositional model is proposed that explains the sand and
clay distribution across the Mahakam delta.

Calculated averaged deposition rates of the sandmud couples indicate that deposition of the tidal-dominated delta-front platform facies is driven by the springneap tide cycles
instead of the daily tidal cycle.

The new
14
C dates fit the eustatic sea level curve of the late Holocene, although the vertical resolution of the dates is in the order of 1 m. However, there is no indication for a late-
Holocene sea level highstand as has been reported in other studies across southeast Asia.

Suspended sediment measurements reveal a surprisingly low concentration of sediment in the water column. The present day sediment load of the Mahakam River is insufficient to
explain the sediment volume of the Mahakam subaerial and subaqueous delta. This may suggest that the hydraulic conditions in the past may have been different. In addition, the
dilute discharge may explain the absence of channel infilling resulting in extremely stable distributary and tidal channel positions.

Comparison with other tropical deltas reveals that the Mahakam delta has a distinct subaqueous delta morphology that is proportional to the subaqueous delta of the Ganges
Brahmaputra.

Acknowledgements
This project was financed by NWO-WOTRO within the framework of the pilot phase of the East Kalimantan Project (EKP). Co -funding came from EUROSTRATAFORM (contract no.
EVK3-CT-2002-00079) for J.E.A. Storms, from the DIOC-program of Delft University of Technology (DUT-DIOC 1.6) for R.M. Hoogendoorn and from the Society for the Advancement of
Research in the Tropics (Treub) for A.J.F. Hoitink.
14
C AMS datings were processed and analysed by Klaas van der Borg at the R.J. Van de Graaff laboratory of Utrecht University. M artin
Konert from the Vrije Universiteit, Amsterdam, assisted with the grain size analysis at the Sedimentology Laboratory. We than k Jan Sopaheluwakan of LIPI, Jakarta for his efforts to
initiate the East Kalimantan Coastal Studies project as well as the management of the Geological Research and Development Cen tre, Bandung for their support to make the
reconnaissance survey possible. Steve Goodbred and an anonymous reviewer and Editor Keith Crook are thanked for their valuabl e suggestions on an earlier version of the manuscript.
Finally we like to mention the crew of the RV Geomarin and our speedboat captain, who guided us through the wilderness of the Mahakam delta.
PETROLEUM Page 7