ABSTRACT

Maximum Flooding Surfaces (MFS) in the Early to mid-Cretaceous mixed carbonate-clastic shelfal systems of the Arabian Plate have been incorporated into a new sequence stratigraphic model that links Kuwait, Iran, Saudi Arabia, Qatar, and the United Arab Emirates, to Oman and Yemen. It is based on regional sequence stratigraphic concepts supported by biostratigraphic, sedimentological and mineralogical data. The model has amended the positions of some existing MFS. The diachronous interplay between large-scale, proximal clastic systems and outboard (down-systems-tract) carbonate platforms was emphasized by concentrating on the depositional history of prodelta areas during delta advance and retreat. The prodelta area of relatively deep water separating the depositional systems has been termed the ‘Migratory Carbonate Suppressed Belt’ (MCSB).

The model proposes that platform limestones expanded back over preceding prodelta areas during transgressions. The most extensive transgressions ultimately led to the demise of MCSBs. The maximum landward retreat of the shoreline coincided with the cessation of clastic input in the most up-systems-tract localities. Thus, the model has predicted that in many places MFS are located in the basal parts of clean carbonates even though these are not the deepest-water sediments. Examples are the Zubair-Shu’aiba (K70 MFS) and the upper Burgan-Maddud (K100) sections of the northern Gulf. Where carbonate platforms did not expand completely across the MCSBs, perhaps because of fault-control, the MCSBs survived and MFS are present within deeper-water, prodelta shales deposited below the most efficient window for carbonate production. Examples are the K40 to K60 MFS in intraformational shales of the Zubair, Biyadh, and Qishn formations of Kuwait, Saudi Arabia, and Yemen, and K100 in the Burgan-Wasia formations of Kuwait and Saudi Arabia. Even in these cases, the MFS are present within limestones deposited further down-systems-tract, notably in Iran (K60—Khalij Member, Gadvan Formation; K100—Dair Limestone Member, Burgan-Kazhdumi formations). Deeper-water dense limestones and shales with accompanying MFS were deposited along the northeastern passive margin of the Arabian Plate, or within intrashelf basins with some limited connection to the open ocean. From a regional perspective it can be seen that eustatic or tectonically forced MFS do not necessarily occur within the deepest-water facies.

A regional understanding is needed for a more precise sequence stratigraphic interpretation of the Early to mid-Cretaceous succession of the Arabian Plate. The identification of the stratigraphic architecture is of major economic importance at the reservoir scale, for instance in recognizing vertical permeability and transmissibility barriers, as well as at the regional-play fairway scale in the distribution of seals and their potential influence on migration pathways. Our interpretations are also relevant to the prediction of source-rock distributions and, in the longer term, may help identify stratigraphic trap potential related to the interplay between clastic and carbonate depositional systems. Although the model proposed relates to the Arabian Plate, general conclusions may be applicable to other regions where mixed carbonate-clastic systems are well developed, for example in many basins of Tertiary age in South East Asia.

INTRODUCTION

This paper presents a new sequence stratigraphic model of the Early to mid-Cretaceous clastic-carbonate sediments across the eastern and southern Arabian Plate from Kuwait to Oman (Figure 1). The chronostratigraphic framework for the model is provided by the identification and correlation of Maximum Flooding Surfaces (MFS) and Sequence Boundaries (SB) from the proximal coarse-grained clastic systems through the intervening prodelta areas to the outboard (down-system tract) carbonate platforms. The model explains why eustatic or tectonically forced MFS do not necessarily occur within the deepest-water facies within a vertical succession, and demonstrates that a regional stratigraphic understanding, especially in a dip direction or along-systems-tract direction, is essential for the accurate correlation of MFS. The paper also briefly examines the clues to relative sea-level change (and hence layering) within thin, high-frequency cycles as seen in outcrop or core, and which are important for reservoir modeling.

The Early to mid-Cretaceous (Hauterivian to Cenomanian) rocks of the Arabian Plate contain some of the most important clastic and carbonate hydrocarbon reservoirs in the world. Clastic reservoirs include the Burgan, Nahr Umr, and Wasia (Safaniya-Khafji members) formations of the northern Gulf, the Zubair Formation in Iraq and Kuwait, and the Qishn Formation of Yemen. Carbonate reservoirs include the Sarvak Formation of Khuzestan Province, Iran; the Shu’aiba, Upper Thamama and Mishrif reservoirs of northeastern Saudi Arabia and the United Arab Emirates; and the Natih reservoirs of northern Oman. These sediments also contain important source rocks, such as the black shales of the Kazhdumi Formation of Iran.

Much of the sedimentary succession includes evidence of the interplay between proximal (western) siliciclastic rocks and distal carbonate rocks (Figure 1). This interplay is particularly evident in a broad belt extending from southern Iraq to Yemen (Murris, 1980; Moshrif and Kelling, 1984; Sharland et al., 2001). At this time, the proximal areas of the Arabian Plate (western Iraq, Kuwait, western Saudi Arabia, and Yemen) were dominated by the major clastic systems referred to above. In contrast, carbonates were deposited in the relatively distal Plate areas of the eastern United Arab Emirates and Oman (Hassan et al., 1975; Harris et al., 1984; Boichard et al., 1995; Pratt and Smewing, 1993; Simmons and Hart, 1987; Simmons, 1994; van Buchem et al., 2002a, this issue), and much of Iran (Gollestaneh, 1974; Motiei, 1993). Northern Iraq was also dominated by carbonate deposition (Al-Shdidi et al., 1995). Intermediate areas such as Qatar have a more mixed sedimentary succession (Sugden and Standring, 1975). The wide variety of coeval lithologies raises significant issues about the sequence stratigraphic interpretations across the region.

Sharland et al. (2001) proposed a sequence stratigraphic scheme for the entire Phanerozoic of the Arabian Plate. They argued that Maximum Flooding Surfaces (MFS) are easier to identify, date and correlate on a regional basis than are sequence boundaries, and identified 63 MFS that can be recognized over large parts of the Arabian Plate (Figure 2). This is certainly the case when only wireline log or limited lithological and biostratigraphic data are available, as in most of the literature examples. Even where seismic is available, the near-horizontal and parallel nature of most of the data makes it extremely difficult to identify all of the major unconformities in the succession. The published biostratigraphic information over large parts of the Arabian Plate is insufficiently detailed to identify some of these unconformities. Only in places where outcrop or extensive core data are available, are the positions of unconformities or sequence boundaries (in the sense of Vail et al., 1991) readily apparent from the vertical lithological succession and evidence of subaerial processes, such as unconformities, soil developments, and vadose diagenetic fabrics. Otherwise, a regional understanding can help in the identification of sequence boundaries by recognizing patterns of facies evolution and geometric relationships, notably by the detection of progradational, aggradational and retrogradational depositional architecture.

This paper examines in detail some Early to mid-Cretaceous MFS from Sharland et al.’s Arabian Plate Tectonostratigraphic Megasequence AP8 (Figure 3; Table 1). It focuses on the Valanginian-Aptian (K40 to K80) and Albian (K90 to K110) Genetic Stratigraphic Sequences (GSS). These include the major Zubair-Biyadh and Burgan-Nahr Umr-Wasia reservoirs of the northern Gulf and Saudi Arabia, the Thamama, Mishrif and Natih reservoirs of northeast Arabia, and the Qishn reservoirs of Yemen. The analysis is also relevant to the understanding of other shelfal depositional systems in the Middle East and elsewhere, where there is a marked interplay between clastic and carbonate components. The examples discussed are all from the central and southern parts of the Arabian Plate (northern Gulf and southwards), as less detail has been published on the northern Plate succession in Syria and northern Iraq.

Methods and Materials

This study was based largely on published data, although the authors’ combined experience includes the biostratigraphical, sedimentological and sequence stratigraphic interpretation of several hundred wells and outcrop sections throughout the area of interest. Several key wells and outcrops have been identified that have good published wireline log data and generally reliable biostratigraphic information. Correlation panels and chronostratigraphic charts have been constructed using the key wells, and conditioned using additional published data and maps. It is acknowledged that the level of published information currently available for large parts of Yemen and Iran is less detailed than that of other parts of the Arabian Plate. Current research by scientists of the Institut Français du Pétrole (IFP), in collaboration with the National Iranian Oil Company (NIOC), should significantly increase our knowledge of the Lower to mid-Cretaceous succession in Khuzestan and Lurestan (Gaumet et al. 2002; van Buchem et al., 2002b), although full results are not yet available.

Note that terms ‘up-systems-tract’ and ‘down-systems tract’ are used throughout this paper. They are applied in the same sense as ‘upstream’ and ‘downstream’ are used in the description of river systems.

MAXIMUM FLOODING SURFACES IN THE VALANGINIAN TO APTIAN DEPOSITIONAL PACKAGE (K40-K80)

The interval described is a second-order Depositional Sequence (DS). It is bounded by the late Valanginian unconformity (LVU) identified in offshore Kuwait (Al-Fares et al., 1998), onshore Saudi Arabia (Le Nindre and Vaslet, 2002), Bahrain (Bassant et al., 2000), Yemen (Ellis et al., 1996; Holden and Kerr, 1997; Brannan et al., 1999), and elsewhere (Sharland et al., 2001), and the intra-Aptian unconformity that corresponds to the top of the Shu’aiba Formation in most places (Alsharhan and Nairn, 1997), but lies within the defined Burgan Formation in the northern Gulf (Al-Fares et al., 1998; Sharland et al., 2001). It encompasses the Maximum Flooding Surfaces K40 to K80 MFS defined by Sharland et al. This interval shows considerable lithological and depositional variation across the region (Figure 3).

For the purposes of the following description, the region has been divided into northern, central and southeastern areas (Figure 1). Each area is described from west to east to follow the depositional systems down-systems-tract. A brief introduction to the succession in Yemen is also provided because of similarities to the northern Gulf and because it may lie up-systems-tract from the outboard carbonate platforms of northern Oman.

Northern Gulf Area

The best stratigraphic and wireline-log information published for the northern Gulf is available from onshore and offshore Kuwait (Al-Fares et al., 1998; Nemcsok et al., 1998; Sharland et al., 2001) and, to a lesser extent, southern Iraq (Al-Rawi 1981; Ali and Nasser, 1989; Khaiwka, 1990; Ali and Aziz, 1993; Al-Ameri and Batten, 1997). Less-detailed information is available for adjacent areas of Saudi Arabia (Moshrif and Kelling, 1984). Clastic deposition is dominant. The best reservoir sandstones represent west- to east-trending estuarine channels cut into north- to south-oriented palaeoshorefaces (Nemcsok et al., 1998). There is a clear down-systems-tract decrease in sediment grain size illustrated by a comparison of the onshore section of north Kuwait (Nemcsok et al., 1998) with the siltstone-dominated succession to the east in well-F of offshore Kuwait (Al-Fares et al., 1998). Similarly, the coarse-grained clastics of the Zubair Formation and Biyadh Sandstone are replaced down-systems-tract by the shale-dominated Gadvan Formation of Khuzestan and offshore Iran (Motiei, 1993; Bordenave and Huc, 1995; Sharland et al., 2001). Even further down-systems-tract, the shale-dominated Gadvan is replaced by undifferentiated carbonates of the Khami Group (Gollestaneh, 1974; Motiei, 1993; Bordenave and Huc, 1995; Sharland et al., 2001).

Sharland et al. (2001) picked the K40, K50 and K60 MFS within the Lower Shale, Middle Shale and Upper Shale respectively of the Zubair Formation of Kuwait (Figure 4). Each corresponds to a major sealing unit in Raudhatain field (Nemcsok et al., 1998). The same sealing units are also identifiable in neighboring areas of southern Iraq (Figure 5; Ali and Nasser, 1989). The K70 MFS was identified in the basal part of the Shu’aiba Formation that transgressed almost to the Arabian Shield following the landward retreat of the Zubair/Biyadh delta system (Sharland et al., 2001).

Within this second-order DS, the distribution of sandstone across Kuwait (Nemcsok et al., 1998; Al-Fares et al., 1998) and Iraq (Ali and Nasser, 1989) suggests that a further, relatively low-order/low-frequency Maximum Regression Surface (MRS) occurs in the Upper Zubair Sand within the K60 GSS. Significantly, the Upper Sand is the best producing reservoir in north Kuwait (Nemcsok et al., 1998). The sandstones within the Upper Shale (Figures 4 and 5), which correspond to the K70 MRS, mark a less extensive regression than the Upper Zubair Sand, as previously noted by Sharland et al. (2001). This is interpreted as clear evidence of punctuated backstepping of the clastic systems within a low-order/low-frequency Transgressive Systems Tract (TST).

In Iran, the widespread Khalij Member limestone of the Gadvan Formation, also known as the Dictyoconnus arabicus limestone, (Mina et al., 1967; Motiei, 1993) contains the K60 MFS (Sharland et al., 2001). The locations of the K40 and K50 MFS are difficult to identify precisely in the Gadvan Formation on the basis of published data. It is even more difficult to identify these MFS further down-systems-tract in the thick undifferentiated limestones of the Khami Group (Gollestaneh, 1974; Motiei, 1993). The K70 MFS is located near the base of the Dariyan Formation (Sharland et al., 2001). The K80 MFS is extremely difficult to identify in the northern Gulf using published data.

In summary, the depositional system between the Lower Valanginian unconformity and the intra-Aptian unconformity in the northern Gulf area comprises a proximal (up-systems-tract), alluvial to deltaic, coarse-grained clastic-dominated unit; a mid-systems tract, shale-dominated prodelta area; and an outboard (down-systems-tract) carbonate platform (Figure 3). Truly distal parts of the systems tract, namely slope and basinal settings, have not been identified. Clear evidence exists of the landward retreat of the coarsegrained proximal facies and the associated landward migration of the prodelta shales and the outboard carbonate platforms during major transgressions. Extensive basinward migration of facies belts is linked to major regressions together with considerable diachroneity (Figure 3).

Central Gulf Area

Published information is limited on the succession in central Saudi Arabia (Moshrif and Kelling, 1984; Vaslet et al., 1991; Robelin et al., 1994) and Qatar (Sugden and Standring, 1975; Schlumberger, 1981; Rubbens et al., 1983; Alsharhan and Nairn, 1994). The Biyadh Sandstone of central Saudi Arabian outcrops is dominated by coarse-grained clastics and is considered to be of fluvial origin (Figure 6, after Moshrif and Kelling, 1984). Paleocurrent data for the Biyadh in this area shows a strong easterly movement (Figure 7; Moshrif and Kelling, 1984). We suggest that by analogy to the time-equivalent Zubair Formation of Kuwait (Nemcsok et al., 1998), the Biyadh to the east in the subsurface of Saudi Arabia probably contains several stacked retrogradational estuarine-dominated packages. The Biyadh Sandstone is more argillaceous eastward (Alsharhan and Nairn, 1997), and becomes more difficult to distinguish from the underlying Buwaib Formation. Indeed, Sharland et al. (2001) identified the K40 MFS in the Buwaib Formation on the basis of evidence in Shebel and Alsharhan (1994), thereby suggesting that the Buwaib-Biyadh boundary is diachronous. Alsharhan and Nairn (1997) described two “relatively thin” limestone members containing orbitolinids within the shales of the Biyadh Sandstone in the Damman and Qatif fields in eastern central Saudi Arabia.

These limestone members are strong candidates for maximum flooding surfaces but, unfortunately, no faunal data have been presented. The anecdotal evidence of orbitolinids suggests that the limestone members are more likely to represent the late Barremian K60 MFS alone (or even K60 and K70) rather than both K60 and the early Barremian K50 MFS. Biostratigraphic analysis would quickly resolve this issue. It may be significant that an ammonite dated as possibly late Aptian has been collected from the upper part of the Biyadh Sandstone of central Saudi Arabia (Vaslet et al. 1991; Le Nindre and Vaslet, 2002). Whereas our current interpretation is that this limestone most likely represents the K70 MFS in a thin remnant of the Shu’aiba Formation in a proximal location (see Figure 6), additional detailed biostratigraphic analysis is required to fully resolve this issue. Further work is planned, although more recent information suggests that the ammonite may be much longer ranging (Denis Vaslet, personal communication, February 2002 and April 2002).

The mixed, though more carbonate-dominated succession of Qatar lies to the east further down-systems-tract. There are lithostratigraphic inconsistencies between the various publications, and precise correlation with other areas is not obvious; for example, compare the lithostratigraphy presented by Schlumberger (1981); Rubbens et al. (1983) and Alsharhan and Nairn (1994). Schlumberger showed that the Shu’aiba Formation directly overlies the Ratawi Formation in Dukhan and offshore western Qatar, whereas this usage of the Ratawi does not match the application of the term in southern Iraq and Kuwait where it underlies the late Valanginian unconformity (Al-Fares et al., 1998). More likely, the term ‘Ratawi’ in Qatar is being used to denote the shale-dominated, down-systems-tract that is equivalent to the Biyadh Sandstone.

Over the Qatar Arch itself, and to the east of the Qatar Peninsula, Schlumberger (1981) showed the equivalent of the Ratawi to be the Kharaib but, following Sugden and Standring (1975), do not identify the Lekhwair Formation as a separate unit. Alsharhan and Nairn (1994, 1997) have followed this convention. The Kharaib and Shu’aiba are separated by the Hawar Shale that, although it has been given formation status in Qatar (Sugden and Standring, 1975; Rubbens et al, 1983), is referred to as the uppermost member of the Kharaib Formation in onshore United Arab Emirates and Oman (Hassan et al., 1975; Hughes-Clarke, 1988). Our interpretation is that it is inappropriate to use the term ‘Hawar Shale’ east of Qatar because it is lithologically distinct from the dense limestones forming the Upper Kharaib of eastern Abu Dhabi and Oman. Rubbens et al. (1983) also distinguished the Lekhwair Formation underlying the Kharaib in northern offshore Qatar. Their stratigraphy is a closer match with that of offshore and onshore Abu Dhabi (Hassan et al., 1975; Boichard et al., 1995; van Buchem et al., 2002a). In particular, they identified the Zakum Member of the Lekhwair that, on the basis of regional seismic data (Haan et al., 1990; Pratt and Smewing, 1993; Aziz and El-Sattar, 1997), clearly overlies the late Valanginian unconformity.

The two orthogonal log correlation panels presented by Rubbens et al. (1983) provided the most detailed well data for the Early to mid-Cretaceous of Qatar (Figure 8). The best candidate for the late Valanginian unconformity (LVU) is at the base of the Lekhwair Formation, below the bottom of the logged section. Schlumberger (1981) clearly identified a widespread unconformity at this level throughout Qatar. Hassan et al. (1975) also recognized that the base of the Lekhwair in the United Arab Emirates is an unconformity, though they considered it to be a minor one.

The overlying Lekhwair, Kharaib and Shu’aiba formations of Qatar consist of alternating units of clean limestone, argillaceous limestones, and shales. There are four argillaceous units, the Lower Zakum Member and the Fahia Shale in the Lekhwair, the Middle Marly Member of the Kharaib, and the Hawar Shale. These separate clean limestones, the Upper Zakum Member of the Lekhwair, and the ‘C’ and ‘B’ Limestone members of the Kharaib. Clean limestones of the Lower Shu’aiba and more radioactive limestones of the Upper Shu’aiba, cap the section below the regional intra-Aptian unconformity (Figure 8). Given that the section crosses one of the most important structural arches on the entire Plate, there are demonstrably variable levels of erosion and/or non-deposition across the Qatar Arch (Figure 8). In the absence in the literature of deeper-water facies in the Upper Shu’aiba of Qatar (Sugden and Standring, 1975), we suspect that its higher gamma-ray character is more likely to be a highstand, landward-attached, argillaceous limestone or a diagenetic signature. This is in spite of its superficial similarity to the gamma-ray signature of the deeper-water Upper Shu’aiba facies of the Bab Basin (compare with illustrations in van Buchem et al., 2002a). The top of the Fahia Shale (i.e. top Lekhwair) is a recognized unconformity in offshore Qatar (Rubbens et al., 1983). Although this is a candidate for the LVU, it is here interpreted as equivalent to the K50 MRS. This is because additional unconformities are more likely to be identified over long-lived, regional highs such as the Qatar Arch, particularly in view of the density of exploration and development wells in such areas (Figure 8).

Although the gamma-ray profiles of Rubbens et al. (1983) have no unit scale, we interpret the higher gamma-ray signatures present in the Hawar Shale, Middle Marly Member of the Kharaib, and the Fahia Shale Member of the Lekhwair to be related to siliciclastic input. We do not interpret them as being due to either high organic content or a diagenetic signature associated with chemical compaction and stylolite formation. Our reasons are as follows:

  • The lithological descriptions are of shale and marl thereby indicating a siliciclastic provenance (from the west).

  • The Hawar Shale (as defined) has a distinctive gamma-ray character in the 11 wells investigated. It is characterized by a sharp basal contact with the highest gamma-ray character immediately overlying the top of the clean ‘B’ Limestone of the Kharaib Formation. By analogy with onshore United Arab Emirates and Oman, this surface is likely to be a sequence boundary (van Buchem et al., 2002a). The Hawar Shale in offshore Qatar displays a gradational cleaning-upward gamma-ray character into the overlying clean limestones that constitute the Lower Member of the Shu’aiba Formation. To a lesser extent, the other argillaceous units show comparable gamma-ray trends.

  • Qatar is in close proximity to areas of northeastern Saudi Arabia containing time-equivalent (Sharland et al. 2001), sandstone-prone sections in the Biyadh Sandstone (Alsharhan and Nairn, 1997).

  • Moshrif and Kelling (1984) identified strongly unimodal, easterly directed paleocurrents in the Biyadh Sandstone of eastern Saudi Arabia (Figure 7), demonstrating that Qatar lies down-systems-tract.

  • A further significant feature of the Hawar Shale (as defined) of offshore Qatar is that there is a marked decrease in shale thickness toward the southeast (Figure 8). This thinning is interpreted as the product of a combination of onlap onto the Qatar Arch and the transgressive reworking of shoreline-attached clastics originally derived from the west.

  • Vahrenkamp (1996) illustrated a clear regional trend from the marl of the Hawar Shale of Dukhan in the west, through dark argillaceous limestone and argillaceous limestone, to the light-colored argillaceous limestone of the Upper Kharaib of Oman in the east.

Thus, we interpret the shales and marls of Qatar as mid-systems-tracts equivalents to the up-systems-tract continental to marginal-marine sandstones and shales of the Biyadh Sandstone (Alsharhan and Nairn, 1997), and coeval with down-systems-tract platform limestones in Abu Dhabi and northern Oman. It is probable that the Hawar Shale is laterally equivalent to the upper part of the Biyadh present in adjacent parts of Saudi Arabia (Alsharhan and Nairn, 1986, 1997).

Having taken into account the observations and regional sequence stratigraphic considerations, we propose that the Hawar Shale of offshore Qatar is a transgressive shale that forms a TST that is bounded by the underlying sequence boundary at the top of the Kharaib Formation, and by an MFS close to the base of the overlying clean limestones of the Lower Shu’aiba Member. This MFS equates with K70 of Sharland et al. (2001). Similarly, we place the K60 and K50 MFS near to the bases of the clean ‘B’ and ‘C’ Limestones of the Kharaib Formation respectively, and the K40 MFS in the clean Upper Zakum Member of the Lekhwair (Figure 8). The position of clean limestones in the basal part of the Kharaib (‘C’ Limestone) immediately overlying the top Lekhwair unconformity identified by Rubbens et al. (1983), is strong evidence supporting our interpretation that clean limestones can host MFS within some mixed carbonate-shale successions.

Understanding the distribution of these marls and shales is important as they act as major intraformational seals that, at least in part, account for the complex distribution of hydrocarbons in this area of offshore Qatar (Rubbens et al., 1983). One final point is that it is difficult to identify the K80 MFS in the Shu’aiba of this area. Our interpretation, as based on regional data from the entire Plate (see Sharland et al., 2001), is that K80 is a higher-frequency surface than K70, though van Buchem et al. (2002a) proposed the opposite. This issue is discussed later in the paper.

Thus, the central Gulf area shows a similar depositional system to that of the northern Gulf. The Biyadh Sandstone represents a proximal (up-systems-tract), alluvial to deltaic, coarse-grained clastic-dominated unit comparable to the Burgan/Nahr Umr of the northern Gulf (Figures 9 and 10). The Qatar area is interpreted as having been transitional between a mid-systems-tract, shale-dominated prodelta area and an outboard (down-systems-tract) carbonate platform. The Qatar Arch probably played an important role in controlling facies distributions and thicknesses (Figure 9), as well as the formation of multiple unconformities (Figure 8). Potentially, these unconformities (sequence boundaries) may correlate with exposure surfaces described from the Lekhwair (Le Bec et al., 2001) and at the top of the Lower and Upper Kharaib in Oman (van Buchem et al., 2002a). There is further evidence of the landward migration of prodelta shales and outboard carbonate platforms during major transgressions (Figures 9 and 10).

Southeastern Gulf Area

For the purposes of this study, the southeastern Gulf is taken to encompass the United Arab Emirates and Oman. As noted above, it is distinct from the northern Gulf area in that it is dominated by carbonate deposition (Hassan et al., 1975; Harris et al., 1984; Alsharhan and Nairn, 1986; Simmons and Hart, 1987; Pratt and Smewing, 1993; Simmons, 1994; Boichard et al., 1995; van Buchem et al., 2002a). The Lekhwair, Kharaib and Shu’aiba formations occur between the late Valanginian unconformity and the intra-Aptian unconformity, and the stratigraphy is thus comparable with that of eastern offshore Qatar (Hassan et al. 1975). The formations can be correlated over a very wide area of the United Arab Emirates and Oman (Simmons and Hart, 1987; Haan et al., 1990; Pratt and Smewing, 1993; Simmons, 1994, Boichard et al., 1995; Vahrenkamp, 1996; Alsharhan and Nairn, 1997; van Buchem et al., 2002a). Strong cyclicity is a feature of the Lekhwair, Kharaib and Shu’aiba formations throughout the area.

The marked seismic discontinuity (toplap?) between the Habshan and Lekhwair formations in eastern Abu Dhabi and Oman (Haan et al., 1990; Pratt and Smewing, 1993; Aziz and El-Sattar, 1997) is further evidence that the late Valanginian unconformity and its correlative conformity can be identified over much of the Arabian Plate. The base of Lekhwair is a major marine flooding surface that can be identified over a very wide area. However, this interpretation should not be extended to the most distal (down-systems-tract) shelf-margin areas where the Lekhwair has a diachronous boundary with the underlying Habshan Formation (Le Bec et al. 2002).

The succession present in offshore western Abu Dhabi (Figure 9) is very similar to that in Qatar, but it is significantly thicker, which suggests a much higher subsidence rate. Hassan et al. (1975) provided details on the stratigraphy and sedimentation of the Lekhwair, Kharaib and Shu’aiba formations. The Zakum Member of the Lekhwair consists of porous limestones separated by dense argillaceous intervals. A notable feature of the Zakum is the basal shale/shaly limestone that is only present in western Abu Dhabi. It is thickest in the west and thins out eastwards. Gamma-ray and neutron logs presented by Hassan et al. (1975) suggested that the argillaceous sections separating reservoir units IV-A, IV-B and IV-C (equivalent to Zones F, G and H of Figure 9) have relatively abrupt bases and clean upward into the overlying porous shallow-water limestones. This pattern is strikingly similar to that observed throughout the section in offshore Qatar (Figure 8). By comparison with the latter, more proximal section, we interpret the Zakum Member to represent a retrogradational package and that the porous zones are a consequence of the punctuated retreat of shoreline-attached, more argillaceous units. Thus, we identify an MFS within the porous limestones of Zone F (Figure 9) and its precise correlation is discussed below. The dense limestone that caps Zone F is an argillaceous lime mudstone that grades up into an argillaceous pellet packstone. We interpret the base of this argillaceous limestone to be a regressive event and that the upward-cleaning gamma-ray log is a transgressive signature. Note that Hassan et al. (1975) related the presence of silt-grade quartz grains in the middle and lower part of the Lekhwair to the Biyadh Sandstone of central Arabia.

Applying similar reasoning as for Qatar suggests that the K70 and K60 MFS should be located in the base of clean limestones in the Shu’aiba Zone A and middle Kharaib Zone B, respectively (Figure 9). Note that these are refinements to the picks presented by Sharland et al. (2001). The position of K50 is less certain though, for similar reasons, our favored pick is in the clean limestone at the base of Kharaib Zone C. It is more difficult to pick K40 with confidence and we present two alternatives on Figure 9.

  1. The higher K40 pick in Zone D of the Lekhwair would imply considerable relief on the late Valanginian unconformity and marked onlap onto the Qatar Arch. In this model, the thicker Lekhwair succession below the K40 MFS represents a well-developed lowstand wedge and associated transgressive systems tract. The marine-flooding surface on top of the Habshan may be close to the K40 MFS in some locations but considerably below it in others. In this model, the MFS within Zone F of the Zakum Member would represent a higher-frequency surface.

  2. The alternative deeper pick in Zone F of the Zakum Member would be more consistent with the interpretation of Sharland et al. (2001). It would indicate much-increased long-term subsidence in western Abu Dhabi compared to the Qatar Arch. In this model, the K40 MFS would always be close to the marine-flooding surface on top of the Habshan. This does not preclude the presence of a higher frequency MFS in Zone D of the Lekhwair Formation.

Regional seismic data allied to detailed biostratigraphic analysis should resolve which of the two models is correct.

A much thinner succession is present in eastern offshore Abu Dhabi (Figure 9). Detailed microfacies data for a Bab Basin well presented by Boichard et al. (1995) provided strong evidence for locating K70 MFS in an open marine shale at the top of the Upper Dense Zone and K60 MFS in tight, orbitolinid-bearing wackestones within the Lower Dense Zone of the Kharaib Formation. The K50 MFS was picked in outer ramp, bioclastic mudstones to wackestones straddling the Kharaib-Lekhwair boundary (Sharland et al., 2001). The location of the K70 and K60 sequence boundaries is clear from the evidence of subaerial exposure described by Boichard et al. (1995). They also identified a maximum flooding surface near the base of the Bab Member (within the Tar Unit of other workers) that corresponds to K80 of Sharland et al. (2001).

This area has particular significance as the well penetration described by Boichard et al. (1995) was selected as the reference section for the Sharland et al. (2001) K70 MFS (Figure 9) even though Boichard et al. (1995) had correctly used the reservoir unit term ‘Kharaib Upper Dense’ (‘KUD’), and not the Hawar Shale for this interval. In contrast to the strict definition of the Hawar Shale in its type locality in Qatar (see above), in the ‘KUD’ the term has been misapplied. It has been used for a dense limestone unit at the top of the Kharaib Formation in the United Arab Emirates and Oman by several authors and in unpublished oil company reports. Sharland et al. now accept that the term ‘Hawar Shale’ was incorrectly used in this regard, and prefer to revert to the ‘KUD’ of Boichard et al. (1995).

The paleoenvironmental data for the well presented by Boichard et al. (1995) identified the ‘KUD’ Member as an upward-deepening section, from inner ramp to open marine. These observations and interpretations are substantiated by the unpublished observations of one of the authors (MDS). The K70 MFS was picked in the thin open-marine shale that caps this upward-deepening section and is coincident with an MFS of Boichard et al. This upward-deepening section within limestones has the reverse lithological succession to the Hawar Shale (as defined in offshore Qatar) where shale grades up into limestone (Rubbens et al., 1983). Both sections are interpreted as transgressive systems tracts capped by the K70 MFS of Sharland et al. (2001). Hence, K70 is an excellent example of an MFS located within the basal part of clean limestones at many locations, but in basinal shales or dense limestones in other more distal settings.

Of the other picks, the K60 MFS is noteworthy because it is in dense limestones in offshore eastern Abu Dhabi (Figure 9) but in porous limestones in Qatar and western offshore Abu Dhabi (Figures 8 and 9). This change reflects the regional systems tracts and the location of the latter two locations closer to the major siliciclastic input from the Biyadh delta system. Regionally, the MFS are more likely to be in dense but clean limestones the farther they are from siliciclastic influence. In locations closer to the siliciclastic systems, clay within the argillaceous limestones is more likely to have promoted stylolitization and associated cementation. In the platform areas, it is the quieter-water, finer-grained carbonates that, in most cases, should have been more prone to cementation than the coarser-grained bioclastic packstones and grainstones that represent higher-energy platform-top facies.

Sharland et al. (2001) provided a sequence stratigraphic interpretation for the section in Lekhwair-7 near the Oman-United Arab Emirates border originally described by Hughes Clarke (1988). This well is located near to the southeastern margin of the Bab Basin (see van Buchem et al., 2002a; this issue), and the succession is interpreted in a similar way to that in the offshore Abu Dhabi well described above (Boichard et al., 1995). It is probable that this area was beyond the influence of siliciclastic systems, and therefore that MFS (with the exception of K80) are more likely to be associated with finegrained dense limestones. Hence, the K50, K60, and K70 MFS are all placed in dense limestones (Figure 9). Note that the K50 MFS has been moved up substantially from Sharland et al.’s pick to take more account of the stratigraphic ages provided by Hughes Clarke and the detailed information presented by van Buchem et al. (2002a). The K40 MFS is located close to the base of the Lekhwair Formation within the equivalent of the Zakum Member of Abu Dhabi (Figure 9). This interpretation is reasonable given the seismic evidence from the eastern United Arab Emirates and Oman Haan et al., 1990; Pratt and Smewing, 1993; Aziz and El-Sattar, 1997). The K80 MFS was picked in argillaceous carbonates within the basinal equivalent of the upper part of the Lower Shu’aiba (Sharland et al., 2001).

It should be noted that van Buchem et al. (2002a) interpreted a much thicker TST in this part of the section, and hence their K80 MFS pick is located significantly higher in the section. Our view is that they have not taken full account of the conflict between regional highstand progradation and increased subsidence in the developing Bab Basin, and also that they mistook the development of the Bab Basin as indicating a plate-wide deepening. Our interpretation of van Buchem et al.’s Figure 8 is that the Oman side shows clinoforms downlapping onto the regional, and hence lower-order K70 MFS, whereas the higher-frequency/higher-order K80 MFS follows a clinoform within the prograding Lower Shu’aiba.

Similar clinoform geometries are present on the opposite side of the Bab Basin (van Buchem et al. 2002a). Work on the Shaybah field in this area by Saudi Aramco identified an MFS with an association of orbitolinids and pelagic foraminifera near the base of the Shu’aiba Formation and overlying the Biyadh Member (Figure 11; Hughes, 2000; Heil, 2001). Its early Aptian age (G.W. Hughes, personal communication, February 2002) clearly identified this as the K70 MFS. We recognized the K80 MFS by a marked landward step of depositional facies within the Middle Shu’aiba of Shaybah (Figure 11), and that the limited highstand progradation following this MFS implied structural pinning of the shelf margin along the edge of the Bab Basin. This ties in with the interpretation by van Buchem et al. (2002a) of a maximum flooding surface within the Shu’aiba Formation although Hughes (personal communication, February 2002) did not identify this as the K80 MFS. According to van Buchem et al., the K80 MFS corresponds to changes in both depositional geometries and facies in the Oman platform and the Shaybah area. They interpreted the depositional geometries as changing from backstepping and aggradational to progradational, whereas the facies changed from microbial (Lithocodium) boundstones to rudist and miliolid-dominated grain-supported limestones.

Sharland et al. (2001) presented an interpretation of the succession in Wadi Mu’aydin in northern Oman. They picked the K70 MFS within dense limestone at the top of the Kharaib Formation, the K60 MFS in a lower dense limestone within the Kharaib, and both the K50 and K40 MFS within the Lekhwair, although the K50 pick was tentative. The K50 pick, in particular, now looks too low, and probably should be moved up to the Lekhwair-Kharaib boundary (Figure 9). This would be in closer agreement with the published age information for this section (Simmons and Hart, 1987; Simmons, 1994), though van Buchem and his co-workers would place the K50 MFS still higher in the section (Frans van Buchem, personal communication, January 2002). It may be that the pick by Simmons (1990) for the Lekhwair-Kharaib boundary in the Wadi Mu’aydin section is deeper than that of van Buchem et al. (2002a). As in Lekhwair-7, Sharland et al.’s K40 MFS pick is placed near to the base of the Lekhwair Formation.

Detailed outcrop logs of the Shu’aiba and Kharaib formations in northern Oman, including Wadi Mu’aydin, presented in Figure 6 of van Buchem et al. (2002a), and also described by Pittet et al (in press), identified three scales of depositional cyclicity. In their large-scale cycles (tens to hundreds of meters thick), they show that there are subtle, but nonetheless distinct, upward decreases in clay content in their ‘Hawar Member’ at the top of the Kharaib Formation, and in the basal part of their Upper Kharaib Member. The influence in both cases is very weak and manifests itself as a trace of clay and a slightly elevated potassium/thorium response on a spectral gamma log, though this is largely obscured by an increased uranium response (van Buchem et al., 2002a). The base of the Lower Kharaib Member has a subtler, though apparently still upward-cleaning character. All of these clay-bearing limestones contain common orbitolinids. It is also noteworthy that the base of the ‘Hawar Member’ (KUD of Boichard et al. 1995) in Oman is interpreted as a sequence boundary (or a regional exposure surface according to van Buchem et al., 2002a). Pittet et al. (in press) reported rootlets in core from the United Arab Emirates that indicate subaerial exposure.

Van Buchem et al. (2002a) argued that there are two control mechanisms for the cyclicity seen in the Kharaib and Shu’aiba formations: these are accommodation (relative sea level) and climate. They, and Pittet et al. (in press) in a companion paper, suggest that climatic variations allied to eustasy account for the formation of the Lower Kharaib, Upper Kharaib and Shu’aiba cycles, and that “spasmodic tectonic events did not play a significant role during sedimentation of the studied succession”. However, none of these authors appear to have considered the high clastic-sediment influx from the west into central areas of the Arabian Plate. We interpret the textural and chemical maturity of the Biyadh and Zubair sandstones (Moshrif and Kelling, 1984; Khalaf et al. 1985; Nemcsok et al. 1998) as indicating reworking of early and/or late Paleozoic clastics. It is most probable that the main control on the high clastic-sediment flux was the tectonically driven uplift of the Paleozoic source areas, as suggested by Sharland et al. (2001), and that the climatic control proposed by Pittet et al. (in press) and van Buchem et al. (2002a) was at best subsidiary. We also suggest that considerable time intervals may be represented by the sequence boundaries at top Lekhwair, top Lower Kharaib and top Upper Kharaib/base Kharaib Upper Dense Zone that could also suggest a potential tectonic control on cyclicity. The best evidence of repeated subaerial exposure is found over a salt dome at Jebel Madar (van Buchem et al., 2002a) and is further evidence of factors other than climatic change and eustasy influencing the cyclicity of these formations.

Pittet et al. (in press) and van Buchem et al. (2002a) do not fully explain the origin of the clay in the sections overlying these three sequence boundaries. They appear to argue for climatic variation having partially controlled the sediment flux that, together with transgressive retreats of the shoreline, caused the upward decrease in the clay content. No positive evidence for this climatic variation exists and we identify two main models for the upward-decreasing clay content in the Kharaib and Shu’aiba formations of the Oman platform (Figure 12):

  • Model 1: Landward retreat and transgressive reworking of ‘weak’ fringing clastic systems (Figure 12a). This model is somewhat similar to that of van Buchem et al. (2002a) but does not require climatic fluctuation to account for the distribution of clay-bearing limestones. According to H. Droste (personal communication, January 2001), the regional data suggested that the clay-bearing horizons are landward-attached, which is consistent with the regional pattern illustrated by Vahrenkamp (1996) but difficult to confirm without access to regional subsurface data. We are not aware of any direct evidence linking the clay-bearing horizons of northern Oman to the shales and marls present in Qatar and western Abu Dhabi, although the up-systems-tract Biyadh Sandstone of Saudi Arabia is the most likely provenance. The depositional systems of Yemen (see below) seems a less likely source given the great distances involved and doubts concerning the continuity of systems tracts between the two areas (see below). However, Redfern and Jones (1995) suggested correlation between the upper Qishn carbonates of Yemen and the Shu’aiba Formation of northern Oman.

This model would appear to require the Arabian Plate to have acted as a single, structureless slab such that each deltaic advance from the western margin would have left a signature across the whole Plate. A consequence would be that the clay-bearing limestones represented TSTs, each capped by an MFS at approximately the base of the overlying clean limestone.

  • Model 2: reworking of paleosols from the subaerially exposed platform tops (Figure 12b). A wind-blown source for the clay of the paleosols is possible. In this model, the large deltaic systems of the Biyadh Sandstone and Qishn Formation did not encroach onto the limestone platforms. Some structural control over the distribution of the clastic and carbonate depositional systems is likely unless the Arabian Plate acted as a single, structureless slab. It might be expected that the clastic systems should have been at their most extensive during lowstands of relative sea level but these are the times when the ‘outboard’ platforms were most likely to have been exposed.

For this reason, our second model proposes that the clastic systems were unlikely to have extended as far as eastern Abu Dhabi and Oman, as no accommodation space would have been available over the exposed platforms. A full Transgressive System Tract (TST) succession would only have been preserved away from the platform areas. The platform areas themselves would have remained positive through most of the TST, only becoming drowned near to, or at the times of, maximum flooding. Reworking during initial transgressions would probably have destroyed the clay-bearing paleosols themselves. Outcrop observations suggest that the clay was redistributed by burrowing organisms common in the low-energy regime of the MFS and very early highstand. In this model, an upward-cleaning character cannot be interpreted as a signature of transgressive reworking, but is a sign of the limited amounts of clay available for reworking by burrowing organisms. Thus, the MFS can be located within the clay-bearing limestones.

There is clear evidence in the Wadi Mu’aydin outcrop of northern Oman of high-frequency cyclicity in the Kharaib Formation that is relevant to this discussion. This cyclicity is described in detail by van Buchem et al. (2002a) and Pittet et al. (in press). The ‘shalier’ orbitolinid-bearing intervals are present in meter-scale cycles with very distinctive vertical successions (Figure 13), which we believe may be used as analogs for the third-order depositional sequences. The cycles consist of a lower, dark, thin-bedded, orbitolinid-bearing argillaceous limestone that grades up into an overlying, lighter-colored, thicker-bedded, cleaner, micritic limestone (Figure 13). The clean limestones contain partially cemented, centimeter-scale vugs strongly suggestive of subaerial exposure (Figure 13). The lower, orbitolinid-bearing, argillaceous limestones have penetrated the underlying limestones as pipe-like features (Figure 13). Although further work is required to establish whether these features are diagenetically modified burrow infills in firmgrounds and/or hardgrounds, the vugs suggest that they are probably karst related. In this case, it is possible to interpret the meter-scale cycles as high-frequency cycles. The upper surfaces of the cycles are minor short-lived hiatuses, whereas the subtle but rapid upward-cleaning represents infaunal (i.e. burrow) reworking following rapid flooding, and is not due to transgressive (i.e. current) reworking. It can be argued that limestones forming these cycles are of high-frequency highstand origin and that the upper surfaces of the cycles probably represent the bulk of time. The source of the clay (shoreline-attached or wind-supplied paleosols) within the more argillaceous limestones remains unproven. A catastrophic source for the clay (see below) is another possibility.

The abundance of large, flat orbitolinids (plus associated forms such as echinoderm debris, Permocalculus algae and Lenticulina) in such close proximity to probable subaerial surfaces is problematic. Many workers would use this data to suggest relatively deep water of several tens of meters (e.g. Arnaud-Vanneau, 1980, 1994; Vilas et al., 1995; Hughes, 2000; Simmons et al., 2000; Figure 14). Others might argue that the faunal association reflects poor light-penetration in very shallow water due to very high suspended sediment loads. However, they fail to account for the origin of this high sediment load or the absence of evidence of associated turbulence such as traction-current energy.

For example, van Buchem et al. (2002a) interpreted orbitolinid-dominated facies as being found in a range of environments from tidal channel to subtidal. They used a similar association of orbitolinid-bearing calcareous shales and limestones with desiccation cracks to argue for a very shallow-water facies deposited in water depths of 0 to 3 m. However, the locality, Jebel Madar in northern Oman, overlies a Huqf salt dome (Mount et al., 1998). Given evidence of structural uplift of the Qatar Arch and Hormuz salt domes during deposition of the latest Lekhwair or earliest Kharaib (Hassan et al., 1975), shallowing at Jebel Madar could be the result of subtle, repeated salt and/or tectonic movement. Pittet et al. (in press) and van Buchem et al. (2002a) argued that the good correlation between their studied sections indicated minimal relief (no more than several meters) between Jebel Madar and the other localities, and that this lithofacies association also represents shallow-water conditions in other areas. To some extent, the observations of meter-scale cyclicity with orbitolinid-dominated facies and repeated probable subaerial exposure in Wadi Mu’aydin (Figure 13) may support van Buchem et al.’s interpretation, but in our view this is an issue that would benefit from further investigation.

An alternative, ‘deeper-water’ model requires that the minor clay-content of the orbitolinid-bearing limestones was derived from the overlying water column and not by transgressive (or infaunal burrow) reworking of a shoreline-attached clastic system or paleosol. This begs the question of how the high volume of suspended load was supplied, given that the major delta systems were located hundreds of kilometers away to the west. Unless a catastrophic source—such as widespread volcanic fallout—is invoked to supply clay-grade material in the form of ash, then it is difficult to identify an alternative origin for the clay content of the limestones. A phase of volcanism related to the ongoing breakup of Gondwana started at about 125 Ma (close to the Hauterivian-Barremian boundary) in the Levant and in northeast India (Rajmahal volcanics) (Segev, 2000), so a volcanic origin could be feasible. X-ray diffraction analyses of the clay-bearing limestones in the Kharaib Formation of Wadi Mu’aydin identified clay-mineral contents of 5 to 7 percent consisting of illite/mica with minor amounts of smectite (analyses performed by the University of Leicester). Although smectite is commonly of volcanic origin (e.g. Chowdhury, 1982), it can also be an indicator of climatic variations (Singer, 1984). In the Kharaib Formation, the low concentration of smectite allied to the absence of other volcanic associates (such as zeolites or volcanic glass) precludes a definitive interpretation in this case, and understanding the origin of the clay needs further work.

We consider that eastern offshore Abu Dhabi and Oman were far enough away from the source of siliciclastic input that the upward decrease in clay content cannot be explained by transgressive reworking of shoreline-attached clastics, and hence in this area there is no signature of a TST. Thus, on the basis of the microfauna, we pick the K70 MFS to be in van Buchem et al.’s (2002a) ‘Hawar Member’ (Upper Dense Member or KUD of Boichard et al., 1995) of the Upper Kharaib in northern Oman as did Sharland et al. (2001). However, our interpretation differs from van Buchem et al. (2002a), who regard the ‘Hawar Member’ (KUD) and lower half of the Shu’aiba Formation as belonging to the same, much longer lived TST. We infer that the micropaleontological evidence shows shallowing-up throughout the Lower Shu’aiba and thus that the lower part of the Shu’aiba is in the early part of the highstand system tract overlying the K70 MFS. This view is consistent with the microfaunal data presented by Boichard et al. (1995). We also suggest that van Buchem et al’s Figure 8 (a-c) showed clear highstand progradation with clinoforms in the Lower Shu’aiba, prior to initiation of the Bab Basin, and that this is inconsistent with their interpretation of the same interval as late TST. However, they argued that the clinoforms represented higher (fourth-order?) progradation and were not as significant as those present higher in the section (Frans van Buchem, personal communication, April 2002). If our updated model is correct, then minor adjustment of some of Sharland et al.’s (2001) MFS within the eastern carbonate platform successions may be required.

Van Buchem et al. (2002a) and Pittet et al. (in press) suggested that the succession of facies associations indicated that the K70 MFS is a fourth-order maximum flooding surface, though they acknowledged that it is identifiable over a wide area. They further suggested that the maximum water depth was attained during development of the Bab Basin and the accumulation of organic matter, and therefore that the succeeding MFS (K80) was a more important stratigraphic surface. Van Buchem et al. interpreted that there was no evidence of tectonic control in the formation of the Bab Basin; a predominantly eustatic control is favored by them. It is a theme of our paper that maximum water depth should not be automatically associated with maximum flooding. In this case, although we agree that maximum deepening is recorded within the organic-rich facies of the Bab Basin, we interpret the development of the Bab Basin as a response to the tectonically driven subsidence of a limited part of the Arabian Plate. However, we accept that more definitive evidence of this subsidence is required. We argue that the maximum deepening cannot be used to interpret the K80 MFS as being a more important plate-wide flooding event than the K70 MFS. Further, we infer that extremely detailed faunal and microfloral analysis of the Shu’aiba Formation on the western side of the Bab Basin (G.W. Hughes personal communication, February 2002) supports our interpretation that the K70 MFS is the lower-order/lower-frequency surface. We urge caution in extrapolating data from tectonically influenced intrashelf basins to the Plate as a whole.

In summary, the main differences between our findings and the interpretation of van Buchem et al. (2002a) are as follows:

  • (a) We propose increased subsidence as controlling the onset of formation of the Bab Basin whereas van Buchem et al. did not recognize evidence for tectonic control. They propose a solely eustatic model. We accept that more evidence is required to demonstrate that increased subsidence was operative, though we find it striking that the development of Bab Basin preceded the formation of the plate-wide intra-Aptian unconformity. Either eustatic sea-level fluctuations were increasing in magnitude or some tectonic influence seems necessary. Perhaps orbital-forcing models may hold the answer to resolving this issue (see Matthews and Frohlich, 2002; this issue).

  • (b) Van Buchem et al. invoked climatic control (fluctuating humid and arid settings) allied to accommodation to account for the stratigraphic and areal distribution of clays in the Kharaib and Shu’aiba formations of the Oman Platform. There is no positive evidence for a dominant climatic control of clay distribution. Instead, our regional data is more consistent with intermittent hinterland uplift being the dominant control, though an alternative catastrophic source cannot be ruled out. We suggest that more evidence is required to substantiate a dominant climatic control.

  • (c) This paper has followed conventional models for the paleoecology of Barremian to Aptian microfossils whereas van Buchem et al. (2002a) and Pittet et al. (in press) have presented important new data that question the environmental interpretation of orbitolinid-bearing argillaceous limestones. We agree that a re-evaluation of orbitolinid paleoecology is desirable. Ultimately this will help decide whether MFS should be picked in the argillaceous, orbitolinid-bearing limestones, as in this paper, or in the overlying limestones, as proposed by van Buchem et al. (2002a). This debate is only relevant to picking MFS in the Oman Platform and easternmost Abu Dhabi, since in more westerly areas, including western Abu Dhabi, we identify landward-attached argillaceous limestones and shales separating clean limestones.

  • (d) Finally, this paper supports the interpretation of Sharland et al. (2001) that the K70 MFS is the more important MFS, being identified over most of the Arabian Plate, whereas van Buchem et al. (2002a) proposed that the succeeding K80 MFS is a more important stratigraphic surface.

It is striking that van Buchem, Pittet et al. and Sharland et al. identified the same number of cycles (third-order?) within the Barremian to Aptian succession based on data from more carbonate and more clastic-dominated areas respectively. The most significant difference between the two groups of workers seems to be in the interpretation of the precise locations of MFS. In this paper, we consistently pick MFS deeper in the stratigraphy than would van Buchem et al. While this paper differs from van Buchem et al. in some details of interpretation, we fully endorse their regional approach to unraveling the detailed stratigraphy of the Arabian Plate.

Yemen

Ellis et al. (1996) and Holden and Kerr (1997) described the succession in Yemen. More detailed paleogeographic information is available for the Central Marib-Shabwa Basin (Brannan et al., 1999). Holden and Kerr (1997) and Brannan et al. (1999) provided key well data for which Sharland et al. (2001) presented sequence stratigraphic interpretations. The late Valanginian unconformity (LVU) is located at the base of the mixed clastic-carbonate Qishn Formation over much of the area (Figure 15; Ellis et al., 1996; Brannan et al., 1999). However, lithostratigraphic nomenclature in Yemen is complicated (Beydoun et al., 1998) and the term Qishn Formation has a variety of usage. Redfern and Jones (1995) interpreted the Qishn clastics as being age-equivalent of the Biyadh Sandstone of Saudi Arabia. They also identified a clear overall transgressive signature through the Qishn clastics and noted that the upper part of the formation typically consists of carbonates. Holden and Kerr (1997) identified an additional shale-dominated Furt Formation in the Sayun-Al Masila Basin that appears to be laterally equivalent to the Lekhwair Formation of the southern Arabian Gulf. Large thickness changes (Figure 16) may be partially due to salt movement and inherited rift topographies, and therefore may indicate considerable relief on the LVU.

A down-systems-tract facies change from coarse-grained clastic deltas in the west to shale and limestone-dominated successions farther east was recognized by Redfern and Jones (1995), Holden and Kerr (1997), and Brannan et al. (1999). Brannan et al. also identified a strong control on sedimentation by inherited synrift topography and mobile Jurassic salt walls in the Marib-Shabwa Basin. The orientation of these rift basins (Ellis et al., 1996) suggests that these important clastic systems were more likely to have been funneled toward the southeast and were separated from the northern Oman systems by the Hadramaut Arch (see Enclosure 1 of Sharland et al., 2001). However, Ellis et al.’s Barremian paleofacies map suggests otherwise. According to Redfern and Jones (1995), carbonates at the top of the Qishn Formation (i.e. Upper Qishn) can be traced across the North Hadramaut Arch. They probably equate with the Shu’aiba Formation of Oman but connections between the Kharaib and Lekhwair of Oman with the Lower Qishn carbonates of Yemen are more speculative (Figure 17). Perhaps this reflects earlier high sediment supply filling the grabens and the establishment of new sediment patterns dominated by carbonates during a subsequent rise in relative sea level. However, there is doubt as to whether the earlier clastic systems could have supplied the traces of clay seen in the northern Oman sections.

Putnam et al. (1997) provided a sedimentological analysis of the Upper Qishn in the Masila region of Yemen. Key interpretations included an overall transgressive pattern of coarse-grained fluvial sandstones overlain by heterogeneous estuarine facies, which are themselves overlain by bioturbated, shelf mudstones. The bioturbated mudstones form part of the seal to the sandstone reservoirs. There are clear similarities to the Zubair of the northern Gulf (Nemcsok et al., 1998) both in the association of depositional facies and in the role of intraformational shales as seals to underlying fluvial and estuarine sandstone reservoirs.

By analogy with the Zubair Formation of the northern Gulf, it seems reasonable to identify the K40, K50, and K60 MFS within intraformational shales of the sandstone-dominated parts of the Qishn Formation (Figures 15 and 16). These intraformational shales can be interpreted as representing landward retreats of the major delta systems during maximum flooding. Farther east, by analogy with Qatar, the K50 and K60 MFS are more likely to be present within clean limestones of the Lower Carbonate Member of the Qishn, whereas interbedded shales indicate highstand advance and probably early stages of retreat of the delta systems (Figure 16). The presence of sandstones associated with shales in the limestone-dominated K60 Genetic Stratigraphic System (GSS) of Al-Furt-1 is good evidence that the shales are associated with highstand delta advance and are not deeper water shales (Figure 16). Note that the sandstones are only present in the K60 GSS, which is consistent with the identification of a relatively low-order/low-frequency maximum regression surface in the Upper Zubair Sand of the northern Gulf. We believe that the models presented in this paper better explain the stratigraphic organization of the Qishn Formation outlined by Ellis et al. (1996). The Qishn Shale Member should be interpreted as a regressive episode, and not a marine transgression and associated MFS as previously suggested by Ellis et al.

The overall but subtle cleaning-upward gamma-ray character of the Furt Formation suggests that the K40 MFS is located near to the base (Figure 16). The K70 MFS can be picked in the (upper) Carbonate Member of the Qishn Formation (Figure 16), which is consistent with the correlation proposed by Redfern and Jones (1995), thereby suggesting that the Oman and Yemen basins were linked by this time. However, the K70 MFS may be missing by erosional truncation by the coarse clastics of the overlying Tawilah Group in the western part of the Marib-Shabwa Basin (Figure 15).

In summary, the succession between the LVU and the intra-Aptian unconformity in Yemen contains a depositional system that is comparable to that of the northern Arabian Gulf (Figure 17). It consists of a proximal (up-systems-tract), alluvial to deltaic, coarse-grained clastic-dominated unit, a mid-systems tract, shale-dominated prodelta, and an outboard (down-systems-tract) carbonate platform (Figure 17). There is clear evidence of landward retreat of the coarse-grained proximal facies and associated landward migration of the prodelta shales and the outboard carbonate platforms during major transgressions (Redfern and Jones, 1995; Ellis et al., 1996; Putnam et al., 1997; Beydoun et al., 1998). Strong basinward migration of facies belts was linked to major regressions. Considerable diachroneity is associated with these migrations (Figure 17).

MAXIMUM FLOODING SURFACES IN THE ALBIAN DEPOSITIONAL PACKAGE (K90 MFS-K110 MFS)

This interval lies between the intra-Aptian unconformity and the late Albian K110 MFS recognized across the whole of the Arabian Plate (Sharland et al., 2001). Hence, it is not bounded top and bottom by a sequence boundary and does not equate to a depositional sequence in the sense of Mitchum et al. (1977).

As noted previously, the intra-Aptian unconformity corresponds to top Shu’aiba Formation in most places (Alsharhan and Nairn, 1997), but lies within the defined Burgan Formation in the northern Gulf (Al-Fares et al., 1998; Kirby et al., 1998; Sharland et al., 2001). All of these authors interpreted the top of the Shu’aiba as being conformable with the lower part of the Burgan. It is therefore significant that Al-Ateeqi and Foster (2001) used 3-D seismic data to interpret the surface of the top Shu’aiba in the area surrounding the Minagish and Umm Gudair fields in southern Kuwait as being strongly karstified. This implies, though it does not prove, that the intra-Aptian unconformity in this area is coincident with the top of the Shu’aiba Formation. However, Kirby et al. (1998) identified the major unconformity within the Fourth Sand (Lower Burgan) of Umm Gudair, though their evidence for this was not presented. This issue is discussed further later in this paper.

This interval contains the K90 and K100 MFS as defined by Sharland et al, although K90 has proved difficult to identify outside the eastern and northern Arabian Plate as there is considerable depositional variation across the region (Figure 3). For the purpose of the following description, the region has been divided into Northern, Central, and Southeastern Gulf areas (see Figure 1). Again, a brief description of the succession in Yemen is provided in order to facilitate comparisons with the northern Gulf and because it may lie up-systems-tract from the outboard carbonate platforms and intrashelf basin areas of north Oman.

Northern Gulf Area

Some of the most detailed descriptions of the Albian successions are available from the northern Gulf, notably by Al-Fares et al. (1998), Kirby et al. (1998) and Al-Eidan et al. (2001). Grant and Al-Humam (1995) presented a useful regional cross-section for the Kuwait-Saudi Arabia Partitioned Neutral Zone and extending into offshore Iran. Denby et al. (2001) presented some logs for the Nowrooz and Soroosh fields of offshore Iran, though these only cover the lower part of the Albian section. Some information is available on southern Iraq (van Bellen et al., 1959; Buday, 1980; Ibrahim, 1981, 1983; Alsharhan, 1994; Al-Ameri et al. 2001), but unfortunately these sources did not provide sufficient detail for our analysis. However, we predict that our interpretations could be applied in this area.

Kuwait

The interval is dominated by the Burgan Formation that has lateral equivalents in the Khafji (equivalent to Lower Burgan) and Safaniya (Upper Burgan) members of the Wasia Formation in adjacent parts of Saudi Arabia (Entsminger, 1981; Grant and Al-Humam, 1995). It also includes the lower part of the Mauddud Formation. The Burgan Formation is the major reservoir in the Burgan, Raudhatain, and Sabiriya fields in Kuwait (Brennan, 1990, 1991). As noted previously, there is good biostratigraphic evidence that the major intra-Aptian unconformity lies within the defined Burgan Formation in offshore Kuwait (Figure 18; Al-Fares et al., 1998) and in the Burgan field (Kirby et al., 1998). This unconformity almost certainly corresponds to the LB25SB surface of Al-Eidan et al. (2001) within the Lower Burgan Member of North Kuwait. Both Kirby et al. and Al-Eidan et al. presented environmental interpretations of the Burgan Formation based on extensive core data, although Al-Eidan et al. only described the Middle and Upper Burgan members.

The Lower Burgan (Fourth Sand of the Burgan Formation) is dominated by very fine to medium-grained sandstone that forms a very high net-to-gross reservoir (Kirby et al., 1998). Kirby et al. interpreted this section as having been deposited in a braided river to braided delta environment within a Lowstand Systems Tract (LST). They noted an upward increase in reservoir heterogeneity and interpreted the upper part of the Lower Burgan as a transgressive system tract consisting of coastal braid-plain sediments in the southwest of Kuwait, and deltaic and estuarine sediments in the northeast.

The Lower Burgan (Fourth Sand) is capped by shoreface sands and marine mudstones that can be correlated throughout the Greater Burgan field and much of Kuwait (Figure 19; Kirby et al., 1998). The equivalent section of marine mudstones forms the lower part of the Middle Burgan Member of North Kuwait (Al-Eidan et al., 2001) and is present in offshore Kuwait (Al-Fares et al., 1998). The K100 MFS is located near to the base of these marine mudstones (Figures 18 and 19). Significantly, Al-Eidan et al. reported thin, orbitolinid-bearing limestone beds associated with these marine mudstones, thereby hinting at greater limestone development down-systems-tract. This would match southern Iraq where van Bellen et al. (1959) described the Dair Limestone Member of the Nahr Umr Formation as thinning westward and southward, and grading into shale and sandstone. Van Bellen et al. also noted that the Dair Limestone Member could not be recognized in the Kuwait fields.

As noted above, the K90 MFS is difficult to identify in this area and may be missing in the northern Gulf due to onlap and/or erosion. The only candidate for the K90 MFS identified so far from information in the public domain is a marine section separating the massive ‘M’ and more heterogeneous ‘L’ Sands of the Lower Burgan in North Kuwait (see Figure 16 of Al-Eidan et al., 2001). This marine section is a major control on vertical connectivity in the Lower Burgan reservoir in this area (Shaikh and Al-Saig, 2002). More details of the Lower Burgan need to be published before a final interpretation can be reached.

More detailed descriptions are available for the Upper Burgan (Third Sand) of Kuwait (Kirby et al., 1998; Al-Eidan et al., 2001). Both sets of workers independently identified a progradational lower section, a more sand-rich section in the middle of the Upper Burgan corresponding to the K100 MRS, and a retrogradational (i.e. punctuated backstepping) upper part. Depositional environments include amalgamated channels, isolated marine/estuarine channels, shoreface sandstones and marine mudstones. A significant lateral facies change is associated with lower net-to-gross to the northeast (Figure 19). Channel orientations from Greater Burgan (Kirby et al., 1998) and North Kuwait (Al-Eidan et al., 2001) indicate that this is down-systems-tract. Significantly, it matches the direction of facies change identified between the Dair Limestone Member and equivalent shales of the Nahr Umr in southern Iraq (van Bellen et al., 1959).

Clastic incursions occurred in the lower part of the Mauddud reservoir of northern Kuwait (Figure 20; Al-Ajmi et al., 2000). The result is such that the lower part of the formation contains clear evidence of continued retrogradation of the Upper Burgan delta system in the form of thick prodelta to marine shales interbedded with limestones (Figure 13 of Al-Eidan et al., 2001; Al-Ajmi et al., 2000). Kirby et al. (1998) presented a correlation panel from Umm Gudair to Bahrah. It shows a lateral change from thick sandstones in the Upper Burgan (and either a very thin or no Mauddud in Umm Gudair to the southwest), to a thick Mauddud Formation overlying a more heterogeneous Upper Burgan in Bahrah field to the northeast (Figure 19). This northeasterly thickening is comparable to that of the Dair Limestone Member described by van Bellen et al. (1959). In northern Kuwait, one of the prodelta shales in the lower half of the Mauddud Formation is capped by a permeable shoreface sandstone (reservoir unit MaH of Al-Ajmi et al., 2000), confirming that these are shoreline-attached shale systems and not offshore basinal shales.

There seems little doubt therefore, that on a regional scale the lithological contact between the top of the Burgan Formation and the base of the Mauddud Formation is a diachronous facies change as illustrated by Kirby et al. (1998) and interpreted in this paper. Strohmenger et al. (2002) independently made the same interpretation based on an extensive subsurface data set from Kuwait. The K110 MFS of Sharland et al. (2001) must thus lie above the mixed carbonate and clastic retrogradational package that forms the lower half of the Mauddud Formation in northern Kuwait (reservoir layers MaF to MaJ of Al-Ajmi et al., 2000). The highest-energy, shallow-water facies described by Al-Ajmi et al. is a rudist floatstone associated with grainstones and skeletal packstones near the top of the Mauddud in reservoir layer MaC. It is therefore reasonable to interpret most of the upper carbonate-dominated section of the Mauddud as a progradational highstand system with the K110 MFS located near its base, probably within Al-Ajmi et al.’s MaE unit (Figure 20).

Note that this is a very different interpretation from that of Al Ajmi et al., who placed the MFS near the base of the Mauddud on the basis of vertical sequence analysis rather than a regional perspective. However, Al-Ajmi et al. labeled several MFS within their progradational package, which suggests that they were referring to marine flooding surfaces bounding individual parasequences. The prodeltaic shales within the low-frequency Transgressive System Tract (TST) have considerable economic significance as they act as major vertical permeability and transmissibility barriers in the northern Kuwait fields (Al Ajmi et al., 2000). They may also have some stratigraphic trapping potential.

Kuwait-Saudi Arabia Partitioned Neutral Zone

The succession in the Partitioned Neutral Zone is predicted to be similar to the Kuwait succession, though less information has been published. Some details are available in Al-Sabti and Al-Bassam (1993) who presented a composite stratigraphic column for the Wasia Formation in the northern offshore area of Saudi Arabia. They showed that both the Khafji Member (equivalent to the Lower Burgan) and the Safaniya Member (equivalent to the Upper Burgan) consist of three units—a lower more heterogeneous section, a middle sandstone dominated section, and an upper heterogeneous section (Figure 21). This is similar to the description of the Upper Burgan in particular (Al-Eidan et al., 2001; Kirby et al., 1998). Al-Sabti and Al-Bassam also described a decrease northeastward in the net-to-gross content of the Safaniya Member, which is consistent with the data from Kuwait. The Khafji and Safaniya members are separated by an interval of shale and limestone known as the Upper Khafji or the Dair Limestone (Figure 21).

Northern Saudi Arabia

Our sequence stratigraphic interpretation of the Wasia Formation of offshore northern Saudi Arabia is presented in Figure 21. By analogy with Kuwait, we place the intra-Aptian unconformity at the base of the main sand of the Khafji Member. The Main Sand and the Upper Khafji Stringers represent a LST and succeeding TST, respectively. As in other parts of the northern Gulf, the K90 MFS has not been recognized. The lateral equivalent to the shale in the middle of the Burgan is the Dair Limestone (Entsminger, 1981; Al-Sabti and Al-Bassam, 1993; Grant and Al-Humam, 1995). More than 20 years ago, Entsminger identified this limestone as a major flooding event separating regressive episodes in the Khafji and Safaniya members of the Wasia Formation. The limestone contains the K100 MFS (Figure 21). The sand-rich middle unit of the Safaniya Member contains the K110 MRS whereas the upper heterogeneous and lower parts of the overlying Mauddud Formation represent the succeeding TST. Again, by analogy to Kuwait we interpret the K110 MFS as being present in the clean Upper Mauddud reservoir above the more heterogeneous lower section of the Lower Mauddud Member (Figure 21).

Iran

The Burgan (Kazhdumi) reservoir and Dair Limestone are also present in parts of offshore Iran (Denby et al., 2001), although the Burgan A, B and C units as illustrated are only equivalent to the Lower Burgan of Kuwait and the Khafji Member of Saudi Arabia. The maximum extent of sands in the northern Gulf approximates to the location of the present Iranian shoreline (Motiei, 1993; Bordenave and Huc, 1995). In the Soroosh and Nowrooz fields, the Dair Limestone is at least 50 m thick (Denby et al., 2001), which is consistent with the northeasterly thickening of this unit in southern Iraq (van Bellen et al., 1959). The Dair Limestone of offshore Iran contains the K100 MFS of Sharland et al. (2001) (Figure 22). The overlying Kazhdumi Shale Member (not illustrated) is widely recognized in the northern Gulf and is laterally equivalent to the Upper Burgan (unpublished observations of two of the authors, RBD and PRS). It is overlain by the thick limestones of the Sarvak Formation (Motiei, 1993). The K110 MFS is located in the basal part of the limestones (Sharland et al., 2001). Again, it was not possible to identify K90 in this area but it could be in shales that separate the Shu’aiba Formation from the sharp base of the first sandstone (Figure 22). By comparison with other areas, we suspect that the sharp base of the Burgan-B sandstone in Soroosh is a better candidate for the intra-Aptian unconformity than the deeper pick of Denby et al. (2001). Biostratigraphic data are required to establish the age of the shales.

Farther to the northeast, the succession above top Dariyan Formation (equivalent to the Shu’aiba) consists of the Kazhdumi shales and the basal part of the Sarvak Formation (Motiei, 1993). These two formations vary considerably in thickness across the area (Motiei, 1993) and we suspect that the Kazhdumi-Sarvak boundary is diachronous. Current work by van Buchem and colleagues should refine our understanding of the precise stratigraphic relationships between the Dariyan, Kazhdumi, and Sarvak formations (van Buchem et al. 2002b).

The Kazhdumi is a major source rock in parts of Khuzestan (Bordenave and Burwood, 1994). The published model for its formation is that a shallow-water area or sill separated the source-rock area from the open ocean and was largely responsible for the restriction of the intrashelf basin in the Dezful Embayment (Bordenave and Huc, 1995). There was probably some structural control over the location of the sill as it was shallow enough at times for dasycladacean algae to grow, which implies water depths of only a few meters. It is currently unclear whether the area between the sill and the open ocean was prone to carbonate or shale deposition. However, it would be reasonable to assume that siliciclastic deposition was favored during regressive phases associated with delta advance, and that carbonate deposition was more likely during periods of delta retreat. The area of deeper water was bounded to the south by the northern limit to the Fars Platform. Another area of shallow-water carbonates, the Bala Rud Shoal, bounded the area to the north (Bordenave and Huc, 1995), and represented at least a partial outboard platform area in a down-systems-tract position. The recent work of van Buchem et al. (2002b) identified this as infill of an intrashelf basin surrounded by carbonate ramps, although this interpretation appears to underestimate the importance of the major contemporary clastic systems of the Nahr Umr, Burgan and Wasia formations to the west.

Summary

Three units dominate the depositional system between the intra-Aptian unconformity and the K110 MFS in the northern Gulf. They are a proximal (up-systems-tract), alluvial to deltaic, coarse-grained clastic-dominated unit; a mid-systems tract, shale-dominated prodelta area prone to source-rock deposition; and a combination of carbonate platform areas and a shallow-water sill (down-systems-tract) separating the source prone area from the open ocean (Figure 23). Carbonates, notably the Dair Limestone Member and the lower Mauddud/Sarvak, developed during major periods of delta retreat. Carbonate deposition dominated on structural highs (Fars Platform, Bala Rud Shoal) away from the main loci of clastic input (Figure 23), although the Burgan-Kazhdumi system may have extended to the basin margin in northeastern Khuzestan. A clear landward retreat of the coarse-grained proximal facies and associated landward migration of the prodelta shales and the outboard carbonate platforms occurred during major transgressions. Strong basinward migration of facies belts was linked to major regressions. Considerable diachroneity was associated with these migrations on the western side of the Gulf (Figure 23), but there is more evidence of structural control on facies boundaries in parts of Iran.

Central Gulf Area

Central Saudi Arabia

The information concerning this interval in central Saudi Arabia and Qatar is more limited. The observations of Entsminger (1981) have been referred to above but may be more relevant to the subsurface of northeastern Saudi Arabia. Moshrif and Kelling (1984) and Sharief et al. (1989) both provided some information on the Wasia Formation in outcrop. Sharief et al. showed that the Khafji and Safaniya members are separated by silty shale. By analogy with the Burgan Formation, this silty unit is interpreted as a prodelta shale that contains K100 MFS and, therefore, is the lateral equivalent of the Dair Limestone Member identified by Entsminger (1981). The shale is difficult to identify in the sections described by Moshrif and Kelling (1984).

Two other observations can be made. Firstly, limited paleocurrent data for the Wasia Formation indicate unimodal easterly flowing fluvial systems (Figure 7). Secondly, the vertical succession for the Safaniya Member illustrated by Sharief et al. (1989) shows the ‘B’ Sand overlain by a shale-dominated succession and then by the overlying orbitolinid limestones of the Mauddud Formation. We interpret this succession to represent a TST comparable to the time-equivalent section forming the upper part of the Upper Burgan and lower part of the Mauddud of Kuwait. Hence, we interpret K110 MFS to be present in the Mauddud Member of the Wasia Formation at outcrop in central Saudi Arabia.

Qatar

The succession in Qatar is markedly different. The lateral equivalent of the Khafji and Safaniya members is the shale-dominated Nahr Umr Formation. According to the paleocurrent data of Moshrif and Kelling (1984), the Nahr Umr should lie down-systems-tract. It is much thinner than the coarser-grained section to the west, probably reflecting a combination of thinning onto the Qatar Arch and its down-systems-tract position. The Nahr Umr consists of shale with a few thin sandstones toward its base that are a minor reservoir in offshore Qatar (Wells, 1987). According to Ibrahim et al. (2000), palynological data from a well in Dukhan field on the west side of Qatar indicated that the lower, sandier A member of the Nahr Umr is prodeltaic and that the overlying, more shale-dominated B and C members represent shelfal-water depths of 10 to 80 m. The K100 MFS probably lies close to the base of the B member. Ibrahim et al. assigned a mid-late Albian age to the Nahr Umr. This suggests that the early Albian K90 MFS may be absent here and confined to the area east of the Qatar Arch, though it should be noted that their lowest two samples contain no diagnostic fauna and could be older. The absence of the K90 MFS would be consistent with its apparent absence in northern Kuwait (see previous discussion) and possibly Yemen (see later discussion).

The most detailed log information for the overlying Mauddud Formation is available in Rubbens et al. (1983). There is an overall upward-cleaning character to the gamma-ray log through the Lower Marly Member into the Upper Limestone Member (Figure 24). High gamma-ray peaks in the Lower Marly Member are probably equivalent to the prodelta shales of the northern Arabian Gulf area. It is proposed that the Lower Marly Member represents a low-frequency TST capped by the K110 MFS present near to the base of the Upper Limestone Member. The remainder of the Upper Limestone Member above K110 belongs to the succeeding HST. Rubbens et al. showed complementary thickness variations between the Lower and Upper members, reflecting the probable diachroneity of this boundary (Figure 24). In detail therefore, the K110 MFS should be close to the base of the Upper Limestone Member in wells with thinner sections (such as NWD-2) but should be much closer to the middle in thicker sections (such as NR-1 to the northwest). While there may be some stratigraphic trap potential associated with such diachronous boundaries, there has been little success so far in exploring for stratigraphic traps in the mid-Cretaceous of the Arabian Plate.

Summary

The central Gulf area has a depositional system similar to that of the northern Gulf. The Wasia Formation clearly represents a proximal (up-systems-tract), alluvial to deltaic, coarse-grained clastic-dominated unit. The Qatar Arch approximates to the depositional limit of westerly derived sandstones and the shales of the Nahr Umr Formation are widespread to the east of the Arch. There is further evidence of landward migration of prodelta shales and the outboard carbonate platforms during the major transgression that culminated in the K110 MFS.

Southeastern Gulf Area

Eastern Saudi Arabia

In eastern Saudi Arabia, the United Arab Emirates and Oman, the top of the Shu’aiba Formation corresponds to the regional intra-Aptian unconformity (Alsharhan and Nairn, 1997; Hughes, 2000; Sharland et al., 2001). Newell and Hennington (1983) described the overlying Wasia Formation of eastern Saudi Arabia using data from the South Kidan and North Kidan fields. The Khafji and Safaniya members were both identified in this area but, unlike in central Saudi Arabia, they are dominated by marine, probably pro-delta shales sourced from the major delta systems to the west. The two members are separated by an organic-rich laminated lime mudstone that Newell and Hennington identified as a major transgressive event. By analogy with the equivalent section in central Saudi Arabia, this organic-rich unit is interpreted as containing the K100 MFS, and therefore that it is the lateral equivalent of the Dair Limestone Member identified by Entsminger (1981). The Mauddud Member of the Wasia Formation overlies the Safaniya Member.

United Arab Emirates and Oman

The Albian in the United Arab Emirates and Oman is dominated by the Nahr Umr Formation (Alsharhan, 1991; Immenhauser et al., 1999). A marked southeasterly thickening of the Nahr Umr takes place away from the Qatar Arch into the area of the Bab Basin (Alsharhan, 1991), implying that the Qatar Arch remained a positive feature during this period of deposition. It is probably significant that sandstone beds are effectively absent in the United Arab Emirates. For this reason, the Nahr Umr is not a reservoir east of the Qatar Arch but instead becomes a major regional seal (Alsharhan, 1991), especially in Oman (Hughes Clarke, 1988; Grantham et al., 1988).

The Nahr Umr is overlain by the Mauddud Formation in the United Arab Emirates and by the Natih Formation in Oman. Biostratigraphic evidence presented by Simmons and Hart (1987), Kennedy and Simmons (1991), Witt and Gokdag (1994), and Immenhauser et al. (1999) suggested that the Nahr Umr in Oman represents a longer period of time than on the western side of Qatar. This may explain why the K90 MFS is identified in the southern Gulf area but is difficult to identify in the northern Gulf. It is noteworthy that several authors (Masse et al., 1997; Immenhauser et al., 2000; Immenhauser et al., 2001), described a platform-margin carbonate succession, the Al Hassanat Formation, in the Saih Hatat area of northern Oman. They all interpreted it as being a lateral equivalent of the Nahr Umr Formation. Their data provided a further example of a carbonate platform situated down-systems-tract from a major clastic-dominated depositional system.

Immenhauser et al. (1999) used evidence from stable isotopes, brackish fluid inclusions, and petrographic evidence of root structures to interpret repeated subaerial exposure in the Nahr Umr of northern Oman, although at outcrop the relevant surfaces have the appearance of hardground. They argued for relative sea-level changes of several tens of meters to account for subaerial exposure in this interpreted southeastern extremity of the Bab Basin (Murris, 1980). This interpretation is difficult to reconcile with that of Sharland et al. (2001) who selected one of Immenhauser et al.’s sections (Wadi Bani Kharus) as the reference locality for their K90 and K100 MFS (Figure 25). It is striking that the K100 MFS in particular was picked in a limestone bed containing ammonites immediately above Marker Bed II. According to Immenhauser et al. Marker Bed II has sequence boundaries (SB6 and SB7) at its top and base; note that K100 pick was incorrectly shown in Figure 4.59 of Sharland et al. (2001).

The work of Immenhauser et al. (1999) implied that there should be evidence of subaerial exposure on both sequence boundaries, although explicit data was not provided. We find it difficult to accept that changes in relative sea level of the order of tens of meters could happen with such regularity without more overt signs of changing sedimentation style; for example, comparable to the evidence described above for the high-frequency cycles of the Kharaib Formation. We note that shallow-water, oxygenated Kazhdumi Formation marls in the Fars Platform to the north of the Bab Basin are reported to contain evidence of temporary subaerial exposures (Bordenave and Huc, 1995), though it should be emphasized that this is a platform area and not an intrashelf basin. The observation does at least raise the possibility of temporary restrictions to the connections with the open ocean, which could have led to salinity fluctuations in the intrashelf basin itself. Whether this could tie into the geochemical evidence of Immenhauser et al. (1999) remains to be investigated. More work is required to reconcile the two models.

The overlying Natih Formation of Oman was thoroughly described by van Buchem et al (1996), and van Buchem et al. (2002c). Sharland et al. (2001) picked the K110 MFS in the lower part of the Natih-G limestones. Likewise they placed K110 in the Mauddud Formation of the United Arab Emirates. Given the lack of detailed published biostratigraphic evidence, we have no cause to change these picks. However, we appreciate that in the Bab Basin it is possible that K110 could occur within the upper part of the Nahr Umr, and that the Mauddud and Natih-F and Natih-G may be highstand carbonates.

Yemen

The succession in Yemen was described by Ellis et al. (1996), with information on individual formations available in Beydoun et al. (1998). Additional details of the Central Marib-Shabwa Basin were presented by Brannan et al. (1999). Given the present sparcity of published data, it is difficult to understand how the Yemen succession links with the Nahr Umr Formation of northern Oman, especially in the carbonate-dominated succession of eastern Yemen. Regional structural cross-sections and a chronostratigraphic diagram presented by Redfern and Jones (1995), showed continuity of the Aptian to Albian section across the North Hadramaut Arch, which suggests that a link between the two areas is possible.

The Albian succession of western Yemen contains most of the Harshiyat Formation, which is the lower part of the Tawilah Group (Ellis et al., 1996) but the Harshiyat was not differentiated in the clastic-dominated section present in the Marib-Shabwa Basin (Brannan et al., 1999). Farther east, in the areas studied by Ellis et al., the Albian section is dominated by sandstones (Harshiyat) in the west and carbonates (the Fartaq Formation) to the east (Figures 17 and 26). Ellis et al. showed two limestone units, the Rays and Dha Sohis members, in the Harshiyat sandstones although, in some areas, the two names have been applied to the same depositional unit (Beydoun et al., 1998).

The interpretation of Ellis et al. (1996) was based on a large subsurface data set, which implies that two separate limestone members are traceable over a wide area. They interpreted the limestones as being the products of transgressive events. Beydoun et al. (1998) also interpreted thin carbonate members in the Harshiyat as representing flooding events. Our favored interpretation is that the Rays and Dha Sohis members are separate and contain the K100 and K110 MFS respectively (Figure 26). We have based our interpretation as follows:

  • Ellis et al. showed two separate limestone members that are both of Albian age;

  • the location of the Yemen basins is to the west of the Qatar Arch; and

  • a long distance separates western Yemen from any proven identification of the K90 MFS.

The overall transgressive character of the Fartaq Formation described by Beydoun et al. (1998) seems comparable with the pattern interpreted for the Burgan/Mauddud section of the northern Gulf (see previous discussion). Further work is needed to test these interpretations of the Yemen succession.

The data of Brannan et al. (1999) identified a strong control by salt movement over deposition at this time in the Marib-Shabwa Basin to the west. There is no published evidence of a link between the Yemen succession and the Nahr Umr of northern Oman, and the details of the Albian systems tracts along the southern part of the Arabian Plate remain uncertain. The data presented by both Ellis et al. (1996) and Beydoun et al. (1998) suggested that there was continual carbonate deposition in eastern Yemen. It also implied that the Nahr Umr shales of northern Oman had no direct link to the Harshiyat Formation of Yemen (Figure 17)—although facies similar to the Nahr Umr are present at outcrop in Dhofar (unpublished observations of author RBD). This is another area that would benefit from further research.

DEFINITION AND IDENTIFICATION OF MAXIMUM FLOODING SURFACES IN MIXED CLASTIC-CARBONATE DEPOSITIONAL SYSTEMS

Sharland et al. (2001), following the original definition of Posamentier et al. (1988), identified a maximum flooding surface as an isochronous surface within a sedimentary succession deposited during the most landward flooding event within a cycle of transgression and regression. They also noted that “it commonly represents the time of deepest water deposition within a basin”. Maximum Flooding Surfaces (MFS) form the upper and lower boundaries of Genetic Stratigraphic Sequences in the manner of Galloway (1989). They separate Transgressive Systems Tracts (TST) below from the early Highstand Systems Tracts (HST) above. In simple terms, the shoreline retreats landward in the TST and advances basinward in the HST.

A key aspect of this definition is that an MFS corresponds to the most landward retreat of the shoreline within a Depositional Sequence of Mitchum et al. (1977). Although Sharland et al. (2001) noted that it commonly coincides with the time of maximum water depth within a basin, this is not always the case. As documented in this paper, in mixed clastic-carbonate systems there are many examples where bathymetry is more complex, maximum water depth is not associated with the MFS, and a regional perspective is required to correctly identify the MFS. This is because the rock record is the product of accommodation space and sediment flux, and is not controlled by relative sea level alone.

Essentially, the bathymetric issues relate to the interaction of carbonate productivity, siliciclastic poisoning, and ongoing steady subsidence. It is generally accepted that carbonate productivity is able to keep up with all but exceptional rates of relative sea level rise (Schlager, 1981). Only when the rate of relative sea level rise is accelerated by exceptional events, such as fault-controlled subsidence or widespread ‘instantaneous’ melting of polar ice caps, should platform drowning become the norm (Jones and Desrochers, 1992). During ‘greenhouse ‘ periods with minimal polar ice caps, such as the Cretaceous, it is much more difficult to invoke polar melting as causing exceptional rises in sea level and cessation of carbonate deposition. It is consistent, therefore, that evidence of drowned platforms replaced by hemipelagic deep-water facies is rare or even unknown in the Early to mid-Cretaceous of the Middle East. However, there are many, well-documented examples on the Arabian Plate where carbonate deposition is replaced by siliciclastics, notably the Zubair and Burgan delta systems. Given the widespread nature of these events, they are difficult to explain by fault-controlled subsidence alone and it is necessary to invoke additional mechanisms.

The phenomenon of carbonate deposition being replaced by relatively shallow-water siliciclastics is commonly termed ‘siliciclastic poisoning’, but it is something of a misnomer as carbonate-secreting organisms can continue to flourish in waters containing a considerable muddy suspension load (Tudhope and Scoffin, 1994). Where replacement occurs it is the result of one of two events.

  • Either the siliciclastics tend to rapidly occupy the accommodation space previously occupied by carbonates, thereby diluting the carbonate content; or

  • The nutrients and freshwater pulses associated with large deltaic systems encourage the proliferation of soft, non-calcareous algae that out-compete, and therefore replace, the previously dominant carbonate-secreting fauna (Hallock and Schlager, 1986).

The non-calcareous algae do not fossilize, and the depositional product is a siliciclastic mudstone. This phenomenon was first identified in the southeastern United States, where the Suwanne Limestone retreats basinward in front of the advancing Paleogene deltaics (McKinney, 1984). It should also apply to the major Lower to mid-Cretaceous delta systems of the Arabian Plate.

In the ‘C’ Sequence of the Jurassic Smackover Formation of the north-central Gulf Coast of the USA, Heydari (1998) identified clear relationships between a high siliciclastic content and a high organic content in laminated mudstones. He ascribed this to a high rate of freshwater discharge and associated increased nutrient supply from a major river system. Plant material of undoubted terrestrial origin is common in shales of the Kazhdumi Formation in Kuh-e-Bangestan in the Zagros fold belt of Iran. It seems probable therefore, that the distribution of organic content in the Kazhdumi Formation is related to nutrient supply and freshwater discharge from the Burgan delta system rather than to condensation associated with relative sea-level maxima or periods of global anoxia. Such an interpretation would be consistent with that of Bordenave and Huc (1995). They suggested that density stratification of a near-surface layer diluted by freshwater runoff from the Arabian Shield overlying more saline oceanic water, was a major factor in the high preservation of organic matter in parts of the Kazhdumi.

Hence, it is conceivable that the distribution of organic matter in prodelta areas may show a greater correspondence with regressions than flooding events. Indeed it might be predicted that flooding events should have led to increased connectivity with the open ocean, and consequently to less-restricted intrashelf settings. An important implication is that the highest gamma-ray signature (if dominantly due to organic and uranium contents) may not equate with the maximum flooding surface. There are clearly opportunities for further research in this area.

MAXIMUM FLOODING SURFACES IN EARLY TO MID-CRETACEOUS DEPOSITIONAL SYSTEMS OF ARABIA

The Early to mid-Cretaceous of the Arabian Plate contains evidence of the advance of several major deltaic systems, notably the Zubair, Burgan/Nahr Umr formations and the Wara sandstones of the northern Arabian Gulf (Figure 3). In more detail, these clastic systems were characterized by extremely rapid progradation, probably driven by hinterland uplifts and plate-wide tilting, producing major unconformities (such as the base of the Lower Burgan Sandstone, Al-Fares et al., 1998, LB25SB of Al-Eidan et al., 2001). This was followed by more gradual, punctuated retreats (i.e. retrogradation) with evidence of estuarine depositional settings dominating relatively thick TST’s in the Zubair and Burgan formations (Nemcsok et al., 1998; Kirby et al., 1998; Al-Eidan et al, 2001; Sharland et al. 2001). Hinterland uplift would seem to be the most viable mechanism to expose the Paleozoic clastics that are the most likely provenance of these major deltaic systems.

The prodelta areas in front (i.e. to the east) of the Zubair and Burgan deltas are represented by thick shale successions of the Zubair/Gadvan and Nahr Umr/Kazhdumi formations, respectively (Figures 3, 9, 10 and 23). Little has been published about these shales and their depositional bathymetry is thus somewhat uncertain and no published regional seismic data are available to address the issue. Al-Fares et al. (1998) used ostracod data to suggest that the prodelta shales in front of the Zubair delta represented water depths of 20 to 40 m. The presence of common nannoplankton also argues for significant water depths, and for some degree of connection between the open ocean and the prodelta area.

Bordenave and Huc (1995) interpreted the Gadvan shales of southwest Khuzestan as having been deposited in water that was slightly deeper than either the up-systems-tract Zubair-Biyadh deltas or the down-systems-tract, shallow-water limestones of northeast Khuzestan. However, they suggested that water depths did not exceed 100 m except in the Garau Basin of Lurestan. Shallow-water limestones were also deposited on the Fars platform. The Gadvan Formation of southwest Khuzestan contains a marginal source-rock facies deposited in a dysoxic environment which is further evidence of somewhat deeper water, possibly of tens of meters. Bordenave and Huc (1995) also inferred some degree of connection to the open ocean. Burrow assemblages in bioturbated shales of the time-equivalent Upper Qishn of Yemen (Putnam et al., 1997) indicated quiet-water, oxygenated environments below storm wave base, which suggests water depths of a few tens of meters.

Immenhauser et al. (1999) argued for fluctuating water depths of up to several tens of meters for the Nahr Umr shales of northern Oman. Al-Eidan et al. (2001) described offshore shales in the Burgan Formation that were bioturbated with an Anconichnus/Chondrites fabric (see Plate 1 of Al-Eidan et al., 2001), suggesting oxygenated bottom conditions in water depths below storm wave base, probably of the order of a few tens of meters. Limestones with discoid orbitolinids, echinoderm debris, Permocalculus algae, and Lenticulina are associated with these shales, which Simmons et al. (2000) suggested indicated similar water depths of a few tens of meters (but see Figure 14). Hence, the evidence suggests that the prodelta shales represent water depths of a few tens of meters. They were significantly deeper than either the deltaic up-systems-tract (Putnam et al., 1997; Kirby et al., 1998; Nemcsok et al., 1998; Al-Eidan et al., 2001) or the shallow, intermittently exposed limestone platforms down-systems-tract (Bordenave and Huc, 1995; Sharland et al., 2001). Unfortunately, there are no published clinoform geometries available on seismic sections that would give an independent check on these interpretations.

In areas where the deltaic successions are most strongly developed, the prodelta shales represent the furthest down-systems-tract facies present. At these locations, MFS must be present within the prodelta shale settings. Examples are K40, K50 and K60 in the Zubair Formation of onshore Kuwait and southern Iraq (Figures 4 and 5; Nemcsok et al., 1998; Sharland et al., 2001); the Qishn/Furt formations of western Yemen (Figures 15 and 16); and K100 in the lower part of the Middle Burgan Member, also in onshore and offshore Kuwait (Figures 18 and 19; Al-Fares et al., 1998; Kirby et al., 1998; Al-Eidan et al., 2001; Sharland et al., 2001). These prodelta shales associated with MFS within clastic-dominated successions can be traced for hundreds of kilometers along depositional strike and for tens of kilometers down-systems-tract. For example, they are almost certainly present in the Zubair Formation of southern Iraq (Figure 5; Ali and Nasser, 1989) and in the Wasia Formation of Saudi Arabia between the Khafji and Safaniya members (Sharief et al., 1989; Grant and Al-Humam, 1995). Understanding their distribution is important at both the reservoir scale, where they are a major control on vertical compartmentalization (see Figure 16 of Al-Eidan et al., 2001), and in regional play fairway analysis where they are likely to be a major controlling factor in hydrocarbon migration and trap effectiveness. Away from the main sources of clastic input, the prodelta shales commonly show down-systems-tract changes to shallow-water limestones. For example, the K100 MFS is in shale in Kuwait (Figures 18 and 19) but is in the Dair Limestone Member of the Wasia/Burgan formations in southern Iraq, offshore Saudi Arabia and offshore Iran (Figures 21 and 22; van Bellen et al., 1959; Entsminger, 1981; Al-Sabti and Al-Bassam, 1993; Grant and Al-Humam, 1995; Denby et al., 2001). Similarly, the K60 MFS is represented by shale in Kuwait but by the Khalij (Limestone) Member of the Gadvan Formation in parts of Khuzestan and offshore Iran (Motiei, 1993; Sharland et al., 2001). Also, Alsharhan and Nairn (1997) noted that whereas the Biyadh Sandstone of eastern Saudi Arabia (Zubair equivalent) contains only two thin limestone members in the Damman and Qatif fields, it becomes mostly limestone toward Qatar and the United Arab Emirates. These relationships are well illustrated in the regional stratigraphic cross-sections presented by Grant and Al-Humam (1995; their figure 2) and Sharland et al. (2001; their figure 3.28).

Thus, in many cases, it seems that large coeval carbonate platforms were present down-systems-tract from the major mid-Cretaceous delta systems and that deeper-water prodelta areas, at times sites of source-rock development, were located between the shoreline clastic systems and the carbonate platforms (Figure 27). The prodelta areas do not seem to have been particularly deep (see discussion above) and, in the absence of siliciclastics, should have been sites of carbonate deposition. Hence, it is interpreted that the suppression of carbonate production in the prodelta areas was due to the siliciclastic input from major delta systems. Nutrient excess and freshwater run-off from the deltas themselves may have played an important role in source-rock development as well as causing reduced carbonate deposition (Hallock and Schlager, 1986). During delta advance as a direct consequence of increased clastic supply, the prodelta area would have migrated basinward accompanied by retreat and shrinkage of the down-systems-tract carbonate platforms, though the location of some of these platforms may have resulted from an initial underlying structural control. Conversely, during periods of rising relative sea level and/or reduced clastic sediment supply, the clastic systems were in retreat and carbonate platforms expanded by landward progradation across the prodelta areas (Figure 28). For this to happen, the prodelta areas would need to be of rather limited bathymetry, as too deep a setting would have precluded landward expansion of the outboard carbonate platforms. Water depths of 20 to 40 m proposed for the prodelta shales of the Zubair system by Al-Fares et al. (1998) should have facilitated the landward carbonate expansion associated with retreat of the delta system during the TST.

During the most regionally extensive flooding events, notably the Shu’aiba and Mauddud of the northern Gulf, the area of carbonate deposition expanded landward almost to the exposed Arabian Shield, their maximum extents being obscured by subsequent erosional truncation. This mechanism explains diachronous depositional relationships between, for example, the Upper Burgan and the overlying Mauddud in the K110 TST (Figure 19), and the upper part of the Zubair to the overlying Shu’aiba in the K70 TST (Figures 9 and 10). It also predicts that the area of suppressed carbonate production in the deeper-water prodelta area migrated back and forth between the delta system and the down-systems-tract carbonate platform (Figures 27 and 28). It was not a fixed facies belt as implied by the paleogeographic maps of Murris (1980) and Moshrif and Kelling (1984). A new term ‘Migratory Carbonate Suppressed Belt’ (MCSB) is proposed for these migrating prodelta areas.

The model predicts that down-systems-tract carbonate platforms should continue to expand landward for the duration of the TST while the delta systems were retreating (Figure 28). This would be due either to a simple eustatic rise in sea level or to reduced sediment supply resulting from gradual denudation of the uplifted hinterland (Sharland et al., 2001). Hence, at the time of maximum retreat of the shoreline, that is of the MFS itself, the carbonate platform had expanded substantially. In consequence, the MFS must be located in the basal parts of carbonate successions that expanded back across the MCSBs (Figure 28). These carbonates are interpreted as representing shallower-water settings than the prodelta shales in the MCSBs. Thus, where a carbonate platform succession conformably overlies a prodelta succession, the MFS do not lie in the deepest-water sediments in the succession (Figure 28). It is worth noting that this is somewhat contrary to conventional models of carbonate sequence stratigraphy (e.g. Sarg, 1988; Handford and Loucks, 1993). In the case of the Lower to mid-Cretaceous systems of the Arabian Plate, the high productivity of carbonate sediments must have been more efficient at filling the developing accommodation space than the previous laterally sourced clastic delta systems. This made it possible, even though at first sight it might seem illogical, for the whole system to shallow-up during major transgressions.

There are also hints in the literature of an underlying structural influence on the distribution of the major clastic systems. For example, many authors have placed the intra-Aptian unconformity at the lithological boundary between the Shu’aiba limestones and the Burgan or Nahr Umr clastics (e.g Alsharhan and Nairn, 1997). In contrast, Al-Fares et al. (1998), Kirby et al. (1998) and Al-Eidan et al. (2001) all depicted this major unconformity to be within the lower part of the Burgan Formation and not at the top of the limestone. Al-Fares et al. presented biostratigraphic data that the basal Burgan clastics are of Aptian age, significantly older than the remainder of the Burgan Formation and denote this section as the “Un-named Clastics”. Their isopach map of the “Un-named Clastics” clearly indicates that they correlated this lithological unit from offshore Kuwait into the Greater Burgan, Raudhatain and Sabiriyah fields, thereby supporting the independent interpretations of Kirby et al. and Al-Eidan et al. It is significant that Al-Ateeqi and Foster (2001) used 3-D seismic data to interpret the top Shu’aiba surface in the area surrounding the Minagish and Umm Gudair fields in southern Kuwait to be a major karstified horizon. This implies, but does not prove, that the intra-Aptian unconformity in this area was coincident with the top of the Shu’aiba Formation. It suggests that southern Kuwait could have been a high that was bypassed by highstand deposition of the “Un-named Clastics” of northern Kuwait, although well data is needed to check this hypothesis. An underlying structural control seems the most tenable explanation of these observations. The distribution of the karst is economically important, not for exploration purposes, but because the karstic interval is being considered for the disposal of produced water in southern Kuwait (Al-Ateeqi and Foster, 2001).

DISCUSSIONS AND IMPLICATIONS

The observations and model presented in this paper raise a number of important issues, some of which have major economic significance. It has been argued that the deepest-water facies in a mixed carbonate-clastic succession does not necessarily represent a MFS, but relates to the delicate balance between siliciclastic input, carbonate productivity, eustatic sea level, and subsidence. Evidence has also been presented that some of the major boundaries between formations are markedly diachronous on the regional scale.

Sharp lithological boundaries between limestones (below) and deeper-water shales (above) do not necessarily mean that depositional systems were retreating shoreward. This is even though in most respects these surfaces match the definition of marine flooding surfaces as a stratigraphic surface separating older from younger strata across which there is evidence of an abrupt increase in water depth (after van Wagoner et al., 1990). In some cases, the surfaces record rapid advance of a siliciclastic system and associated suppression of carbonate deposition, to form a deeper-water prodelta area located between a delta complex and a shallower-water, outboard carbonate platform. Reworking of sediment during the subsequent transgression may have obscured the original regressive nature of these surfaces. A key point is that any time-equivalent carbonate platform to the shales overlying the surface is located basinward, not landward of the shale. This is a factor that might not be predicted by routine interpretation of the vertical sequence seen in a single well or outcrop section. Hence, conventional interpretation of the vertical succession in an area containing a mixed carbonate-clastic succession can be a poor indicator of the regional paleogeography.

Only a combination of detailed subsurface and outcrop studies allied to a plate-wide sequence stratigraphic understanding can resolve such issues. This has major implications for the prediction of regional facies patterns, in particular the extent of potential major seal units, and also the location and distribution of potential source-rock prodelta shales, commonly interpreted as occupying intrashelf basins. The diachronous nature of contacts may set up stratigraphic traps. Although these are likely to be small and economically unattractive for large corporations, they may prove to be commercially important for smaller companies.

It should also be noted that not all inputs of siliciclastics causes such retreats of the outboard carbonate systems. There are well-documented examples on the Arabian Plate, such as in the Garagu Formation of northern Iraq (equivalent to the Yamama), where the depositional system is truly mixed-carbonate/clastic (van Bellen et al., 1959), rather than an alternation of carbonates and clastics. In such cases, it is likely that the siliciclastics were derived locally and were mixed into the carbonates, probably by transporting the coarse fraction along the coast by longshore drift. In this case, the total amount of siliciclastics involved is small when compared to the enormous volumes of sand in the Zubair and Burgan delta systems. In addition, many of the Late Cretaceous and Tertiary foreland basin carbonates (Aqra, Bekhme, Sinjar, Avanah, Asmari formations and the Kirkuk Group) survived in front of significant input of molasse and flysch derived from an ophiolitic/carbonate provenance. It is possible that clastic bypassing of the carbonate shelf was responsible for the co-existence of these two sediment systems in this setting. The tenacity of these carbonate systems in such a foreland basin setting is important in that they contain a significant proportion of the hydrocarbon reserves of the Middle East.

Furthermore, in other areas of the world, for example the Mahakam delta in Kalimantan, Indonesia, carbonates co-exist with deltaic clastics with no ‘zone of suppression’ developed between the two systems, although the carbonates are developed only on abandoned or inactive delta lobes (Bishop, 1980; Nuay et al., 1985). Similarly, smaller-sized fan deltas in the Red Sea often have carbonate reefs developed in their shallow-marine parts. In these, the carbonates exist because for 99 percent of the time there is no significant clastic input until wadi-flooding results in the carbonate system being overwhelmed by pulses of clastic material (Hayward, 1982; Friedman, 1988). Also, in the Permian Basin of the USA and the Dinantian of Anglesey, North Wales, UK, clastic shoreline facies are found associated with carbonate shelf facies. However, in these cases, no basins have developed between the clastics and the carbonates (Melim and Scholle, 1995; Walkden and Davies, 1983). In both examples, fluvial to eolian clastics overlie paleokarsts that had developed on the shelf carbonates during lowstands. Hence, the periods of clastic and carbonate deposition seem to be temporally distinct, which is in contrast to the Lower to mid-Cretaceous examples of the Arabian Plate. Whereas this may reflect a real distinction between sediment patterns in ice-house and greenhouse periods, a simpler explanation would be that the larger scale of the Arabian Plate depositional systems permitted the operation of contemporaneous carbonate and clastic systems.

Other examples where clastic entry did result in retreat of the carbonate system, similar to that observed in the Lower to mid-Cretaceous of the Arabian Plate, appear to be related to the development of large, point-sourced clastic deltas, particularly if they prograded into relatively shallow water. These include the Suwanne Limestone of Oligocene age in the southeast USA (McKinney, 1984), the Yoredale clastic-carbonate cyclicity of the late Dinantian of Northern England (Moore, 1958; Pickard, 1994), as well as Pennsylvanian cycles of the cratonic interior of the USA (Heckel, 1984; Goldhammer et al., 1991). The progressive drowning of the Luconia pinnacles of Neogene age on the northwestern coast of Sarawak (Epting, 1980; 1989); the advance of the paleo-Pearl River delta across the Liuhua area of China (Tyrrel and Christian, 1992); and the retreat of Oligocene carbonate platforms in Kalimantan, Indonesia (Saller et al., 1993), all appear to have occurred in what were generally deeper-water settings.

One issue that remains to be examined in more detail is the faunal changes that accompanied the advance of the siliciclastic systems. Many of the carbonates in the siliciclastics show fully open-marine benthic biotas, typically dominated by organisms such as sponges and benthic larger foraminifera. These cannot be explained as drifted-in biota that had become trapped in the deeper-water ‘lagoons’ (although such planktonics and epiphytic organisms are commonly present). They imply that although deeper-water conditions prevailed, the salinity and light penetration must have been reasonably similar to open-marine shelf environments. Similarly the presence of evaporites in the Albian Jawan Formation of northern Iraq (van Bellen et al., 1959) remains to be explained.

Another issue is the extent to which some of these deeper-water areas may have had an intrashelf-basin formation component related to restriction and carbonate suppression within the carbonate shelf area, which would have been completely independent of siliciclastic input. We consider that true intrashelf basins, for example those in the Jurassic platform systems of Arabia, are more likely to be characterized by carbonate sediments that contain restricted biota, whereas carbonate sediments in MCSBs are more likely to contain predominantly open-marine biotas. Similarly, MFS in true intrashelf basins correlate with the deepest water depths, but those in MCSBs do not. However, disentangling the processes of restriction caused by delta advance and restriction caused by the building of major platform-margin barrier systems, is not always easy. Clearly, much more work is needed before the Lower to mid-Cretaceous stratigraphy of the Arabian Plate is fully understood.

CONCLUSIONS

The main conclusions of this study are:

  1. A regional picture extending well outside field, license area, and even country boundaries is required, especially in a down-systems-tract direction, for the confident sequence stratigraphic understanding of the mixed carbonate-clastic shelfal intervals in the mid-Cretaceous of the Arabian Plate. We propose some minor amendments to the MFS identified by Sharland et al. (2001) and some additional examples.

  2. The mid-Cretaceous systems tracts commonly show a tripartite arrangement of proximal coarsegrained deltaic clastics, a deeper-water mud-prone prodelta area, and an outboard carbonate platform area. The mud-prone prodelta areas clearly migrated back and forth across the Arabian Plate in response to changes in sediment supply, carbonate productivity, and relative sea level. We propose a new term, ‘Migratory Carbonate Suppressed Belt’ (MCSB) for these migrating prodelta areas. Our model predicts that down-systems-tract carbonate platforms should continue to expand landward for the duration of the Transgressive System Tract (TST) while the delta systems are retreating, whether this is due to simple eustatic sea level rise or reduced sediment supply due to a gradual denudation of the uplifted hinterland. In successions dominated by coarse-grained clastics, MCSB shales represent the most distal (down-systems-tract) facies and contain the MFS. Examples are K40, K50, and K60 in the Zubair/Biyadh/Qishn of Kuwait, Saudi Arabia, and Yemen, and K100 in the Burgan/Wasia of Kuwait and Saudi Arabia.

  3. During the most extensive flooding events, the area of carbonate deposition expanded landward almost to the outcrop of the Arabian Shield. Depositional packages consisting of alternating units of limestone and clastics represent retrogradational TST. Low-order/low-frequency MFS are located within the bases of clean limestones overlying these mixed carbonate-clastic packages. Good examples are the K70 and K110 MFS in the Shu’aiba and Mauddud limestones of the northern Gulf and Qatar. Thick, clean limestone sections are predominantly the product of highstand system tracts.

  4. MFS are not always located within the deepest-water sediments within a single vertical succession. Furthermore, they should not be picked simply in the sediments with the highest gamma-ray response.

  5. Sharp lithological boundaries between limestone formations (below) and deeper-water shales (above) do not necessarily indicate transgressions. In some cases, the surface records the rapid advance of a siliciclastic system and the associated suppression of carbonate deposition to form a deeper-water prodelta area located between a clastic delta complex and a shallower-water, outboard carbonate platform. A key point is that any time-equivalent carbonate platform to the shales overlying the surface is located basinward, not landward, a factor that might not be predicted by simplistic interpretation of the vertical succession seen in a single well or outcrop section.

  6. These issues are of major economic importance at both the reservoir scale, for instance in the distribution of vertical permeability barriers, and at the regional-play fairway scale in the distribution of seals and the location of the potential source-rock ‘intrashelf basins’. The interpretations presented here are therefore relevant to the prediction of reservoir, source, and seal distributions at all scales from detailed reservoir modeling to regional exploration programs. In the longer term they may lead to the identification of stratigraphic traps.

ACKNOWLEDGMENTS

The authors thank Henk Droste, Pieter Spaak, and Peter Osterloff of Petroleum Development Oman for sharing their understanding of the Wadi Mu’aydin section, and Frans van Buchem of Institut Française du Pétrol for discussions on the enigmatic ‘Hawar Shale’, though the interpretations presented are solely the responsibility of the authors. Denis Vaslet and Wyn Hughes provided valuable detail on outcrop sections and the Shaybah field in Saudi Arabia. The thorough reviews by Frans van Buchem and an anonymous reviewer raised many further questions that helped to clarify our ideas. We are particularly grateful to Frans van Buchem, Bernard Pittet and their fellow workers for the free exchange of ideas. Finally, we thank the staff of Gulf Petrolink, particularly David Grainger for editing the paper, and Nino Buhay and Napoleon Afuang for their assistance in designing and drafting the figures.

ABOUT THE AUTHORS

Roger Davies is Director Field Evaluation, Neftex Petroleum Consultants Ltd., and is a co-author of Arabian Plate Sequence Stratigraphy. He has a PhD in carbonate sedimentology from Southampton University. Roger has 22 years oil industry experience, having worked on regional exploration, field appraisal, and development projects worldwide for BP and as an independent consultant. Projects for BP included reservoir characterization offshore and onshore U.A.E., and a major regional study of the Khuff Formation. His current interests include the application of sedimentology and sequence stratigraphy to improve reservoir description, Cretaceous and Jurassic reservoirs of the Arabian Plate, especially Kuwait and Iran, and Permo-Triassic carbonates plate-wide.

roger.davies@neftex.com

David (Dave) Casey is a Geologist with Neftex Petroleum Consultants Ltd., and is a co-author of Arabian Plate Sequence Stratigraphy. He has a BSc in Geology and an MSc in Hydrogeology from Reading University, and a PhD in Geology from Oxford University. Dave has 17 years oil industry experience, mostly in the Middle East and Caspian region, having worked for 11 years with BP Exploration and subsequently as an independent consultant. His current research interests include the tectonic and sequence stratigraphic development of the Arabian Plate petroleum systems.

dave.casey@nefltex.com

Andrew Horbury is a Carbonate Geologist with Cambridge Carbonates Ltd., which he co-founded in 1992 after spending six years working in the Middle East with BP Exploration. He has a BSc from Bristol University and a PhD in Carbonate Sedimentology from Manchester University. Andrew is a co-author of Arabian Plate Sequence Stratigraphy. His present interests include the sequence stratigraphic evolution, sedimentology and reservoir geology of northern Arabia, complex fractured carbonate and dolomite reservoirs in southern Italy, Mesozoic play systems and reservoirs in Mexico, and the development of an image-analysis system for characterizing porosity systems and diagenetic textures in carbonates.

a.horbury@hotmail.com

Peter Sharland is Managing Director of Neftex Petroleum Consultants Ltd., and was lead author of Arabian Plate Sequence Stratigraphy. He received his BSc in Geology from London University in 1983 and has 19 years of international oil industry experience, having worked for LL&E Inc., BP Exploration, and latterly as Head of Subsurface for the Middle East and Caspian Business Unit in LASMO. Peter is a member of the Editorial Advisory Board for GeoArabia. His professional interests include the construction of sequence stratigraphic models to better understand and constrain risk.

peter.sharland@neftex.com

Michael (Mike) Simmons is Managing Director and Chief Geologist of CASP, Cambridge University, UK. He has a BSc and PhD from Plymouth University. He was formerly the Head of the Department of Geology and Petroleum Geology at the University of Aberdeen. He previously spent 11 years with BP Exploration working as a Senior Geologist/Biostratigrapher specializing in the Middle East and Former Soviet Union. Mike is a coauthor of Arabian Plate Sequence Stratigraphy. His research interests include applied biostratigraphy, the geology of the Tethyan region, and the use of outcrop analogs in understanding subsurface reservoirs.

mike.simmons@casp.ac.uk