Ediacaran–Cambrian Sirab Formation of the Al Huqf region, Sultanate of Oman

The Ara Group represents a thick package of at least six carbonate-evaporite cycles deposited in the latest Ediacaran to early Cambrian in the South Oman Salt Basin (SOSB) in the Sultanate of Oman (Gorin et al., 1982; Hughes Clarke, 1988; Amthor et al., 2003, 2005; Figure 1 and Enclosure I-1). The cycles are informally referred to as A0 to A6 and are subdivided into carbonate ‘C’, or evaporite ‘E’ units, for instance A4C and A4E. The evaporite units consist of ca. 10–20 m-thick gypsum and anhydrite intervals associated with halite deposits. The salt was probably in the order of 10s to 100s of metres thick prior to subsequent halokinesis (Schröder et al., 2003). The evaporites accumulated during periods of basinal lowstand with restricted circulation and marine replenishment. Carbonate cycles occur as isolated platforms known as ‘stringers’, varying from low-gradient ramps to rimmed shelves. These formed during transgression to highstand in the basins, which allowed at least periodic connection with the open ocean (Grotzinger and Amthor, 2002; Amthor et al., 2003; Schröder et al., 2004, 2005). Typical shallow ‘stringer’ carbonate facies consist of above-fair-weather wave-base ooidal grainstones and oncolite packstones with microbial barrier build-ups of laminated stromatolites and thrombolites, with a characteristic ‘clotted’ fabric (Grotzinger and Amthor, 2002; Amthor et al., 2003). High primary porosity in intervals of microbial thrombolite framestone means that together with the Ara salt seal, the ‘stringers’ form a hydrocarbon play system which is one of the oldest in the world, and provides some of the most important hydrocarbon reserves in Oman (Al-Siyabi, 2005).

The Ara Group and its facies were defined in the SOSB using subsurface exploration wells. In the SOSB the group overlies the Buah Formation of the Nafun Group and is overlain by the Nimr Group (Enclosure I-1 inset; Droste, 1997; Forbes et al., 2010). With the SOSB buried to a depth of over 2 km under the desert in southern Oman, the only likely actual exposures of Ara carbonates are fragments contained in salt-piercing domes in central Oman (Peters et al., 2003). As such, Ara Group lithostratigraphic nomenclature should not be applied to any surface units unless they can be demonstrated to be identical in facies and age to the Ara. However, given the importance of hydrocarbon play systems within the Ara Group, then clearly the recognition of an outcrop equivalent could yield critical information in the future on the evolution of Ara-type facies architecture and geometry in potential source, reservoir and seal units both spatially and through time.

In the Al Jabal al-Akhdar of the Oman Mountains, the Buah Formation is overlain by about 600 m of carbonate olistostromes, volcaniclastics and siliciclastics known as the Fara Formation (Figure 1) [UTM 547358, 2575519]. Following work by Rabu (1988), McCarron (2000) undertook reconnaissance work on the Fara Formation sampling a volcanic interval, which subsequently yielded a U-Pb age date of 544 ± 3.3 Ma (Brasier et al., 2000). This confirmed that the Fara Formation straddles the Ediacaran/Cambrian boundary and is a chronostratigraphic equivalent of the subsurface Ara Group. However, the lithofacies of the Ara Group in the SOSB little resemble those of the Fara. This difference might be expected given that the onset of Ara deposition is marked across Oman by a period of tectonic activity and subdivision into salt basin depocentres, plus the fact that the Fara Formation lies some 400 km north of the SOSB.

A more promising area for potential equivalents of Ara Group facies to be exposed is the Haushi-Huqf region of east central Oman as it is only ca. 130 km northeast of the present-day edge to the SOSB (Enclosure I-1 inset). However, the geological survey of the Al Huqf by the Bureau de Recherches Géologiques et Minières (BRGM) (Dubreuilh et al., 1992; Platel et al., 1992) was not originally conducted at a scale which would recognise an Ara equivalent; thus on published geological maps the Buah Formation is shown at the top of the Huqf Supergroup, truncated by the regionally extensive Angudan Unconformity, at the base of the Haima Supergroup, and overlain by the Thumaylah Formation (now referred to as the Amin and Miqrat formations in Petroleum Development Oman – PDO – nomenclature, Forbes et al., 2010).

Research on the Buah Formation in the Al Huqf during the late 1990s raised interest in the depositional environments of the upper Buah. In a detailed field study of the Nafun Group as a whole, McCarron (2000) summarised and rationalised previous work on internal facies and subdivisions within the Buah Formation. She recognised that in the Al Huqf it appeared to be comprised of two members. The lower member is composed of dolostones passing up into stromatolites and grainstones, and is generally uniform in thickness across the region. The upper member consists of peritidal - sabkha cycles. Each member typically forms a prominent topographic ridge above the otherwise generally low-lying topography.

In addition to the recognition of evaporitic units at the top of the Buah, as it was defined at that time, unpublished internal PDO reports in the mid 1990s identified sandstones within the top of the Buah Formation as mapped by the BRGM near the village of Sirab [UTM 582912, 2234881]. This raised the possibility that these thin red sandstone beds might be equivalent to later Ara deposition in the SOSB where there is a terrigenous clastic input (subsurface Ara Cycle A6). Subsequently, field surveying was undertaken by the first author between 1999 and 2001 to establish whether potential Ara equivalents existed in outcrop. The unpublished PDO report from this work suggested for the first time that a distinct stratigraphic unit could be recognised lying between the Buah grainstone ridges and the Nimr conglomerates or Amin sandstones in exposures around the Buah Dome of the Al Huqf (Enclosure I-1). However, until now the only published record of these findings has remained that of Nicholas and Brasier (2000).

Subsequent studies on the Buah Formation in the Al Huqf by Cozzi and Al-Siyabi (2004) and Cozzi et al. (2004) confirmed that McCarron’s two members of the Buah Formation could be recognised in the Al Huqf. In contrast to their works (A. Cozzi, 2009, written communication) the present paper follows Nicholas and Brasier (2000), in which the top of the Buah Formation is defined at the top of McCarron’s lower member, beneath the first peritidal-evaporite unit. In effect this lower member is now the Buah Formation sensu stricto, with a thickness of ca. 130–190 m below potential Ara equivalents. This redefinition also brought the lithofacies of the Al Huqf in line with the laterally persistent facies of the entire Buah thickness subsurface and in the Al Jabal al-Akhdar to the north. In this scheme there are now no recorded occurrences of evaporite beds in the Buah Formation.

Here we present results from the original unpublished survey of potential Ara equivalents conducted between 1999 and 2001, integrated with new field data from a recommencement of surveying in 2006. A re-appraisal of former sections combined with fresh localities now provides a robust regional lithostratigraphic framework for recognition of a new unit at the top of the Huqf Supergroup in the Al Huqf, which we define here as the Sirab Formation (Enclosure I-1). As such, it occupies the corresponding litho-stratigraphic position of the subsurface Ara Group; overlying the Buah Formation but underlying the middle Cambrian Angudan Unconformity, the basal regional unconformity of the Haima Supergroup clastics.

False-coloured Landsat and ‘Quickbird’ satellite imagery was provided for interpretation by PDO. All large-scale geological maps presented here have a superimposed Universal Trans-Mercator (UTM) grid using the WGS 84 datum and are based on our own field surveying between 1999–2001 and 2006–2008. Thus, in previous sections and those which follow, all grid references quoted in square [ ] brackets refer to UTM square 40Q using WGS 84, and have an estimated position error of ≤ 5 m. All stratigraphic sections were logged by us using a laser Abney Level and Jacob’s Staff to ensure accurate true stratigraphic thicknesses were measured. Lithological colour variations described in sedimentary logs were made using the Geological Society of America rock colour chart (based on Munsell Soil colours). All field photographs illustrated here were taken by C. J. Nicholas.

This study was a dominantly field-based survey and initial description of the Sirab Formation, as a precursor to more detailed sedimentological and petrographic research in the future. Consequently, descriptions concentrate on field observations and textures in outcrop. In order to formally define a new formation, a lithostratigraphic approach was taken at the outset of surveying. However, within these constraints, a distinct suite of lithofacies were identified with their lithofacies associations (Table 1).

The Sirab Formation proposed here represents a discrete sedimentary package consisting of lagoonal to peritidal carbonates, sabkhas and associated evaporites, with a component of continental-derived clastics. As such, it recognisably differs from the grainstone-dominated inner- and mid-ramp facies of the underlying Buah Formation and the ferruginous fluvial conglomerates and sandstones of the overlying Haima Supergroup (Figure 2). The base of the Sirab Formation can be observed in the majority of localities to be conformable upon the underlying Buah Formation. However, in December 2007 we discovered new sections at Wadi Salutiyyat and Wadi Shuram, which demonstrate an angular unconformity between the lower Sirab Formation and either the Buah or Shuram formations underneath (Enclosure I-1). At its upper contact, the Sirab Formation is separated from the overlying Haima Supergroup by a regional angular unconformity, which incises down into the Sirab Formation at many localities. As such, and because the Sirab Formation remains carbonate-dominated, rather than clastic-dominated like the overlying Haima, we suggest that this new formation is incorporated into the upper part of the Huqf Supergroup lithostratigraphic scheme for outcrop (Figures 1 and 2).

We divide the Sirab Formation into three principal members; a lower Ramayli Member, a middle Shital Member and an upper Aswad Member (Figure 2). A fourth member, the Salutiyyat, can be recognised where the Sirab Formation lithostratigraphic section below the easily recognisable Shital Member peritidal carbonate cycles is condensed and rests with angular discordance upon eroded remnants of Buah Formation or Shuram Formation as at Wadi Salutiyyat itself. At these localities, a reddened basal conglomerate marks the contact with the underlying Nafun Group. Across the region, the Shital Member appears to disconformably overlie the Ramayli Member and this significant bounding surface is the most easily traced boundary throughout the entire Sirab Formation outcrop. Higher in the lithostratigraphy, the youngest Aswad Member is unfortunately rarely seen in contact with the underlying Shital Member. However, at Wadi Aswad, fault slivers expose fragments of this upper Sirab stratigraphy with the Aswad Member apparently conformably overlying tufted mats and conophyton beds of the upper Shital Member (Enclosure I-1).

Minimum thicknesses measured for the principal three members across the region were 45.3 m for the Ramayli Member, 149 m for the Shital Member and 10.75 m for the Aswad Member. This indicates that the Sirab Formation is at least 205 m thick. However, this does not take into account the many sections observed throughout the Al Huqf, which are considerably thicker but we have not included in this study as they are fault bounded, leaving their precise lithostratigraphic position within members unclear at present. High-resolution, stable-isotope chemostratigraphy is currently underway in order to incorporate such sections in the future into the Sirab Formation stratigraphic framework. Thus a more realistic general estimate would have the Shital Member with a thickness of over 250 m and the Sirab Formation well over 300 m thick.

The Arabic word ‘Sirab’ (or alternatively spelled ‘Sarab’) loosely means ‘no life here’. As a geographic name, it is used locally in the central Al Huqf region to include the desert from Wadi Shital just west of the Al Maha petrol station on the main road, eastwards and southwards to the sprawling village of Sirab by the coast, and is labelled on some Tourist maps of Oman (Enclosures I-1 and I-2). Prior to construction of the tarmac road and filling station, a small petrol dump was located at the turn-off to Sirab [UTM 576098, 2233244]. This is close to the base of the most complete single stratigraphic section through the formation we define here [UTM 578843, 2231645]. Sirab is the only area so far discovered in which all three members are exposed in close proximity and with a minimum of fault disturbance. Hence we take the name ‘Sirab’ for the formation, and the region of desert that it covers is considered as the type area (Enclosures I-1 and I-2), with a composite stratotype section consisting of the type locality of Wadi Shital ST-1 combined with the short upper section at ST-2 (Enclosure I-2 and Figure 3). The meaning of the name ‘Sirab’ also highlights the current lack of any trace or body fossils within the formation.

The age of the subsurface Ara Group is now well constrained (Amthor et al., 2003; Bowring et al., 2007). U-Pb zircon geochronology, carbon-isotope chemostratigraphy and biostratigraphy together indicate that the Ara Group spans the latest Ediacaran to earliest Cambrian, with the Ediacaran/Cambrian boundary at or close to the base of the A4 Cycle which is dated at 542 ± 0.6 Ma (Amthor et el., 2003) (Figure 1). Ash from the underlying A3C carbonate has also yielded an age date of 542.6 ± 0.3 Ma (Amthor et al., 2003). The presence of the earliest calcareous tubular fossil Cloudina and also possible Namacalathus in the Ara below A4 also supports the conclusion that the lower units of the Ara Group are terminal Neoproterozoic in age. The top of the underlying Buah Formation sensu stricto both subsurface and in outcrop has therefore been estimated to be of ca. 550 Ma (Cozzi and Al-Siyabi, 2004).

Cloudina and Namacalathus are intimately associated with thrombolite framestones in the Ara carbonate ‘stringers’ subsurface and also in exposed analogous Ediacaran reefs in Namibia (Schröder, 2000; McCormick and Grotzinger, 2001; Grotzinger et al., 2005; Dibenedetto and Grotzinger, 2005). Although we have discovered clotted thrombolite framestones and stromatolitic thrombolites in the upper Shital Member and Aswad Member extensively across the Al Huqf area, we have so far not observed any body or trace fossils in the Sirab Formation. The absence of Cloudina in the thrombolitic upper units of the Sirab Formation could indicate that they are of Cambrian or younger age as Cloudina became extinct at the Ediacaran/Cambrian boundary in Oman (Amthor et al., 2003). Alternatively, they may be present and we have simply not yet observed them, or the restricted depositional environment with limited water exchange and evaporation excluded Cloudina and Namacalathus from the environment.

Prior to this study the occurrence of thrombolites in Oman was only known in the salt-piercing domes in central Oman (Peters et al., 2003), and the subsurface Ara Group in the SOSB. This in itself may suggest an age-equivalence between the Ara Group and the upper Sirab members. However, strictly speaking, thrombolites are not age-determinant in themselves, having a stratigraphic age range from Neoproterozoic to Phanerozoic. They are particularly common in the Palaeozoic (Kennard and James, 1986) and also occur in the Permian–Triassic Saiq Formation of the Saiq Plateau in Oman (Koehrer et al., 2010).

The age of Haima Supergroup units exposed in the Al Huqf remains unresolved. It has been reported to be constrained by a rare occurrence of upper Cambrian trilobites in the Al Bashair and lower Barik Sandstone formations of the Andam Group (Fortey, 1995; Droste, 1997; Forbes et al., 2010) (Figure 1). The unfossiliferous Haima units below the Andam Formation could therefore be of any age from early Cambrian to earliest late Cambrian. Droste (1997) and Al-Husseini (2010) correlated the sandstones of the Amin Formation with the Lalun Sandstone sequence of Iran, which is overlain by marine middle Cambrian beds. If this were the case, then the basal Nimr Group would be of an early Cambrian age. The regional Angudan Unconformity is indicated by Droste (1997) to be between the Nimr Group and Amin Formation, as the former are clearly truncated and onlapped by the latter subsurface. In exposures around the Buah Dome in the Al Huqf, this major angular unconformity can clearly be seen to underlie the conglomerates and this suggests that there is no unequivocal Nimr Group exposed in the Al Huqf.

The age estimates for the top of the Buah Formation sensu stricto of ca. 550 Ma, and base of the Haima Supergroup in early Cambrian at ca. 530–520 Ma, constrain the age of the Sirab Formation to a ca. 20–30 million years (My) interval spanning the late Ediacaran to early Cambrian. Whilst we cannot be more precise concerning age dates within the Sirab Formation at present, and accepting that we have identified at least one disconformity or unconformity within the succession separating the Shital Member from units below, it is still a distinct possibility that the Al Huqf area contains exposed sections across the Ediacaran/Cambrian boundary. Detailed stable-isotope chemostratigraphy might be expected to better constrain the timing of Sirab Formation deposition. However, attempts at regional or global correlation of carbon isotope excursions has also remained inconclusive (Gold, 2010).

The base of the Sirab Formation (Buah Formation/Ramayli Member boundary) is well exposed at the stratotype section at Wadi Shital (Enclosure I-2 and Figure 3). However, this lower interval of the formation, the Ramayli Member, is actually best exposed at a series of sections exposed around the western flanks of the Buah Dome faulted pericline (Enclosures I-3, I-4 and I-5). The Buah Dome structure consists of an inner zone with prominent Khufai Formation dolostone ridges dipping out away from the core of the fold, and an outer zone on the western flank where a single prominent ridge forms the only real topography above the desert (Enclosures I-3 and I-4). This ridge is composed of the dolostone-dominant ooidal/peloidal grainstone and interbedded silts of the upper Shuram Formation transitional facies (McCarron, 2000) and the entire thickness of Buah Formation sensu stricto as defined above (Enclosures I-3 and I-4).

At the stratotype section of the Ramayli Member at B-NW3, the upper beds of the Buah Formation dip ca. 30° to the west-northwest off the main topographic ridge (Enclosures I-3 and I-5; Figure 4) [UTM 569371, 2253038]. They consist of cross-bedded medium dolomitic grainstones of inner to mid-ramp tidal shoals and form massive weathering, low, rounded exposures at about waist height. As the topography dies away off the grainstone ridge to the west-northwest, there is a lithofacies change to finer-grained dolostones, with a mud component (Figures 4a and 4b). The exposures become less massive and more irregular. This change to packstones and wackestones suggests a seaward-stepping pattern with deposition of facies from closer to the palaeo-shoreline, and from more shallow lagoonal environments. Pseudomorphed gypsum laths are common in some horizons (Figure 4c), overlain by at least one hardground and this is followed by a switch to calcarenites with a dispersed detrital quartz component.

There is evidence at B-NW3 of fissures, breccias and infilled karst hollows around this interval (Enclosure I-5). We originally suspected that there may be a karst surface marking the top of the Buah Formation. However, the field evidence now seems more complex and fragmentary. This zone seems to expose an interplay between in situ brecciation and partial dissolution from fault fracturing and fluid movement up from the underlying Shuram Formation (an original suggestion by Joerg Mattner, which we concur with), and sporadic karst hollows penetrating down from the highly angular Angudan Unconformity surface above. Haematite cementation seems to be a particular feature of the Amin at the base of the Haima, and haematite crusts can be found draping the surface of these suspected karst hollows. This karst penetration down into the Huqf Supergroup can be observed at several localities on a larger scale around the outer Buah Dome (Enclosure I-3). Therefore, we now consider the Buah/Ramayli boundary to be conformable and transitional.

The bounding surface at the base of the Sirab Formation could be defined at either of two lithostratigraphic changes. It could be taken at the level at which clear shallowing comes-on up-section, indicated by a change from grainstones to either packstones or wackestones. This shallowing trend does not seem to be significantly reversed at any point through the Ramayli Member and is therefore a correlatable surface. Secondly, the boundary could be defined by the first appearance of dispersed pseudomorphed evaporitic minerals; either gypsum or anhydrite.

At the stratotype section for the Shital Member in the type area at Sirab, the Buah/Ramayli boundary is exposed at the base of section ST-1 close to the tarmac road (Enclosure I-2 and Figure 5a). The transition is similar to that at B-NW3, but helps resolve where to place the base of the Sirab Formation. Again, cross-bedded grainstone beds of the upper Buah Formation (Figure 5b) are overlain by low-lying, weathered packstones. However, here the change is across a single bedding surface which is marked by scattered partly chert replaced dolomicrite rip-up clasts (Figure 5c). As such it may represent a minor non-sequence. For ca. 9 m below this surface, scattered pseudomorphed gypsum laths and anhydrite rosettes can be observed in the Buah grainstones. This suggests restriction of the inner ramp and lagoon area had begun prior to a change in sediment supply at this locality. However, as seafloor evaporite mineral growth could penetrate down through a thickness of unconsolidated beds from the seawater - sediment interface, the first appearance of evaporite minerals at ST-1 could be younger than the bed in which they are observed. Therefore, we consider this to be an unreliable criterion upon which to define the base of the formation.

Instead we propose that the base of the Sirab Formation (the base of the Ramayli Member) is taken at the lithostratigraphic surface which marks the change from grainstones to either packstones or wackestones. This surface is relatively easy to locate across the Al Huqf area where this stratigraphic level is exposed. Around the western flank of the Buah Dome, this boundary is always within an interval ca. 5 m to the west of the outer surface of the main topographic ridge. It can be followed almost continually from north to southwest and we have logged this transition at various points to confirm its presence (Enclosure I-3). At Wadi Shital, again there is a topographic change up-section, coming down off massive-weathering Buah to low-lying Ramayli Member. This easily constrains the boundary at a distance to within a few metres on the ground (Figure 5a). At an additional section just south of the Mukhaibah Dome structure at Wadi Sidr, the Buah/Ramayli transition is also exposed and the change from grainstones to packstones occurs across a single bedding surface (Enclosure I-1) [UTM 573243, 2202667]. Therefore, we can demonstrate that this basal bounding surface to the Sirab Formation can be successfully located in sporadic exposures from the northern tip of the Buah Dome south to Wadi Sidr, a distance of approximately 56 km.

The Ramayli Member is the oldest of the four members, which compose the Sirab Formation defined here. It represents a marked shift to shallower, lagoonal and sabkha depositional environments across the Al Huqf region and the first evidence in Huqf Supergroup exposures of widespread evaporite deposition. The base of the member has already been defined above as the base of the Sirab Formation. The top of the member is the base of the first peritidal carbonate-evaporite cycle of the overlying Shital Member, which typically forms prominent topographic ridges and which will be discussed further below in relation to the Shital Member stratotype at ST-1 (Enclosure I-2). The Ramayli Member can be informally divided into lower and upper units, dependant on quality of exposure at a particular locality. The lower unit is composed of fine-grained carbonates shallowing into lagoonal and peritidal environments, passing up into evaporite dissolution and collapse beds. The upper unit is typically composed of red and purple mudstones, siltstones and dolomicrites with common evaporite dissolution and collapse beds and emergent silcrete crusts and breccias.

The best exposures of the Ramayli Member, which demonstrate these key characteristics are found along the west flank of the Buah Dome, centred around the stratotype section at B-NW3 (Enclosures I-1, I-3 and I-5; Figure 6). There are few geographical names available for use in defining new lithostratigraphic units in this area. Since the tarmac road was built through the Al Huqf, the turn-off to the Buah Dome structure has been marked by a road sign to ‘Ar Ramayli’. This name refers to the area of shifting sands that must be negotiated on entering the area of Buah Dome outcrop [UTM 578848, 2248639]. Since it helps mark access to the stratotype section and, appropriately, this member is typified by low-lying, poor exposure beneath the desert sands, we adopt this name for the basal conformable unit in the Sirab Formation.

The Ramayli Member stratotype section at Buah Dome B-NW3 [UTM 569371, 2253038] affords the most complete exposure through this unit so far discovered in the field (Figures 6, 7 and 8). The section is situated on the western flank of the outer Buah Dome ridge composed of the Buah and Shuram formations of the Nafun Group (Enclosures I-3, I-5 and Figure 4). The base of the Ramayli Member also defines the base of the Sirab Formation, as discussed above, and is marked by the first occurrence up-section of packstones observable in the field. At B-NW3, this occurs at a height of 8 m above our arbitrary base of section anchored in Buah Formation grainstones below.

The basal 2.68 m of section B-NW3 are composed of packages of fining-upward medium to fine-grained, yellowish grey (5Y 7/2), dolomitic grainstones set in a fine to medium dolospar cement (dusky yellow 5Y 6/4) (Figure 6). Localised reddening and mosaic brecciation in patches or along veins and joints appears to be due to post-depositional fluid flow rather than intra-formational subaerial exposure and karst development. The overlying 1.82 m consists of a medium to coarse dolomitic grainstone with sub-angular grains, perhaps suggesting peloids, and crypt-oolitic patches where later cement has not occluded the original fabric. At 4.5 m above the base of the section a 1.03 m thick, graded bed of fine grainstones has a coarser basal zone containing clear ooids, coated peloids and elongate fine dolospar (?after dolomicrite) rip-up clasts set in medium grainstones. Overlying this unit are 2.47 m of fine, cross-bedded, yellowish grey grainstones. Cross-sets in places appear to show bi-directional E-W current directions, but in most sets the palaeo-current direction was to the west.

At 8 m above the base of section a slight colour change occurs which reflects both a grain size change and the appearance of a finer, dolomicritic mud component to the sediment interpreted to indicate that deposition shifted from inner mid ramp to more sheltered lagoonal palaeo-environments. This surface is the base of the Ramayli Member. From 8–10 m, the basal 2 m of Ramayli Member is composed of a fine, very pale orange to dark yellowish orange (10YR 8/2 to 10YR 6/6) packstone, which weathers to a pale yellowish orange (10YR 8/6). Throughout this interval there is patchy cementation and mosaic brecciation associated with joints and veins. A thin bed of fine, greyish orange (10YR 7/4) packstone between 9.96 m and 10.6 m displays well-developed small fenestrae with some larger bird’s eye and laminoid fenestrae within the central zone (Figure 7a). This is overlain by a 0.35 m-thick bed of greyish orange packstone-wackestone containing dispersed lath-shaped dolomite pseudomorphs of gypsum and is the first indication of evaporite mineral precipitation within the section. The sequence from ca. 7 m to this point at 10.95 m above the base of the section indicates a gradual relative shallowing in depositional environment from higher energy cross-bedded grainstones to packstone-wackestones with gypsum pseudomorphs.

These greyish orange packstones exhibit patches of well-preserved ooids and peloids up to 0.5 mm in diameter on weathered surfaces, again with some peloids coated by concentric micritic layers. Following this, relative shallowing-up appears to have occurred again as the interval from 12.89–15.73 m consists of dolomicrites. These greyish orange dolomicrites contain dispersed ooids, peloids and oncoids, but also have elongate dolomicrite rip-up clasts, with rounded terminations, partly replaced by chert (Figure 7b). Between 14.28 and 14.72 m the finely laminated, greyish orange dolomicrites exhibit dolomite pseudomorphs of anhydrite rosettes (Figure 7c) and gypsum lath pseudomorphs. Such rosettes are typical of diagenetic anhydrite forming in modern peritidal evaporitic environments (for instance see Cussey, 1979). They mark a second shallowing-up, or decreasing energy, depositional environment. However, unlike the one below, this younger interval appears to be capped by a potential palaeosol horizon. A sharp surface at 15.73 m marks the top of a bed, which has undergone pervasive dissolution-recrystallization events to destroy much of the original fabric and is now composed almost entirely of a medium calcite spar. The basal surface of this bed is irregular and appears to cut down into the underlying fine dolospar pelleted muds. These mudstones display more patchy dissolution and recrystallization in association with extensive chert replacement of dolomicrite rip-up clasts and dolomite pseudomorphs of gypsum laths, probably as part of palaeo-caliche development.

The interval from 15.73–25 m consists of fine, very pale orange to pale yellowish orange (10YR 8/2 to 10YR 8/6) calcitic packstones-wackestones. Bedding is on a 5–10 cm scale and there is a quartz sand component throughout. In the basal 1.38 m of this unit the quartz is a medium sand with well-rounded grains, some stained with haematite and occasional small chert concretions. This then passes up into the remainder of the unit containing a fine quartz sand mixed in with the carbonate grains. This unit marks the first indication of terrigenous clastic material being delivered to these shallow-marine palaeo-environments and may suggest that these beds were deposited in shallow salinas or sabkhas at, or close to, the palaeo-shoreline.

Towards the top of the unit, between ca. 23 and 25 m, the beds appear to be a fine-grained calcite-dolomite mix, with pelleted muds exhibiting dolomicrite envelopes and sutured contacts from soft sediment compaction (Figure 7d). This interval is also marked by laterally discontinuous chert replacement along laminae, forming silicified crusts where well developed (Figure 7e). Immediately above an interval of such crusts at 24.25 m, a thin lens of dolomitic ooids occurs. The pelleted muds suggest a low-energy lagoonal-type palaeo-environment with occasional, probably storm-induced, encroachment of higher-energy ooid shoals from the mid-ramp area. This suggests that a gradual increase in water energy occurred during deposition of this calcareous unit. As apparent confirmation of the increase in water energy up-section at this part of the section, from 25–25.67 m a packed dol-oolite bed occurs, with well-preserved internal concentric laminations to ooids, but later fractured and offset by the growth of gypsum laths in the sediment (and subsequently replaced by dolomite) (Figure 7f).

The surface at 25.67 m marks a change in the depositional style towards that of a more open lagoon with a tidal or storm influence and perhaps slightly greater water depth. Here, from 25.67–26.75 m, a prominent lens of finely crystalline, greyish orange (10YR 7/4) dolospar contains well-developed low stromatolite domes, with a ca. 20 cm diameter. This lens is set in a pale yellowish orange to very pale orange (10YR 8/6 to 10YR 8/2) fine packstone with thin, undulating microbial laminations, which are not quite coherent enough to form identifiable domal structures. Irregular cavities have developed along laminae. From 26.75–28.82 m, this microbially laminated unit also contains large, irregularly-shaped chert concretions up to ca. 30 cm in diameter, weathering to a moderate-light olive brown (5Y 4/4 to 5Y 5/6) (Figure 8a). The most likely explanation for these is that they are replaced stromatolite domes. However, no clear stromatolitic structures have so far been found still preserved in the chert concretions. These chert nodules form a marker horizon and weathered-out chert concretions from it can be found intermittently exposed at the surface at this stratigraphic interval from B-NW3 all the way to B-N1 at the northern tip of the Buah Dome.

Overlying the chert marker horizon, from 28.82–31 m, is a 2.18 m-thick slumped and brecciated unit (Figure 8b). The base of this unit is marked by a laterally persistent, 6 cm thick, banded chert horizon. Above this chert are chaotic, apparently partially convolute or liquefied, laterally discontinuous, thin beds composed of microbially laminated, very fine dolospar with dispersed sub-rounded quartz grains. Present within these dolospars are large inter-laminar cavities, which have been followed by partial chert replacement. Between microbial dolospar beds are thin, 2–3 cm thick, very pale orange (10YR 8/2) dolomicrite beds. In some instances, these beds can be observed to have broken and peeled away downwards from overlying bedding surfaces whilst still plastic. Typically these terminate in a pocket of brecciated dolomicrite clasts, some of which appear partially rounded. These brecciated dolomicrites are only present in the lower half of the unit, and the general coherence of the beds increases upwards towards its top as the deformation decreases. Given that the beds below are undisturbed, we interpret this as a collapse unit having formed from the early burial dissolution or withdrawal of an evaporite bed, or beds, immediately above the banded chert but below the convolute dolospar and brecciated dolomicrite beds, prior to their complete lithification. This best explains the downward collapse, brecciation and partial dissolution of dolomicrite beds, the convolute bedding of horizons overlying this lower zone indicating fluid loss and/or loading, and the pervasive chert replacement in the unit.

In this interpretation, the basal banded chert may thus represent diagenetic silica replacement of floor gypsum or anhydrite. Exposure dies away at the top of this unit and the logged section steps south by ca. 20 m. Positive correlation of this convolute unit with beds to the south where exposure continues allows the stratotype section to continue unbroken (Enclosure I-5). From 31 to 32.5 m, another convolute or slumped unit occurs. However, this is thinner and the disturbance to bedding laminations is less well-developed than in the underlying example. The thin beds are composed of greyish orange, fine-to medium-grained dolomitic packstones, with dispersed sub-rounded quartz grains throughout. There is an undulating microbial lamination in the lower half of the unit, which develops into well-formed stromatolite domes ca. 15 cm wide in the top half. Again, irregular laminoid fenestrae have developed and have been partly replaced by silica. A thin, dolomicritic solution collapse breccia is present within the unit at 31.8 m, suggesting some minor dissolution or withdrawal of evaporites at this level. Overall, there is a sense that this unit was more lithified prior to evaporite withdrawal and/or that the evaporites were originally much thinner and thus fluid loss from the unit was less. From the top of this unit at 32.5 m there is a break in exposure as the topography dies away, and the rocks are covered by the desert sands.

Low-lying, intermittent poor exposure reappears at 35.5 m as a 0.6 m thick, moderate red (5R 5/4) fine, tight packstone and thin, 0.5–1.0 mm pervasive laminations. Both the colour and tight laminations are characteristically different from underlying units and this bed marks a change that is typical of the remainder of the section. As such, the base of this bed at 35.5 m is taken as the bedding surface, which informally divides the Ramayli Member into lower and upper units. At B-NW3 and at other sections around the Buah Dome, the upper unit is dominated by poorly exposed red, fine grained siliciclastics and carbonates (Figure 2; Enclosures I-3 and I-5). From 37.9–38.7 m, a thin unit of pale reddish brown to very pale orange (10R 5/4 to 10YR 8/2) dolomicrite with patches of very fine dolospar is exposed. It contains very thin microbial laminations with associated laminoid fenestrae and fine dispersed quartz grains. The laminations develop up through the bed into small stromatolite domes ca. 5–10 cm in diameter at the top. The upper surface of these stromatolites and their interdomal areas are also accentuated by the development of pseudomorphed anhydrite rosettes (Figure 8c) and fragments of dolomicrite breccia. This thin, red, stromatolite bed is also a marker horizon (the ‘squat stromatolite marker’), which can be followed laterally around the Buah Dome.

Another interval of no exposure occurs between 38.7 and 41 m. From 41–41.7 m the beds consist of pale red to very pale orange (10R 6/2 to 10YR 8/2) mudstone with very small traces of a dolomite cement, but mostly silica cemented indicated by a slight cherty sinter on fracture (Figure 8d). The unit is thin-bedded on a ca. 5 cm scale but with internal slight wavy laminations and thin partings of very fine quartz or carbonate grains. There are no fenestrae or other porosity. The basal 20 cm appears to be a thin collapse breccia horizon, with elongate mud clasts held in a mud and sand matrix. Pseudomorphed anhydrite rosettes are present at the top of the unit at 41.6 m. The remainder of the interval between 41.7 and 46 m is composed of similar low-lying, intermittently exposed red mudstones. They are typically weathered to become fissile, breaking-up along the thin parallel or slightly wavy laminations. Also, interspersed through these mudstones are silicified, pustular crusts which on closer inspection appear to be horizons of poorly preserved anhydrite rosette pseudomorphs.

Within this unit, at 43.15 m, an apparent exposure surface is marked by a reddened chert breccia. Here, angular fragments of chert-replaced mudstone are cemented in an interstitial mud matrix with quartz blebs (probably after anhydrite pseudomorphs). The dark reddish brown (10R 3/4) chert of the top surface is easily observed from a distance and this brecciated surface can be followed, where exposed, from B-NW3 to B-N1, a distance of nearly 3.3 km away to the north.

The interval from 46.1–50.5 m is composed of pale red to pale greyish yellow (10R 6/2 to 10Y 8/2) interlaminated mudstones, weathering to a powder in situ and with a pervasive calcite cement throughout and calcite pseudomorphs of pustular anhydrite crusts. The key characteristic of these beds is that they all display convolute bedding or slump structures similar to the dolospar interval between 28.82 and 31 m. This includes clear examples of downward collapse of thin beds, whilst still partially lithified, into areas of dissolution breccias as described above (Figure 8e). Superficially, these beds are reminiscent of ‘teepee’ - type structures, however, there is a pervasive sense of downward collapse and no clear ‘teepees’ were observed. The upper 1.2 m of this interval, from 49.3–50.5 m, consists of slightly coarser sediment, either composed of fine dolomitic or quartz grains. The topography of this 4.5 m-thick unit suggests two main collapse units are present, the lower one being 1.9 m thick and the upper 2.5 m (Enclosure I-5 and Figure 6). This unit of convolute red mudstones passing up into red sandstones suggests a gradual relative shallowing and restriction of marine conditions from probable lagoonal carbonates and stromatolites below. Perhaps the best depositional environment envisaged for these beds is one of sabkha or salina pools, possibly stranded above the mean high tide mark, with only periodic marine replenishment and at least some terrestrial clastic input.

After an interval of no exposure for 2 m, the uppermost unit of the Ramayli Member stratotype section is exposed from 52–56.5 m. This final 4.5 m-thick unit is entirely composed of pale red, thinly laminated mudstones with relatively undisturbed laminations, in contrast to the units immediately below. However, here, haematitic crusts have developed along laminations throughout the entire interval and as a whole it has taken on the dark brown desert varnish typical of the overlying Nimr Group ferruginous conglomerates. As discussed above, the top of this interval of mudstones at 56.5 m marks the top of the Ramayli Member at B-NW3 and is only separated from the basal Haima angular unconformable surface by ca. 0.5 m of completely brecciated blocks of white chert-replaced dolostones. We interpret these to be the basal bed(s) of the overlying Shital Member, as the onset of cyclical carbonate deposition can be seen to come on abruptly in the comparable section at Wadi Shital (ST-1). Therefore, it appears that erosion during the Angudan Unconformity managed to remove whatever thickness of Shital Member or younger units had been deposited here right down to the basal beds. Subaerial exposure accompanied by tropical weathering during the unconformity, prior to deposition of the Haima Supergroup, would have allowed the weathering zone to penetrate down through the remaining basal bed of Shital Member and into the underlying top 4.5 m of Ramayli Member. As such, it is likely that these haematitic crusts are a feature of palaeo-caliche development within the Ramayli Member mudstones rather than a significant change in lithology at the top of the unit.

During field surveying which attempted to find further measurable sections through the Sirab Formation, we noticed that a series of linear ridges just east of the roadside in the area of Wadi Salutiyyat appeared from a distance to be Shital Member peritidal cycles (Enclosure I-1). On further investigation, it was not the ridges themselves which were of particular interest, but the exposures at their base. On leaving the black-top and driving towards the ridges, low-lying brick-red siltstones of the Shuram Formation can be observed weathering-out of the desert floor and dipping 45° to the east. These are overlain by thinly laminated microbial dolostones, which rise to form the base of the first prominent ridge, and which could represent either the upper Shuram dolostone member or Buah Formation. In either case, these Nafun Group units are then seen to be truncated along an uneven surface upon which a coarse and poorly sorted dolostone angular conglomerate is developed. Overlying this conglomerate with marked angular discordance are thin, poorly-developed beds of Shital Member peritidal cycles (Figures 9a and 9b) [UTM 582395, 2245731]. Thus, at this locality, the Sirab Formation clearly rests unconformably upon the Shuram/Buah formations. Even more intriguing, is that when this unconformable surface is followed north by about 20 m, a sliver of Buah Formation grainstones is observed to lie beneath it, but is itself then separated from the Shuram Formation below by a second conglomerate horizon, albeit then faulted at a later date. This Buah Formation grainstone sliver is thus bounded top and bottom by unconformities.

Overlying the conglomerate at the top of the grainstones is a short ca. 25 m section with large, metre-scale stromatolite domes developing small digitate columns on their top surfaces. These pass upwards into an interval of characteristic, weathered-back, pale yellow, marly packstones to wackestones which are exposed beneath the first peritidal cycle of the Shital Member. As such, this short section is visibly distinct from the Ramayli Member seen at either the Buah Dome or Wadi Shital. Thus the section here at Wadi Salutiyyat is important for two reasons; firstly, that it provides clear evidence of an unconformable base to the Sirab Formation at this locality. Secondly, that the unit directly overlying the basal conglomerate is of a markedly different lithofacies to either the Ramayli or Shital members.

The section at Wadi Salutiyyat in fact helps place in context a logged section from the earlier fieldwork conducted between 1999 and 2000 which has always remained problematic. Logged as Wadi Shuram-8 (WS-8), this section also exposes large domal, digitate stromatolites overlain by pale yellow marly packstones beneath the first, clear Shital Member cycles (Enclosure I-1) [base of logged section at UTM 557889, 2225046]. However, the field relationships at the base of WS-8 are complicated by later faulting. There appears to be a sliver of Buah Formation grainstones, which can be followed laterally southwestwards and is pinched-out between the overlying Sirab Formation and the underlying Shuram Formation. The top surface of the small Buah ridge at this locality has well-developed karst cavities filled with large gypsum lath pseudomorphs. There is a red, angular conglomerate present between the Sirab and Buah formations, but unfortunately this has been faulted. Thus, it could be argued that the conglomerate is actually a fault breccia and that the Buah has been faulted against the Sirab and truncated against the Shuram. This locality has been re-visited on a number of occasions with a variety of different geologists and continues to provide different interpretations of the field evidence on each occasion. However, the lithofacies of the Sirab section at Wadi Salutiyyat correlate directly with those at Wadi Shuram-8. This suggests that the conglomerate between Buah and Sirab is indeed a basal conglomerate developed along an unconformable surface, faulted at a later date.

Approximately 8 km southwest of Wadi Shuram-8, further evidence of a basal unconformity to the Sirab Formation is exposed on a much larger scale (Figures 9c and 9d) [UTM 550725, 2220630]. At this locality, sub-horizontal Shital Member cycles appear to directly overlie westerly-dipping Shuram Formation red siltstones and dolostone intervals (Figures 9c and 9d). Abundant scree covers the unconformable contact across most of the hillside, but there are one or two poor exposures of reddened Shuram dolostones overlain by in situ brecciated dolomicrites reminiscent of the Shital Member. Following the Shital Member beds westwards, they gradually begin to dip down to the northwest before exposure disappears just short of Haima Supergroup siliciclastics. Thus all of the Buah Formation is missing here, presumably eroded prior to Sirab Formation deposition. A NW-trending fault passes along the northeasterly edge of this ridge and offsets the Shuram dolostone ridges out on the flat plain below (Figures 9c and 9d). Interestingly, the succession on the northern side of this fault includes several tens of metres of Buah Formation, which is then overlain by red, recrystallized fine dolospar and light grey dolomicrites below the Angudan Unconformity. This suggests that syn-depositional faulting occurred during Sirab Formation deposition at this locality, with subsidence on the northern hangingwall causing the preservation of Buah Formation and lower Sirab Formation beds. South of the fault, the footwall remained subaerially exposed as a palaeo-topographic high until the onset of Shital Member deposition.

Linking the evidence observed at wadis Shuram and Salutiyyat, it seems that there is a clear unit that can be correlated from northeast to southwest across the Al Huqf and which rests unconformably upon the Nafun Group but lies below the Shital Member cycles. This can be recognised as another unit within the Sirab Formation, and since the most complete exposure of it is at Wadi Salutiyyat, we use this name for the member.

Driving south from Ar Ramayli on the tarmac road, a large wadi is crossed before reaching a relatively straight interval (Enclosure I-1). A few hundred metres east of the road here are prominent ridges topped by Shital Member cyclical carbonates. However, the flat desert floor is composed of Shuram Formation. The logged section of Wadi Salutiyyat SU-1 [UTM 582329, 2245812], which includes the stratotype section for the Salutiyyat Member begins on the wadi floor approximately 20 m short of the first ridge as seen from the road, and has an arbitrary base anchored in Shuram Formation on the flat plain (Figure 10).

The first 40 m of section, up to the base of the first prominent ridge, is composed of typical Shuram Formation, thinly laminated on a centimetre to millimetre scale, fissile and micaceous, fine sandstones to siltstones. The laminations are occasionally crinkly or wavy, indicating microbial mat growth. The beds dip at 45° to 102°. At 40 m above the base of section, a variable thickness interval of angular breccia can be observed to have developed on the top surface of the Shuram fine sandstones and siltstones. At the point of the logged section this breccia zone is 6 m thick, but it pinches and swells laterally either side and has been subsequently faulted to the south. At the base of the breccia interval, a 2.5 m-thick unit is present consisting of poorly sorted, angular, pink red to dark grey, coarsely crystalline dolospar clasts which range in size from ca. 20 cm down to sub-millimetre size. These are supported in a yellow matrix composed of medium to fine grains of dolospar. The breccia is cross-cut by reddened veins which exploit a crude bedding developed in the unit. Overlying this is a second breccia unit. However, this differs from that below in that the clasts have undergone in situ brecciation, not moving far from their original position and causing a ‘jigsaw’ effect. The clasts themselves are composed of coarse dolospar rhombs and there is pervasive veining and fracturing between clasts. Again, there is a crude overall 0.5 m scale bedding developed in the unit. Towards the top, a slightly more weathered-back horizon has a distinctive red colour and finer dolospar rhombs. The upper bounding surface to the breccia unit is irregular and overlain by 1.5 m of fine dolomitic packstone to wackestones. These are laminated on a millimetre to centimetre scale, with cross-bedding visible in places. These finer-grained carbonates pass up into an interval of 19 m consisting mostly of well-laminated, cross-bedded fine to medium grainstones, to the top of the first ridge on the line of section. There are occasional thin beds of packstones or recrystallized dolospar. Towards the top of this interval, dolomicrite rip-up clasts are present at some horizons and occasional ooids can be seen. Sporadic dissolution hollows filled with collapse breccia are also present.

Grainstone beds continue over the first ridge. However, both rip-up clasts and small vuggy cavities are commonly replaced or filled by chert. At 70.1 m above the base of the section, the grainstone beds are again brecciated in situ and weathered-back into the hillside. From 71.8 m to 74.5 m, a second breccia or angular conglomerate horizon is present. Clasts of grainstones are typically between 1 and 10 cm in size, but whilst most are angular, there are some which appear partly sub-rounded (Figure 9b). Many clasts are replaced by chert. Clasts are supported in a matrix consisting of fine- to medium-sized grains. The top surface of this unit has developed a chert crust over extensively chert-replaced clasts. Thus there is a clear grainstone unit present in this section, bounded top and bottom by a breccia unit. However, it is only this upper breccia, which is associated with extensive chert replacement.

There are two options for the lithostratigraphic correlation of the grainstone unit here. Firstly, that it is a sliver of Buah Formation preserved between the basal Sirab unconformity and the Shuram Formation. Secondly, it is possible that this is in fact the basal unit of the Sirab Formation at this locality. The lack of chert replacement until the very top of this unit might suggest the former is the case, as this can be used as a criterion for the Buah/Sirab boundary elsewhere (for instance at Wadi Shital; Figures 3 and 5c). Chemostratigraphy might help resolve this issue in the future. However, at present, the top surface of the upper breccia horizon will be taken as the base of the Sirab Formation and consequently the base of the Salutiyyat Member.

The basal 6.1 m of the Salutiyyat Member consists of thinly, often wavy, laminated beds of fine packstones to wackestones with abundant chert replacement along laminae. Small pockets of dissolution breccias are also common throughout the beds. Overlying this interval is a characteristic 0.9 m-thick unit consisting of large, metre-scale stromatolite domes, with well-developed internal laminae. The top few centimetres of these domes are covered in small 2–3 cm sized digitate columns, partly chert replaced. There is a 2.6 m interval of almost no exposure directly above the stromatolite domes. Only two, thin, poorly exposed beds are present, composed of fine, orange packstones to wackestones. The lowermost bed of a second characteristic unit within the Salutiyyat Member is exposed 84.1 m above the base of section. This mainly consists of thin beds of a powdering, marly, fine, pale yellow packstone to wackestone. These typically weather-back into the hillside, but the pale yellow colour is prominent at a distance. This marly unit is intermittently exposed for 7.8 m before exposure disappears again.

The interval between 91.9 and 102.5 m only has one thin bed exposed within it. This is a 0.35 m-thick bed of coarsely crystalline dolospar rhombs. When seen in other sections, such as at Wadi Shital, individual beds of these dolospar rhombs are usually associated with evaporites, and the lack of exposure within this interval would also support this interpretation. The first Shital Member dolomicrite-evaporite bed couplet is present 102.5 m above the base of section. It is composed of 2.2 m of fenestral, thinly laminated dolomicrite with common dissolution pockets and collapse breccias. Chert replacement along laminae increases towards the top. A 10 cm cleft between the top of this bed and the base of the next dolomicrite has no exposure, but again in all other sections these weathered-back intervals typically mark thin evaporite beds. These Shital Member cycles continue with good exposure over the next 61.5 m to the top of the logged section at 164 m. Thus the top of the Salutiyyat Member can be defined as the base of the first fenestral dolomicrite-evaporite rhythm. This gives the Salutiyyat Member at SU-1 a total thickness of just 28 m.

The Shital Member is the central unit of the Sirab Formation and is by far the most important member in terms of its stratigraphic thickness. It represents a significant change in depositional style from the underlying Ramayli or Salutiyyat members. The characteristic feature of the Shital Member is the tendency for extensive repetition of bed couplets through large intervals of stratigraphy (Figures 11 and 12). At the stratotype section at Wadi Shital ST-1 (Enclosure I-2; Figures 3 and 12), 1–2 m-thick beds of fenestral, laminated dolomicrites, overlain by thin 10–20 cm intervals of replaced anyhydrite or chicken-wire dolomicrite, repeat over and over again for up to 100 m through the section (Figures 11, 12 and 13). Similar intervals are present in other sections of this member, and in the field ridges composed of these ‘cycles’ are perhaps one of the most obvious features of the Sirab Formation to identify at a distance. The base of the Shital Member then, can be defined as the bedding surface at the base of the first peritidal carbonate-evaporite rhythm (Figures 5a and 13a). This significant bounding surface can be traced regionally, for instance it is particularly clear at the northern tip of the Buah Dome at B-N1.

The presence of a buff-weathering quartz sandstone unit within the otherwise carbonate-dominated Shital Member at ST-1 is a potentially useful marker horizon (Figures 3 and 12b) [UTM 576839, 2233696]. In the beds immediately above this sandstone are the first occurrences of conophyton bioherms in the section. The bedding dips gradually level out in the upper reaches of ST-1 and the exposure becomes a series of mounds, some of which are entirely composed of conophyton build-ups (Figures 3, 12c and 12d) [UTM 577525, 2233831]. The presence of the sandstone unit plus the development of conophyton above it allow the Shital Member to be informally sub-divided at localities with good exposure into lower and upper units separated by the sandstone marker horizon. In support of this, an important point to note here is that we have not encountered any well-developed conophyton to date below the sandstone marker in other sections in the region.

The top of the Shital Member is more difficult to define due to the poor exposure at the top of the stratotype section at ST-1 and the dips becoming sub-horizontal. The overlying Aswad Member is exposed across the road to Sirab to the southeast of ST-1, but a low-lying sandy area covers the boundary between the two (Enclosure I-2). Consequently, the only locality where we have seen the two members in direct contact with each other is at Wadi Aswad to the north at section WA-1 (Enclosure I-1 and I-6). Here, the top of the Shital Member has a locally developed sub-Haima karst pocket [UTM 591131, 2277591], apparently conformably overlain by the oncolite and thrombolite beds of the Aswad Member. The field evidence at this boundary will be considered in more detail in sections below when defining the Aswad Member.

The most complete exposure of the Shital Member where its base can clearly be seen overlying the Ramayli Member is in the roadside section at the eastern edge of Wadi Shital in the central Al Huqf area (Enclosure I-2). Wadi Shital itself is the sandy area west of the main tarmac road as it passes through the area of Sirab by the Al Maha petrol station. The name does not appear on any road signs but is marked on a Tourist Map of Oman as ‘Wadi Shitall’. However, this seems to be a mis-transliteration and local pronunciation of the wadi name is either as Shital, Sta’l or Shtall. The former seems the closest to the way it is spoken in the area but the official Government list of Wadi names for Oman presents the name as Wadi Shital (pronounced S’hital, with a silent ‘h’) and thus we adopt it here for the name of the middle member in the Sirab Formation.

The arbitrary base of section Wadi Shital ST-1 [UTM 576808, 2234196], which incorporates the stratotype section of the Shital Member, is anchored in Buah Formation cross-bedded grainstones about 300 m northeast of the Al Maha petrol station on the main tarmac road to Duqm and Salalah (Enclosure I-2). Buah Formation low-lying grainstone ridges close to the roadside are overlain to the southeast by poorly exposed Ramayli Member units in a level area mostly covered by desert sands. The Ramayli/Shital boundary is located where the topography begins to rise again from the desert floor about halfway up the first visible prominent ridge from the road (Figures 3 and 13a). Thus, the base of the Shital Member stratotype section is located 63.3 m above the base of ST-1 at [UTM 576903, 2234116] (Figure 3).

An evaporite unit within the upper Ramayli Member is exposed at the bottom of the first topographic ridge, between 50.25 and 58.5 m above the base of ST-1 (Figure 3). This includes beds of coarse dark yellowish orange (10YR 6/6) and light brown (5YR 5/6) dolospar, which still retain preserved patches of halite. This unit passes up into 3 m of no exposure. From 61.5 to 62.1 m, two thin beds of coarse light brown dolomite rhombs, with halite patches, are exposed. After another thin interval of 0.6 m of no exposure, a 0.3 m-thick bed of greyish orange pink (5YR 7/2) coarse, laminated dolospar contains a thin dissolution and collapse breccia horizon at its centre. After another 0.3 m of no exposure, the base of the first peritidal cyclical unit occurs at 63.3 m. Over the next 5.5 m of stratigraphy, these cyclical beds remain weathered-back into the ridge in a similar fashion to the upper Ramayli dolospar beds below. However, above 68.5 m, the beds are massive and weather prominently from the hillside. Despite this difference in weathering, there is continuous exposure of the beds upwards from the base of the Shital Member, which we define at 63.3 m.

The basal bed of the Shital Member from 63.3–64 m is composed of a very pale orange (10YR 8/2) dolomicrite. It is thinly laminated in its basal half on a 1–2 cm scale, but this passes up into the upper portion of the bed where laminations are less well pronounced and are replaced by bird’s eye fenestrae. Small patches of dissolution and collapse breccia are dispersed throughout the bed. There is a small weathered-back cleft between the top of this bed and the next exposed at 64.1 m. We interpret this 10 cm interval to represent the evaporite bed within the couplet, which is a feature clearly seen at some localities such as at the Buah Dome (Figures 11a to 11e). As such, this interval from 63.3–64.1 m represents a characteristic carbonate-evaporite rhythm of the lower Shital Member (Figure 11a). Clearly, however, the evaporitic top to the couplets did not always develop, or was not preserved, and the interval from 64.1–67.9 m consists of three dolomicrite beds all identical to the basal one, but with no weathered-back chicken-wire anhydrite or dolomicrite between them.

In the field, a dolomicrite bed within a couplet, well-laminated at the base and passing up into a fenestral zone with breccia pockets, was informally recognised as ‘Facies α’. These formed the typical carbonate unit within a couplet. However, a second subtle variation was also observed, ‘Facies β’. The interval from 67.9–69.75 m at ST-1 contains three such beds. These consist of a very pale orange (10YR 8/2) mix of interbedded coarsely crystalline dolomite with chert replacement along laminae and irregular chert concretions, with dolomicritic beds ca. 2–5 cm-thick sandwiched in-between (Figure 13b). The dissolution-reprecipitation of dolomite appears associated with gypsum lath pseudomorphs and may suggest Facies β was deposited in a more restricted water mass than Facies α, favouring evaporite mineral growth (Figure 13c). From 69.75–70 m, a thin 25 cm-thick bed of white, powdering dolomicrite weathers back into the ridge and probably represents the top evaporitic unit to the upper bed.

From 70–136.2 m, these repetitive beds of Facies α and β and occasional thin, weathered-back interbeds of fissile chicken-wire dolomicirite continue with only minor loss of exposure over this and then the next prominent ridge to the southeast (Figures 3, 12a, and 13d). Indeed, the monotonous nature of this cyclicity is the most characteristic feature of this lower unit of the member. Also within this interval, the thin laminations within many beds can be observed to have been wavy microbial mats or developed into recognisable, low stromatolite domes.

From 136.2–142 m there is a marked break in exposure. Overlying this interval is a 0.5 m-thick bed of a pale red (5R 6/2), well-laminated, coarse dolospar interbedded with thin dolomicrites. Exposure is missing again from 142.5–143.25 m, where a similar red dolospar bed is then exposed, as is another between 144 and 144.25 m. This younger bed is also brecciated and possibly marks a palaeo-exposure surface. Where similar red dolospar beds were encountered at the top of the underlying Ramayli Member in ST-1, they were intimately associated with halite and it is reasonable to infer that they also form part of an evaporite unit (‘Evaporite unit 2’) here between 136.2 and 144.25 m. This would also correlate with a similar interval at approximately this stratigraphic level in sections at the Buah Dome (B-N1) and Wadi Sidr (SD-1) (Enclosure I-1). There is no exposure between 144.25 and 146 m, above which Facies α and β are intermittently and then poorly exposed from 146–163.1 m. Exposure then disappears, coinciding with the topography levelling-out into a broad, flat plain of desert sand.

Exposure reappears at the base of the next ridge to the southeast, at 178.8 m above the base of ST-1, with a 1.5 m-thick poorly exposed dolomicrite rhythm, overlain by a further 4.5 m of no exposure. However, from 184.8–188.4 m there is an abrupt change in lithology. Here, laminated moderate reddish brown to yellowish brown (10R 4/6 to 10YR 5/4) fine quartz arenites in the lower 2 m are overlain by more massive quartz arenites with a bi-modal fine- and medium-coarse grain-size distribution in the top 1.6 m (Figure 14a). These quartz arenites are mostly silica and haematite cemented, but there is a minor dolomitic component. The more massive upper unit appears to have undergone some localised soft-sediment deformation, which in patches is reminiscent of bioturbation. Referred to in the field as the ‘buff sandstone marker horizon’, this unit can be traced at this stratigraphic level from Wadi Aswad in the north of the Al Huqf all the way to Wadi Sidr in the south, a distance of ca. 85 km (Enclosure I-1). It also occurs to the west in Wadi Shuram as part of a fault sliver of Shital Member. The lateral extent of such a unit is perhaps surprising and the possibility raises itself that there is more than one of these buff sandstones. However, no more than one has so far been found in any section and thus it remains a good general marker for lithostratigraphic correlation in the field.

The buff sandstone marker is overlain by a thin break in exposure of 0.6 m. Then, between 189 and 192 m a laterally extensive breccia unit is exposed. This 3 m interval consists of a lower, irregular zone of brecciated and collapsed dolomicrites, overlain by more coherent, laminated pale yellowish orange (10YR 8/6) fine dolospar beds with some relict halite, passing up into a second zone of brecciation and bed collapse (Figure 14b). This upper unit also contains the first recognisable fragments of brecciated conophyton (Figure 3). Given the similarities with evaporite-associated dolospar beds and breccias lower in the sequence, we interpret this interval as another evaporite interval with dissolution and withdrawal of coherent evaporite beds having led to brecciation and collapse in the overlying dolospar beds. This unit is labelled as ‘Evaporite unit 3’ on Figure 3 and correlates with fine red muds at this interval in Wadi Aswad WA-1 to the north.

The brecciated top surface of Evaporite unit 3 at ST-1 forms the crest of a low ridge and has patches of in situ conophyton developed directly on the irregular top surface of the breccia. This indicates that although evaporite dissolution was post-lithification, it took place during very early burial. Transgressive drowning of the evaporite unit would have allowed conophyton build-ups to begin establishing themselves on its top bed. Dissolution and collapse must then have occurred, brecciating these earlier conophyton domes, before more extensive conophyton mounds re-established themselves on top of the now brecciated horizons below.

The remaining 23 m of the Shital Member from 192–215 m are composed of wavy-laminated microbial, pale yellowish orange to greyish orange pink (10YR 8/6 to 10R 8/2) dolomicrites which at intervals can be seen to have developed into pointed conical conophyton domes ca. 10–20 cm high (Figures 12d and 14c). The beds often have sporadic dissolution breccia pockets developed within them and many of the top-bedding surfaces are marked by a thin breccia horizon. Further to the southeast up-section, the topography levels out as the bedding dips become sub-horizontal. Here the Shital Member statotype section ends amongst the dissected mounds of the uppermost exposed beds. These can be observed to be large, exhumed palaeo-microbial mounds or ‘reefs’ of conophyton, often flanked with a breccia apron (Figure 12c). Some of the weathered surfaces in and around these mounds also appear to show a crypto-clotted or mottled fabric similar in size and scale to thrombolites in the overlying Aswad Member (Figure 14d).

The Aswad Member is the youngest exposed unit in the Sirab Formation. In the type area of the formation at Sirab, the dip of the beds at the top of section ST-1 gradually level out and exposure ends north of the tarmac road to Sirab in a series of low mounds with well-developed conophyton mounds (Enclosure I-2; Figures 12c and 12d). However, to the southeast of the road, a plateau of low hills can be seen which lies slightly higher both topographically and stratigraphically than the last exposures in ST-1. Upon investigation we were surprised to find that these hills are composed of horizontal interbeds of oncolites and thrombolites. The oncoids are typically 0.5–1.0 cm in diameter. As the beds have little or no dip, the subaerial exposure of this unit is large in area (Enclosure I-2). However, mainly because of this we were unable to locate the base or top of this unit at Sirab and estimate its total stratigraphic thickness to be ≤ 30 m. Further south, just off the main road before reaching the Wadi Sidr turn-off, a small knoll of Cretaceous white limestones with Thalassinoides burrows unconformably overlies the oncolite beds, but that remains at present the only age constraint for them in this area (Enclosure I-1) [UTM 575089, 2223216].

The key to placing this unit in a lithostratigraphic context was found at Wadi Aswad to the north (Enclosures I-1 and I-6). Again unexpectedly whilst surveying, north this time along the Shital Member/Haima Supergroup contact, we began encountering oncolite and thrombolite mounds identical in lithofacies to the unit southeast of the Sirab road. Although faults at Wadi Aswad cross-cut and truncate the stratigraphy into a series of fault slivers, we managed to locate one coherent sliver which included the mid-Shital Member sandstone marker unit at its base and logged this section up towards the oncolite mounds (section WA-1; Enclosures I-1 and I-6). Just below the first appearance of oncolites, the Shital Member dolomicrites form a prominent ridge about 2 m thick consisting of about two to three massive beds. These are penetrated by karst fissures filled with dolomitic breccias and matrix (Enclosure I-6; Figures 15a and 15b) [UTM 591131, 2277591]. At first glance this could of course be a modern karst development and there is certainly some modern aeolianite cemented across the surface of some of the blocks. The karst only seems to be developed locally along this ridge and down from the top surface of the beds which compose it.

Elsewhere in this area we have observed basal Haima sandstones piped down into hollows developed on a palaeo-karstic surface at the top of the Sirab Formation (see below). Therefore, we take this karst surface to have developed during the Angudan Unconformity and that the Aswad Member rests conformably upon the Shital Member. As the section at WA-1 is clearly the key to defining this member and the stratotype section for it runs northwest across the hills on the western side of Wadi Aswad, we propose to name the member after the wadi. The definition and concept of the member is simple. It is composed of a laterally extensive, but internally variable, mix of intricately associated oncolite, thrombolite and mixed oncoid-thromboid interbeds. The base of the member can be defined as the first appearance up-section of such a mix of lithofacies. The top-bounding surface to the Aswad Member is either truncated by overlying Haima Supergroup clastics or is horizontally bedded and not overlain by any other surviving units as in the type area south of Sirab.

The base of section WA-1 [UTM 591131, 2277591], which incorporates the stratotype section of the Aswad Member, begins at the base of the buff sandstone marker horizon exposed in the centre of the ridges forming the western flank of Wadi Aswad (Enclosures I-1 and I-6; Figure 16). The succession overlying the sandstone marker is dominated by monotonous, repetitive, fine dolospar beds, which represent minor dissolution and recrystallization events in former dolomicrites of the upper Shital Member. A faulted sliver of the upper portion of this section repeats west of Aswad Member oncolite mounds and contains tufted mats and conophyton mounds characteristic of this upper Shital Member unit at Wadi Shital (Figures 15a and 15b). The base of the Aswad Member is defined by the first appearance of thrombolite/oncolite lithofacies and at Wadi Aswad WA-1 this is a relatively abrupt transition (Figure 15). We were only able to measure 4.5 m of Aswad Member facies at WA-1 before bedding dips flattened out to horizontal and became a series of dissected mounds across the plain (Figure 16).

The basal 4.3 m of the Aswad Member consists of light grey to medium light grey weathering (very pale orange when fresh; 10YR 8/2) dolomicrites, bedded on a 20–40 cm scale (Figure 16). Internally, these beds have intervals of coherent wavy microbial lamination and pockets of intrabed breccias. However, the characteristic fabric of these beds are the thromboid mesoclots which occur in the zones between breccia pockets. From ca. 4.3–4.5 m, the ‘cauliflower’ thrombolite heads begin to grade up into more coherent, disarticulated balls so that the uppermost bed of the unit is composed entirely of oncoids, with diameters similar to the width of thrombolite heads (Figure 15c). This oncoid bed marks the top of the Aswad Member stratotype section at WA 1.

The Sirab Formation is overlain with marked angular unconformity by the Haima Supergroup. Amin sandstones or conglomerates at the base of the Haima can be found overlying any of the three Sirab members at various localities around the Al Huqf (Figures 17 and 18). This regional angular unconformity is referred to as Angudan Unconformity (Droste, 1997). Depending on the age dates for the base of the Amin Formation (and Nimr Group subsurface) as discussed above, this could represent a hiatus of anything up to ca. 40 My between the termination of Sirab Formation accumulation and the recommencement of deposition with the siliciclastics of the basal Haima Supergroup.

Erosion of the Sirab Formation during this period was highly variable. On the western edge of the Buah Dome the Sirab Formation has been completely eroded and the Amin rests unconformably upon a thinned Buah Formation (Enclosure I-3; Figures 17a and 17b) [UTM 568525, 2250173]. Evidence that the Sirab Formation did exist here prior to erosion is provided by an isolated fault sliver of Shital Member preserved under the Amin sandstones (Enclosure I-3; Figure 17c) [UTM 568709, 2251022]. Elsewhere, such as at Wadi Aswad to the north, the Amin sandstones variably overlie the oncolite/thrombolite beds of the upper Sirab Formation Aswad Member [UTM 590681, 2277124] or dolomicrites of the upper Shital Member [UTM 591082, 2278095] (Figures 18a to 18e). At one locality, the Amin sandstones can be seen to lie upon Aswad Member thrombolites with angular unconformity [UTM 591740, 2280540] (Figures 18c and 18e). Thus, how much of the originally deposited Sirab Formation was eroded prior to the onset of Haima deposition is difficult to estimate, but clearly nowhere do we see the original thickness of the unit.

The nature of the Sirab/Haima unconformable surface is interesting in itself, and is best seen at exposures around the western flank of the Buah Dome (Enclosure I-3). At one of our logged sections on the northwest flank of the Buah Dome, B-NW4, a sliver of Shital Member appears to have been faulted down against underlying Ramayli Member and preserved under the erosive base of the overlying Amin Formation sandstones (Enclosure I-4). In addition, the lowermost beds of Amin sandstone can be found forming a series of small exposures, which gently dip in towards the centre of a hollow, cutting across the strike of the Shital Member beds adjacent to them. It is possible that this could be basal Amin wadi fill. However, the sandstone dips of this structure in relation to the overlying beds which plane over the top suggest that this was a palaeo-bowl-shaped depression as might be expected from karst development and collapse in the carbonates below. The beds have subsequently been tilted to dip gently to the northwest so that this structure is now eroded in oblique section.

Some of the faults at this locality cross-cut the lower Amin beds; but others, such as that throwing the Shital Member block down, do not and clearly pre-date Amin deposition. It is therefore possible that not only did karst dissolution develop along Sirab Formation and Buah Formation joints, but also preferentially exploited pre-existing fault lines and suggests a period of faulting post-Sirab, but pre-Amin deposition. Other localities around the Buah Dome where Haima clastics appear to locally cut down into the carbonates below also seem to be associated with faults on satellite imagery (Enclosure I-3). Just north of section B-SW1 [UTM 568582, 2249669], the Buah Formation ridge is incised by obliquely orientated Amin sandstones with convolute bedding and conglomerates (Enclosure I-3; Figures 17a and 17b) [UTM 568624, 2249825]. Again this suggests either a linear karst collapse or a wadi in oblique section.

The Haima Supergroup is only preserved in exposures along the western edge of the Haushi - Al Huqf region. It was eroded in the central and eastern coastal parts of the area prior to the Cretaceous (Debreuilh et al., 1992; Platel et al., 1992). Here, isolated knolls of white limestones with colonial coral bioherms [UTM 583329, 2235672] and large Thalassinoides burrows can be found unconformably overlying Sirab Formation exposures in the Sirab area and to the south (Enclosure I-2). Thus, it is unfortunate that in the type area of the Sirab Formation itself, the true nature of the top-bounding surface cannot be seen. However, detailed lithostratigraphic correlation of units laterally confirms that the three Sirab Formation members exposed at Sirab and Wadi Sidr to the south are the same units that crop out below the Haima clastics to the north and west.

The original section where the Sirab Formation was first recognised in 1999 is at the old section of B-SW1 (Enclosure I-3). Here the Buah grainstones ridge is separated by an interval of only about 20 m of low-lying, poor exposures of red dolomitic muds and ‘squat’ stromatolites from a prominent ridge of ferruginous Amin conglomerates and sandstones (Figure 19a). This is the most southerly exposure of the Rumayli Member in the Buah Dome, but where not incised by the overlying Haima, the member can be traced intermittently from here all the way to the northern tip of the structure at another of our early sections at B-N1 (Enclosure I-3; Figure 19b) [UTM 571720, 2256193]. Indeed, the red mudstones and ‘squat’ stromatolite horizon appear to be laterally continuous around the Buah Dome as they can be located at B-N1 (Figure 19b and Enclosure II-1). No unequivocal Ramayli Member has so far been recognised north of the Buah Dome. Red mudstones do occur in isolated fault blocks at Wadi Aswad, but the section at WA-1 demonstrated that an interval of mudstones is present above the buff sandstone marker horizon in the mid-upper Shital Member (Enclosure I-6 and Figure 16).

The Ramayli Member is reasonably well exposed at Wadi Shital ST-1 (Enclosure I-2). However, above another ‘squat’ stromatolite bed (which may be the same as in the Buah Dome) there is a visible difference in facies from that at B-NW3. Dissolution breccias and collapsed red mudstone beds were characteristic at the Ramayli Member stratotype section. At ST-1, the mudstones and siltstones in the upper part of the member are not heavily stained with haematite and form a very finely laminated interval reminiscent of lacustrine varves (possibly a salina), overlain by halite beds. In these beds, yellow dolospar breccia blocks are supported by a matrix which seems to be a mix of recrystallized dolomite and halite (Figures 19c and 19d). There is sufficient halite preserved for these beds to taste salty in patches.

To the south of Sirab, the section at Wadi Sidr south of the Mukhaibah Dome structure exposes an excellent continuous section through the Ramayli and lower Shital members (Enclosures I-1 and II-1). Here, the siltstones and mudstones in the upper half of the member also pass up into an interval of yellow recrystallized dolostones immediately below the base of the Shital Member [UTM 573283, 2202592]. To the west, in Wadi Shuram, red mudstones are present in fault-bounded slivers [UTM 556323, 2226346] but until chemostratigraphy can aid lateral correlation, these are best considered to be part of the Shital Member as at WA-1. In all occurrences where the complete Ramayli Member has been observed it seems to have a remarkably consistent thickness of around 45–50 m, although within this the thickness of upper and lower divisions seems more variable. Overall, there is a general shallowing-up trend in the Ramayli Member, marked by a fining-up in the lower carbonates until continental siliciclastics become a more dominant component of the sediment and with emergent surfaces common. Within the upper half of the upper unit there is evidence at three of our main sections that evaporite beds were deposited, including halite precipitation, thus allowing recognition of ‘Evaporite unit 1’ within the Sirab Formation (Enclosure II-1).

Further along the road from ST-1 to the village of Sirab, a series of exposures in the Shital Member give an impression of lateral facies variation or persistence at specific stratigraphic levels (Enclosure I-2; Figure 20a). These localities are approximately 6 km from ST-1. Above a typical interval of cyclical peritidal carbonates, there is a thin bed of white, ‘marble’ calcite, which appears to have replaced the original mineral and texture (Figure 20b) [UTM 583164, 2235052]. In places the bed consists of a series of well-developed, 0.5 m-sized domes, whilst elsewhere laterally these are not present and the bed is more massive with a slightly undulating top surface. Immediately below this bed are friable, red recrystallized dolospars apparently after an anhydrite chicken-wire fabric. At the base of the replaced anhydrite bed is a top-bedding surface with well-developed mud desiccation cracks (Figure 20c). The replaced white calcite bed can be followed for several kilometres west towards ST-1 (for instance, it is exposed at UTM 581289, 2235213), but it is not present at ST-1 itself and must pinch-out laterally.

At a slightly higher level at the Sirab localities, is a thin interval of ‘tea-green’ marly clays (Figure 20d). This lithology in the field is very reminiscent of the evaporite-dominated units of the Upper Triassic of England. These clays are exposed on the Sirab roadside [UTM 582632, 2234877] just below an interval of red quartz sandstones on a small hill (Figure 20e). This is the original locality where the possibility of Ara Group clastics was raised. If this is the same sandstone marker unit as seen approximately half-way through the section at ST-1, then this section on the Sirab road should correlate with the mid to upper interval of the lower Shital Member. In support of this correlation, overlying the red sandstones at Sirab are conophyton mounds (Figure 20e).

The thin sandstone marker unit has also been located in other sections at about this stratigraphic level within the Shital Member and we consider these to be the same beds. At section WA-1 at Wadi Aswad, an interval of red mudstones and recrystallized dolostones overlie the sandstone marker (Figures 16 and 20f) [UTM 591346, 2277490]. This suggests that a second evaporite unit is present immediately overlying the sandstone marker horizon, and it persists laterally from Sirab to Wadi Shital to Wadi Aswad (Enclosure II-1). Further up the WA-1 section, conophyton bioherms are well developed and in some cases appear to be underlain by an interval of microbial ‘tufted’ mats (Enclosure II-1; Figure 20g) [UTM 591082, 2278095]. To the west of Sirab, at Wadi Shuram (WS-1) (Enclosures I-1 and II-1), even though our logged section does not have the sandstone marker exposed, the upper part of the section contains conophyton and excellent stromatolitic thrombolites (Figure 20h) [UTM 554800, 2223955]. Thus, based on these lithostratigraphic arguments, we would currently assign this part of the section to the upper Shital (Enclosure II-1). Also at the northern tip of Shital Member ridges in Wadi Shuram, is an enigmatic area of deformation and bed collapse which is underlain by remnants of salty-tasting halite beds similar to those at the top of the Ramayli Member at Wadi Shital [UTM 556531, 2226509]. Here, however, they are clearly amongst peritidal cycles of the Shital Member, but suggests that at least some salt deposition occurred at this higher stratigraphic interval (Enclosure I-1).

The type area of the Sirab Formation provides a second, slightly more extensive section through dissected mounds of horizontally bedded Aswad Member, ca. 2 km to the southeast of the top exposed Shital Member beds (Enclosure I-2). We estimate that the low-lying mounds in this area south of the road to Sirab yield less than 30 m of Aswad stratigraphy. Despite this, they provide ample opportunity to examine lateral facies change within the member. The upper oncoid bed at WA-1 consists of a medium grey matrix, which suggests a reasonable component of organic matter is still preserved, and that such oncolites could be a potential oil source rock (Figure 15c). In contrast, in the Sirab area, oncoid beds are white to cream in colour and do not appear to have any significant organic matter now present (Figure 21a). Indeed their weathering in this area actually accentuates the sedimentary textures particularly well. We logged a short section southeast of ST-1 to include as a composite stratotype of the Sirab Formation, but as such, it can also be used as a parastratotype for the Aswad Member itself (section ST-2, [UTM 578843, 2231645]; Enclosure I-2 and II-1; Figures 3, and 21a to 21d).

Section ST-2 was recorded through a representative mound of the many exposed in this area. The base of the log is pinned at desert sand level, where there is no exposure for the first 0.7 m (Figure 3). The lowermost unit at ST-2, from 0.7–1.3 m, consists of a pinkish grey weathering to a very light grey when fresh dolomicrite (5YR 8/1 to N8). Irregular microbial laminations on a 1–2 mm scale in the basal 40 cm of this unit pass laterally into patches of thromboid clots (Figures 22a and 22b). There is also a vertical gradational change within this unit from laminated dolomicrite at the base into a thrombolitic fabric within the top 20 cm zone (Figure 21d). Wherever a typical thrombolitic fabric is developed (facies ‘T₁’ on Figure 3), a thin white fringing dolomite cement often coats thromboid ‘heads’ and this is particularly well-developed in the upper 20 cm (‘fd’ in Figure 22b). In addition, a later, zoned dolomite cement is typically present plugging original depositional cavities in the framework (‘zd’ in Figure 22b).

From 1.3 to 1.6 m, a very well-developed thrombolite bed occurs similar to that immediately below. Between ca. 1.6 and 1.65 m there is an apparent mix of oncoids and thromboid heads ‘pinched’ at the neck (facies ‘T₂’ on Figure 3), and this appears to be a gradation within the thrombolites towards higher energy, fully disarticulated microbial balls. These ‘pinched’ thromboids pass upwards into more typical framework fabrics at the top of the bed at 1.75 m (Figure 21b). ‘Pinched’ thromboids occur again between 1.75 and 2.05 m, overlain by more typical clotted thrombolite fabrics, which dominate for the remainder of the section (Figures 22c and 22d). In some instances, zoned cavity-fill cements themselves have been seeded upon by late stage dolomite cements with dolomicrite debris cemented within the centre of the pore space (Figure 22d). This suggests that several phases of dolomitic cements precipitated and gradually restricted the framework cavities during growth of the bioherm and that in this instance they were not post-depositional diagenetic cements. The possibility that some of these thrombolite units may have retained open effective porosity during burial has important implications for them as potential petroleum reservoirs. As such, these Aswad Member facies are a direct analogy for the subsurface Ara Group ‘stringer’ reservoirs. ‘Pinched neck’ thromboids (facies ‘T₂’) begin again to develop at ca. 4.5 m (Figures 22e and 22f), but otherwise the typical clotted fabric is maintained to the top of the interval at 5.65 m. Finally at ST-2, the uppermost bed from 5.65–6 m, consists of fully disarticulated oncoids set in a fine-grained dolomicrite (Figures 22g and 22h).

Other, dissected, lateral-equivalent mounds exposed around ST-2 exhibit gradational facies between individual oncoids, oncoids microbially bound together in patches, laminated (stromatolitic) thrombolites and well-developed thrombolite clotted textures, for instance, at [UTM 578993, 2231422]. There are even partly lithified slabs of microbial dolomicrite that were rolled by currents and subsequently microbially re-bound (Figure 21c) [UTM 578664, 2231841]. These ‘giant’ oncoids are particularly reminiscent of similar structures found in the Jurassic of the south Dorset coast, known locally as ‘snuff boxes’.

The presence of persistent thrombolites within the Aswad Member indicates in general a more open marine depositional environment than the lower units in the Sirab Formation. However, given this control on their growth, thrombolite lithofacies could occur lower in the Sirab stratigraphy if there had been sufficient localised subsidence to cause an increase in water depth and circulation. Such a deepening event appears to have taken place, at least temporarily, in the upper Shital unit at Wadi Shuram WS-1 (Enclosures I-1 and II-1). Here an interval of excellent domal stromatolitic thrombolites (Figure 20h) is overlain by a single oncolite bed only ca. 10 cm thick, but identical in oncoid size and packing as in typical Aswad Member facies. Further to the northeast along the Shital Member ridges in Wadi Shuram, small, isolated thrombolite reefs can also be found amongst the conophyton [UTM 556554, 2226603] (Figure 21g). These observations in Wadi Shuram indicate that the transition from Shital Member lithofacies to those of the Aswad Member are gradational and controlled by water depth in the lagoon.

Finally in relation to the Aswad Member, and a further indication that water depths had increased during deposition of the younger Sirab Formation units, we discovered a single locality where thrombolitic build-ups are capped by crinkly laminites [UTM 577840, 2221938] (Figure 21e). In cores from the South Oman Salt Basin, such microbial laminites are organic rich and assumed to be deeper water source rocks within ‘stringers’.

There is still some debate as to the exact palaeogeographic position and structural regime of Oman during the Ediacaran–Cambrian transition. However, at present it seems likely that Oman as part of the Arabian Plate occupied a marginal position in an accreting Gondwana at ca. 600 Ma juxtaposed with India and Pakistan to the southeast (Powell and Pisarevsky, 2002). The Al Huqf region today is cross-cut by a series of dominantly NS-trending faults, which have had a complex history of movement. These faults can be demonstrated to have been active both prior to, and post-Sirab Formation deposition. Less common are NW-trending faults, which are a relic from the late Neoproterozoic Pan-African Orogeny, or ‘Najd Event’. NE-trending structures, such as the Khufai Dome, and faults across the area appear to follow basement terranes and their sutures respectively. This framework of faults provides the basis upon which an initial attempt can be made to reconstruct the depositional palaeogeography of the Sirab Formation during the Ediacaran–Cambrian (Enclosures II-2 to II-5). In the following sections, we synthesise our field observations to produce conceptual models for Sirab depositional setting and facies evolution (Enclosures II-6 to II-9) and place them in this palaeogeographic context.

Relatively uniform mid- to upper- ramp grainstone, ooid and peloid shoal lithofacies occur regionally throughout the top of the Buah Formation, as outlined above when discussing the conformable base of the Sirab Formation. These are overlain in all exposed sections of the Ramayli Member by finer-grained carbonates grading into carbonate muds with dispersed evaporite mineral growth. Thus the lower part of the Ramayli Member represents a shallowing and restriction in water circulation, with facies changing to suggest shallow, low-energy lagoonal environments. Occasional marine incursions into the lagoon(s) are marked by the brief influx of ooids, peloids, oncoids and the increase in water energy also caused the rip-up of partially lithified dolomicritic mud from the lagoon floor. Conversely, periodic draw-down of sea level at various times caused chert replacement along some horizons, localised dissolution and brecciation, and in some instances subaerial exposure with karst developing (Figure 6). Silica is a common evaporite replacement mineral in the mixing zone between meteoric and marine pore waters in peritidal environments (Milliken, 1979). The influx of terrigenous siliciclastics is also present in some sections close to the base of the formation. For instance, at B-NW3, dispersed rounded quartz grains are found in beds only ca. 9 m above the top of the Buah Formation (Enclosure I-5). Therefore, where it is exposed, the lower unit of the Ramayli Member records a fine oscillation of eustatic sea-level rise and fall towards an overall shallowing trend (Enclosures II-2 and II-3).

Shallowing up through the Ramayli Member is also accompanied by extensive microbial mat growth, with patches of stromatolite domes. The onset of siliciclastic siltstone and mudstone deposition in the upper unit of the member marks a further shallowing. The growth of low, ‘squat’ stromatolite domes in a single marker horizon around the Buah Dome indicates that water depths were extremely restricted, limiting vertical growth of the mat communities to only a couple of centimetres. At B-NW3, the interval above the ‘squat’ stromatolite marker horizon is dominated by fissile red mudstones with pseudomorphed anhydrite rosettes, few microbial mat beds and a regular occurrence of emergent surfaces and chert breccia crusts. Together these suggest that depositional environments had become supratidal and a sabkha setting is the most obvious interpretation for this interval. Added to this are the presence of dissolution and collapse beds at B-NW3 and lateral equivalent halite beds at Wadi Shital. Thus at the top of the Ramayli Member it can be demonstrated that extremely restricted evaporite deposition occurred in at least one period of marine inundation into these sabkha environments, forming ‘Evaporite unit 1’ (Enclosures II-1, II-2 and II-6). More importantly, this also demonstrates that salt deposition did occur during the Ediacaran–Cambrian interval as far east as the Al Huqf and was not restricted to north and south of the region as previously thought (Mattes and Conway Morris, 1990).

Sediment progradation caused a shallowing-up sequence to develop from above fair weather wave base grainstones through peritidal lagoonal micrites to supratidal sabkha muds and possibly salinas. Essentially, the pre-existing palaeotopography in and around the top surface of the Buah Formation was filled to the limit of the available accommodation space. The inferred presence of palaeo-topographic highs and the shallow, restricted embayments that would have been present around them helps explain the subtle lateral lithofacies differences between the Ramayli Member at Buah Dome, Sirab and Wadi Sidr sections (Enclosures II-2 and II-3). Finally, it is worth noting that in subsurface exploration wells close to the Al Huqf area, such as Miqrat-1 to the north, the interval between the Buah Formation and what is often interpreted as truncated Ara Group has a high gamma log response. It is possible in the light of our outcrop work that this corresponds to the red mudstones and siltstones of the upper Ramayli unit, as these represent an interval dominated by fine siliciclastics and not carbonates.

The Salutiyyat Member crops out in a NE-trending zone across the Al Huqf from Wadi Salutiyyat to Wadi Shuram. The existence of the member and its basal unconformity indicates that there must have been a series of palaeo-highs across the Al Huqf at the onset of Sirab deposition, which exposed the Nafun Group to subaerial erosion. By definition, this also therefore outlines the Nafun Group palaeo-high upon which it was deposited (Enclosure II-2). This linear geometry might suggest that the palaeo-high Nafun ridge was fault-bounded, and indeed it exactly coincides with a basement terrane and sutures. Movement on these NE-SW ridge-flanking faults, after the onset of Ramayli deposition, to produce a shallow trough along this axis would have allowed temporary flooding and accumulation of the Salutiyyat Member until all accommodation space was filled (Enclosues II-2 and II-3). Subsequent basin-wide fault subsidence would then have allowed the rhythmic cycles of the Shital Member to be deposited across all previous stratigraphy as outlined below for the deposition of the lower Shital unit (Enclosure II-4). As such, the Salutiyyat Member could thus represent the localised initial stages of what became regional fault-controlled basin subsidence. This would fit well with the interpretation of the Ramayli Member as a highstand fill around existing palaeo-topography.

Whilst the carbon-isotope chemostratigraphy currently underway should help in future to resolve this issue, there are some further field observations, which might be used in support of this model. Wadi Shuram contains some of the thickest sequences of the Sirab Formation, clearly indicating that it was a significant depocentre during accumulation of the formation (for instance Wadi Shuram-1; Enclosure II-1). However, the thick ridges of Sirab Formation, which crop out from WS-8 to WS-1 and further southwest, were deposited on the Nafun Group palaeo-ridge, indicating that what was a palaeo-high prior to the onset of Sirab deposition, subsequently became a significant palaeo-low during its accumulation. In fact, far to the southwest along these ridges, the Shital Member appears to unconformably overlie ridges of Shuram Formation with little or no Salutiyyat Member in-between [UTM 550725, 2220630]. The dramatic topographic reversal from the beginning of Sirab deposition to its end at this locality is most likely to have been caused by fault movement on the ridge flanks. We also located further outcrops just north of the Mukhaibah Dome in the south of the area, where upper Shital Member beds, including the buff sandstone marker horizon, directly overlie Shuram Formation laminated siltstones and sandstones (Enclosure I-1). This provides evidence that the presence of such Nafun Group palaeo-highs even persisted during lower Shital Member deposition.

At Wadi Salutiyyat, there is an interval of poor exposure directly below the first appearance of Shital Member cycles (Figure 10). A single bed within this interval is composed of dolospar rhombs, which in other sections have always been found in association with evaporite units such as the halite beds at Wadi Shital (Enclosure II-1). If this does indeed correspond to an evaporite unit, it limits the lithological correlative options as we have only so far recorded three such intervals within the Sirab Formation. Evaporite unit 3 directly overlies the sandstone marker horizon, and Evaporite unit 2 occurs in sections after a significant sequence of Shital cycles. As explained above, it is unlikely that the sequence below Shital Member cycles at Wadi Salutiyyat represents the lateral time-equivalent of cycles in neighbouring sections. Therefore, the best option is that the evaporite unit in the Salutiyyat Member is Evaporite unit 1, which equates to beds in the upper Ramayli unit. However, this potentially produces a second paradox in the Salutiyyat Member, this time in relation to subsidence rather than its age. Elsewhere across the Al Huqf, the Ramayli Member was deposited as a shallowing-up sequence during tectonic quiescence. If the Salutiyyat Member is partly time-equivalent to the Ramayli Member as suggested, then the topographic high upon which it was deposited must have undergone rapid subsidence to create accommodation space when simultaneously elsewhere (such as the Buah Dome to the northwest) that space was all but filled. This again suggests that the principal depositional controls acting at this time were not eustatic sea-level changes but fault movement causing localised hangingwall subsidence and footwall uplift.

The most characteristic feature of the Shital Member are the bed couplets which repeat on a 1–2 m scale for 10s to 100s of metres of stratigraphy, particularly in the lower half of the member (Figure 11a; Enclosures II-1 and II-7). The lithologies of these couplets suggests that each one represents a shallowing-up sequence from peritidal carbonates to evaporites to temporary emergence. The thicker, main bed in each couplet is composed of a massive weathering, well-laminated (often clearly wavy microbially laminated) light grey or toffee-coloured dolomicrite. Passing up through the bed, large bird’s eye fenestrae are often common, probably generated from gases released from the microbial mats lower in the bed. This would indicate that lithification occurred only after each bed of mud had accumulated. The fine carbonate mud of which the beds are composed, plus the development of microbial mats and stromatolite domes in some cases, are indicative of a shallow lagoonal to intertidal environment. Chert replacement along laminae is often seen throughout the beds and often particularly towards the top. Thus the development of chert-replaced laminae and chert replacement breccias suggest that these beds were not only deposited in shallow lagoonal settings, but were in the shallowest parts of the lagoon between high and low tides. The associated pockets of dissolution and collapse breccias within these beds also indicates repeated, post-lithification, periodic emergence of the top bedding surfaces to allow the onset of karst dissolution.

The second bed in each couplet can vary depending on localised environments. Typically this is a thin, 10–20 cm-thick weathered-back bed and composed of fine red dolospar after chicken-wire anhydrite, or chicken-wire dolomicrite from the dissolution of gypsum. In either case, these are thin evaporite caps to the main, more massive bed below in each couplet. These evaporite caps are generally poorly exposed, but are clearly missing altogether in some sections that presumably either had less accommodation space or failed to become restricted enough for the bicarbonate supply to be exhausted (for instance the top half of section WA-1). Conversely, at Wadi Sidr to the south, these evaporite caps are about 20–40 cm-thick and composed of yellow, recrystallized dolospar. At Wadi Shital, this type of coarse, blocky dolomite was found in association with halite and it is possible that the evaporite caps were better developed and thicker in the south of the Al Huqf area.

The nature of the basal bounding surface of the Shital Member and the bedform style prevalent in the lower part of the member together have far-reaching repercussions for the evolution of the Sirab Formation and the Al Huqf area as a whole during the Ediacaran–Cambrian. As described above, the lithofacies developed at the top of the Ramayli Member are highly variable laterally as all remaining accommodation space was filled around Nafun Group palaeo-topographic highs. However, the lower Shital Member, wherever it is encountered in the field, is always composed of repetitive cycles of peritidal carbonate-evaporite bed couplets, and is thus remarkably uniform in lithofacies across the entire Al Huqf area. Each rhythm pair is not only remarkably uniform in thickness as it repeats up through the stratigraphy, but they are also laterally continuous with little or no change in bed thickness. There seems to be no evidence in the field of gradual, diachronous eustatic transgressive flooding across the top of the Ramayli Member. Instead, the flooding appears to have occurred simultaneously, in geological terms, in all sections. Perhaps the best explanation of this phenomenon is that of a finely balanced interplay between incremental fault-controlled basin subsidence and minor fluctuating eustatic sea levels from sub-Milankovitch cyclicity, such as Dansgaard-Oeschger events or Bond Cycles (Enclosure II-7). Each time that 1–2 m of accommodation space was created, deposition resumed and filled the space with a wedge of peritidal dolomicrite, followed with evaporites as the shrinking water mass became isolated from marine influence. Once the water had evaporated, the top surface of the bed was exposed until the next flooding and the next cycle of deposition began.

Thus, the stratigraphy suggests that regional, basin-bounding controlling faults became active at the onset of Shital Member deposition (Enclosure II-4). The fact that the member is at least 150 m thick (and in areas such as Wadi Shuram easily over 250 m thick), and that elements of facies cyclicity can be found in sections right to the top of the member at some sections (for instance WA-1; Enclosure II-1), indicate that these faults remained active and controlled deposition across the area (Enclosures II-4 and II-7). It also means that the Al Huqf area, whilst remaining a structural high during the Ediacaran–Cambrian, was not subaerially exposed as has generally been thought. Within this regional high, Shital Member deposition indicates development of a localised symmetric or slightly asymmetric graben. The presence of cyclical carbonates of the lower Shital Member from Wadi Aswad in the north all the way to Wadi Sidr in the south, a distance of ca. 85 km, makes a broadly north-south orientation most likely, and this fits well with the trend of pre-existing faults across the Al Huqf. This trough was inverted post-Sirab Formation deposition, presumably during the Angudan Unconformity, accompanied by karstification and then subsequent incision by the basal Haima Supergroup clastics. The erosion of the entire Sirab Formation at the Buah Dome during this time means that 250 m+ of carbonates were removed prior to Haima deposition.

In sections from the Buah Dome to Wadi Sidr, red recrystallized dolospar and mudstones can be found in association with dissolution breccias about 70–100 m up from the base of the Shital Member (Enclosure II-1). These lithologies and bedforms are clearly associated with evaporite deposition and emergence where seen in good exposures, such as at the top of the Ramayli Member at Wadi Shital ST-1, and indicates that there is a second evaporite unit present at this level across the area (Enclosure II-7). Enough accommodation space was created for a restricted water mass to develop and deposit an evaporite unit along the length of the trough, perhaps due to eustatic sea level rise and highstand flooding over barriers. The remaining interval to the base of the quartz sandstone marker horizon is poorly exposed in many sections but it appears that the deposition of peritidal carbonate - evaporite cycles resumed after deposition of ‘Evaporite unit 2’.

The presence of a thin quartz sandstone unit ca. 2–4 m thick from north to south across the Al Huqf (Enclosure II-1), is best explained by an episode of subsidence across the region providing enough accommodation space for accummulation and preservation of this thin unit. It is overlain in Wadi Aswad and Wadi Shital by an evaporite unit and this would seem to confirm the creation of space similar to that for ‘Evaporite unit 2’. Thus the base of the sandstone is effectively a correlatable flooding surface. The unique influx of continental-derived siliciclastics at this point in Sirab Formation deposition in an otherwise carbonate-dominated depositional environment means that this unit provides a robust marker horizon in the lithostratigraphic framework of the area.

Again, there is a change in the style of subsidence post-’Evaporite unit 2’. At Wadi Aswad to the north, the deposition of cyclical peritidal carbonates resumed until passing into an interval of ‘tufted mats’ at the very top (Enclosure II-1). However, at Wadi Shital and Wadi Shuram there seems to have been significant deepening, allowing the development of microbial mounds or reefs. At Wadi Shital, these take the form of conophyton build-ups, possibly topped by thrombolites (Figures 14c and 14d), which would have formed in subtidal environments (Donaldson, 1976; Grotzinger, 1989). At Wadi Shuram there are good examples of small, individual thrombolite mounds (Figure 21g), as well as large domal stromatolitic thrombolites (Figure 20h). All of this is evidence of more open-marine conditions not envisaged for the peritidal-evaporite couplets, and provides good evidence for more increased general subsidence in this area during upper Shital deposition (Enclosure II-5). Thus, spatially at this time, the Sirab trough may have developed into fringing low-angle carbonate ramps dipping in towards a depocentre across a NE-SW axis between the Wadi-Shital - Wadi Shuram areas (Enclosures II-5 and II-8).

Thus, four carbonate-evaporite cycles can be recognised from the base of the Sirab Formation to the top of the Shital Member. The lowermost of these corresponds to the Ramayli and Salutiyyat members themselves. However, where the Salutiyyat is extremely thin or missing over the crest of Nafun Group palaeo-highs, such as to the southwest in Wadi Shuram (Figure 9c), this lowermost cycle may also be absent (Figure 23) and the cycle marking the base of the Shital Member will rest unconformably upon subaerially eroded Nafun Group and appear to form the basal cycle in the succession. Evaporite units 2 and 3 effectively subdivide the Shital Member into two more carbonate-evaporite cycles, with the beginnings of a fourth represented by the upper Shital Member (Figure 23).

The upper Shital Member perhaps demonstrates a greater differentiation in subsidence across the Al Huqf than the lower half of the member (Enclosure II-8). However, the Aswad Member seems to represent a relatively uniform facies across the whole area (Enclosure II-9). Thus the increased differential subsidence of the upper Shital Member appears to have gradually become a more general, uniform regional flooding. Although the surviving stratigraphy of the Aswad Member is relatively thin (≤ 30 m), there is little change in its gross litho- and bio-facies from Wadi Aswad to the southern tip of its exposure south of the Sirab area, a distance of ca. 60 km (Enclosure II-1).

The association between framestone thrombolites, laminated thrombolites and oncoids is complex and gradational both laterally and vertically. Laminated, stromatolitic thrombolites can be seen to grade laterally and upwards into more vuggy, thrombolite reef beds and as such are presumably the stage prior to full reef development (Schröder, 2000) (Figure 21d). The thrombolite reef beds themselves form a framestone with a high primary porosity between mesoclots. In some this has clearly been plugged by coarse sparry calcite at a much later stage (Figure 21f). There is also a gradation from clotted fabric into oncoids, with patches where oncoids appear to have stuck together and been followed by the development of a clotted fabric (Figure 21b). The development of extensive oncolite beds at this time indicates that the Al Huqf area effectively developed into a geographically extensive shallow-marine ramp where tidal swash continuously rolled shoals of microbial balls in-between ‘fields’ of thrombolite framestone beds and more localised bioherms (Enclosures II-5 and II-9). In at least one locality, south of Wadi Shital, water depth became sufficient for microbial crinkly laminites to form around the flanks of one such localised thrombolite mound. Conversely, at Wadi Aswad, oncolite beds contain a component of organic matter not seen in the Wadi Shital area. It is possible that these oncoids were forming in slightly deeper, but current dominated water than those at Wadi Shital.

This study defines a new formation, the Sirab Formation, in the Haushi-Huqf region of Oman. It generally lies conformably on the Ediacaran Buah Formation and underlies the siliciclastics of the Amin Formation of the Haima Supergroup. It therefore occupies the equivalent lithostratigraphic position of the Ediacaran–early Cambrian Ara Group of Oman. The formation is subdivided into three principal members. At the base, the lower section of the Ramayli Member is a transitional unit overlying the Buah Formation, and demonstrates a lithological change from grainstones to the packstones and wackstones typical of the Sirab Formation. The upper Ramayli Member is an evaporitic mudstone unit. The Shital Member lies disconformably upon the Ramayli Member and is characterized by a thick succession of fault and/or eustatic sea level controlled deposition of cyclical peritidal carbonates and evaporites in a shallow trough or graben on the regional Huqf High at that time. At the top of the Sirab Formation, the Aswad Member is significant in that it contains excellent examples of microbial build-ups including thrombolite framestone reefs, laterally extensive thrombolite ‘fields’, oncolites and rare ‘crinkly’ laminites. A fourth member, the Salutiyyat Member, can be recognised locally where the Sirab Formation lies on eroded Nafun Group palaeo-topographic highs, and is probably the chrono-stratigraphic lateral equivalent at least in part of the upper Ramayli Member.

If it is accepted that the Sirab Formation occupies the equivalent lithostratigraphic position as the Ara Group, and indeed contains some similar lithofacies such as evaporites and thrombolites, then the next logical step would be to attempt high-resolution correlation between the two. However, outstanding stratigraphical issues remain with the Sirab Formation, which cannot at present be answered by field relationships alone. The best solution at this time is to identify three models, each of which would resolve the chronostratigraphic relationship of the Sirab Formation to the Ara Group. The first two of these simply accept that whilst there are some similar lithofacies shared between both Sirab and Ara, they are in fact of two quite distinct ages.

Model 1, would place the Sirab Formation older than any Ara stratigraphy; essentially a late Neoproterozoic unit directly overlying the Buah in the Huqf, which is missing subsurface, but would have a projected lithostratigraphic position beneath the Ara (pre-A0 stringer). Model 2 is the diametric opposite, with a marked unconformity on top of the Buah in the Huqf representing Ara deposition subsurface, and with the onset of Sirab depositon post-Ara Cycle A6. In this scheme the Sirab Formation would be entirely younger than the Ara and would have been deposited during the early Cambrian. Finally, Model 3 is a compromise whereby at least some, if not all, of the Sirab Formation is a chronostratigraphic equivalent of the Ara Group.

These three simple scenarios are in fact complicated further by the presence of a marked disconformity within the Sirab Formation at the base of the Shital Member. This raises the possibility that the Ramayli Member could be a chronostratigraphic Ara equivalent disconformably overlain by Shital and Aswad members that are younger than any Ara. Or, conversely, that the Ramayli Member is older than any Ara but the Shital and Aswad are chronostratigraphic equivalents. Either of these latter two scenarios are essentially more complex versions of Model 3. In any event, the presence of the Sirab Formation in the Huqf region can now be recognized, but clearly the more detailed stratigraphic relationships between Ediacaran–early Cambrian in outcrop and subsurface remain to be resolved.

This research was funded by Petroleum Development Oman (PDO). The Ministry of Oil and Gas and the Exploration Management of Petroleum Development Oman L.L.C. are thanked for their permission to publish this paper. We would also like to thank Salim Al Maskery and Shuram Oil & Gas L.L.C. for logistical support during field surveying. This research has benefited from fruitful discussions in the field with many geologists over the years, including more recently Joerg Mattner, John Graham and George Sevastopulo. In particular, we would like to thank the field assistants from 1999–2001; Liz Hawkins, Paul Bellingham and Claire Marie Mulhall. PDO geologists past and present have also engaged in helpful discussions and encouragement, and of these we would particularly like to thank Joachim Amthor, John Millson, Jan Schreurs and Aly Brandenburg. This manuscript was significantly improved by the comments and suggestions of Moujahed Al-Husseini following two anonymous reviewers. The authors thank GeoArabia Designer, Nestor “Niño” Buhay IV, for designing the paper for press.

Chris Nicholas is Lecturer in Geology at Trinity College, The University of Dublin. He graduated in Geological Sciences from the University of Birmingham, UK in 1989 and was awarded his PhD in 1994 on ‘Strontium isotopes across the Precambrian/Cambrian boundary’ by the University of Cambridge, UK. As a postgraduate and post-doctoral researcher he investigated issues of Ediacaran–Cambrian palaeontology, stratigraphy, sedimentology and geochemistry in Arctic Canada, Newfoundland, Kazakhstan, eastern Siberia and Oman. Chris moved to a permanent Lectureship position at Trinity College Dublin in 2001. Since 1998, he has conducted extensive field work on the Mesozoic and Tertiary geology of East Africa, including publication of the first geological maps of some regions, and formed the ‘Tropical Geology & Exploration’ Research Group at TCD. This group specialises in using applied field geology to help direct subsurface exploration in problematic petroleum frontier areas and has operated in Tanzania, Uganda, Democratic Republic of Congo, Mozambique, Oman and Java. From 2008–2010, Chris acted as Chief Geologist for Dominion Petroleum in East Africa and continues to pursue a dual role linking Academia with the Petroleum Industry.

nicholyj@tcd.ie

Sophia Gold graduated with a first class degree in Geology from Trinity College, The University of Dublin in May 2006. That same year she began her PhD also at Trinity College Dublin, supervised by Chris Nicholas, on a project entitled ‘Stratigraphy and correlation of the Sirab Formation, Al Huqf area, with the subsurface Ara Group play systems of Oman”, funded by Petroleum Development Oman. Her PhD was awarded in 2010 and Sophia now works for Tullow Oil in London.

sophia.gold@tullowoil.com