Al Khlata Formation is an oil-bearing reservoir in Oman. Its origin during the Late Palaeozoic Gondwana glaciation in the southern part of the Arabian Peninsula is demonstrated by the glaciogenic deposits of the formation in Oman. Outcrops of the Al Khlata Formation occur in a belt parallel to the Huqf fold axis, the best outcrops being found in the two wadis Al Khlata North and South. In the southern wadi, glacial deposits rest directly on dolomites of the Precambrian Khufai Formation showing northeast-southwest trending glacial striations.

Earlier workers proposed that the direction of ice flow was from the southwest to northeast. This is not accepted in this paper, where evidence is presented to support an ice-flow direction from northeast to southwest. Later fluvial flow towards the northeast possibly resulted from the collapse of the continental margin towards the new Proto-Arabian Sea and its subsequent subsidence.

We propose that the Early Permian glaciation in Oman resulted from crustal uplift just prior to the calving of a microcontinent along the northeast Arabia margin of Gondwana and the creation of Neo-Tethys. It is suggested that the triple-junction area where the Neo-Tethys and the Proto-Arabian Sea were later to meet, was a site of sufficient thermal uplift to become a center of mountain glaciation.


The Late Palaeozoic glaciation of Gondwana lasted for over 100 million years. Several small and isolated ice caps (Crowell, 1983) were distributed within the supercontinent, and nowadays glaciogenic deposits can be found on all the former Gondwana continents. The world-wide occurrences of the Permo-Carboniferous glacial deposits have been extensively reviewed in Hambrey and Harland (1981) and Eyles (1993).

The earliest evidence for the presence of Permo-Carboniferous tillites in southern Arabia (Figure 1) was given first by Morton (1956) and then by Helal (1965) and confirmed by McClure (1980). Kruk and Thiele (1983) report the occurrence of tillites overlying a polished and striated pavement in Yemen. McGillivray and Husseini (1992) discuss the Unayzah Formation reservoir rock (Al-Laboun, 1987) of Saudi Arabia; they mention that Hughes-Clarke (1988) had inferred the existence of a peri-glacial environment in central Arabia at that time, a suggestion that was not refuted.

Confirmation that Permo-Carboniferous glacial rocks also occur in southern Oman can be found in Braakman et al. (1982). Granville (1982) described the occurrence of oil in glaciogenic rocks of that age in Dhofar, and a description of the depositional facies, leading to a sophisticated model of the Al Khlata environments of deposition, can be found in Levell et al. (1988). The discovery of striated pavements underlying the Al Khlata Formation in the Huqf area indicated that the glaciation was centered close to the Arabian Sea coast.

This paper aims to present a new hypothesis for the origin of the Al Khlata Formation glaciation as a result of ongoing study of the outcrops of the formation in Oman. Before this, the paper will address briefly the different facies that are exposed in the outcrops.


Oman once formed a part of Gondwanaland, and is underlain by continental crustal rocks that are over 700 million years old. In the late Precambrian, around 620 million years before the present, it was close to the South Pole and was covered by ice. The occurrence of dropstones indicates the presence of floating ice during or towards the end of this glaciation. Sea level rose following the melting of the ice, and the climate became warmer as Oman moved away from the Pole. During the Permo-Carboniferous (Besems and Schuurman, 1987), or perhaps more precisely the Early Permian (Eyles, 1993), Oman experienced another Ice Age in the Huqf area of southeast Oman, even though only 40-45° from the equator (Scotese and Langford, 1995). This glaciation resulted in deposition of the Al Khlata Formation (Levell et al., 1988), the subject of this study.

Over much of Arabia, deep erosion coincided with the time of the Permo-Carboniferous glaciations, much of it probably resulting from a glacially-induced lowering of global sea level. The erosional period was followed by a mid to Late Permian transgression, which in the Oman Mountains resulted in shallow-marine carbonate rocks onlapping to the north over a land surface comprising Lower Palaeozoic and Precambrian strata.

It is suggested here that the Early Permian glaciation in Oman resulted from crustal uplift just prior to the calving of a microcontinent along the northeast Arabian margin of Gondwana and the creation of Neo-Tethys: similar uplift may have occurred along the southeast margin of Arabia where India (and the intervening Helmand Block, now South Afghanistan) could already have been approaching separation from Afro-Arabia (e.g. Glennie, 1995). Thus the Early Permian glaciation in southern Arabia probably coincided with some of the earliest movements that brought about the eventual break-up of Gondwana into a series of individual continents and microcontinents.


The Al Khlata Formation is known in the subsurface of much of southern Oman, and as far north as the Fahud field. The formation crops out along the western flank of the Huqf fold axis over an area of about 150 by 10 to 20 km (Platel et al., 1992 (two papers); Dubreuilh et al., 1992). The best outcrops are found in the southern Huqf in wadis Al Khlata North and Al Khlata South (Figure 1). Other newly-described outcrops occur a little further north in three more wadis, Shab Nakad North, Central and South. All other outcrops to the north of this latter area are generally poorly preserved because of extreme weathering and deflation associated with flanking Quaternary sabkhas, and do not contribute much additional information; they still have uses, however, for the simplest of comparisons. Only one locality in the southern part of this northern area is important; Ain Hindel. (This locality has been reported as Ain Hindi in previous published literature. The correct name is Ain Hindel.) Here is an excellently preserved grooved and striated pavement; unfortunately, no Al Khlata sediments are present, the pavement being overlain by cemented gravels of probable Holocene age.

The Al Khlata Formation comprises four major facies-indicative lithological units, each of which can be subdivided further; diamictites, conglomerates, sandstones and fine-grained siltstones and mudstones (Figure 2, see also Levell et al., 1988).

Diamictite Facies

Massive Diamictite (tillite)

This facies normally forms the lowest unit of the Al Khlata overlying the striated surface of the Precambrian Khufai Formation. It crops out mainly in Wadi Al Khlata South but also occurs in Shab Nakad North. This facies contains well rounded boulders and cobbles up to 1 meter in diameter, of which the most characteristic is granite, in a matrix of reddish-brown clay, and lacks any stratification. Some boulders are covered with multiply oriented striations. The upper contact may be either sharp or a complex erosional surface with injection structures (Figure 3). Because of its stratigraphical position directly overlying the striated pavement it has been interpreted as a lodgement or melt-out till (Levell et al., 1988) deposited beneath or at the melting snout of a retreating glacier.

The origin of the granite boulders found in the diamictite is not known. There are three known areas of granite, all relatively close to the Arabian Sea coastline. One area is the Murbat-Halaaniyat (former Kuria Muria) Islands about 280 km southwest of Wadi Al Khlata, and another in Jebel Ja’alan, about 330 km to the northeast of Wadi Al Khlata, together with a smaller area on the Gulf of Oman coast near Qalhat (Gass et al., 1990). Significantly, there is also a small occurrence of granite/granodiorite in the northern Haushi-Huqf area (Dubreuilh et al., 1992; see Figure 1). These sites of Precambrian granite occur in the middle and at either end of an ancient positive area known as the Huqf Arch, so the granite boulders found in the Huqf area could have come from similar sites exposed during glaciation anywhere between these extremes. In this respect, it is pertinent that the largest known boulder of granite within the Al Khlata Formation, about the size of a four by four wheel drive vehicle, is exposed above the present deflation surface about 100 km north of Wadi Al Khlata, about half way between Wadi Al Khlata and the northern Huqf outcrop of granite.

Stratified Diamictite

The best exposures of this facies are found as isolated bodies in the upper reaches of Wadi Al Khlata North and Wadi Al Khlata South. It also occurs in the lower reaches of Wadi Al Khlata South. The stratified diamictite has a matrix of mixed sand and claystone (Figure 4) with isolated boulders similar to those found in the massive diamictite. Its thickness varies from 1 to 3 m. It may have originated as a sub-glacial melt-out.

Conglomerate Facies

Cross-stratified Sandy Conglomerate

This facies is common to all localities reported here. It is found predominantly in Wadis Al Khlata North and South where it attains a maximum thickness in excess of 6 m. The palaeocurrents in this facies are towards the northeast. This facies is overlain by the Cross-bedded Sandstone and Pebbly Sandstone Facies and is underlain by the Gravel Lag Facies. It is interpreted as the product of fluvial reworking underlying diamictites and other facies.

Gravel Lag

The Gravel Lag Facies varies in thickness from 20 to 80 cm and directly overlies the Massive Diamictite Facies. It contains lots of reworked and rip-up clasts of diamictite in a matrix of medium to coarse, iron-stained sandstone. It is overlain by the Cross-bedded Sandstone and Pebbly Sandstone Facies.

Massive Gravel

This facies consist mainly of gravels in a sandy matrix. It is found in all the localities described in this study. The best outcrops, however, are found in Wadi Al Khlata North and Shab Nakad North. A typical thickness of this facies is about 3 m. It is overlain by pebbly sandstone but the base is not seen.

Sandstone Facies

Rippled Sandstone

The Rippled Sandstone Facies dominates the outcrops in Shab Nakad South but also occurs in Wadi Al Khlata North, Wadi Al Khlata South and Shab Nakad North. The sandstone is fine- to medium-grained with climbing-ripple structures. The flow direction is consistently towards the southwest. The climbing ripples may indicate a degree of vertical accretion as the sands were slowed on reaching a body of standing water such as an ice-margin lake. Locally, this sandstone unit is also graded, forming fining-upward layers about 6 cm, or more rarely, up to 15 cm thick. Load structures are well developed in some horizons where they overlie claystone.

Cross-bedded Sandstone and Pebbly Sandstone

The Cross-bedded Sandstone and Pebbly Sandstone Facies crops out at all the localities described in this paper. The sandstones vary in thickness from a common 1 or 2 m to 5 m at the northern end of Wadi Al Khlata South (Figure 5). This facies is underlain by the Cross-stratified Sandy Conglomerate Facies and overlain by a 1 to 2 m cover of cemented Quaternary gravels and sand. The palaeocurrent directions are consistently towards the northeast. Earlier workers (Levell et al., 1988; Braakman et al., 1982) have interpreted this facies to be of glacio-fluvial origin, deposited during the deglaciation process; thus they believe ice movement to have been from southwest to northeast. We refute that interpretation, believing firstly, that the direction of ice movement was in the opposite direction, from northeast to southwest; and secondly, that there was a considerably time gap, possibly of a few million years, between the end of glaciation and deposition of these sands. Some of the evidence to support our interpretation will be presented under ‘Striated Pavements’.

Fine-Grained Facies

Deformed Silt/Mudstone

A unit of deformed siltstone and mudstone is present in the central area of Wadi Al Khlata South. It overlies the Massive Sandy Diamictite Facies and in places directly overlies the striated Khufai Formation. The unit is slumped, stratified, deformed and shows plenty of microfaults which dominantly dip towards the southwest. Some excellent examples of dropstones are also present in this facies. In the central area of the wadi, the deformed siltstone unit directly overlies the stepped and striated Khufai dolomites. This can only be explained by the fact that the Khufai dolomites at this locality is cut by many faults, which probably implies that the earlier deposited diamictite was prior to the deposition of the siltstone unit. Deformation in the siltstone implies that it had been deposited on a plastic substrate.

The siltstone unit has a thickness of up to 3 m. The unit has a thin cover of cemented Quaternary gravels and sand and overlies the Massive Diamictite Facies but at other localities it overlies the Khufai pavement. It is interpreted as having been deposited at the edge of a glacial lake.

Laminated Siltstone/Mudstone

This facies occupies part of the northern wall of the lower reaches of Wadi Al Khlata North and consists of a coarsening-upward sequence of dark gray siltstones at the base grading up to interbedded silt and very fine sandstone (Figure 6). The whole sequence is laminated and has been highly weathered, especially in the lower part, due to the effects of modern sabkha conditions. It is overlain by a thin cover of cemented Quaternary gravels and sand. The base of this facies is not exposed.

Black Shale/Mudstone

The Black Shale/Mudstone Facies is found in a small outcrop near the mouth (downstream end) of Wadi Al Khlata South and Shab Nakad North. The facies is thinly laminated and gradually passes upwards into gray laminated siltstone.

Graded Rhythmic Siltstone

This facies comprises cm-scale couplets mainly of silt and clay, and has been recorded in small outcrops in Wadi Al Khlata South, where it is about 2 m thick, and Shab Nakad North where it attains a thickness of up to 3 m. The small-scale couplets are interpreted as varves, the result of deposition in a peri-glacial lake that had its supply of silt-size sediment cut off each winter because of freezing conditions. The facies is overlain by the Stratified Diamictite Facies and underlain by the Laminated Siltstone.


Striated glacial pavements provide the best evidence of the former presence of moving ice, whether in the form of a sheet or a valley glacier. The trend of the striations conforms to the alignment of ice movement but does not give the sense of movement. Differences in the way that striations terminate have been used as evidence of the direction of movement, but rarely are they fool-proof on flat surfaces. At two localities, however, we believe that the sense of movement is displayed without ambiguity.

In Wadi Al Khlata South and Shab Nakad North, the glacial pavement underlying the Al Khlata has several unfaulted steps, some 10 to 30 cm high and about 2 m apart (Figure 7) in the former locality and 1.5 m high in the latter, that descend to the southwest; the striations are preserved at all levels, so the steps are not an artifact of recent exposure and erosion. The striation distribution at Shab Nakad North, is especially pertinent in this respect (Figure 8). Such steps would be preserved by the lee-side ‘plucking’ action of descending ice movement towards the southwest. Had the ice moved up the steps to the northeast, however, the edges of the steps would have been beveled and strongly grooved, rather than merely striated, by the boulders and pebbles held in the ice.

At Ain Hindel is an exposure of spectacular grooves cutting the polished dolomite surface of the Khufai Formation. The grooves trend northeast-southwest (45-225°) and are up to 1 m deep. Overprinting these grooves are another two sets of finer striations, each of which have a slightly different orientation, one trending 219° and the other 242°. The change in trend is interpreted as a lateral shift in the ice direction with time. The grooves would represent the original direction of ice movement, and the sets of fine striations probably indicate a shift in direction of thinner and lighter ice towards the end of the glaciation; alternatively, and much less likely, the grooves and striations represent three different glaciations.


By not accepting a direction of ice flow from southwest to northeast, as indicated by foresets in what we interpret as post-glacial fluvial sandstones, the writers have set themselves a problem in explaining a reversal of sediment transport direction in sequences that are physically in contact with each other. If we accept for the moment that ice movement was from northeast to southwest, as strongly indicated by the stepped striated pavements, then a plausible hypothesis is essential to account for the apparently conflicting direction of later fluvial flow.

The Early Permian glaciation in Oman took place no farther from the Equator than about 45° (Scotese and Langford, 1995). Crowell (1995) mentions but does not illustrate the presence of an Early Permian glaciation anywhere in Arabia, but maps a major area over northern India parallel with the presumed coast of Tethys. He also shows mountain glaciation parallel to the western margin of South America reaching to 30° from the Equator.

Plate reconstructions (Scotese and Langford, 1995) indicate that adjacent to the Afro-Arabian part of Gondwana during the Late Carboniferous-Early Permian were several large land masses that are no longer contiguous with it. These land masses include India, with the intervening Helmand Block (now South Afghanistan) to the southeast of Arabia, and Anatolia, Central Iran and possibly central Afghanistan-Tibet to the northeast (Figure 9). By the mid to Late Permian, some of these land masses had already broken away from northeast Arabia to create the intervening narrow Neo-Tethys Ocean (Glennie, 1995). Less certainly, a proto-Arabian Sea, now represented by the fragment of 145 million year old oceanic crust that is Masirah Island, may already have been uplifted over a future spreading ridge prior to crustal separation.

It is suggested that the triple-junction area just off the northeast corner of Oman, where Neo-Tethys and the Proto-Arabian Sea were later to meet, was a site of sufficient thermal uplift to become a center of mountain glaciation (Figure 10). A tentative glacial bias towards the southwest is suggested by increasing proximity to other centers of glaciation in southwest Arabia and in Ethiopia. A proto-Arabian Sea receives support from Madagascar which, apparently beginning in the Permian, formed a microcontinent between East Africa and India; there, a small area of tillites and fluvio-glacial deposits followed by 1,400 m of coal measures and red beds, is overlain by a middle Permian brachiopod fauna (Kent, 1974). Glaciation in Madagascar probably preceded that in Oman (Besems and Schuurman, 1987).

It is pertinent that the Huqf lies to the southwest of the proposed area of uplift northeast of Oman. How high the triple-junction area would have had to be to induce glaciers that could extend as far away as the Huqf is not known. The northerly direction of onlap of the mid to Late Permian shallow-marine limestones in the Oman Mountains indicates that at that time, the Arabian continental margin adjacent to Neo-Tethys was still elevated in a manner perhaps similar to the mountains of western Arabia flanking the modern Red Sea. The absence of glacial rocks of this age in the Oman Mountains could be because uplift on that scale induced erosion rather than deposition. It is known that the later Permian marine transgression began earlier in southern Oman than in the mountains.

To find a plausible site for the center of glaciation was relatively easy. To reverse the sense of sediment transport in the southern Huqf requires an understanding of the short-term history following continental break-up.

Once continental separation had been achieved, the marine flooding and opening of a through-going Neo-Tethys would have had a strong warming effect on the local climate; a rapid collapse of the adjacent centers of glaciation could be expected.

The time span separating the end of glaciation and the onset of fluvial erosion and deposition in the Huqf area is not known, but it could have been several million years. Following thermal uplift and crustal separation, the new and unsupported continental margin is generally unstable and collapses via a series of listric faults (Gorin et al., 1982) towards the newly forming oceanic crust. This process is very clearly illustrated in cross-sections of the North Sea’s Viking Graben (Ziegler, 1990), for instance, an aborted crustal separation.

If rapid subsidence took place along the new continental margin, then the reversal of relief could well have resulted in fluvial sands from the southwest replacing glacial ice from the northeast. The example of the currently uplifted flanks of the Red Sea some 20 to 30 million years after crustal separation cannot be taken as an analogue; their continued elevation is probably an artifact of the Red Sea’s inability to spread as it would wish. In Oman, a marine transgression followed deposition of the Al Khlata fluvio-lacustrine sequences both in the oil fields area and in the Huqf (Rahab Formation), so interior Oman could not have been elevated. The flow of fluvial sediments of the Al Khlata Formation towards the sea was possibly the result of a fall in sea level induced by another glaciation closer to the South Pole, but not matched in Oman because of the changed geography.


The Al Khlata Formation crops out along the western flank of the Huqf fold axis. The best exposures are found in the southern part of the outcropping area in wadis Al Khlata North and South, and Shab Nakad North, Central and South. The Al Khlata Formation comprises four major facies associations: diamictites, conglomerates, sandstones and fine-grained siltstones and mudstones. Each of these units can be subdivided into different facies. There are two known areas of granite from which the granite boulders found in the diamictite could have originated. These are the Murbat-Al Halaaniyat Islands area and the Jebel Ja’alan and Qalhat area. Limited evidence of boulder size suggests that the northern area is the more plausible.

The Early Permian glaciation in Oman resulted from crustal uplift just prior to the calving of a microcontinent along the Arabian northeast Margin of Gondwana and the creation of Neo-Tethys. Ice flow from the southwest to the northeast is not accepted and instead evidence was presented to support an ice flow direction from northeast to southwest. The triple-junction area just off the northeast corner of Oman, where Neo-Tethys and the Proto-Arabian Sea were later to meet, was likely to have been a site of sufficient thermal uplift to become a center of mountain glaciation responsible for the deposition of the Al Khlata Formation. Later fluvial flow towards the northeast is perhaps due to collapse of the continental margin towards the new proto-Arabian Sea and to lowering of global sea level because of other glaciations over Gondwana closer to the South Pole.


Juma Al-Belushi would like to thank Petroleum Development Oman LLC for sponsoring this project, which forms part of a Ph.D. project carried out at the University of Aberdeen. We also would like to thank Dr. J. Le Métour (BRGM), Dr. Richard Steel (PDO) and an anonymous reviewer for their comments on improving the quality of this manuscript. Thanks are also due to K. Al-Riyami and Omar Al-Jaidi (PDO) for their assistance in the field. Finally, we thank Petroleum Development Oman LLC and the Oman Ministry of Petroleum and Minerals for permission to publish this paper. The graphics for this paper were drafted by Gulf PetroLink.


JumaD.Al-Belushi is currently a final year PhD student at the University of Aberdeen working on the HC- producing glaciogenic sediments (Permo-Carboniferous) of Oman and Australia. He has been employed by Petroleum Development Oman since 1992 as an Exploration Geologist. Juma is a member of the AAPG and SPE and has participated in many international conferences. He has a MSc in Petroleum Geology from the University of Aberdeen and is scheduled to graduate with a PhD in October, 1996.

Kenneth W. Glennie received his BSc and MSc degrees from Edinburgh University. He then spent the next 32 years as an Exploration Geologist with Shell, working in New Zealand, Canada, Nepal, Oman, Iran and Turkey before spending 15 years on the North Sea geology based in London. He retired from Shell in 1987, and he is an Honorary Professor at Aberdeen University, spending part of each year since 1990 working on the desert of south-eastern Arabia and maintaining his interest in the Oman Mountains. He is a member of the AAPG and Geological Society of London, Edinburgh, Aberdeen and The Netherlands.

Brian P.J. Williams obtained a BSc (Honors) in Geology and a PhD in Sedimentology from the University of Wales. From 1964 to 1970 he worked as a Postdoctoral Research Fellow at the University of Ottawa, Research Fellow at the University of Wales, and a Senior Hydrogeologist for the Water Resources board in Reading. From 1970 onwards Brian was a Lecturer, Senior Lecturer and Reader in Sedimentology in the Department of Geology, University of Bristol. In 1988, Brian became Professor in Petroleum Geology at the University of Aberdeen where he is the director of the MSc course in Petroleum Geology. His current research interests include hydrocarbon reservoirs in Australia, Canada, Texas and the North Sea; non-marine clastic sedimentology and basin analysis. Brian is a member of the Institute of Petroleum, SPE, PESGB, AAPG, Society for Sedimentary Geology, International Association of Sedimentologists and the Geological Society of London.

First Page Preview

First page PDF preview