“Tuning” High-Frequency Cyclic Carbonate Platform Successions Using Omission Surfaces: Lower Jurassic of the U.A.E. and Oman
Published:January 01, 2000
- PDF LinkChapter PDF
Gordon M. Walkden, Jose De Matos, 2000. "“Tuning” High-Frequency Cyclic Carbonate Platform Successions Using Omission Surfaces: Lower Jurassic of the U.A.E. and Oman", Middle East Models of Jurassic/Cretaceous Carbonate Systems, Abdulrahman S. Alsharhan, Robert W. Scott
Download citation file:
Meter-scale cyclicity is developed on a number of Lower Jurassic peri-Tethyan carbonate platforms, and comparable cyclicity is known from basinal successions in parts of Europe. Analysis of the 234-m-thick Hettangian to Early Pliensbachian section of a recently constructed reference section through inner- to mid-shelf Jurassic carbonates in Wadi Naqab, U.A.E./Oman border, shows that facies patterns alone give an inaccurate and even misleading record of this cyclicity. Omission surfaces, however, can provide far more reliable evidence for sea-level change and enable the “tuning” of cyclicity data derived from facies studies. Clear distinctions need to be made, in the field, between a number of different marine and subaerial omission surfaces, including firmgrounds, hardgrounds, karsts, corrosion surfaces, and paleosols. Some of these surfaces display similar or ambiguous characteristics in the field, and difficulties can be compounded by the re-texruring and remodeling effects of burial diagenesis. High-resolution data, incorporating systematically observed and logged omission surfaces, have potentially high value in correlation and greatly improve the reliability of interpretations. In the Wadi Naqab succession the incorporation of carefully observed cycle-boundary data more than doubles the number of cycles seen. The data show that average duration of the high-frequency cycles was ca. 90,000 years, well within the range of Milankovitch forcing factors but outside the 40,000 years obtained from basinal Milankovitch successions of the same age in Europe. The difference reflects the incompleteness of the shallow-water succession, and it is likely that both types of cyclic successions in the Lower Jurassic share the same climatic and/or eusta tic causes. Fischer-plot analysis of the 149 cycles reveals the additional operation of two strong third-order sea-level fluctuations, and possibly two further smaller ones, between late Hettangian and early Pliensbachian. Subaerial omission surfaces are best developed and most numerous on the falling limbs of these, where cycles are thin. These points would correspond to the falling stage of third-order sea-level change when there was limited creation of new accommodation and maximum subaerial exposure. Omission surface data therefore support the case for both fifth-order and third-order sea-level change in the Jurassic. A particular characteristic of this and other shallow-water Milankovitch successions is their thin-bedded nature. This is likely to impart a strong flow anisotropy in reservoir settings, enhanced by stratiform weathering and diagenetic features, but many of the definitive features are likely to be missed in seismic sections and well logs. Actual omission surfaces can be hard to detect even in core.
Cyclicity within the range of fourth- to sixth-order frequency (10–100 thousand years) is a mark of Jurassic sedimentary patterns world-wide (e.g., Valdes et al., 1995; House, 1985). In the Lower Jurassic such cyclicity is noted in both basinal successions (House, 1985) and shallow carbonate platform successions (Crevello, 1991; Bassoulet and Bergougnan, 1981; de Matos et al., 1994). It has been attributed to orbital forcing of climate (e.g., Crevello, 1991; Weedon, 1993), although tectonic causes have also been invoked (Moses, 1995). In the peri-Tethyan carbonate successions of Morocco and the Arabian Peninsula the key evidence for a causative mechanism is the meter scale of the cycles, their widespread development, and the presence of subaerial discontinuity surfaces (including karsts, calerete, and soils) separating normal marine facies. Eustatic change of at least a few meters is, thus, the prime candidate (de Matos et al., 1994), providing a causative link with the very similar but much better documented eustatic cyclicity of the Middle to Late Triassic in the same province (e.g., Goldhammer et al., 1987, 1990; Fischer and Botrjer, 1991).
Cyclic changes with a 10–100 thousand years periodicity that are synchronous and worldwide provide exciting potential for high-resolution correlation, not only locally but globally, and not only between platforms but also between these and their basinal equivalents. This would be particularly valuable in shallow carbonate successions in which some of the bestbiostratigraphic markers, such as ammonoids, are persistently absent. The likeli hood that the great majority of cycle boundaries in high-frequency cyclic successions might represent precise, unique, and potentially correlatable ticks of the great geological clock has ensured that such successions throughout the geological column have received much attention in the last decade. Despite this, there remain many problems in the use of high-frequency cycles in correlation (e.g., Cotillon, 1995). These include:
separating eustatic and tectonic effects;
recognizing gaps in the shallow-water record caused by prolonged sea-level falls;
homogenization of the record in deep-water sequences;
no systematic means of reliably calculating cycle magnitude or period;
cycles may be incomplete or hard to distinguish.
The purpose of this paper is to address the last of these problems, that of improving discrimination within cyclic sequences. In many shallow-water cyclic successions the cycles (more appropriately, but seldom, termed cyclothems; e.g., Walkden, 1987) are commonly incomplete, lacking crucial evidence of a succession of depth-related facies. In some cases very similar facies, but representing separate cycles, are built directly on top of one another (cycle amalgamation). Reliable discrimination thus breaks down, reducing potential for correlation and for identifying causative mechanisms. This is particularly inhibiting where statistical techniques are being attempted, for example power spectral analysis in cases where astronomical forcing factors may be involved (e.g., Weedon, 1993). In sequence stratigraphic terms, the cycles produced by fifth-order cycles of sea-level change could be regarded as parasequences, but it takes confidence to designate a thin cycle, perhaps even comprising a single bed, as a parasequence.
Our study is based in the Liassic of the Musandam Peninsula (U.A.E./Oman, southern Arabian Gulf) (Fig. 1). Here we have established that careful study of omission surfaces is an independent tool in defining cycles, determining the nature of early Jurassic sea-level change, recognizing parasequence sets, and identifying lower-order (longer-term) patterns of relative sea-level change. This use of omission surfaces depends upon reliably distinguishing between a number of depositional, erosional, and diagenetic effects at bedding discontinuities that can create similar and sometimes easily confused phenomena (Fig. 2). Commonly these effects are difficult to resolve even under ideal outcrop conditions, and their recognition in core is particularly difficult.
The Lower Jurassic of the Musandam Peninsula
The Jurassic of the Musandam Peninsula has been studied in detail in Wadi Naqab, ca. 16 km SE of Ras Al Kaimah, U.A.E. (Fig. 1). The first reference section through the Jurassic in the region has recently been established here, comprising 1310 m of inner- to mid-shelf carbonates (de Matos, 1997; de Matos and Walkden, this volume), of which the Lower Jurassic is at least 470 m thick. This is built of meter-scale peritidal cycles, and we have logged in particular detail the first 234 m, up to a prominent and widespread lithological marker, a 10 m thick white micritic peritidal laminite. The Lower Jurassic here is poorly constrained biostratigraphically, especially towards the base, where the macrofauna is sparse and microfauna and flora lack diversity. The Hettangian definition is unsatisfactory, and although we have tried to constrain it paleontologically no forms were found that reliably prove its extent. The Triassic-Jurassic boundary is taken at a condensed horizon comprising a crinoidal phosphatic grainstone with an identifiable vertebrate and invertebrate fauna (de Matos et al., 1994; de Matos, 1997). The Sinemurian is more reliably constrained on the basis of foraminifera and algae, and the “white band” (peritidal micrite) described above is mid-Pliensbachian (de Matos, 1997).
Carbonate sediments in the Lower Jurassic of Wadi Naqab have been assigned to 23 microfacies types, which are described in detail in de Matos (1997). Facies recognition is informed by a combination of criteria, including grain types, grain size and texture, rock texture, rock color, biota (microfossils, macrofossils and algae), and sedimentary structures. The contribution from biota is variable, reflecting the patchy distribution of reliable faunas, but in the higher parts of the logged section foraminifera, bivalves, gastropods, sponges, echinoderms, ostracods, and dasycladaceans are increasingly common. The first identifiable corals appear in the Pliensbachian but ammonites remain absent.
For the purposes of cycle recognition, microfacies can be grouped into three marine facies associations, viz: low-energy subtidal, high-energy subtidal/shoal, and intertidal. In addition, a number of supratidal subfacies and phenomena have been recognized. These are discussed in detail below. Commonly occurring and characteristic representatives of the three facies associations are: nodular clay/carbonate pseudobreccia and spicular lime mudstones (low-energy subtidal); oncoidal wackestones, Lithiotis-rich (bivalve) foraminiferal packstones, Orbitopsella (foraminifera) wackestones and oolitic packs tones to grainstones (high-energy subtidal); stromatolitic and cryptalgal laminites, birdseye mudstones to packstones and ostracod-rich grainstones to mudstones (mainly intertidal). The supratidal environment is characterized by karst, rubbly soils, rhizolith horizons, collapse breccias, and rare calcretes. In the Wadi Naqab succession the majority of carbonates relate to low intertidal to shallow subtidal environment, with high intertidal algal mat and low-energy subtidal facies less well represented.
Cyclic Patterns and Cycle Boundaries
Asymmetry of facies stacking patterns, with a dominant shallowing-up regressive trend, is normal in meter-scale to decameter-scale cycles in the geological record (e.g. Goodwin and Anderson, 1985; Walkden and Walkden, 1990). Many authors prefer to attribute such facies asymmetry to a rapid generative sea-level rise that outpaces sedimentation, although others consider that carbonate “startup” is slow, leading to a distinctive transgressive lag (Ginsburg 1971; Read et al., 1986; Goldhammer et al., 1987; Walkden and Walkden, 1990) (Fig. 3A). No doubt both are important, but another possible cause of asymmetry is storm reworking. Probability ensures that the earliest deposited facies of any developing cycle are automatically at greatest risk of removal or severe modification because they are in place for the greatest period of time. Most inner-shelf to mid-shelf sites are probably most prone to storm reworking at the deepest, most open-water phase, so that replacement facies are more likely to be regressive than transgressive. The effects of such reworking upon a hypothetical cycle are shown in Figure 3B. Early lithification through subaberial exposure might further reinforce such a bias by protecting regressive facies from erosion during the next transgressive phase.
Facies patterns in shallow-water carbonate settings are therefore unlikely to deliver a full record of the sea-level changes that have built them. Similar facies deposited directly above one another, but from differing cycles, might be taken to represent a single cycle, whilst a couplet of regressive facies, separated only by partial reworking through the effects of a single storm, might be taken as separate cycles. In Wadi Naqab only three cycles out of 149 have been found to contain representatives of all three depth-defined carbonate facies associations plus a subaerial cap. Distinction between one cycle and the next based only upon facies stacking pattern is therefore hazardous. Cycles are easily misidentified at the outset, cycle numbers and thicknesses will consequently be wrong, and overall cyclic stacking patterns will be misconstrued. All these are crucial in the application of high-frequency cyclic successions to correlation and in their interpretation in terms of autocyclic or allocyclic processes. In such incomplete successions proper study of omission surfaces, starting with the accurate identification and logging of these in the field, provides an invaluable aid to cycle definition.
Omission Surfaces as Cycle Boundary Markers
An omission surface is any sedimentary interface (discontinuity; diastem; time-elapsed surface) where there is evidence for the effects of one or more nondepositional processes that may account for a substantial period of time. In shallow marine environments a very high proportion of cycle time may be represented by omission surfaces (Koerschner and Read, 1989; Wilkinson et al., 1991; Sadler, 1994), so that features they display may reflect conditions representative of a very significant but otherwise unrecorded part of geological history (e.g., Figs. 3A, 3B). They can also reveal important geochemical, sedimentological, biological, or climatic information. Examples are subaerial dissolution surfaces (including paleokarsts, pebbly soils, and intertidal corrosion zones), diagenetically modified surfaces (including dolomite caps, calcretes, hardgrounds, and mineralized crusts), or ganically colonized surfaces (including rhizolith mats, bored or encrusted surfaces, and burrowed firmgrounds), and submarine erosion or nondeposition surfaces (including lag surfaces, ravinements, and sea-floor corrosion surfaces). In theory, any bedding plane is a candidate omission surface, but evidence of at least one of the above effects is needed.
Any one of the surface types listed might mark a cycle boundary, but some surfaces can be formed at any point within a cycle and so provide equivocal evidence. For example, a bored hardground, a marine dissolution surface, or a glauconite and/or phosphate crust could form at or around a maximum flooding surface, but a karst and paleosol can form only during emergence. Some omission surfaces may have a compound history, such as a hardground exhumed by marine reworking and subsequently subaerially exposed (e.g., Immenhauser et al., this volume). Furthermore, the effects of diagenesis may exaggerate features (a stylolitized hardground might be impossible to distinguish from a karst) or may destroy distinguishing textures (a stylobreccia might look like a fossil soil). Many omission surfaces have probably been subjected to more than one type of contemporaneous modification, and most are affected to a greater or lesser extent by shallow to deep diagenesis during burial.
The Wadi Naqab succession presents an ideal case study for developing “ground rules” for recognizing the varied characteristics of omission surfaces and in particular for identifying omission surfaces that are also cycle boundaries. The succession is extensive, the cycles are well exposed, and the inner- to mid-shelf setting ensured that only relatively small sea-level changes were sufficient to exert a substantial effect on depositional environment, in most cases leading to emergence of the shelf. Set against this advantageous shallow shelf setting is the near certainty that superimposed longer-term sea-level changes, such as third-order sea-level lows, could have caused nondepositional gaps in the record with the loss of cycles or groups of cycles. Such lost cycles (e.g., Fig. 3A) might be picked up downdip, upon as yet unrecognized more distal parts of the shelf, or they may be present only in the basin beyond. Nonetheless, closer attention to omission surfaces in the delineation of cycles inevitably leads to improved data and the refinement of our knowledge of the overall cycle stacking pattern. A more accurate picture of the distribution of subaerial omission surfaces could reveal the location of longer-term sea-level lows through the vertical clustering and/or relative maturity of subaerial surfaces.
Characteristics of Subaerial Omission Surfaces
Subaerial effects in carbonate depositional environments are reviewed in Esteban and Klappa (1983) and Demicco and Hardie (1994), and good case studies are in Wright (1986) and James and Choquette (1988). Reviews of the sequence stratigraphie context of emergent surfaces are in Tucker (1993), Loucks and Sarg (1993), Read and Horbury (1993), and Budd et al. (1995). Three principal types of emergent surface have been recognized in the Lower Jurassic of the Musandam Peninsula, viz.: paleokarsts, corrosion surfaces, and paleosols. The features of these surfaces are intergradational, and the distinction especially between karsts and corrosion surfaces is probably genetically artificial. Calerete textures have been noted, although surfaces extensively altered to calerete are absent. Surfaces commonly display red staining through precipitation of iron oxide, probably at a soil/rock interface. Geopetal silts and contemporaneous speleothem-like calcite cements also appear directly beneath some surfaces but are not easily identified in the field. Rhizoliths are commonly associated with each of these types of surface. Eolian facies have not been recognized.
These are here recognized as small-scale dissolution surfaces with rolling, rounded, hummocky, or pitted surfaces that are essentially smooth rather than fretted. Their appearance is very similar to paleokarsts described by Walkden (1974), and surface relief up to 0.5 m has been observed. These surfaces clearly incise underlying sediments, and they are normally filled by horizontally bedded sediments that onlap the karstic irregularities (Fig. 4). Associated irregular cracks or fissures are also filled with carbonate (Fig. 5). Paleokarsts of this type probably formed in porous carbonates beneath a thin layer of mineral or organic soil, locally assisted by plants (e.g., Herwitz, 1993). If so, any direct evidence of soil or substantial plant material is usually lost.
The distinguishing features of corrosion surfaces are a scalloped or fretted surface morphology, usually underlain by sediment-filled cavities. Corrosion surfaces sometimes underlie a pebbly paleosol (see below), but they can also be buried by carbonate mud and show up as a result of contrasting color or lithology. In either case randomly shaped dissolution pipes and cavities may extend well beneath the corrosion surface, commonly opening into geopetally filled microcaves up to tens of centimeters across (Fig. 6). Key evidence of a dissolutional origin (as opposed to a burrowed origin) is that sediment removal was mineral-selective and texture-selective, leaving, for example, sediment-filled mollusc molds (Fig. 7) or algal thalli. More resistant material such as oncoids and large forams survive, although the sediment-fossil interface has become the focus of dissolution, with the gap later filled by infiltered mud. Demicco and Hardie (1994) note that such surfaces could result from bioerosion of rocky surfaces, but our clear evidence for dissolution discounts this. Nonetheless some Lower Jurassic corrosion surfaces can display burrow-like features. In many cases these are probably root casts, but vertical bioturbation (and even desiccation) could well have opened up some sediment surfaces to penetrative dissolution, thereby contributing to the creation of this distinct category of subaerial omission surface.
Corrosion surfaces can be formed in the intertidal region (e.g., Read and Grover, 1977; Esteban and Klappa, 1983), but the broad, shallow, dissolution hollows of Recent intertidal surfaces are in marked contrast to the penetrative and microcave-riddled features noted in the Lower Jurassic of Wadi Naqab. In numerous examples, a corrosion surface has evidently finally evolved, after prolonged dissolution, into a corroded breccia, but diagenetic effects may also contribute to this texture (see below).
These are structureless and unfossiliferous clay horizons with a gray, ochre, or red color (original color retention is unlikely in this modern arid environment). They may incorporate pebbles (made of recognizable carbonate sediment) or nodules (formed in randomly shaped micritic carbonate). They are seldom more than a few centimeters thick, but thicken appreciably where they overlie and fill a karst or corrosion surface (Fig. 8). X-ray diffractograms reveal poorly ordered mixed-layer clays with dominant smectite/illite and kaolinite.
Features Providing Key Confirmation of Emergence
Subaerial surfaces can appear in unexpected places, capping incomplete cycles or separating otherwise very similar facies. This means that they can provide key but unique evidence of 200,000 years sudden and probably very brief sea-level falls. In these instances, where surface geometry is the only criterion available for emergence, their identification in the field needs supporting evidence. In a few cases, probable karst surfaces have had to be ignored because they lack much surface relief and display no other evidence of identity. In other cases breccia horizons have provided ambiguous evidence for soil-forming processes. Associated features such as calerete textures, rhizoliths, mineral crusts, or early meteoric cements therefore provide vital corroboration.
Many subaerial omission surfaces show locally concentrated festoons of micron- to millimeter-sized, branching and mainly vertically arranged tubules, as well as less common larger (up to 10 mm) meandering vertical tubes. These extend tens of centimeters beneath the surface and are normally obvious, being filled mainly with mottled brown calcite. Where the texture is particularly dense there may be abundant millimeter-scale fracturing of the sediment. In thin section the tubes may be centered within, or contain concentric laminae of brown micrite and may show geopetai blockages of clear equant micrite or silt (vadose silts). Larger tubes have blockages of sediment that commonly show mottled and anastomosing networks of apparently collapsed tubules. Crystal silt, comprising evidently corroded calcite fragments in a matrix of marine carbonate, is also present within some of the larger tubes. The tubules and tubes might be burrows, but the association with karsts and corrosion surfaces, together with crystal silt, brown laminated calcite fills, and fracturing suggests that they are root molds and rhizoliths.
On exposed surfaces in arid environments care needs to be taken when identifying calcrete, because it may be Recent. Unequivocally Jurassic calcretes are rare, and a full caliche profile has not been seen. Evidence is confined to patchy brown-stained laminated, brecciated, and tubule-rich alteration fabrics and superficial crusts. These coat surfaces and line cavities and may relate to rhizolith formation rather than meteoric alteration. Pebbles within soil horizons locally show calcrete-like brown staining.
Red and ochre (iron oxide and hydroxide) crusts on discontinuities are common. Those present on dissolution surfaces beneath probable paleosols may be an original feature, representing, for example, a redox interface beneath a gleyed soil. Such crusts could survive reworking and removal of a soil during transgression and, therefore, merely lack a confirmatory soil cover. However, oxide crusts appear in pelagic successions (Tucker and Wright, 1990) and are also common beneath pyritic horizons where shallow diagenetic or recent weathering reactions cannot be ruled out. They are, therefore, not a reliable diagnostic feature of emergence.
Dolomite pervades some karst and corrosion surfaces, weathering to a characteristic buff or brown color rather than the normal gray (lost on fresh surfaces). Cracks in such surfaces have been observed filled with normal undolomitized sediment (Fig. 5), evidently confirming early alteration of the hostrock. Supratidal dolomitization (e.g., Shinn, 1983) was probably responsible, although later selective alteration cannot be discounted. Sediments infilling the distal parts of dissolution tubes can also be dolomitic, in contrast to the host sediment. These may demonstrate that the sediment was already (at least partly) dolomitic when it was emplaced.
This is a poor criterion, because meteoric effects can be pervasive for some distance below emergent surfaces, spanning several cycles (e.g., Walkden, 1987; Read and Horbury, 1993). They are also hard to recognize in the field. Vadose silts are found below some Musandam omission surfaces, including within fenestral peritidal deposits. Together with rare dark brown cements (full of presumed organic inclusions) they occur as the earliest obviously crystalline precipitates within rhizoliths and calcretes. More easily recognized in the field are brown cements (but clear in thin section) that line all types of cavity, but these are common throughout the Lower Jurassic of Wadi Naqab and are unlikely to relate to local subaerial surfaces.
Characteristics of Submarine Omission Surfaces
Submarine omission surfaces include hardgrounds, firmgrounds, and ravinements. Hardgrounds and firmgrounds are described in Scholle et al. (1983) and reviewed in James and Choquette (1990) and Demicco and Hardie (1994). They differ only in original degree of cementation and so are difficult to distinguish in lithified sediments. Marine corrosion of hardgrounds is common in deeper-water settings (Scholle et al., 1983), but this is not further considered here. Ravinements have not been identified as such in the Lower Jurassic, although intracyclic erosion is common (Fig. 3A). Submarine omission surfaces are characteristically intracyclic, potentially occurring anywhere within a cycle, so that they are not necessarily useful cycle boundaries. However their proper identification is essential because they have characteristics that might be confused with, or even overstamp, those of subaerial surfaces. For example, irregular hardgrounds and burrow-scalloped firmgrounds, given appropriate fill and later modification by pressure dissolution, might look like corrosion surfaces, and subtidal burrow systems might be taken for supra tidal root molds. Key evidence of marine diagenesis comes from features such as marine cements, organic encrustation or borings, mineral coatings, erosion-intersected fossils, derived pebbles, and fractured overhangs in cavernous hardgrounds. All these are accessory features and may be absent. Hardgrounds form at any water depth, but they commonly occur with lag deposits and may be used to define transgressive surfaces and maximum flooding surfaces within parasequence sets (e.g., Emery and Myers, 1996). Modern examples are seen in sand shoals of the Bahamas (Demicco and Hardie, 1994) and in shallow subtidal settings in the Arabian Gulf (e.g., Walkden and Williams, 1998). Firmgrounds (e.g., Bromley, 1996) lack early lithification, but candidates must display evidence of resistance to erosion and self-supporting characteristics.
Only one unequivocal biologically encrusted hardground has been recognized in the Lower Jurassic of Musandam, comprising small compound corals on a burrow-modified discontinuity (Fig. 9). No examples of bored hardgrounds have been seen, but they are common in the Middle Jurassic of Musandam (de Matos and Walkden, this volume). Their absence is probably linked with the general impoverishment of the Lower Jurassic fauna. Otherwise, hardgrounds have been identified only on the basis of irregular, eroded discontinuities, sometimes burrowed, showing sharp lithological contrasts. These tend to be grouped in multiples of two or three, a feature we have also seen in Recent hardgrounds beneath the Arabian Gulf and previously noted by Shinn (1969).
Top-filled or empty burrow systems truncated by planar erosion surfaces, and burrow-scalloped surfaces with overhangs might qualify as firmgrounds. These are common in the Musandam succession, but in the absence of positive evidence for lithification, many surfaces dismissed as probable firmgrounds may actually be hardgrounds. Similarly, many surfaces such as sharply defined bedding planes that lack any other features might have originated either as firmgrounds or as hardgrounds.
Omission Surfaces Modified by Diagenesis
A number of diagenetic processes are capable of destroying, mimicking, or enhancing key characteristics of omission surfaces, thereby removing evidence or creating ambiguity (Fig. 2). For example, meteoric alteration of Pleistocene limestones presently beneath the Arabian Gulf has removed aragonitic cements from marine hardgrounds, rendering them almost indistinguishable within subsea cores (personal observations). On the other hand, pressure dissolution acting at a textural interface can enhance unimportant discontinuities in more deeply buried limestones, giving them undue prominence. Pressure-dissolution seams (stylolites) create sutured interfaces that might be taken for corrosion zones, and burrowed intervals are reduced to rubble by pressure dissolution so that they resemble soil or corrosion zones (Fig. 2). This diagenetic retexturing or masking is a major barrier to the recognition of omission surfaces and highlights the danger of overreliance on surface morphology alone.
Breccias, conglomerates, and pebbly deposits can provide important corroboration for omission surfaces where they accumulated as transgressive lags, soils, and solution breccias. Confusion arises, though, in trying to distinguish between these and so-called “pseudobreccias”, which are diagenetically produced nodular fabrics, including those created by localized cementation (e.g., Scholle et al., 1983), and those created or enhanced by pressure dissolution (e.g., Logan and Semeniuk, 1976). Nodular fabrics can be formed through selective lithification of burrowed carbonates both at the surface as successive hardgrounds and in the subsurface accompanying differential compaction. Petrocalcic soils can also create nodular horizons. Stylobreccias and stylonodular fabrics form through pressure dissolution acting along texturally defined or randomly distributed interfaces and can modify any of the primary and secondary textures already noted. Such textural convergence means that any nodular or brecciated horizon showing evidence for pressure dissolution might have originated as a cycle boundary, and it demands careful attention. This problem is common in the Musandam succession, and several criteria are needed to distinguish between “stylonodular”, “burrownodular” and “pedonodular” categories. These include the morphology and associated features of the underlying discontinuity, the lithology and internal texture of nodules (e.g., calcreted clasts, burrow, fabrics, growth fabrics), the morphology of nodules (e.g., randomness, fitted fabric), the lithology, volume, texture, and the color of matrix, and the presence or absence of fossils within it.
Omission Surfaces and Climate
Climate plays a major role in determining the type and nature of omission surfaces that can develop. In the marine environment, for example, hardground development is encouraged by aridity and evaporation (Tucker, 1993; Walkden and Williams, 1998). In the case of subaerial surfaces, aridity inhibits dissolution and encourages the development of calcrete (e.g., Hird and Tucker, 1988; Read and Horbury, 1993). Aridity also promotes the development of dolomite crusts (Shinn, 1983). A more humid regime is more likely to lead to a dense plant cover, enhancing soil development and encouraging the growth of pitted karsts. These are broad generalizations, though, and a strongly seasonal climate is likely to encourage a range of contradictory phenomena.
In the Musandam succession, however, there are few contradictions. Dissolution features at cycle tops are common throughout the Hettangian, Sinemurian, and Lower Pliensbachian whilst calcretes and dolomite crusts are uncommon. In the Upper Pliensbachian dissolution features become less prominent at cycle tops, and collapse breccias and desiccation features become common. Dolomitization is also more common in the higher parts of the succession, although not confined to cycle tops. These lines of evidence point strongly to the prevalence of moderately humid conditions for most of the early Jurassic, giving way to more general aridity in the Late Pliensbachian.
Omission-Surface Analysis Applied to the Cyclic Lower Jurassic
The use of omission surfaces, particularly those of subaerial origin, has provided a valuable tool in the recognition and placing of cycle boundaries, leading to a substantial increase in resolution of cycles and, hence, creating a more accurate cycle database. This increases the value of the section for statistical analysis of cycles and of cycle stacking patterns, for example through the use of Fischer plots. The relative completeness of the database also permits accurate determination of the average duration of cycles, but in such a stratigraphically incomplete proximal shelf setting, where erosion and nondeposition account for more than half of the sedimentary record, the results of methods such as power spectral analysis are less likely to be meaningful and have not been attempted. Of 149 cycles recognized in the 234 m succession, more than half are topped by subaerial omission surfaces. Paleokarsts or corrosion surfaces account for 77 cycle tops whilst prominent soils, with or without breccia or pebbles and sometimes lacking an obvious underlying dissolution surface, cap 10 cycles. At least 17 of the subaerial surfaces display penetrative dissolution features between 20 cm and 50 cm deep. An additional nine surfaces have been logged as cycle boundaries but lack clear evidence of a subaerial origin. As regards marine omission surfaces, only 15 probable hardgrounds have been recognized, mostly in clusters, and few provide sufficient evidence (e.g., sharply backstepping facies) to invoke a cycle boundary.
Many cycle boundaries produce an outcrop bench, so that little excavation was necessary to identify crucial features and very few could not be examined in detail owing to poor exposure. It is worth noting, however, that in a less ideal setting such as a freshly quarried outcrop, a more fully vegetated area, or a less accessible setting, many cycle boundaries would have gone unnoticed and fewer cycles would have been identified. Furthermore, many of the characteristics used to define cycle tops in this outcrop study are on a scale or pattern of distribution that would render them difficult to recognize in core. Even in core from the floor of the present-day Arabian Gulf, Pleistocene marine and subaerial omission surfaces are difficult to recognize owing to subsequent and pervasive meteoric alteration (personal observation and in preparation).
Interpretation of Lower Jurassic Cycles and Cycle Stacking Patterns
The probable time span of Jurassic stages Hettangian-Sinemurian-mid-Pliensbachian was about 14 million years (Hardenbol et al., 1998; Gradstein et al., 1994). This gives an average duration of around 90,000 years each for the 149 lower Jurassic cycles covering this time span. The figure is comfortably within the range of Milankovitch orbital forcing periods (e.g., House, 1995) and close to the strongest of these in the present day, the 95,000 year eccentricity cycle. With an unknown number of cycles missing through nondeposition or erosion in this proximal shelf setting, however, it is predictable that the calculated figure for cycle duration should be greater than the true average during the Early Jurassic. In certain European Lower Jurassic hemipelagic mud successions, where breaks in sedimentation are less deleterious to the sedimentary record, Milankovitch orbital forcing is regarded as the main cause of minor cyclicity (e.g., House, 1985; Weedon, 1993; Weedon and Jenkyns, 1990). Average durations of around 40,000 years have been obtained for such minor cycles by House (1985) and Weedon et al. (2000), and it is likely that basins and platforms share the same orbitally forced influences. Continental massing in the Jurassic may have favored limited ice buildup (Crowley et al., 1992; Valdes et al., 1995) and the abundant omission surfaces in the Wadi Naqab succession could be the first direct evidence of glacioeustatic sea-level change in the Early Jurassic. However, other possible orbitally forced links with sea-level change have been discussed, such as varying storage of groundwater on the continents (Jacobs and Sahagian, 1993) and thermal expansion of the oceans (Schulz and Schafer-Neth, 1997).
Thickness of cycles ranges from less than 1 m to around 10 m, with an average of 1.5 m. In Figure 10, thicknesses of all 149 cycles are plotted in histograms against the corresponding Fischer plot. Fischer plots use thickness data to show the cumulative departure from cycle mean thickness (e.g., Sadler et al., 1993). They are a method of smoothing raw thickness data and, by assuming constant cycle periodicity, are an analog of changes in the rate of creation of accommodation. The plot reveals some clear trends in cycle thickness and stacking patterns that may reflect at least two distinct third-order sea-level variations superimposed upon the fifth-order variations represented by the cycles themselves. The first of these, a peak around cycles 8–12, may be incomplete, the initial cycles lying beneath our measured section and not included in our database. The second major Fischer peak, around cycles 82–88, reflects a steady buildup in cycle thickness followed by an abrupt reduction to near average thickness.
Because each of the cycles can be regarded as a parasequence, the pattern produced is a classic example of the variations in parasequence stacking pattern caused by third-order relative sea-level change noted by numerous authors (e.g., papers in Loucks and Sarg, 1993; Tucker, 1993). Assuming that accommodation was always quickly filled, the thicker parasequences marked by the rising Fischer curve show where background tectonic subsidence and third-order sea-level rise worked together in the creation of accommodation, so that each superimposed fifth-order eustatic rise had maximum effect. This was followed during third-order sea-level fall by the canceling out of tectonic subsidence so that superimposed fifth-order sea-level rises could create little to no additional accommodation. Parasequences are correspondingly thin and some may be missing. Tucker (1993) notes how dominant facies change in tandem with the changing parasequence stacking patterns.
Amongst others, Read et al. (1986) and Walkden and Walkden (1990) have used original computer simulations to generate variations in cycle stacking patterns. They show how, during the transition from highstand to lowstand, when parasequences are thin, much time is taken up with exposure rather than deposition, providing an extended opportunity for subaerial modification at cycle boundaries. This prediction from computer modeling points to an important application for data on omission surfaces derived from outcrop. Given that subaerial dissolution features are found throughout the Lower Jurassic, their variable development, particularly the depth of dissolutional penetration, could provide a proxy for the duration of exposure. Clearly there are numerous factors that can influence the development of karsts and corrosion surfaces, including lithology, vegetation cover, rainfall, prevailing wind direction, and depth of the meteoric vadose zone. Neverthless, length of time exposed and elevation during sea-level falls, both of which are directly related to sea-level change, must be amongst the most important influences on the amount of dissolution that can take place.
In Figure 10B all cycles with firmly identified subaerial omission surfaces are plotted, together with a general indication of the depth of dissolution at each of these. Clearly, subaerial omission surfaces, especially the best developed of these, are not evenly distributed. In particular there is an out-of-phase relationship between periods of maximum cycle thickness and periods of maximum subaerial modification. The main Fischer peaks in Figure 10A, showing where third order sea-level rise had its maximum additive effect upon cycle thickness, are around cycles 8 and 84. These are directly followed by thinner cycles, where the reverse effect was felt, and it is in these positions, in the regions of cycles 20 and 96 (trends 1 and 3 in Figure 10B), that omission surfaces tend to be best developed or most numerous. Evidently the best developed omission surfaces, and to some extent also their clustered distribution, reveal the parts of the succession where third-order sea-level fall caused a reduction in the period of deposition for each minor cycle and a consequent lengthening of the intervening periods of exposure. This pattern is consistent with the predicted one, and provides independent support for the interpretation of the Fischer peaks as reflecting long-term secular sea-level change. Of added interest, therefore, is the similar trend between cycles 38 and 60 (trend 2 in Figure 10B), and the less distinct example between cycles 112 and 140 (trend 4 in Figure 10B). Evidently the relatively minor mid-Sinemurian rise in the Fischer plot around cycle 40, which might otherwise be dismissed as a statistical artefact, was a real third-order sea-level event. A similar case can be made for the indistinct “hump” close to cycle 112, which lies on the back of the main Pliensbachian third-order fall.
High-frequency (fifth-order) cyclicity can be recognized in both basinal and carbonate platform successions in the peri-Tethyan Lower Jurassic, providing exciting potential for high-resolution correlation and detailed characterization of sea-level change. However, many problems remain in ensuring accuracy in the correlation of such successions. One of the most serious of these relates to the use of facies patterns alone for defining cycles in high-frequency shallow-water carbonate successions. These are characteristically incomplete, and the diversity and forced rate of change in facies are so great that ideal regressive patterns are rarely developed. We are commonly uncertain about where cycle boundaries lie and about how many cycles are contained within a given interval.
Closer study and systematic logging of omission surfaces offers a solution. Omission surfaces are likely to be abundant in shallow-water carbonate successions, and subaerial surfaces, in particular, are valuable direct indicators of sea-level change. Consequently these have considerable potential for improving cycle discrimination, and they provide us with a means of “tuning” high-frequency cyclic successions.
A recently described, biostratigraphically constrained, Lower Jurassic Hettangian to lower Pliensbachian reference section close to the Oman/U.A.E. border includes marine hardgrounds, burrowed firmgrounds, karsts, corrosion surfaces, and paleosols. These account for a substantial proportion of elapsed time, but they share certain key physical characteristics and their recognition is not always straightforward. In particular, identification of omission surface types is hampered by diagenetic changes, which render some of their characteristics ambiguous.
Systematic study and logging of subaerial omission surfaces has substantially improved discrimination of cycles in our reference section and has increased known numbers by more than 100%. This improved database enables high-confidence analysis of the Lower Jurassic cyclic record and reveals an average cycle period within the Milankovitch range of around 90,000 years. This is longer than the 40,000 years obtained from equivalent basinal successions in Europe, and reflects the inherent incompleteness of shallow-water successions in which some cycles fail to be recorded. The recognition of a Milankovitch periodicity strengthens the likelihood that the sea-level change responsible for the fifth-order platform cycles was climatically generated, and a small-scale glacioeustaric mechanism might have been the cause.
Fischer analysis of cycle data reveals two major accommodation peaks that may relate to third-order sea-level change, one at the Hettangian-Sinemurian boundary and one at the Sinemurian-Pliensbachian boundary. Subaerial omission surfaces are most common and best developed on the falling limbs of the Fischer plot, which corresponds to a time when the third-order sea-level “downwave” would maximize subaerial exposure at fifth-order cycle boundaries. Extending this analysis to two smaller fluctuations on the Fischer plot, one mid-Sinemurian and the other mid-Pliensbachian, the distribution and intensity of omission surfaces suggests that these fluctuations are also likely also to relate to longer-term sea-level change. The inclusion of reliable omission surface data therefore not only improves discrimination of cycles but it increases confidence in the use of Fischer analysis both as a means of detecting secular change in cylicity and as a tool in the interpretation of cycle stacking patterns. Evidently the pattern of sea-level change on this Early Jurassic Tethyan carbonate platform was complex, but not random.
Milankovitch-related cyclicity exerts a strong control on facies development and formation architecture in shallow-water carbonates and was pervasive during the early Mesozoic. It is likely that similar patterns will be found globally and in other parts of the Mesozoic where there was an appropriate combination of climatic and palaeogeographic conditions. In the subsurface, the resultant characteristics, including sub-meter-scale bedding, weathered zones, soils, and early diagenetic caps, can strongly influence water and hydrocarbon movement, creating potentially strong anisotropy, particularly in nonfractured reservoirs. Nevertheless, many of the features by which such formations could be characterized are very unlikely to be revealed in seismic, will be missed in most well logs, and can be difficult to recognize in core.
The authors would like to thank ADCO and PARTEX, who supported the original field work upon which this paper is based. Parexpro and the University of Aberdeen have provided facilities for its completion. In particular, we gratefully acknowledge Barry Fulton (UoA), who finalized the diagrams. Thanks also go to Charlie Kerans and Dennis Prezbindowski, who reviewed the original manuscript and made helpful suggestions that undoubtedly led to some improvements.
Figures & Tables
Middle East Models of Jurassic/Cretaceous Carbonate Systems
This volume will interest tectonic modelers, stratigraphers, sedimentologists, and explorationists. It is the product of the international conference of “Jurassic/Cretaceous Carbonate Platform-Basin Systems, Middle East Models” that was convened in December 1997 jointly by SEPM (Society for Sedimentary Geology) and the United Arab Emirates University in Al Ain, United Arab Emirates. The twenty-three papers present new data and interpretations arranged in three sections: 1) sequence stratigraphy, cyclostratigraphy, chronostratigraphy, and tectonic influences, 2) depositional and diagenetic models of carbonate platforms, and 3) hydrocarbon habitat and exploration/development case studies. New tectonic models of the Arabian Basin, new stratigraphic and sequence stratigraphic reference sections, new geochemical and source rock data, and new reservoir data are presented. New geologic models make this set of papers relevant to geoscientists working outside of Arabia also.