[Back to Full Graphics Version]

Anatomy of a source rock: environmental, climatic and stratigraphic signatures in the type Kimmeridge Clay

This project, under the aegis of the Rapid Global Geological Events Special Topic, aimed to develop a high-resolution stratigraphy for the Kimmeridgian Stage comparable to that already produced for the Pleistocene. The work involved the examination of three cores drilled at two locations close to the Kimmeridge Clay type section in Dorset: Swanworth Quarry and Metherhills, thereby making best use of the palaeontological, geochemical and sedimentological work already completed from outcrop studies. A programme of comprehensive micropalaeontological, sedimentary, geochemical and mineralogical analysis on the core material was undertaken to characterize the succession at all relevant scales, and to provide a data base for cyclostratigraphic (i.e astronomically based) and isotope-stratigraphic (delta13C(OM)) studies.

The data generated from the core was used to test models of climatic change and climatic forcing of the Mesozoic sedimentary record, and to evaluate the inter-relationships of these phenomena with relative sea-level change. Published data bear on the principal controls of organic-matter accumulation in a major hydrocarbon source-rock.


Programme of Research:

Background to the Proposal

This multi-disciplinary and multi-institutional (Luton, Newcastle, Open, Oxford, Reading and Southampton Universities) project aimed to understand the principal controls on a key episode of petroleum source-rock formation, namely that of the Kimmeridge Clay. This formation has long been recognized as the major source rock for North Sea oil, although at its type locality in Dorset the succession is immature, having suffered only modest burial. The relative lack of post-depositional alteration rendered the sequence particularly suitable for study by a range of analytical techniques. We investigated the stratigraphic, palaeogeographic and palaeoceanographic context of this unit at all possible scales of analysis. As such, the project encompassed stratigraphy, sedimentology, geochemistry and palaeontology. The inter-relationships between the various consortium institutions are shown as a flow diagram in Fig. 1.

The location of three boreholes drilled by the project close to the outcrop near Kimmeridge Bay in Dorset enabled the bio- and lithostratigraphy of the type section, worked out in detail over the last century, to be directly applied to the cored material. The type section is the most stratigraphically complete section through the Kimmeridgian Stage in Europe. There was no need to address the palaeontological aspects of correlation since, by siting the boreholes close to the type locality, we could use the preferable means of an extant and well-established biostratigraphy (Figs 2-3).

A key element of this project was the use of a combination of extant biostratigraphy, lithostratigraphy, newly generated cyclostratigraphy and magnetostratigraphy to place all other data within a secure and well-calibrated chronometric framework. To this end, some geochemical parameters were  determined at a sample interval of 10 cm, which allowed  recognition of the main astronomically controlled climatic variables.

Specific Objectives - All Institutions

The main questions addressed were as follows:

Co-ordination Meetings

One-day meetings were held twice per year for all members of the consortium aimed at ensuring efficient collaboration and exchange of ideas.

Method and Approach

In order to tackle these problems we obtained an overview of the total succession by analyzing some quantities at a regular interval of 10 cm (e.g. TOC, CaCO3, major and trace elements). In addition, other studies concentrated on key intervals of particular interest.


At the outset of the project, it was decided that the most appropriate drilling program would be to attempt to obtain two continuous cores through the entire Kimmeridge Clay at a single site as close to the type section of Kimmeridge Bay cliffs as possible. The type section is cut by a number of small faults, both normal and reverse, and to avoid such structural complications it was necessary to locate the drill site on the near-horizontal part of the fold limb of the Purbeck - Isle of Wight monocline (i.e near the eastern end of the outcrop at St Alban's Head). Six sites were suggested and Swanworth Quarry [SY 9675 7823] near Worth Matravers was chosen on the basis of its logistical and environmental advantages. In addition to facilitating core-outcrop correlation at a high degree of resolution, this location also allowed recovery of intervals that are otherwise very poorly exposed (baylei-mutabilis ammonite zones) and/or located in structural settings (e.g. on the footwall of the Purbeck-Isle of Wight disturbance) different from that of the type locality.

The initial borehole drilled in Swanworth Quarry was completed in satisfactory time and yielded good results, which motivated the project to drill a second borehole nearby in the Quarry, together with a third at Metherhills [SY9112 7911], Kimmeridge. Together, these three boreholes recovered the Kimmeridge Clay in its entirety, and their offset double-coring provided an overlapping section between coring breaks which gave ample sample material. A collection of fresh samples from the boreholes was essential for meaningful geochemical work because some intervals exposed along the coast are very badly weathered (e.g. autissiodorensis, elegans and scitulus ammonite zones). Additionally, a full suite of borehole logs was run, including tools such as the formation microscanner and geochemical tool.

Programme of Research - Oxford, Open University (and others)

Plan of Research: Project Co-ordination, Correlation and Regional/Global Setting

Our programme of research was directed principally towards determining the rates of change of stratigraphically significant phenomena such as relative sea-level cycles and the chemistry of sea water. The role of coordinating the analytical efforts of different laboratories to produce an overall synthesis of the component studies was the responsibility of H.C. Jenkyns, S.P. Hesselbo, A.L. Coe, and PDRA H. Morgans-Bell based in Oxford.

Our specific aims can be described as Follows, but more technical details are available in the accompanying proposals from other institutions.

Role of the PDRA

In addition to co-ordinating the project, the scope of the PDRA's research included: 1) a detailed visual core description of both holes and detailed comparative logging of the type section, extending the work of Cox and Gallois (1981), to produce a full synthesis of previous work and its integration with the new data, and 2) the description, interpretation and correlation of cores and geophysical log data from the North Sea Kimmeridgian made available to us by BP; 3) to understand the regional sedimentological, palaeoceanographic and stratigraphic context of the Kimmeridge Clay. Furthermore, the PDRA and others made a bed-scale, outcrop-based spectral gamma-ray log from the type section (cf. Myers and Wignall, 1987; Parkinson, in press) to aid in cross-correlation to the boreholes and as a basis for investigating the use of gamma-ray logs in sequence stratigraphic analysis (cf. Underhill and Partington, 1993; Tyson, in press; Hesselbo, in press).

Palaeomagnetic studies

Correlation of the Kimmeridgian magnetostratigraphy developed during this project with the Geomagnetic Polarity Time Scale (GPTS) was a desired aim of the project but most of the investigated sediments proved unsuitable for this type of study.


Programme of Research - Reading (and others)

Palaeoclimate Change, Clay Mineralogy and Chemostratigraphic Variation

Palaeoclimate modelling at Reading University, employing UGAMP GCM experiments (Valdes and Sellwood 1992; Sellwood and Price 1993; Valdes 1993; Valdes et al 1995) has thrown light on the way in which orbitally induced variations in solar energy might be translated into a Kimmeridgian climate response. In northwestern Europe the 100 Ka eccentricity signal would have been received in terms of significant temperature variation and minor changes in rainfall. Times of "minimum seasonal forcing" (equivalent to Quaternary ice age times) would have been cooler and slightly wetter whereas times of "maximum seasonal forcing" (equivalent to early Holocene deglaciation) were warmer and drier. This gives a predictable signal which can be sought, through the application of advanced Palaeobiological, mineralogical and geochemical techniques. These results give an ideas-driven basis for evaluating Milankovitch cyclicity, and the recognition of rapid global geological events.

As part of this programme it was our aim to evaluate the climatic changes in the Kimmeridge Clay depositional regime likely to have resulted from Milankovitch forcing. Subtle geochemical and clay mineralogical changes are to be expected. The "normal" pattern of eccentricity-forced changes would be for a progressive and long-term (tens of thousand of years) cooling associated with a slight increase in humidity. This should be reflected in secular changes in the weathering and soil-forming processes in adjacent areas (as recorded in expected clay mineralogical changes).

Work with Valdes (Meteorology Department, Reading University) involved the running of GCM experiments to evaluate the possible impact of precession (19-23 ka) and obliquity cycles (41 ka) on the Kimmeridgian Earth and a refinement of the current model for eccentricity cycles (100 ka).

It was the aim of our analytical programme to provide an overview of the clay mineralogical variation by performing analyses of the < 2 mm fraction every 10 m throughout the core and, by reference to the natural gamma-ray log of the nearby Encombe Borehole (Sellwood et al. 1990) to perform very detailed sampling (15 cm spacing) over a number of intervals. The interval between the Washing Ledge Stone Band and Flats Stone Band in the of the eudoxus Zone  provided a control for much of the Kimmeridge succession, through which the gamma-ray response shows small-scale rhythmicity and the overall gamma activity is moderate. Clay mineralogical analyses was determined through a set of upward-declining gamma-ray cycles in the scitulus Zone and the immediately succeeding wheatleyensis-Zone succession (approaching the Black Stone Band) through which the natural gamma signature becomes very much less regular. These features (together with the high TOC) provided a significant target unit for detailed investigation. In the hudlestoni Zone, because of the complex gamma-ray pattern, it was essential to sample in detail approaching the White Stone Band. Detailed analyses continued up into the lower pectinatus Zone, where gamma-ray intensity declines overall and where patterns become more regular. 

These are all intervals in which pronounced small-scale cyclicity is particularly well developed. The results were integrated with the data generated by the other Groups.

Method and Approach

The clays were separated by standard extraction techniques of disaggregation and sedimentation. In addition, because of the calcareous nature of some of the shale intervals, it was  necessary to obtain a control in the analytical programme by removing the carbonate from duplicate samples (without affecting the clay minerals). Selected samples were analysed using the method developed at PRIS by Cook (1992).

The separated clay fractions were analysed by standard XRD techniques and quantified using the method of Weir et al (1971). In addition, a portion of the separated < 2 mm fraction was sent to the University of Luton (Weedon) so that the rare-earth content of the clays could be analysed. Recent research completed at PRIS by Parker on North Sea Cretaceous clays has shown that the non-exchangeable rare- earth composition of clay minerals can be related to both clay type and provenance. Such an approach allowed subtle changes in mineralogy, as predicted by the palaeoclimate modelling, to be more readily discerned.

A principal, and particularly significant part of this programme was the definition of the major- and minor-element variation through the entire Kimmeridgian succession at 10 cm intervals (5,000 samples). This analysis was undertaken at PRIS using XRF and provided a geochemical base-line against which many of the other geochemical and mineralogical parameters could be evaluated.

XRF provides a means of rapid and accurate characterisation of the sediment. The 5 g samples were powdered by the dedicated technician at Southampton University at 10 cm intervals throughout the core and analysed at PRIS.

Programme of Research - Southampton (and others)

Any study of the cyclicity in a sequence such as the Kimmeridge Clay involves discussion of the mechanism causing the changes in relative abundance of depositional components. Work carried out at Southampton linked the metre-scale cyclicity, as seen in the outcrop and well section, with the individual processes acting during sedimentary deposition. This approach allowed an understanding of how variations in climate can be expressed by changes in mineral and organic sedimentation.

It is now increasingly accepted that many cyclic sedimentary sequences result from orbitally forced changes in sedimentation (see Luton proposal). Much has been made of the overall pattern of cyclicity through such sequences but even less is understood of the parameters which control the pattern of sedimentation in single cycles. A novel approach has been that adopted by Waterhouse (1995) who, from studying Jurassic sediments with cyclicity, showed how the characterization of the types and absolute abundances of the palynological organic matter within individual cycles could be used to understand the relationship between climate change and sedimentation.

A further approach to the study of organic-matter-rich sediments is that of integrated BSEM and palynofacies studies at the lamination scale within Jurassic mudstones. Research at Southampton pioneered the use of back-scattered electron imagery (BSEI) of resin-impregnated sediment and demonstrated that this technique alone gives the necessary resolution and compositional information to analyze unbioturbated pelagic and hemipelagic sediments, particularly where laminations or other fine-scale compositional changes are preserved (Kemp, 1990; Kemp, 1991; Kemp and Baldauf, 1993; Brodie and Kemp, 1994; Kemp et al., 1995; Pearce et al., 1995; Pike and Kemp, 1995, a,b; Bull and Kemp, 1995; Kemp, 1995). The studies cited above demonstrated that BSEI techniques are capable of resolving individual flux events preserved in laminated sediments as thin as 30 mm.

Building on the expertise developed at Southampton, we  used integrated BSEM and palynofacies studies at the lamination scale to characterize and interpret the processes of formation of a small number of Kimmeridge Clay cycles, thus determining the controlling factors on the different depositional environments represented during the deposition of these cycles.

In addition, selected aspects of the cycle-scale palynofacies studies as carried out by Waterhouse were continued in conjunction with the study of organic productivity and preservation by Tyson (Newcastle).

Method and Approach

Lamina-scale analysis

Following drilling, logging and initial core description, four Kimmeridge Clay cycles were selected for multi-disciplinary studies. Representative facies of each of these cycles were sampled and slabbed. For lamina-scale studies the 'part' was polished for BSEM and the 'counterpart' split or cut along individual laminae, using the Westbourne wire saw, and prepared for palynological study. The wire saw removes only 15 mm of rock with each cut and so is capable of sectioning individual laminae.

All palynological samples were Lycopodium-spiked to permit determination of abundances per gram rock and per gram total organic carbon. In addition, preparations were made in which an ultrasonic probe was used to remove the amorphous organic matter that is the dominant kerogen component. This process enables accurate quantification of  particulate organic matter and determination of the environmentally most sensitive kerogen components.

In turn, the polished slab was characterized using the SEM, including element mapping, with the construction of a photomosaic that effectively represents an extraordinarily detailed sedimentary log. This technique allowed lamina-scale sedimentary units to be recognised and depositional processes inferred (e.g. variation in input of the different categories of terrestrial and marine organic matter including calcareous microfossils). Where possible, varve laminae were identified, enabling a direct measure of sedimentation rate and estimates of particle flux. Such varves were subjected to time-series analysis which allowed quantification of sub-millenial time and cycles and provided  an essential higher frequency cycle hierarchy and time base within the decimetric Kimmeridgian cycles. The SEM observations on organic input were substantiated and quantified by lamina-scale palynofacies analysis which allowed determination of abundances of these different categories of terrestrial and marine organic matter.

Individual cycle analysis

In addition to the lamina-scale study, some further work was carried out at a coarser scale on the Kimmeridge Clay cycles in order to develop certain aspects of the work of Waterhouse (1995). The first, in conjunction with the organic geochemical and productivity studies of the Newcastle group, investigated the dinoflagellate cysts both in absolute numbers and species composition. This study  enabled some additional element of productivity within the water column to be determined. A parallel study was carried out on the foraminifera by C.D. Jenkins, who undertook a PhD project on the benthic foraminifera within the Kimmeridge Clay cycles to show how variation in species composition, abundance and diversity of foraminifera within cycles correlates with changing bottom conditions to provide a proxy record of climate change. The additional Kimmeridge Clay core material was particularly important for providing unweathered continuous cored section through individual cycles.

Plan of Research

Programme of Research - Luton (and others)

Spectral Analysis and Milankovitch Control

Work at Luton aimed to investigate the detailed high-resolution stratigraphic time series generated from measurements of percentage total organic carbon (%TOC), %CaCO3, and bulk-rock concentrations of the principal major and trace elements at sample intervals of 10 cm. Additionally, multi-sensor core data collected at 5cm intervals was used for cyclostratigraphic and correlation purposes.

Several earlier studies have implied orbital-climatic (Milankovitch) forcing of sedimentation during the Kimmeridgian (e.g. Dunn, 1974; Oschmann, 1988; Wignall and Hallam, 1991). However, despite recent evidence for metre- and decimetre-scale sedimentary cyclicity, spectral analyses to date have been restricted to just 20 m of the type section (Dunn, 1974; Herbin et al., 1991). The 11,000 point time series generated by the core logging are amenable to spectral analysis in order to test rigorously for the presence of the regular lithological cyclicity which has repeatedly been inferred to relate to orbital-climatic (Milankovitch) forcing. If regular cyclicity, as a function of rock thickness, is detected spectrally, then candidate orbital frequencies need to be unambiguously recognized in order for the "Milankovitch chronometer" to be applied. Successful recognition is most likely to depend upon cycle wavelength ratios as radiometric dating constraints remain too poor for the Late Jurassic. If particular orbital-climatic cycles are identified, evolutionary spectra and band-pass filtering will be used to monitor stratigraphic variations in sedimentation rates at the site of the borehole. Interval dating between approximate time planes (e.g. biostratigraphic datum levels and the coccolithic limestones) can be achieved using counts of regular cycles isolated using filtering (Weedon and Jenkyns, 1990). Where the time planes are recognised elsewhere, especially across the North Sea using geophysical log picks and palynological correlation, the interval dating can be used to determine variations in sedimentation rates across basins.

Method and Approach

It was essential that the true depths of individual sections of the two cores were known precisely and accurately, bearing in mind the possibility of minor faulting and gaps between successive core sections. On the basis of experience on Legs 117 and 154 of the Ocean Drilling Program, the best procedure was to use very closely spaced core-logging methods to generate high-resolution time series. Accordingly, magnetic susceptibility, GRAPE, p-wave velocity and colour reflectance were measured at 5-cm intervals using the Institute of Oceanographical Sciences "Multi-sensor Core Logging System" at Southampton by the consortium's technician. These parameters relate to dilution of paramagnetic clays by carbonate and organic matter; bulk density; sonic velocity and mineralogy respectively. The sample interval was designed to be the most effective given the wavelength of the lithological cyclicity (70-150 cm) at the type section. The resulting four time-series can be then compared between the two holes and used to establish cm-scale lithological correlations, estimate coring gaps and used to design a composite depth scale as the basis for sampling between the two holes (cf. procedures on ODP Leg 154).

Previous field- and laboratory-based work on the Lower Jurassic Belemnite Marls and Blue Lias in southern Britain has established that magnetic susceptibility measurements in the field using the Bartington system provides a rapid lithological logging tool (Weedon et al. 1999). In these rocks, and by analogy in the Kimmeridge Clay, early diagenetic dissolution of ferrimagnetic phases such as magnetite by H2S leaves paramagnetic clays as the predominant susceptibility carriers. Magnetic susceptibility logging thus gives an index of clay content and hence an inverse measure of calcium carbonate. Accordingly magnetic susceptibility logging in the field was used to allow direct lithological correlation with the susceptibility logging of the core material. One advantage of this procedure over natural gamma-ray measurement is that very closely spaced values can be obtained (minimum independent measurement spacing equals 2 cm). These data were used to check that the complete section was drilled and aided direct application of the micro- and macro-fossil biostratigraphy of the type section to the cores.

As described above, Reading investigated clay mineralogy using 250 samples at 10 cm intervals and an additional 50 samples at 10 m intervals throughout the section.  Work by Parker at Reading had found that variation in the concentration of rare-earth elements can reveal clay mineralogy and changes in clay provenance (see Reading section) Thus 300 ICP analyses for REE concentrations was undertaken at Luton on the same samples used for clay-mineral determination. The resulting high-resolution time-series was compared to the clay-mineralogical data in order to reveal Milankovitch-scale variations which could be related to climatic rather than purely diagenetic factors. Time series of all the major analytical data sets were  subjected to spectral analyses. Those parameters which reveal regular cyclicity were compared with the geophysical log and multi-sensor track data using cross-spectral methods. The coherency spectra were used to test which parameters are correlated at which spectral frequencies. Coherent parameters are likely to be related to the same climatic or diagenetic forcing mechanism. For the coherent pairs of parameters, phase spectra can reveal the timing of variations (i.e. which parameter changes before another).

Programme of Research - Newcastle (and others)

Palaeoproductivity and Preservational Controls on the TOC Cyclicity

The decimetre- to metre-scale cyclicity of the onshore Kimmeridge Clay Formation is primarily expressed by the variation in the biogenic component of the sediments, namely organic matter (kerogen) and carbonate (largely coccolithic). The Newcastle effort was directed towards elucidating the key mechanism responsible for the production of the organic-rich´┐Żorganic-poor cycles. Determining a Milankovitch origin did not solve this problem in itself. The questions addressed were: Are the cycles produced by variations in palaeoproductivity, palaeoxygenation, or dilution, or some combination of these, and, if the latter, how do the relative roles of these parameters vary within the formation?

The IFP Yorkim Group and associates recently addressed these issues in some detail, and have generally concluded that palaeoproductivity is the key control. The main arguments they have used to support this view are:

Although apparently compelling, these arguments are not definitive (Tyson, 1996). (a) The relative uniformity of the non-carbonate mineralogy cannot be considered surprising at the decimetre to metre scales in such comparatively distal facies. (b) The presence of lamination does not alone adequately distinguish the significant difference between suboxic and anoxic conditions (Tyson and Pearson, 1991). (c) The relatively constant proportion of the terrestrial component is not what would be expected from palaeoproductivity controlled increased fluxes of marine (AOM) material and a background stable terrestrial supply: a greater marine AOM flux ought to dilute the terrestrial component. The relatively constant terrestrial fraction implies that the variations in TOC are likely to be mostly due to changes in sediment dilution. The actual concentration of terrestrial particles (no./g) increases in the organic-rich part of the cycles (Waterhouse, 1995), which is probably due to lower sediment dilution (Tyson, in press). (e) The SRI index is probably strongly influenced by environment-related variations in "iron limitation" (Tyson, 1995). (f) The amorphous kerogen fraction that is most correlated with the peak TOC values is not the apparently most refractory ultralaminar variety (whose abundance might be expected to best reflect the original flux), but the truly nanoscopically amorphous material, whose distribution would appear to be more preservation controlled. (g) The asymptotic HI value (approx. 600-800) does not occur at a constant TOC, but can range from 6% up to 20% or more (e.g. Huc et al., 1992). Considered in the context of the other available information, such differences are likely mostly due to differences in sediment dilution rather than palaeoproductivity (Tyson, 1996).

Method and Approach

This study adopted several complementary approaches to assess the controls on the TOC cyclicity, which involved the critical testing of several models recently presented in the literature.

Application of the approach

The first approach used 'whole sequence' data generated at Southampton, and also supplementary higher resolution TOC and carbonate data generated at Newcastle. Because of its time-intensive nature, the second approach has only been carried out on cycles from the four key intervals within the formation that have been selected because of their sequence stratigraphic or other significance (cf. Tyson, 1996). The last approach analysed particular cycles in detail. In order to enhance the interpretation, toward the end of the project the results of the above approaches were integrated with information from the palaeoclimatic modelling (Reading) in order to gain a better understanding of the hydrographic behaviour of the Late Jurassic shelf seas, and to evaluate the "seasonal anoxia" models of Oschmann (1988) and Tyson and Pearson (1991).

References (and additional publications)

Search Form

Last updated: Saturday 25th July 2015

Design: Squiz Live Site Text Only