Paul Markwick

Descent toward the Icehouse: Eocene sea surface cooling inferred from GDGT distributions

The TEX86 proxy, based on the distribution of marine isoprenoidal glycerol dialkyl glycerol tetraether lipids (GDGTs), is increasingly used to reconstruct sea surface temperature (SST) during the Eocene epoch (56.0–33.9 Ma). Here we compile published TEX86 records, critically reevaluate them in light of new understandings in TEX86 palaeothermometry, and supplement them with new data in order to evaluate long-term temperature trends in the Eocene. We investigate the effect of archaea other than marine Thaumarchaeota upon TEX86 values using the branched-to-isoprenoid tetraether index (BIT), the abundance of GDGT-0 relative to crenarchaeol (%GDGT-0), and the Methane Index (MI). We also introduce a new ratio, %GDGTRS, which may help identify Red Sea-type GDGT distributions in the geological record. Using the offset between TEX86H and TEX86L (ΔH-L) and the ratio between GDGT-2 and GDGT-3 ([2]/[3]), we evaluate different TEX86 calibrations and present the first integrated SST compilation for the Eocene (55 to 34 Ma). Although the available data are still sparse some geographic trends can now be resolved. In the high latitudes (>55°), there was substantial cooling during the Eocene (~6°C). Our compiled record also indicates tropical cooling of ~2.5°C during the same interval. Using an ensemble of climate model simulations that span the Eocene, our results indicate that only a small percentage (~10%) of the reconstructed temperature change can be ascribed to ocean gateway reorganization or paleogeographic change. Collectively, this indicates that atmospheric carbon dioxide (pCO2) was the likely driver of surface water cooling during the descent toward the icehouse.

Late Cretaceous and Cenozoic global palaeogeographies: Mapping the transition from a “hot-house” to an “ice-house” world

A set of late Cretaceous stage level and Cenozoic sub-epoch level palaeogeographic maps have been constructed for 20 intervals representing the last 100 million years of Earth history. Maps include five levels of terrestrial elevation: 0 m (shoreline), 200 m, 1000 m, 2000 m, 3000 m; two levels of bathymetry: –50 m and –200 m; and terrestrial ice extent. These maps provide boundary conditions for GCM (General Circulation Model) experiments and a backdrop for terrestrial palaeoecological, palaeoclimatological, and palaeobiogeographical studies, which are essential for understanding the palaeoclimatic transition from the “hot-house” world of the Mesozoic and early Cenozoic, to the “ice-house” world of the Recent. The method of constructing these maps follows that of the Paleogeographic Atlas Project as described in Ziegler et al. (1985), in which the observed relationship between modern elevation and tectonophysiographic and environmental settings are used to assign values to similar settings in the geological record. Assumptions include that the modern relationship between elevation and tectonics is acceptable for reconstructing past topography (i.e. a uniformitarianist approach), and that mountain belts are generally long-lived features. Additional information, including sedimentological and palaeobiogeographical evidence, is used where available. The data are derived entirely from the published literature, utilizing the large lithological and reference databases of the Paleogeographic Atlas Project (Chicago), the databases of Markwick (1996), and the libraries of the University of Chicago and University of Reading. The maps were initially drawn at Chicago on basemaps at approximately 1:100,000,000 scale, on which were printed reconstructed plate outlines (the plate reconstructions are those of Rowley 1995 unpubl.), a simplified present coastline, and a modern five degree grid (all rotated to their appropriate past position). These maps were subsequently redrawn at a scale of 1:45,000,000 (Mollweide) and 1:30,000,000 (Polar orthographic), and then digitized as ARC/INFO coverages using polar projections for areas poleward of about 50–60°, and the Mollweide projection for the tropical and equatorial regions (ARC/ INFO is a desktop Geographic Information System, GIS). Consequently these maps may be considered to be precise at about 1:30,000,000, but it is not recommended that they be enlarged above this scale. Revisions to these maps are now being made at 1:30,000,000 scale for equatorial regions and 1:20,000,000 scale for polar regions above 30° north and south, and include the addition of the following contours: 3000 m, 4000 m, 5000 m, –1000 m, –2000 m, and –4000 m. The general drainage courses of major rivers have also been added where appropriate. The resulting maps provide the opportunity for a better understanding of not only the interaction between geography, topography, and climate, but also of the role of geography in delimiting past biogeographic and biodiversity patterns. They highlight the coincidence of orographic changes with the known palaeoclimatic cooling of the middle and late Eocene: the uplift of the Transantarctic Mountains, Rocky Mountains, and those of central Asia, together with increasing isolation of the Antarctic continent. Changes in the distribution of major mid-latitude seaways are also suggestive of a link with climate changes on at least a regional scale. The presence of highland in East Antarctica (Gamburtsev subglacial Mountains) continues to be problematic, since their great height might be assumed to be the locus of ice accumulation, even during ‘hot-house’ intervals. However, Markwick & Rowley (1998) have shown that even if these mountains were glaciated above 1500 or 2000 m, it would only accommodate an icesheet comparable to a eustatic sea-level change today of about 8 or 2 m, respectively, which would probably not be resolvable on a global scale in the geological record. It is also interesting to note that the initiation of glaciation of Antarctica in the middle Eocene coincides with the continent moving slightly off of the pole. Other features of interest include the isolation of the Arctic Ocean in the middle Cretaceous and early Palaeogene, and the blocking effect of northwards moving India on equatorial ocean circulation. The palaeogeographic maps generated in this study provide the essential spatial context for understanding the transition from the “hot-house” to the “ice-house” world. The use of GIS technology enhances their utility, by providing digital links between spatial features and the data upon which those features are based, making the maps more accessible to users, while also facilitating the ease with which future changes can be made to them. It is anticipated that they will be of interest and use to researchers in a variety of geological fields.

The impact of Cenozoic cooling on assemblage diversity in planktonic foraminifera

The Cenozoic planktonic foraminifera (PF) (calcareous zooplankton) have arguably the most detailed fossil record of any group. The quality of this record allows models of environmental controls on macroecology, developed for Recent assemblages, to be tested on intervals with profoundly different climatic conditions. These analyses shed light on the role of long-term global cooling in establishing the modern latitudinal diversity gradient (LDG)—one of the most powerful generalizations in biogeography and macroecology. Here, we test the transferability of environment-diversity models developed for modern PF assemblages to the Eocene epoch (approx. 56–34 Ma), a time of pronounced global warmth. Environmental variables from global climate models are combined with Recent environment–diversity models to predict Eocene richness gradients, which are then compared with observed patterns. The results indicate the modern LDG—lower richness towards the poles—developed through the Eocene. Three possible causes are suggested for the mismatch between statistical model predictions and data in the Early Eocene: the environmental estimates are inaccurate, the statistical model misses a relevant variable, or the intercorrelations among facets of diversity—e.g. richness, evenness, functional diversity—have changed over geological time. By the Late Eocene, environment–diversity relationships were much more similar to those found today.

Atmospheric and oceanic impacts of Antarctic glaciation across the Eocene-Oligocene transition.

The glaciation of Antarctica at the Eocene–Oligocene transition (approx. 34 million years ago) was a major shift in the Earth’s climate system, but the mechanisms that caused the glaciation, and its effects, remain highly debated. A number of recent studies have used coupled atmosphere–ocean climate models to assess the climatic effects of Antarctic glacial inception, with often contrasting results. Here, using the HadCM3L model, we show that the global atmosphere and ocean response to growth of the Antarctic ice sheet is sensitive to subtle variations in palaeogeography, using two reconstructions representing Eocene and Oligocene geological stages. The earlier stage (Eocene; Priabonian), which has a relatively constricted Tasman Seaway, shows a major increase in sea surface temperature over the Pacific sector of the Southern Ocean in response to the ice sheet. This response does not occur for the later stage (Oligocene; Rupelian), which has a more open Tasman Seaway. This difference in temperature response is attributed to reorganization of ocean currents between the stages. Following ice sheet expansion in the earlier stage, the large Ross Sea gyre circulation decreases in size. Stronger zonal flow through the Tasman Seaway allows salinities to increase in the Ross Sea, deep-water formation initiates and multiple feedbacks then occur amplifying the temperature response. This is potentially a model-dependent result, but it highlights the sensitive nature of model simulations to subtle variations in palaeogeography, and highlights the need for coupled ice sheet–climate simulations to properly represent and investigate feedback processes acting on these time scales.

The cause of Late Cretaceous cooling: A multimodel-proxy comparison

Proxy temperature reconstructions indicate a dramatic cooling from the Cenomanian to Maastrichtian. However, the spatial extent of and mechanisms responsible for this cooling remain uncertain, given simultaneous climatic influences of tectonic and greenhouse gas changes through the Late Cretaceous. Here we compare several climate simulations of the Cretaceous using two different Earth system models with a compilation of sea-surface temperature proxies from the Cenomanian and Maastrichtian to better understand Late Cretaceous climate change. In general, surface temperature responses are consistent between models, lending confidence to our findings. Our comparison of proxies and models confirms that Late Cretaceous cooling was a widespread phenomenon and likely due to a reduction in greenhouse gas concentrations in excess of a halving of CO 2 , not changes in paleogeography.

Late Cretaceous climate simulations with different CO2 levels and subarctic gateway configurations: A model-data comparison

We investigate the impact of different CO2 levels and different subarctic gateway configurations on the surface temperatures during the latest Cretaceous using the Earth System Model COSMOS. The simulated temperatures are compared with the surface temperature reconstructions based on a recent compilation of the latest Cretaceous proxies. In our numerical experiments, the CO2 level ranges from 1 to 6 times the preindustrial (PI) CO2 level of 280 ppm. On a global scale, the most reasonable match between modeling and proxy data is obtained for the experiments with 3 to 5 × PI CO2 concentrations. However, the simulated low- (high-) latitude temperatures are too high (low) as compared to the proxy data. The moderate CO2 levels scenarios might be more realistic, if we take into account proxy data and the dead zone effect criterion. Furthermore, we test if the model-data discrepancies can be caused by too simplistic proxy-data interpretations. This is distinctly seen at high latitudes, where most proxies are biased toward summer temperatures. Additional sensitivity experiments with different ocean gateway configurations and constant CO2 level indicate only minor surface temperatures changes (<~1°C) on a global scale, with higher values (up to ~8°C) on a regional scale. These findings imply that modeled and reconstructed temperature gradients are to a large degree only qualitatively comparable, providing challenges for the interpretation of proxy data and/or model sensitivity. With respect to the latter, our results suggest that an assessment of greenhouse worlds is best constrained by temperatures in the midlatitudes.

‘Equability’ in an unequal world: The early Eocene revisited

The Mesozoic and early Cenozoic are often referred to as “hothouse” worlds in reference to the interpreted greater warmth of these intervals, relative to the Recent, and the lack of appreciable polar ice-sheets. Associated with this warmth is often the interpretation of thermal ‘equability’, which computer climate modelling experiments have consistently had problems replicating, especially in mid-high latitude continental interiors. In this study we re-examine the early Eocene ‘hot-house’ world using two different General Circulation Models (GCMs): the UK Universities Modelling Project (UGAMP) GCM and the more recent Hadley Centre (UKMO) GCM. In each case the following boundary conditions were set: 4 × CO2; a sea surface temperature gradient defined by 27cos(latitude); palaeogeography from Markwick et al. (this volume p. 103); a uniform, shrub-like vegetation; no permanent ice sheets; all other parameters as per the present day. GCMs provide the best means of understanding the dynamics responsible for past climates, but can only be assessed for their veracity through comparisons with observations. For palaeoclimate these observations are derived from the geological record. Palaeoclimate interpreted from geological data is invariably based on analogy with the Recent, the validity of which depends on corroboration from multiple lines of evidence. This requires the compilation and investigation of large, global datasets of well-constrained geological climate proxies. However, in too many data-model studies the comparisons are qualitative with a modelling experiment considered successful if a ‘warm’ prediction coincides with a climate proxy that indicates ‘warmth’. Although the qualitative comparison of model results against observations provides a broad indication of model success or failure, there is, however, a need for a more quantitative approach. In this paper, we present a quantitative method derived by one of the authors (PJM), in which model results and data are directly compared one with another. In order to illustrate this method we concentrate on only one proxy (fossil crocodilians), while recognizing that it is only through intercomparison between the interpretations from multiple and diverse climate proxies that we can more confidently ‘ground-truth’ modelling experiments. The climate space defined by modern crocodilians has been found to be delimited by a minimum coldest month mean temperature (CMM) of 5°C and the presence of water bodies (Markwick 1998). Today this coincides with a mean annual temperature (MAT) of about 16°C, but whether this is also true for the past cannot be assumed. The method used is as follows. First, the climate space occupied by each climate proxy is defined for the Recent, using a dataset of 1060 globally distributed climate stations: each proxy is ‘recorded’ for all climate stations that fall within the present geographic distribution of that proxy. Then the geological occurrences of each proxy (in this study, fossil crocodilians) are overlain onto the gridded GCM output, and for each proxy location the model output value extracted. These values are then replotted in climate space where they can be directly compared with that based on the present day (for this study of ‘equability’ that climate space is plotted on a graph of MAT versus mean annual range of temperature, MART; CMM can be illustrated on the same diagram; Fig.1). A statistical comparison between the distributions can then be made (using the Mann Witney nonparametric test, for example), although account must be made of the differing sampling densities between recent and fossil records. This method also facilitates model–model comparisons: the position of fossil crocodilians for each GCM run are plotted in the same climate space and a vector drawn between corresponding crocodilian localities. This vector is then a measure of the difference between model results (Fig. 1). For the current study we have used ArcView GIS Spatial Analyst for extracting the GCM values for each proxy location, which has the advantage that the underlying palaeogeographic context of the data can simultaneously be displayed, and StatView and StatisticaMac for the statistical analysis of results. Our results show the following. For both models about half of the early Eocene fossil crocodilian localities fall outside of the climate space defined by modern crocodilians. These fossil localities are concentrated in mid-latitude continental settings, where the modelled CMMs are <5°C. This suggests that both models are underestimating CMMs in these areas. This may be due to a number of reasons including errors in specified GCM boundary conditions, such as incorrect orography (that CMM is too low in mid-latitude interiors might be due to the orography being too high), land–sea distribution (under-estimating seaway extents), or vegetation (ignoring the potential ameliorating af-