As anthropogenic atmospheric warming is forecasted to exceed 2C above preindustrial temperatures by 2100, a key uncertainty in predicting the impact of this change is the quantitative understanding of how this warming will be distributed in the oceans and atmosphere. One means of assessing this is to look to the geological past, especially the late Cretaceous to Eocene (100-34 Ma ago), when atmospheric pCO2 levels were last as high as the 700 ppmv forecasted for 2100, and global mean annual temperatures (MAT) were up to 8C warmer than today (the so-called "Greenhouse World"). Fossil data suggest that temperature-sensitive organisms, such as reptilians, were living in the Arctic-circle during this period, and led to the emergence of the "Equable Earth" hypothesis – a scenario that invokes near total collapse of the meridional, equator-to-pole temperature gradient at this time.

This indicates a climate system that operated in a fundamentally different way to the modern "Icehouse World", with a different/enhanced means of transporting heat from the tropics to the poles. A fundamental problem for scientists aiming to predict future climate change, is that state-of-the-art models are not able to reproduce the degree of collapse of the global meridional temperature gradient suggested by fossil data, reflecting a problem with either the "Equable Earth" hypothesis, or with climate modelling. Either way, this uncertainty impedes our ability to confidently predict the impact of future climate change with far-reaching implications. This research will be the first robust test of the "Equable Earth" hypothesis. We will reconstruct meridional variation in land surface MAT in a transect along the North American Continent, spanning mid- to high-palaeolatitude for several discrete time-equivalent instantaneous time-slices spanning the Cretaceous-Palaeogene (K-Pg) boundary – an interval in the middle of the "Greenhouse World". The MATs will be reconstructed using the brGDGT palaeotemperature proxy from collected coal samples. brGDGTs are lipids produced by bacteria thriving in terrestrial environments, whose distribution is a function of land surface MAT and can be used to reconstruct land surface MATs. We have identified ten separate sites, spanning 47-75N of palaeolatitude, where coals (fossil peats) were demonstrably accumulating coevally, by the occurrence within each of the coals of the globally synchronous Iridium (Ir)-enriched layer that settled from the atmosphere after the impact of a meteorite at the K-Pg boundary. In addition to the Ir-enriched layer, the coals contain datable tephra horizons, which will constrain vertical rates of change of MAT from time-slice to time-slice. They also contain distinctive carbon isotopic events before, during and after the Ir- enriched layer, which provide additional correlatable time lines between all locations. Combined, this provides an unique opportunity to generate serial time-slice reconstructions of meridional land surface MAT gradients, spaced at sub-orbital durations, at this critical period in Earth history. This will provide us with the opportunity to critically test the "Equable Earth" hypothesis, by placing numerical bounds on meridional MAT gradients for a series of time slices in continental interiors at this time. By generating meridional MAT gradients for multiple intervals, and by generating a tephrochronologically-based time-series through the succession, it will be possible to place bounds on the rates of change of MAT in time, from mid- to high- latitude. This will also reveal, for the first time, the dynamics in space and time of the "Greenhouse Earth" climate system, and will also allow us to assess MAT in the aftermath of meteorite impact at the K-Pg boundary, giving insight into the response of the climate system to catastrophic change, and allowing us to test competing hypotheses of climate change as the driver for the mass extinction at the K-Pg boundary.

Grant reference
Natural Environment Research Council
Total awarded
£582,024 GBP
Start date
1 Mar 2019
3 years 8 months 29 days
End date
30 Nov 2022