With ongoing global warming, the climate system is quickly moving outside the range humans have observed and adapted to. Because we lack direct observations for a world with more greenhouse gases, it is challenging to predict the precise nature of future changes. But we can turn to the past, as the Earth has previously experienced various levels of atmospheric CO2 concentrations. By reconstructing and simulating climate under such conditions, we can learn how the climate system operates under high CO2 conditions, including possible thresholds for changes in ocean circulation or ice sheet stability.
In the DOTpaleo project, we will study a time period of particularly high levels of CO2, the early Eocene (~55-50 million years ago), and the time leading up to and following this interval. During this time, CO2 concentrations were 4-8 times higher than before the industrial revolution; levels than could be reached in the next century if emissions continue unabated. Climate was much hotter and ice sheets did not exist, as shown by fossils of plants and even crocodiles on Antarctica.
Reconstructing how warm it was exactly and how the climate system operated this far back in time is challenging. For determining temperature in this ancient ocean, we rely on indirect signals from the composition of small fossil shells preserved in ocean sediments from that time. In DOTpaleo, we will use a new approach that will yield more reliable reconstructions than previously possible.
Another challenge is that in addition to the different CO2 concentrations, the geographic configuration of continents this far back in time was different from today, also affecting global climate. We will use climate model experiments to understand the climate effects of geography, atmospheric CO2, and the combination of both. Our results will enhance our understanding of the climate system under greenhouse conditions, and thus improve our predictions for future climate change.
The reconstructed temperatures we have so far show that the deep Atlantic was very warm, up to 20 °C (compared to around 2°C today). These results were published in Meckler et al. (Science, 2022). The temperatures are warmer than had been suggested by earlier reconstructions, which are associated with more uncertainties. We have recently concluded similar measurements in the Pacific Ocean, the results of which are now being prepared for publication. In addition to the warm temperatures, we have found large changes in temperature over time, which in some cases happened relatively quickly. These changes do not necessarily represent global temperature swings but could also be due to water masses of different temperature and salinity changing place in the deep ocean.
Our temperature reconstructions cover also time intervals known for large changes in climate, such as the very end of the Eocene, when the Antarctic ice sheet first grew to a considerable size. For that interval we have found that deep ocean temperature in the deep Pacific decreased right before the ice growth, but then increased again right afterwards to similar levels as before (published in Taylor et al., 2022). This finding is surprising and has recently been complemented by similar measurements from the North Atlantic, showing a very different pattern. Both the large changes in deep ocean temperature and the heterogeneity of the deep ocean water masses have not been recognized before and suggest that the ocean circulation operates in a fundamentally different way from today in a greenhouse world with less or no ice.
We have also been running climate model simulations for the early, mid, and late Eocene with different CO2 concentrations and geography, with special focus on the connection between the Atlantic and the Arctic Ocean. The results suggest that both factors play an important role for ocean circulation and thus for heat transport in the ocean. In addition, we have analyzed existing model simulations from different research groups worldwide to better understand the connection between surface and deep ocean temperatures and link the model results to our reconstructed deep ocean temperatures.
Global climate is undergoing rapid change due to rising levels of atmospheric CO2 concentration (pCO2). The scale of the projected change is far outside of what has been experienced by humankind, so we have to turn to the geologic record to understand how the climate system responds to such levels of greenhouse gases. The last time pCO2 approached and exceeded 1000 ppm (a level projected for the end of this century in some scenarios), was during the Eocene, 56-34 million years ago (Ma). This period and the preceding Paleocene (65-56 Ma) will be studied in this project.
When studying past climates, one has to rely on indirect information from so-called climate proxies. Unfortunately, these proxies are often influenced by multiple factors, increasing the uncertainty of the reconstructions. In DOTpaleo, we are applying a more recently developed proxy, called clumped isotope thermometry, which does not depend on other factors besides temperature. While analytically challenging and time consuming, this method yields more reliable time series of ocean temperature changes. First results show that deep ocean temperatures were warmer than previously thought during the warmest part of the Eocene, implying that ocean temperature is more sensitive to pCO2 than previously thought. In DOTpaleo we will expand this work and derive more detailed temperature records from different locations in the global ocean. The data will be used to test climate models, including our in-house NorESM, and the comparison with model simulations can in turn help us understand the processes leading to the observed deep ocean temperatures. We will furthermore conduct targeted numerical experiments with NorESM to understand the sensitivity of climate to pCO2 under different background climate and ocean circulation states. Our results will lead to an improved understanding of the climate system under greenhouse conditions, such as those potentially awaiting us in the future.
