全球变化知识资源中心

The Earth system changed dramatically across the Eocene-Oligocene Transition (EOT) on a variety of spatial and temporal scales. Understanding the many complex and interacting factors affecting the Earth's atmosphere and oceans at the EOT requires the combination of both data and modeling approaches and an understanding of the uncertainty in both of these elements. Here uncertainty in the Earth system response to various imposed forcings typical of changes at the EOT is assessed. By using an ensemble of simulations from the fully coupled general circulation model, HadCM3L, the uncertainty due to differences in the boundary conditions and insufficient model spin-up is quantified. The surface temperature response in high-latitude ocean regions, particularly where deep water formation occurs, is found to be highly sensitive to differences in boundary conditions (i.e., have the greatest magnitude of uncertainty), while low-latitude oceans are the most insensitive to differences in boundary conditions (i.e., have the lowest magnitude of uncertainty). The length of spin-up (or how far the model is from equilibrium) can have a significant effect on the response to some forcings and on the magnitude of uncertainty due to differences in boundary conditions. These findings are important to consider for future modeling work and for interpreting previous published simulations.

Plain Language Summary Around 34 million years ago, the first major ice sheet formed over Antarctica. Understanding why the Earth cooled at this time and how this would have impacted on global climate requires the use of geologic data and computer models. Here multiple coupled climate-ocean model simulations are used to assess changes in modeled climate around this period in the Earth's history. Specifically, the climatic sensitivity to subtle changes in the model setup and how long the simulations are run for (the model spin-up) are evaluated. The temperature response in certain regions is found to be particularly sensitive to changes in the model setup and lack of spin-up. This is the case particularly for high-latitude ocean regions, making it harder to build up a consistent picture of how the climate of these regions changed at this time. Lower latitude oceans are much less sensitive, giving us higher confidence in model results for these regions. In these simulations, the global changes that occurred around 34 million years ago are best described by declining atmospheric pCO(2).