Coupling Between Air and Ground Temperatures in PMIP3/CMIP5 Last Millennium Simulations and the Implications for Climate Reconstructions from Borehole Temperature Profiles

Abstract:

The continental energy storage for the second half of the $20^{th}$ century has been estimated from geothermal data to be about $7 \pm 1 \times 10^{21} J$ under the assumption that there exists a long-term coupling between the lower atmosphere and the continental subsurface. For General Circulation Models (GCMs) to simulate the continental energy storage of the Earth's energy budget, however, it is crucial that they correctly capture the processes that partition energy across the land-atmosphere boundary. We evaluate herein the characteristics of these processes as simulated by models in the third phase of the Paleoclimate Modelling Intercomparison Project and the fifth phase of the Coupled Model Intercomparison Project (PMIP$3$/CMIP$5$). We examine the seasonal differences between air and ground temperatures within PMIP3 last-millennium simulations concatenated with historical simulations from the CMIP5 archive. We find a strong air-ground coupling during the summer from $850$ to $2000$ CE. During the winter, the insulating effect of snow and latent heat exchanges produce a decoupling between air and ground temperatures in the northern high latitudes. These seasonal differences decrease with depth, supporting the central assumption of climate reconstructions from borehole temperature profiles. Additionally, we use the simulated temperature trends as an upper boundary condition to force a one-dimensional conductive model to derive synthetic temperature-depth profiles for each PMIP3/CMIP5 simulation. The inversions of these subsurface profiles yield temperature trends that retain the surface temperature variations of the last millennium for all the PMIP3/CMIP5 simulations. These results support the use of underground temperatures to reconstruct past changes in ground surface temperature and to estimate the continental energy storage. Results also provide guidance for improving the land-surface components of GCMs.