Self-consistent modeling of ice-sheet evolution and paleoclimate

Abstract

A self-consistent approach has been developed to determine past climate histories while simultaneously determining the past ice-sheet evolution. We recognize that multiple physical processes are affected by the same climate history and ice-sheet evolution. By combining several processes into one self-consistent model based on physics of ice-sheet flow, heat flow, grain growth, and firn compaction, I can infer the climate history (accumulation rate and delta-age) and ice-sheet evolution (thickness and divide position), that match data sets from ice cores and ice-penetrating radar layers. Ice-sheet behavior has not previously been modeled to enforce self consistency. The self-consistent approach consists of modules, or subroutines, representing physical processes. I have developed forward models to simulate firn densification, grain growth, heat transfer, and ice flow, and inverse models to infer histories including the spatial pattern of accumulation, and the depth-age relationship for ice cores and radar layers. While individual modules can be replaced with modules based on a variety of physical approximations, I provide here proof of the concept that multiple data sets and multiple processes can be combined to provide improved estimates of ice-sheet histories that cannot be directly measured. This new approach provides a way to improve ice-core chronologies from Greenland and Antarctica, and to infer self-consistent histories of climate and ice-sheet evolution at those locations.