Using an Earth System Model to Better Understand Ice Sheet Variability Through the Pleistocene

Abstract:

We use an Earth System model with a dynamic land-ice component to explore several inconsistencies between traditional Milankovitch theory and δ¹⁸O sediment records of the Pleistocene. Our model results show that a combination of albedo feedbacks, seasonal offset of precession forcing, and orbital cycle duration differences can explain much of the 41-kyr glacial cycles that characterize the early Pleistocene. The obliquity-controlled changes in annual average high-latitude insolation produce large variations in arctic vegetation-type and sea-ice cover, which amplify the land-ice response. In contrast, the seasonal nature of the precession insolation signal dampens net ice-melt. For instance, when precession enhances ice melt in the spring, it reduces ice melt in the fall, and vice versa. The lower frequency of obliquity cycles in combination with amplified climate sensitivity due to albedo feedbacks help produce a larger ice-volume response to cycles of obliquity compared to precession, despite precession contributing more to variations in high-latitude summer insolation.

In addition, we can simulate the appearance of a 100-kyr ice-volume signal by reducing basal sliding in the ice sheet model. Model experiments with enhanced basal drag have greater ice sheet elevation because the ice sheets are not able to flow as quickly, leading to increased ice thickness at the expense of ice extent. These thicker ice sheets have colder surface temperatures, receive more snowfall, and do not readily advance past the ice equilibrium line. Greater high-latitude summer insolation from the combination of high obliquity and precession/eccentricity is then necessary to cause complete ice sheet retreat. This research lends support to the regolith hypothesis, which proposes gradual erosion of high-latitude northern hemisphere regolith by multiple cycles of glaciation helped cause the mid-Pleistocene transition.