Modeling the timing and extent of glaciations over southeastern Tibet during the last glacial stage

Deciphering the variations of past glaciations over southeastern Tibet which holds the largest mass of contemporary temperate glaciers over Asia helps to advance our understanding of the ice-climate interactions and how glaciers may respond to future climate change. Using an ice sheet model forced by the reconstructed climatic change during the past 88 ka, we investigate the timing and extent of glaciations over southeastern Tibet and explore the potential influences of model parameters and horizontal resolution. Our results illustrate a long-term decreasing trend in the glacierized area during the past 88 ka, suggesting progressively restricted glaciations over southeastern Tibet. Apart from the long-term trend, there are multiple glacier advance events on millennial timescales. The maximum extent of glaciation occurs at late MIS 4 (∼58 ka) and early MIS 3 (∼51 ka), with the minimum glacier extent predicted in the late Holocene (∼4 ka). The evolution of glaciations during the past 88 ka is dominated by the change in air temperature in our simulations in which temperature and precipitation serve as the climatic forcing. Regarding modeling uncertainty, we show that the pattern of glacier evolution is independent of the values of model parameters and horizontal resolution used, but the amplitude of glacier extent is significantly affected, especially for model parameters. Compared with glacial geological evidence, the timing of the modeled widespread glaciations broadly matches with the ages of known moraines, but the mapped glacier bounds across specific mountain ranges are generally underestimated. The adoption of a spatially uniform climatic forcing and the related uncertainty may contribute to the model-data discrepancy. Our results aid in the establishment of a testable link between climate change and glacier response during the last glacial stage and have important implications for glacier behavior in the future.