Abstract
Abstract. Simulations of the glacial–interglacial history of the Antarctic Ice Sheet provide insights into dynamic threshold behavior and estimates of the ice sheet's contributions to global sea-level changes for the past, present and future. However, boundary conditions are weakly constrained, in particular at the interface of the ice sheet and the bedrock. Also climatic forcing covering the last glacial cycles is uncertain, as it is based on sparse proxy data. We use the Parallel Ice Sheet Model (PISM) to investigate the dynamic effects of different choices of input data, e.g., for modern basal heat flux or reconstructions of past changes of sea level and surface temperature. As computational resources are limited, glacial-cycle simulations are performed using a comparably coarse model grid of 16 km and various parameterizations, e.g., for basal sliding, iceberg calving, or for past variations in precipitation and ocean temperatures. In this study we evaluate the model's transient sensitivity to corresponding parameter choices and to different boundary conditions over the last two glacial cycles and provide estimates of involved uncertainties. We also discuss isolated and combined effects of climate and sea-level forcing. Hence, this study serves as a “cookbook” for the growing community of PISM users and paleo-ice sheet modelers in general. For each of the different model uncertainties with regard to climatic forcing, ice and Earth dynamics, and basal processes, we select one representative model parameter that captures relevant uncertainties and motivates corresponding parameter ranges that bound the observed ice volume at present. The four selected parameters are systematically varied in a parameter ensemble analysis, which is described in a companion paper.
Highlights
Process-based models provide the tools to reconstruct the history of the Antarctic Ice Sheet, leading to a better understanding of involved processes and thresholds regarding the ice sheet’s evolution in the past as well as in the future (Pattyn, 2018)
The modeled ice volumes converge at the Last Glacial Maximum (LGM; 15 kyr BP) with less than 0.2 m SLE difference among the simulations, but they differ for present-day conditions by up to 2 m SLE
As no WAIS collapse is induced in the simulations, we find very similar ice volumes for the Last Interglacial (LIG) and present-day periods
Summary
Process-based models provide the tools to reconstruct the history of the Antarctic Ice Sheet, leading to a better understanding of involved processes and thresholds regarding the ice sheet’s evolution in the past as well as in the future (Pattyn, 2018). This study motivates choices of boundary conditions and climatic parameterizations for application in large-scale paleo-ice sheet simulations and provides an assessment of the associated sensitivity of the model’s response. We run simulations of the entire Antarctic Ice Sheet with the Parallel Ice Sheet Model (Winkelmann et al, 2011; The PISM authors, 2020b) and describe a spin-up procedure for uncertain state variables, such as the three-dimensional enthalpy field or the till friction angle at the base. PISM paleo-simulations are initiated with a spin-up procedure using a fixed ice sheet geometry, in which the threedimensional enthalpy field can adjust to mean modern climate boundary conditions over a 200 kyr period. The sensitivity of the modeled ice volume above flotation to different choices of parameters and boundary conditions is evaluated as the difference to a baseline simulation (see movie in the Supplement) that is consistent with the model configuration of the best-fit ensemble simulation presented in a companion paper (Albrecht et al, 2020, see plots in Sect. 3.3 therein)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
