The role of climate forcing in modeling the evolution of northern hemisphere ice sheets

GCM-derived climate differences between the last glacial maximum (LGM) and thepresent day play a strong role when used as forcing in ice sheet models in determiningkey glaciological predictors, such as ice volume, ice sheet surface area andthe timing of maximum ice volume. Given this strong dependen...

Full description

Bibliographic Details
Main Authors: Zweck, C., Huybrechts, Philippe
Format: Conference Object
Language:unknown
Published: 2004
Subjects:
Online Access:https://epic.awi.de/id/eprint/10335/
https://hdl.handle.net/10013/epic.20822
Description
Summary:GCM-derived climate differences between the last glacial maximum (LGM) and thepresent day play a strong role when used as forcing in ice sheet models in determiningkey glaciological predictors, such as ice volume, ice sheet surface area andthe timing of maximum ice volume. Given this strong dependence, the role of climateforcing on the results of ice sheet models warrants close examination. Using theECBILT/CLIO coupled atmosphere-ocean-sea ice model we have generated climatedata spanning the period from LGM to present day at 200 yr resolution. The atmosphericcomponent of this model generates the ice sheet model forcing and is a T21,3 level, quasi-geostrophic model. Time dependent forcing for ECBILT/CLIO modelconsists of changes in ice sheet elevation and albedo, solar insolation, and the concentrationsof greenhouse gases. The climate data generated in this manner is thenused in anomaly mode from the present day as a high temporal resolution forcing inthe ice sheet model. This procedure generates an ice sheet history more in agreementwith the geomorphological evidence than when generated using simple two time slice(LGM and present day) climate forcing using GRIP ice core data to interpolate climatestates over a glacial cycle. To examine the specific effect of the role of ECBILT/CLIOderived climate forcing for the ice sheet model, model results are presented using reducedtemporal resolution forcing and also using linear interpolation instead of GRIPdata.