Hindcasting to measure ice sheet model sensitivity

Published: Aug 1, 2013 by The PISM Authors

Validation and assessment of model performance is critical, but it is notoriously-challenging in ice sheet modeling. This paper couples PISM to the HIRHAM5 regional climate model for simulations of the Greenland ice sheet. The results are compared to observations in the 1989-2011 period (hindcasting), in which ice geometry, ice surface velocity, gravitationally-derived mass time-series, and surface elevation change observations are all available. The simulations reproduce the seasonal signal and decadal trends in mass loss but they show deficiencies compared to observed changes in ice discharge. The paper concludes that it is important to use //multiple// data sets for model validation, and it identifies rates of change of spatially-dense observations as preferred validation metrics.