ASSESSING THE RELIABILITY OF GLOBAL MODELS USING COMPARISONS WITH GRACE SATELLITE DATA (Invited Presentation)

Global modeling is widely used to evaluate past and projected future climate impacts and human stresses on water resources. The objective of our analysis was to examine the reliability of global models by comparing decadal trends (2002–2014) in land water storage from seven global models (WGHM, PCR-GLOBWB, GLDAS NOAH, MOSAIC, VIC, CLM, and CLSM) to trends from three Gravity Recovery and Climate Experiment (GRACE) satellite solutions in 186 river basins, which cover ∼60% of the global land surface. Results show that modeled water storage trends in river basins greatly underestimate trends derived from GRACE satellites, both large decreasing and increasing trends. Decreasing water storage trends from GRACE are related to human water use, primarily irrigation, and climate variations, whereas increasing storage trends are related primarily to climate variations. The models do not necessarily perform better in large basins relative to small basins. For example, water storage in the Amazon Basin from GRACE increased by ∼43 km³/yr; however, most models simulated decreasing trends, ranging from -71 to 11 km³/yr. Summing land water storage trends over all basins results in positive values for GRACE (∼71–82 km³/yr) but negative values for models (−450 to −12 km³/yr), with opposite impacts on global mean sea level change. The models suggest that the effects of climate forcing are about two times greater than those of human intervention on decadal land water storage trends. This comparison of models and GRACE satellite data underscore the need to improve simulated land water storages in the future. The implications of this study are that model projections may underestimate the impacts of future climate and human induced water storage trends.