Soil moisture simulations in revised AMIP models

Soil moisture is important both in its influence on climate and for assessi ng impacts of future climate change. It is therefore necessary to simulate it correctly in global climate models. We have used the revisit simulations contributed by six of the Atmospheric Model Intercomparison Project 1 part icipating modeling groups to examine the impacts of model revisions, partic ularly the land surface representations, on soil moisture simulations, by c omparing the simulations to actual soil moisture observations. The revised models do not show any systematic improvement in their ability to simulate observed seasonal variations of soil moisture over the regions studied. Man y of the revised models continue to have a strong tendency toward dry soil conditions during Northern Hemisphere summer months, both globally and regi onally. There are no indications of conceptually more realistic land surfac e representations producing better soil moisture simulations in the revised climate models. As the revised models continue to produce incorrect simula tions of the seasonal cycle of regional precipitation, it is not possible t o isolate the effect of land surface schemes on the simulations. The Europe an Centre for Medium Range Weather Forecasts and National Centers for Envir onmental Prediction/National Center for Atmospheric Research reanalyses, ho wever, which are driven by observed precipitation, do capture some of the o bserved interannual variability of soil moisture over Illinois.