STATISTICAL VALIDATION OF GCM-SIMULATED CLIMATES FOR THE UNITED-STATES GREAT-LAKES AND THE CIS EMBA AND URAL RIVER BASINS

Many researchers use outputs from large-scale global circulation model s of the atmosphere to assess hydrological and other impacts associate d with climate change. However, these models cannot capture all climat e variations since the physical processes are imperfectly understood a nd are poorly represented at smaller regional scales. This paper stati stically compares model outputs from the global circulation model of t he Geophysical Fluid Dynamics Laboratory to historical data for the Un ited States' Laurentian Great Lakes and for the Emba and Ural River ba sins in the Commonwealth of Independent States (C.I.S.). We use maximu m entropy spectral analysis to compare model and data time series, all owing us to both assess statistical predictabilities and to describe t he time series in both time and frequency domains. This comparison ini tiates assessments of the model's representation of the real world and suggests areas of model improvement.