Global retrospective estimation of soil moisture using the variable infiltration capacity land surface model, 1980-93

A daily set of surface meteorological forcings, model-derived surface moist ure fluxes, and state variables for global land areas for the period of 197 9-93 is described. The forcing dataset facilitates global simulations and e valuation of land surface parameterizations without relying heavily on GCM output. Daily precipitation and temperature are based on station observatio ns, daily wind speeds are based on National Centers for Environmental Predi ction-National Center for Atmospheric Research reanalysis data, and the rem aining meteorological forcing variables (shortwave radiation, longwave radi ation, and vapor pressure) are derived from the precipitation and temperatu re series. The Variable Infiltration Capacity (VIC) land surface model is u sed to produce a set of derived fluxes and state variables, including snow water equivalent, evapotranspiration, runoff, and soil moisture storage. Th e main differences between the new dataset and other, similar datasets are the daily time step, the use of a specified simulation period as opposed to climatological averages, the length of the simulation period, the use of o bserved meteorological data, and the use of a more realistic hydrological m odel. Comparison with observations and existing climatologies shows that 1) the interannual variation in simulated snow cover extent is similar to obs ervations in Eurasia but is somewhat underpredicted in North America; 2) th e components of the global and continental water balance are similar to tho se in previously produced climatologies, although runoff is somewhat lower; 3) patterns of simulated soil moisture storage are similar to the climatol ogy of Mintz and Serafini, but the more sophisticated VIC hydrological mode l produces a larger range in soil moisture; and 4) the annual cycle and spa tial patterns in soil moisture compare well with soil moisture observations in Illinois and in central Eurasia, but mean modeled soil moisture is some what lower than observed, and observed soil moisture shows a greater intera nnual persistence than do the simulations.