| description abstract | Currently, observations of key components of the earth's large-scale water and energy budgets are sparse or even nonexistent. One key component, precipitation minus evapotranspiration (P ? ET), remains largely unmeasured due to the absence of observations of ET. Precipitation minus evapotranspiration describes the flux of water between the atmosphere and the earth's surface, and therefore provides important information regarding the interaction of the atmosphere with the land surface. In this paper, large-scale changes in continental water storage derived from satellite gravity data from the Gravity Recovery and Climate Experiment (GRACE) project are combined with river discharge data to obtain estimates of areally averaged P ? ET. After constructing an equation describing the large-scale terrestrial water balance reflecting the temporal sampling of GRACE water storage estimates, GRACE-derived P ? ET estimates are compared to those obtained from a reanalysis dataset [NCEP/Department of Energy (DOE) reanalysis-2] and a land surface model driven with observation-based forcing [Global Land Data Assimilation System (GLDAS)/Noah] for two large U.S. river basins. GRACE-derived P ? ET compares quite favorably with the reanalysis-2 output, while P ? ET from the Noah model shows significant differences. Because the uncertainties in the GRACE results can be computed rigorously, this comparison may be considered as a validation of the models. In addition to showing how GRACE P ? ET estimates may be used to validate model output, the accuracy of GRACE estimates of both the seasonal cycle and the monthly averaged rate of P ? ET is examined. Finally, the potential for estimating seasonal evapotranspiration is demonstrated by combining GRACE seasonal P ? ET estimates with independent estimates of the seasonal cycle of precipitation. | |
