Satellite estimates of surface radiative fluxes for the extended San Pedro Basin: sensitivity to aerosols

Surface downwelling and upwelling radiative fluxes are important inputs into hydrologic models that evaluate water budgets, and into land surface data assimilation schemes which are driven with radiative fluxes. For large-scale needs, only remote sensing methods can provide such information. The accuracy of the derived fluxes depends on the inference schemes and on the quality of auxiliary input parameters. At present, information on surface short-wave (SW) radiative fluxes over the United States is produced in real time by the National Oceanic and Atmospheric Administration (NOAA)/National Environmental Satellite Data and Information Service (NESDIS) at 0.5° resolution, at hourly time intervals, using independently derived auxiliary inputs. Information on aerosol properties and their temporal variability is not available, and at best, is only estimated. During 1997 information on aerosol optical properties was collected at the USDA-Agricultural Research Service Walnut Gulch Experimental Watershed, Arizona, in preparation for future validation efforts in support of new satellite observations (e.g., ADEOS-II). This data set was used to test the sensitivity of a radiation inference scheme to aerosols, in particular, on the determination of clear sky fluxes and the surface albedo. Data from the Arizona meteorological network (AZMET) have been utilized to evaluate the satellite estimates for 1997. It was found that the current satellite estimates are within 70 W m−2 of the ground observations on an hourly time scale and within 24 W m−2 on a daily time scale. In the latter case this is less than 10% of the mean. Use of actual observations of aerosols, as compared to climatological values, reduces the bias substantially, while less significant changes in the r.m.s. were found.