84th AMS Annual Meeting

Tuesday, 13 January 2004
Use of GOES solar radiation data to improve long-term retrospective land surface simulations
Hall 4AB
Nathalie Voisin, University of Washington, Seattle, WA; and D. P. Lettenmaier and R. Pinker
Solar radiation is one of the primary determinants of land-atmosphere interactions. Long-term retrospective simulations using macroscale hydrologic models like the Variable Infiltration Capacity (VIC) model have proved useful for various purposes including examination of the role of land surface variables in long-term climate predictability, and initialization of regional climate models. We have performed such simulations for a 50-year period (1950-2000), which is now being extended to the early 1900s. Over most of this period, there were few or no direct observations of downward solar radiation, so we have instead estimated daily total insolation using algorithms developed by others based on the daily temperature range, and rescaled the diurnal cycle (3-hourly time resolution in our case) to theoretical clear sky insolation for the given time of year. The availability of a multi-year (1996-1999) GOES-based hourly solar radiation data set for the continental U.S. now offers the opportunity to evaluate and adjust the temperature index results for the overlap period, which we extend to the entire retrospective period of record. We show the geographic distribution of adjustments both to the total daily insolation, and its diurnal distribution, over the continental U.S. In addition, we compare both the GOES and adjusted temperature range-based values to six Surface Radiation Budget Network (SURFRAD) sites distributed across the continental U.S.

Supplementary URL: