Estimation of Shelter Temperatures from Operational Satellite Sounder Data

Abstract

As a first step in the development of a technique for estimating daily maximum and minimum shelter temperatures for agricultural monitoring, this study made use of operational satellite sounder data to estimate shelter temperature. Linear regression methods were used with a ground-truth data set of surface observations matched against data from operationally derived soundings. Regression estimates based solely on temperature predictors from the satellite soundings yielded residual standard deviations of 1.6?2.6 K for the NOAA-6 and NOAA-7 clear and partly cloudy retrievals. Regressions for cloudy conditions based only on the microwave retrievals had errors ranging from 2.9?4.0 K. Shelter temperatures estimated by regression have errors somewhat smaller than those reported for the lower levels of the atmospheric soundings. This suggests that satellite temperature soundings near the surface are more accurate than previous studies indicate.