Effects of the Choice of Meteorological Data on a Radiation Model Simulation of the NOAA Technique for Estimating Outgoing Longwave Radiation from Satellite Radiance Observations
The technique used by NOAA to estimate the outgoing longwave flux from 10 μm window radiance observations has been reexamined because the data that result from the application of the empirically determined regression equation are systematically lower than those obtained from regression models based on earlier radiative transfer calculations. A new set of radiation calculations was made from a set of 1600 atmospheric soundings and the resulting regression equation gives flux differences from the empirical model that are of the order of ±5 W m−2 over the range of 150 to 300 W m−2, as compared to the +10 W m−2 systematic differences from previous studies. The differences are attributed to the size and representativeness of the sample of soundings used in the radiation calculations. The results also show that although the explained variance of the regression is of the order of 95%, this type of estimation technique may make errors of ±20 W m−2 or larger for a given radiance observation. This will lea... Abstract The technique used by NOAA to estimate the outgoing longwave flux from 10 μm window radiance observations has been reexamined because the data that result from the application of the empirically determined regression equation are systematically lower than those obtained from regression models based on earlier radiative transfer calculations. A new set of radiation calculations was made from a set of 1600 atmospheric soundings and the resulting regression equation gives flux differences from the empirical model that are of the order of ±5 W m−2 over the range of 150 to 300 W m−2, as compared to the +10 W m−2 systematic differences from previous studies. The differences are attributed to the size and representativeness of the sample of soundings used in the radiation calculations. The results also show that although the explained variance of the regression is of the order of 95%, this type of estimation technique may make errors of ±20 W m−2 or larger for a given radiance observation. This will lea...