The SEASAT altimeter wet tropospheric range correction revisited

Abstract

The algorithm used to calculate the wet tropospheric range correction from SEASAT scanning multichannel microwave radiometer (SMMR) brightness temperatures assumed a linear relationship between the brightness temperatures and the range correction. In a previous investigation, the coefficients in this algorithm were adjusted to fit a set of radiosonde observations to coincident SMMR measurements with a standard deviation of 2.79 cm. For the study described here, a significantly increased data base of radiosonde observations was used. From this set of observations, the previously developed algorithm was found to have a bias of about 1.0 cm and a standard deviation of about 2.8 cm. To further improve the wet tropospheric range correction algorithm, the complete set of radiosonde observations was used to obtain the best estimate of parameters for linear, quadratic, and logarithmic models for the brightness temperature to range correction relationships. Using various combinations of these new parameters, the standard deviation between SMMR and radiosonde observations was reduced to approximately 2.1 cm, which is equivalent to the accuracy of the radiosonde measurements. In addition, the performance of algorithms using only two of the three SEASAT radiometer frequencies was evaluated. Little performance degradation was observed for the two‐channel algorithms as long as one of the two channels was 21 GHz.