Development of the NIMBUS-7 SMMR retrieval algorithm

An iterative algorithm for the extraction of geophysical parameters from the sea surface and intervening atmosphere from the Nimbus 7 microwave radiometer was developed. Surface and atmospheric truth data for the North Atlantic were used to assess the algorithm. Results show that satellite data from the most northerly positioned ship appears to be contaminated by sidelobe pick-up. The derived sea surface temperature exceeds ground truth data for all ships. Insertion of biases for the satellite data considerably reduces the residuals and improves the correlations. The required biases suggest that the calibrated channel at 37 GHz for horizontal polarization, in particular, yields outputs which are too high. Changes from the windspeed model of the original program reduce the correlation of derived parameters with the ground truth. Sea surface humidity predictions are particularly good.