An intercomparison of hurricane forecasts using SSM/I and TRMM rain rate algorithm(s)

Summary

Three different Special Sensor Microwave Imager (SSM/I) rain rate algorithms are evaluated as a means of improving both the physical initialization and the hurricane forecast output of the Florida State University Global Spectral Model (FSU GSM). These SSM/I rain rate algorithms are known as Cal/Val, NESDIS, and GPROF 4.0. In addition the Tropical Rainfall Measuring Mission (TRMM) TMI – 2A12 rain rate algorithm is validated, and its impact on FSU GSM hurricane forecasts is also studied.

Validation results of the Cal/Val rain rate algorithm show a bias toward gross underestimation. Both the NESDIS and GPROF 4.0 algorithms produce robust rain rates, in agreement with surface based observations. However, the NESDIS SSM/I rain rate algorithm proves to be the most consistent and accurate in this study. Surface rain rates as estimated by the TRMM/TMI – 2A12 algorithm can be inconsistent, mainly due to satellite observational coverage gaps.

The impact of different magnitudes of rain on the FSU GSM is significant. In theory, the application of more accurate and consistent rain rates should produce an improvement in model-calculated latent heat release and cumulus parameterization. The net effect is a more representative, modeled global circulation and improved hurricane track prediction. This research has shown that the use of NESDIS SSM/I rain rates in the physical initialization of the FSU GSM provides the most accurate hurricane track forecasts.