Search

Knowledge of three-dimensional distribution of moisture is crucial for successful numerical weather prediction in the tropics. Sparsity of moisture data over the Indian Ocean area is one of the main reasons for the poor representation of tropical diabatic forcing during the summer monsoon season. Due to the poor analysis of moisture and divergence fields over data-sparse oceanic regions in the tropics, the rainfall amounts predicted by numerical models often do not compare well with observed estimates. Prediction of flow pattern is also very sensitive to the initial moisture fields. Cloudiness cleared satellite radiances, if included in the framework of four-dimensional variational assimilation, provide the best possible source of moisture information in clear areas or above cloud top over data-sparse oceanic regions. But this may require very high computing resources, which are not available at present. An alternative is to generate synthetic moisture data from the satellite radiances using empirical statistical methods. In the present study, four different schemes of deriving moisture profiles from Indian Satellite (INSAT) and National Oceanic Atmospheric Administration (NOAA) satellite data were tested for the Indian Ocean area (30° N to 40° S, 50° E to 130° E), for two different periods of August 1993 and June 1994 selected from the summer monsoon season. The bias and root mean square error (RMSE) values of the derived profiles are discussed in relation to available radiosonde data (island and coastal stations). The impact of this synthetic moisture on medium-range analysis-forecast system has been found to be encouraging. [ABSTRACT FROM AUTHOR]