Surface radar studies of rain estimation in correlation with infrared and visible techniques of INSAT and GOES-E geostationary satellites

This study explores further the application of the radar reflectivity and the derivation of the empirical relationship linking the rainfall parameters to the radar observables reflectivity (Z), differential reflectivity {dollar}(Zsb{lcub}DR{rcub}){dollar} and attenuation (K), using exponential raindrop size distribution N(D).; The IR data in the range of 10.5-12.5 {dollar}mu m{dollar} from INSAT-IB and 10.5-12.6 {dollar}mu m{dollar} from GOES-E geosynchronous satellites have been used to estimate the rainfall rate using the technique by Stout, Martin and Sikdar (1979) in enhanced IR images. In case of INSAT, four months were processed to show the different variations in rainfall. The months chosen were July 85, October 85, January 86 and April 86. The rainfall estimations from GOES-E over the Florida peninsula were compared with raingage measurements and the accuracy of the rain estimations were discussed. We have observed that in general the satellite estimates are suggesting volumetric rainfall which is slightly less than that inferred from ground truth by actual rainfall observations as determined from raingages. A possible explanation is that the convective cycle for clouds over tropical lands may not be thoroughly inferred from satellite images due to 30-minutes interval between two images and because of the diurnal heating which tends to produce quite explosive convective development.; The results suggested that the satellite convective daily rainfall estimation technique by Stout and Martin can provide rain estimates of considerably high utility on the convective scale when combined with the Scofield/Oliver (1977, 1978, and 1980) technique, especially for large and deep convective systems over tropical oceans as well as lands.