On the use of the log-normal particle size distribution to characterize global rain

摘要

Radar and microwave radiometer measurements from low-earth orbit provide a means to characterize rainfall over much of the globe. In the case of the TRMM Precipitation Radar, monthly accumulations of data in 5{sup}0×5{sup}0 and 0.5{sup}0×0.5{sup}0 degree boxes serve as the basis for calculating histograms and first-order statistics of rain rate and reflectivity factor at several heights. Apart from calculations of diurnal variations of rainfall and important studies on the Madden-Julian oscillation and ENSO, spatial and temporal correlations of rainfall at smaller space-time scales have received less attention. Nevertheless, these second-order statistics are important for a number of studies including algorithm testing and development, partial-beam filling studies, satellite sampling studies, and fade statistics along satellite communication links. Perhaps most importantly, the measurements may provide information from which to construct a model of global rainfall. With advances in ground-based polarimetric and air- and satellite-based dual-wavelength radars, which offer the potential of providing characteristics of the particle size distribution, continued investigation of space-time models of both rainfall and drop size distribution (DSD) is warranted. Particular questions of interest are: how can such models be constructed, what parameters are required for their specification, and how well do various measurements and instruments provide such information. More generally, can such models, derived from radar data, provide a means by which to characterize rainfall and size distributions over a range of space-time scales.