Remote Sensing of Terrestrial Rainfall from Ku-Band Scatterometers

Abstract

Rainfall is the most fundamental variable of the terrestrial hydrological cycle. However, in many regions of the world, ground observations are still very scarce or even missing. Recently, a bottom-up approach, named SM2RAIN, for terrestrial rainfall estimation from satellite soil moisture (SM) products was proposed and successfully applied to C-and L-band products from scatterometers and radiometers. Thanks to the multiple Ku-band scatterometers launched in the recent years and a number of new sensors expected in the near future, accurate rainfall estimation at subdaily time scale could be obtained. We present here a first attempt to estimate terrestrial rainfall from Ku-band scatterometers using SM2RAIN. To this end, backscattering data (sigma-0) collected in central Italy from the RapidScat instrument on board the International Space Station are compared with the Advanced SCATterometer (ASCAT, C-band) SM product and in situ observations for assessing its sensitivity to SM variations. Then, RapidScat sigma-0 is used for rainfall retrieval and compared with ground observations over a regular grid of 15-km spacing. The 8-month period from Nov 2014 to Jun 2015 is considered. Results show a very good agreement between ASCAT SM and RapidScat SM index with a median temporal correlation coefficient R of ∼0.9 and a reasonable performance (R > 0.52) against in situ data. More interestingly, the performance of RapidScat in 1-day rainfall estimation is found to be satisfactory with median R-values equal to ∼0.6. These promising results highlight the large potential of using the constellation of scatterometers for providing an accurate rainfall product with high spatial-Temporal resolution.