Potential of high-resolution detection and retrieval of precipitation fields from X-band spaceborne synthetic aperture radar over land

摘要

X-band Synthetic Aperture Radars (X-SARs), able to image the Earth's surfaceat metric resolution, may provide a unique opportunity to measure rainfallover land with spatial resolution of about few hundred meters, due to theatmospheric moving-target degradation effects. This capability has becomevery appealing due to the recent launch of several X-SAR satellites, eventhough several remote sensing issues are still open. This work is devotedto: (i) explore the potential of X-band high-resolution detection andretrieval of rainfall fields from space using X-SAR signal backscatteringamplitude and interferometric phase; (ii) evaluate the effects of spatialresolution degradation by precipitation and inhomogeneous beam filling whencomparing to other satellite-based sensors. Our X-SAR analysis ofprecipitation effects has been carried out using both a TerraSAR-X (TSX)case study of Hurricane "Gustav" in 2008 over Mississippi (USA) and aCOSMO-SkyMed (CSK) X-SAR case study of orographic rainfall over CentralItaly in 2009. For the TSX case study the near-surface rain rate has beenretrieved from the normalized radar cross section by means of a modifiedregression empirical algorithm (MREA). A relatively simple method to accountfor the geometric effect of X-SAR observation on estimated rainfall rate andfirst-order volumetric effects has been developed and applied. TheTSX-retrieved rain fields have been compared to those estimated from theNext Generation Weather Radar (NEXRAD) in Mobile (AL, USA). The rainfalldetection capability of X-SAR has been tested on the CSK case study usingthe repeat-pass coherence response and qualitatively comparing its signaturewith ground-based Mt. Midia C-band radar in central Italy. A numericalsimulator to represent the effect of the spatial resolution and the antennapattern of TRMM satellite Precipitation Radar (PR) and Microwave Imager(TMI), using high-resolution TSX-retrieved rain images, has been also set upin order to evaluate the rainfall beam filling phenomenon. As expected, thespatial average can modify the statistics of the high-resolutionprecipitation fields, strongly reducing its dynamics in a way non-linearlydependent on the rain rate local average value.