On evaluation of ShARP passive rainfall retrievals over snow-covered land surfaces and coastal zones

TL;DR

This study evaluates the ShARP algorithm for satellite-based rainfall retrieval over snow-covered and coastal regions, showing it improves accuracy and reduces overestimation compared to standard methods, especially in complex terrains.

Contribution

The paper demonstrates the effectiveness of the ShARP algorithm in improving rainfall retrieval accuracy over challenging snow-covered and coastal terrains using TRMM data.

Findings

ShARP significantly reduces rainfall overestimation caused by snow contamination.

It markedly improves rainfall detection near coastlines.

Root mean squared difference decreases by up to 38% annually and 70% in cold months.

Abstract

For precipitation retrievals over land, using satellite measurements in microwave bands, it is important to properly discriminate the weak rainfall signals from strong and highly variable background surface emission. Traditionally, land rainfall retrieval methods often rely on a weak signal of rainfall scattering on high-frequency channels (85 GHz) and make use of empirical thresholding and regression-based techniques. Due to the increased ground surface signal interference, precipitation retrieval over radiometrically complex land surfaces, especially over snow-covered lands, deserts and coastal areas, is of particular challenge for this class of retrieval techniques. This paper evaluates the results by the recently proposed Shrunken locally linear embedding Algorithm for Retrieval of Precipitation (ShARP), over a radiometrically complex terrain and coastal areas using the data provided by the Tropical Rainfall Measuring Mission (TRMM) satellite. To this end, the ShARP retrieval experiments are performed over a region in Southeast Asia, partly covering the Tibetan Highlands, Himalayas, Ganges-Brahmaputra-Meghna river basins and its delta. We elucidate promising results by ShARP over snow covered land surfaces and at the vicinity of coastlines, in comparison with the land rainfall retrievals of the standard TRMM-2A12 product. Specifically, using the TRMM-2A25 radar product as a reference, we provide evidence that the ShARP algorithm can significantly reduce the rainfall over estimation due to the background snow contamination and markedly improve detection and retrieval of rainfall at the vicinity of coastlines. During the calendar year 2013, we demonstrate that over the study domain the root mean squared difference can be reduced up to 38% annually, while the reduction can reach up to 70% during the cold months.