Insight into the precipitation behavior of gridded precipitation data in the Sina basin.

Abstract

This paper examines the performance of three gridded precipitation data sets, namely, Global Precipitation Climatology Centre (GPCC), Tropical Precipitation Measuring Mission (TRMM), and the Modern-Era Retrospective Analysis for Research and Applications (MERRA), for the duration of 25years using 9 rain gauge data sets of the Sina basin, India. Statistical measures were employed to measure the performance in reproducing the rainfall and to assess its ability to detect the rainfall/no rainfall events, its structure, pattern, and spatio-temporal variations in the monthly and annual time scales. Compromise programming (CP) is used to rank the statistical performances of selected gridded precipitation data sets and found that TRMM attained first rank for the 8 stations followed by MERRA. The precipitation concentration index (PCI) checks the pattern and distribution of rainfall and found that observed data shows a uniform distribution in the basin; however, all the three gridded data sets failed to demonstrate uniform distribution. Categorical metrics like Probability of Detection (POD) and False Alarm Ratio (FAR) revealed that TRMM followed by MERRA and GPCC have good capabilities to detect rainfall/no rainfall events at different thresholds. All the trends drawn between observed data set and gridded precipitation data sets revealed that the MERRA data tend to underestimate and the TRMM and GPCC data tend to overestimate the values and intensities of rainfall data sets at most of the stations for both monthly and annual time scales. The data analysis of extreme rainfall points at monthly and annual time scales exhibits better performance of TRMM data sets. Overall, the TRMM data set is capable in replicating different characteristics of the observed data in the study area and could be used for hydro-meteorological and climatic studies when continuous observed data set is not available.