Comparison and Evaluation of Remotely Sensed Gridded and Observed Precipitation Data

Abstract

Spatial variability of precipitation directly affects socio-economic and environmental conditions, therefore accurate and valid measurements of precipitation are needed. Due to the free availability and easy access to gridded precipitation data from multiple sources, questions arise about its credibility like which is a better-gridded dataset and for which precipitation patterns. This study focused on these questions by carrying out the comparison and evaluation of three different gridded datasets and the observed data. These gridded and observed datasets include “The Integrated Multi-Satellite Retrievals for GPM (IMERG), The Tropical Rainfall Measuring Mission (TRMM) products 3B42-V7 & 3B42-RT and Pakistan Meteorological Department (PMD)’’. These datasets were analyzed for three different seasons (monsoon, pre-monsoon, winter), and annually for multiple weather stations. The weather stations were classified into ‘‘Monsoon, Western Distribution (WD), and Hybrid’’ according to the precipitation patterns. Seven comprehensive statistical parameters (mean, Standard deviation, 5-days max rainfall, 95% percentile, percentage wet days, max dry spell length, and max wet spell length,) were calculated for each season and for each precipitation distribution pattern. Pearson correlation of daily data of 20-years at each station was also calculated as an additional parameter. Furthermore, the IDW interpolation technique was used for further analysis of the results of the above-mentioned parameters. Results suggested that in monsoon and WD stations GPM IMERG has the highest performance among the three (r = 0.882 & r =0.543). While at Hybrid stations 3B42_V7 performance is better (r = 0.840). TRMM 3B42_RT was the least accurate in precipitations estimation in monsoon, WD and as well as hybrid stations (r = 0.523, r =0.350, r = 0.708). The results suggest that these datasets can be used as an alternative source for various hydro-meteorological and hydro-climatological applications after accuracy assessment.