A Preliminary Assessment of Global Precipitation Measurement Estimates over the Three Major Watersheds Situated in the Hindukush Mountains, Northern Pakistan

Abstract

Acquisition of accurate precipitation data is essential for water resources management, hydro-climatic research, and prediction of drought and flood scenarios. Sparse and uneven distribution of rain-gauges in the river basins situated in the Hindukush Mountainous range makes it difficult to study the simultaneity of precipitation patterns. Satellite-based precipitation products have remarkably contributed to obtaining continuous precipitation data on fine spatio-temporal resolutions. In this regard, the Tropical Rainfall Measuring Mission (TRMM) has launched several multi-satellite-based precipitation products. The Global Precipitation Mission (GPM) was launched in 2014 as a successor of TRMM. A comprehensive evaluation of these newly launched GPM-based precipitation data sets is essential for better understandings of their uncertainties in different time-periods, topographic conditions, and precipitation patterns and intensities. Therefore, this study was designed to evaluate performance of the Integrated Mulitisatellite Precipitation Retrievals for GPM (GPM IMERG) using in-situ gauging stations data from April 2014 to December 2016 in three major watersheds (Swat, Gilgit, and Chitral) situated in the Hindukush Mountains, northern Pakistan. In order to quantify the improvements in the latest satellite-based precipitation retrieval algorithm used in the IMERG product compared to its predecessor, the performance of TRMM Multisatellite Precipitation Analysis product (TMPA-3B42V7) was also assessed. Both satellite-based precipitation products were assessed on daily, monthly and annual scales for westerlies and monsoon precipitation seasons. The analysis shows that the IMERG product performed better than its predecessor (3B42V7), although both products overestimated (8.9 % by IMERG and 17.9% by 3B42V7) the precipitation compared to the ground-based observations. Results also reveal that the precipitation estimates from both satellite-based products were able to represent the spatial patterns of precipitation during both seasons. Precipitation estimates from IMERG product represent the mean westerlies, monsoon and annual precipitation and its spatial distribution more accurately than the 3B42V7. The IMERG product exhibited substantial improvements in low (0-1 mm/day) intensity precipitation estimates compared to the 3B42V7 product. Overall, IMERG showed clear improvements in the probability of detection (POD) and critical success index (CSI) over the selected watersheds. Present evaluation of the IMERG product and its comparison with its predecessor 3B42V7 is one of the earliest studies of its nature in northern Pakistan. These preliminary findings of the latest satellite-based precipitation data might be useful for water resources managers, meteorologists, and hydrologists.