Performance of satellite-based precipitation products using dense network of rain-gauge observations in Tamil Nadu

The traditional observations of precipitation data is unavailable at many locations in Tamil Nadu. Hence, the performance of four gridded Precipitation datasets; Integrated Multi-satellite Retrievals for GPM (IMERG), TRMM 3B43v7 (Tropical Rainfall Measuring Mission-3B43 version 7), Climate Hazards Group Infra Red Precipitation with Station data (CHIRPS), and PERSIANN (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks) were validated using descriptive statistics against the high density of rain-gauge observations (385 rain-gauge stations) over Tamil Nadu. Monthly precipitation was compared with a point-to-pixel analysis from October to December for the periods of 2015-2017 that were calculated in each month. From the overall analysis of monthly temporal scale, all four (GPM-IMERG, TRMM 3B43, CHIRPS, and PERSIANN) precipitation products exhibited better performance in comparison to TAWN rain gauge stations. An evaluation based on continuous statistics indicated that the highest correlation coefficient ranges more than 0.7 at each month except in the month of October 2015 (CC=0.50). The lowest RMSE has shown in PERSIANN products (3.31 mm/month) during December month of 2017. The overall monthly comparisons between the four high resolution precipitation products (HRPPs) and the rain gauge observations have shown a better agreement.