Spatio-temporal Evaluation of Satellite Precipitation Products in Northwestern Iran

In the analysis of climatic and hydrological events, precipitation is considered as a main parameter and therefore, measuring precipitation data with high Spatio-temporal resolution is very important in predicting weather patterns. Accurate measurement of precipitation on the land surface is very challenging due to the scattering of rain gauge networks, temporal and spatial diversity, wind effects, and topography. In recent decades, the use and development of satellite products and remote sensing techniques have become widespread which is used in precipitation estimation. The aim of this study was to evaluate the satellite precipitation data of TRMM, CHIRPS, Persiann-CDR and GPM-IMERG and compare them with rain gauge data in the north and northwestern region of the country (including Gilan, Ardabil, East Azerbaijan, and West Azerbaijan provinces). For this purpose, after receiving the satellite data series and pre-processing them, an evaluation was performed between the satellite data on a daily, monthly and seasonal time scale with the observational data. Evaluation of the results is performed using definite indicators including POD, CSI, FAR, Bias and statistical criteria including correlation coefficient (Corr) and Normalized Root Mean Square Error (nRMSE). The study period was selected from January 1, 2017 to December 31, 2021 on 56 synoptic stations. The results of most indicators and statistical criteria (Corr, nRMSE, POD, and CSI) showed that in all products the lowest error is related to the southwest of the study area which increases (south of West Azerbaijan province) by moving toward the east of the region and the Caspian coast.  In assessing the regional average precipitation, the results of IMERG, CHIRPS and Persiann-CDR were close to each other and with a slight difference (except in the nRMSE criterion) the IMERG product is superior. Also, in the study of seasonal estimates, the results of CHIRPS and Persiann-CDR were more reliable, but in order to use IMERG and TRMM, it is suggested that the estimates be accurized using different error correction methods. Finally, according to the results of this study, each product based on the type of topography and climate of the region provides a different result in estimating rainfall and there is a need for further studies according to the type of events in each region and a more detailed study of each product.