Performance assessment of multi-source, satellite-based and reanalysis precipitation products over variable climate of Turkey

Precipitation with high spatio-temporal resolution is essential for hydro-meteorological studies on global and regional scales. In recent decades, the scarcity and infrequent distribution of gauge networks have been a great challenge for precipitation observation over different climate zones in Turkey. Conversely, precipitation products (PPs) from various sources can be an alternative to fulfill this shortcoming. In this study, the spatio-temporal consistency of five precipitation products, including MSWXv100, MERRA-2, CFSR, SM2RAIN-ASCAT, and PDIR-Now, is evaluated over the variable climate of Turkey, considering two distinct temporal windows as daily and monthly. The Kling-Gupta efficiency (KGE), including its three components (Pearson correlation coefficient, ratio of bias, and variability ratio), is exploited to assess the consistency of PPs over seven climate zones. Moreover, the True Skill Statistic (TSS) is utilized to evaluate the detectability strength of each PP for different precipitation intensities during five hydrologic years (2015–2019). The results indicate that five PPs differed in their behavior over the variable climate of Turkey and different time steps, while showing higher performance in monthly terms compared to daily. Overall, MSWXv100 is able to present a reliable precipitation estimate over time and space and display the highest performance (median KGE; 0.75) for the entire region, where MERRA-2 comes with the second-best performance (median KGE; 0.64) followed by CFSR (median KGE; 0.61), SM2RAIN-ASCAT (median KGE; 56), and PDIR-Now (median KGE; 53). All PPs show diverse performance over each climate zone, and their best performance (except CFSR) is obtained over Southeastern Anatolia (SAN) for the monthly time step, while CFSR showed its best performance (median KGE; 0.70) over the Mediterranean (MED) climate zone.