Regional precipitation regimes and evaluation of national precipitation datasets against satellite-based precipitation estimates, Republic of Georgia

Abstract

Analyzing water resources in areas with few hydrometeorological stations, such as those in post-Soviet countries, is difficult due to station closures after 1989. In Caucasus, evaluations often rely on outdated data from nearby rivers. We evaluated one national-level precipitation dataset, the Water Balance of Georgia (WBG) with two satellite-based precipitation products from 1981 to 2021, including the Climate Hazards Group Infrared Precipitation with station data (CHIRPS), and CHIRPS blended with a dense rain gauge network (geoCHIRPS). We modelled mean annual precipitation from geoCHIRPS as a function of coastal distance and elevation. CHIRPS underestimated precipitation in the cold and wet seasons (R2 = 0.74, r = 0.86), and overestimated dry season precipitation, while geoCHIRPS performed well in all seasons (R2 = 0.86, r = 0.92). Distance from the coast was a more important predictor of precipitation than elevation in Western Georgia, while precipitation correlated positively with elevation in the East. At four hydroelectric plants, the underperformance as a percentage of capacity (∼37%) corresponds with the percentage difference between difference in precipitation products (∼38%), suggesting that plants designed based on WBG may be systematically over-designed, but further work is needed to determine the reasons for the underperformance of the plants and frequency. We conclude that 1) existing WBG does not accurately reflect elevation-precipitation relationships near the coast and 2) for accurate analysis of spatiotemporal precipitation variability and its impacts on hydropower generation, environmental and sustainable water resource management, it is essential to calibrate satellite-based precipitation estimates with additional rain gauge data.