Rainfall variability based on the Climate Hazards Group InfraRed Precipitation with Station Data (CHIRPS) in Lesti watershed, Java Island, Indonesia

Abstract

Rainfall fluctuates spatiotemporally. Availability of rainfall data is essential in watershed management planning. The lack of accessibility causes watersheds in Indonesia to have insufficient rain gauges. The Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) provides free global rainfall data sets available on the web ranging from 1981 to the near-present. This paper aims to validate the CHIRPS model and analyze dry and wet event severity in Lesti watershed, the upper part of Brantas watershed. This study compared data during 1990-2011 obtained from the CHIRPS and three observed rainfall stations. The validation was examined by calculating the Nash and Sutcliffe Efficiency (NSE), while dry and wet event severity was analyzed using the Standardized Precipitation Index (SPI). The validation showed that the CHIRPS model has a good performance in representing the actual data. The R2 value was 0.656, which means the rainfall data obtained from the CHIRPS can be used for further analysis in a similar location. The rainfall temporal distribution analysis showed over 91% of the annual rainfall occurs during the rainy season. It means only less than 9% of the annual rainfall occurs during the dry season. These significant difference variations potentially lead to hydrometeorological disasters, such as floods and drought. Lesti watershed experienced the driest year in 1997 and vice versa in 2010. Repetition of both wet and dry events needs to be identified to reduce the risk of potential disasters.