Assessing the Accuracy of NASA POWER Data in Estimating Weather Variables (Precipitation and Temperature) for Khassa Chai River Watershed

This study intends to evaluate the accuracy and reliability of satellite data from the NASA POWER project in capturing important climatic variables, which are precipitation and air temperature, compared with data collected from the National Climate Data Center (NCDC) ground-based weather stations located in the Khassa Chai River basin. The basin has a topographic gradient beginning on the northeastern highlands and sloping towards the southwest. The Khassa Chai River Basin experiences a semi-arid climate, with hot, dry summers and relatively cold and wet winters. Daily precipitation and temperature records (2010–2024) were collected from four weather stations, in addition to records from the NASA POWER recording at the nearest grid points. To evaluate the reliability of the datasets from NASA POWER, this study used several relevant statistical indicators (i.e., the coefficient of determination (R²), correlation coefficient (CC), Nash-Sutcliffe efficiency (NSE), mean bias error (MBE), and root mean square error (RMSE)) at daily, monthly, and yearly time scales. The analysis of precipitation indicated an excellent fit between the NASA POWER satellite data and in-situ data. For monthly comparison, the R² was 0.89, and the CC was 0.94, while for annual comparison, the R² was 0.81, and the CC was 0.88. Overall, the fit improved with longer time scales, which indicates the ability of the satellite data to accurately capture precipitation trends over time. The range of NSE values from 0.72 to 0.87 also reinforces the ability of the data to reproduce precipitation changes over time.