Natural phenomena are basically derived from the pattern of rain that changes due to climate change. Usually in Indonesia there are floods in the early rainy season, while in Indonesia there is a drought when the rainy season is over. Rainfall data can be obtained by utilizing satellite data and field rainfall data / ARR (Automatic Rainfall Recorder) on rivers. This study applies validation analysis of field rainfall data with satellite rain data with bivariate and R square correlation methods with field rainfall data and MERRA-2 Model and TRMM satellite rainfall data on 9 rain stations in the Yogyakarta and Central Java watersheds with a period of 2015- 2018. The results showed that the analysis used bivariate correlation for ARR-MERRA-2 data. The model from uncorrelated was 0 to strongly correlated with 0.844, analysis using daily data obtained an average yield of 0.1921 while the ARR-TRMM data from uncorrelated ie 0 to correlated strong is 0.845, analysis using daily data obtained an average yield of 0.4692. Based on using the coefficient of determination analysis for ARR data MERRA-2 Model has a relationship value of 0 - 0.7118, analysis using daily data obtained an average yield of 0.0811 while the ARR-TRMM data has a relationship value 0 - 0.7133, analysis using daily data obtained an average result average 0.2374. Satellite data can be used as a determination of rainfall data but it is better if only as complementary data for ARR rain station data, both for MERRA-2 Model and TRMM satellite data. If satellite data will be used as data for determining rainfall data, it is better to use it in the form of daily data rather than time and time data.