Rainfall Forecast Using Machine Learning with High Spatiotemporal Satellite Imagery Every 10 Minutes

Increasing the accuracy of rainfall forecasts is crucial as an effort to prevent hydrometeorological disasters. Weather changes that can occur suddenly and in a local scope make fast precise weather increasingly difficult inform. Additionally, results numerical model used by Indonesia Agency for Meteorology, Climatology, Geophysics are only able predict with temporal resolution 1–3 h cannot yet address need information high spatial resolution. Therefore, this study aims provide forecast spatiotemporal using Himawari-8 GPM IMERG (Global Precipitation Measurement: The Integrated Multi-satellite Retrievals) data. multivariate LSTM (long short-term memory) forecasting employed cloud brightness temperature selected bands input training For rain rate regression, we random forest technique identify non-rainfall pixels from data advance. showed low values mean error root square 0.71 1.54 mm/3 h, respectively, compared observation data, indicating proposed may help meteorological stations aviation purposes.