Forecasting of Rainfall in Sumatera Barat: Singular Spectrum Analysis (SSA) Application

Abstract

Abstract West Sumatra has 2 peak of rainy seasons, October-November and March to May. The high intensity of rainfall causes West Sumatra has potential to floods. Precise forecasting is needed to be a reference for the government. One of the method can be used is SSA. This method is flexible because it does not require specific form of time series data, as well as parametric assumptions. Thus, accurate forecasting results is expected to be provided from the SSA method. From the forecasting process that has been done, obtained MAPE of 17% with the tracking signal value within the tolerance range.