Forecasting for monthly rainfall in Gondang Reservoir Lamongan using Singular Spectrum Analysis (SSA) method.

Bengawan Solo is a watershed in Indonesia with an area of ±16,100 km2. The Gondang Reservoir Lamongan Post is one of the posts passed by the Bengawan Solo River Basin. Gondang Reservoir has benefits such as supplying field reservoirs and irrigation, providing water, controlling the flood or drought, and tourism objects. However, the flow of the Gondang Reservoir river has a negative side that has the potential to damage, so that it can cause disasters in the form of floods, drought, or broken embankments. Therefore it is necessary to have a control system to reduce the risk of disasters in the future. One of the important components in the watershed is rainfall data. Rainfall is very useful in determining the beginning of the season and evaluating a model. Rainfall forecasting is very important for workers in various sectors to determine the direction of a policy to be drawn up and be able to control the flows of the river in conditions that are not destructive. Forecasting techniques is one of the statistical methods that can be used to predict rainfall data because it provides significant results in predicting time series data in the long term. In this research, the Singular Spectrum Analysis (SSA) method was applied for rainfall forecasting in the Gondang Reservoir. The SSA method consists of two stages, decomposition, and reconstruction. Each stage consists of several parts. The decomposition stage consists of embedding and singular value decomposition. Meanwhile, the reconstruction stage consists of grouping and diagonal averaging. Forecasting was conducted on Gondang Reservoir rainfall data from January 2010 to December 2020. From this analysis, it can be concluded that the SSA method could predict well enough rainfall data in the Gondang Reservoir Lamongan post using L = 32 and 5 groups with an MSE value of 43.87 and MAPE value of 21.28%. [ABSTRACT FROM AUTHOR]