Monthly Rainfall Prediction Using Multiple Linear Regression Method in West Nusa Tenggara Region

Minardi, Suhayat
Baihaqi, Anas
Wazni, Haerul

Frontier Advances in Applied Science and Engineering
Vol. 3 No. 1 (2025)

Rainfall is a crucial meteorological element in tropical regions like Indonesia. The significant influence of rainfall on various sectors of life means that rainfall predictions are necessary for making plans. This research aims to determine the accuracy of rainfall predictions using the Multiple Linear Regression method and what local factors influence rainfall in the West Nusa Tenggara Region. Multiple Linear Regression is a method that can predict monthly rainfall using more than one independent variable. There are inconsistencies in the regression analysis process, and to overcome this in this study, DMC (Double Mass Curve) was used. The data used is BMKG data from the West Nusa Tenggara (NTB) Climatology Station for 2013 - 2022. In general, the level of prediction accuracy ranges between 54.10% - 87.50%. The best correlation coefficient value for the Lombok Island Region is r = 0.79. The Sumbawa Island region is r = 0.88, and the Bima region is r = 0.83. Based on the multiple linear regression equation model obtained, the most dominant local factors influencing rainfall in the NTB region are air and sea surface temperatures.