PEMILIHAN PARAMETER OPTIMUM EXPONENTIAL SMOOTHING DENGAN GOLDEN SECTION UNTUK MERAMALKAN CURAH HUJAN DI KOTA BALIKPAPAN

Abstract

Rainfall is a crucial component of the climate system with significant impacts across various sectors, especially in disaster mitigation such as flooding. Balikpapan City, as a coastal region, is highly vulnerable to extreme rainfall patterns. This study aims to optimize the parameters in Exponential Smoothing methods—Single, Double, and Triple Exponential Smoothing (SES, DES, TES)—using the Golden Section approach to improve the accuracy of rainfall forecasting in Balikpapan. The study utilizes secondary data on monthly average rainfall from January 2019 to December 2023 obtained from the Central Bureau of Statistics (BPS) of Balikpapan. The results indicate that the TES method achieves the lowest Mean Absolute Percentage Error (MAPE) at 0.0456, outperforming DES and SES. The Golden Section method proves effective in systematically identifying the optimal parameters. Therefore, the optimized TES approach is the most suitable forecasting method for supporting Cdisaster mitigation and data-driven decision-making in the region.