PERAMALAN SUHU MINIMUM BULANAN KABUPATEN MALANG MENGGUNAKAN METODE HOLT–WINTERS UNTUK ANALISIS POLA MUSIM DINGIN TROPIS

Abstract

Monthly minimum temperature is an important indicator in climate analysis, particularly for identifying tropical winter patterns in Malang Regency. This study aims to forecast monthly minimum temperature using the additive Holt–Winters method and to analyze the resulting tropical winter pattern. The data used consist of monthly minimum temperature from 2019 to 2025. The model parameters were optimized, resulting in = 0.23314, = 0.0, and = 0.59085. Forecast accuracy was evaluated using Mean Absolute Deviation (MAD), Mean Squared Error (MSE), and Mean Absolute Percentage Error (MAPE). The results show that the model achieves a very high level of accuracy with a MAPE value of 8.00626%. The 2026 forecast indicates that the lowest minimum temperatures occur in July and August, reflecting a tropical winter pattern in mid-year. Therefore, the additive Holt–Winters method is capable of representing the annual seasonal pattern and is suitable for forecasting monthly minimum temperature in Malang Regency.