Rainfall Prediction in Central Maluku Regency using The Autoregressive Moving Average (ARMA) Model Approach

Indri Kezia Latupeirissa (Universitas Pattimura, Indonesia)
Fadila Rumbaroa (Universitas Pattimura, Indonesia)
Henry Junus Wattimanela (Universitas Pattimura)
Wa Ode Ma’arifah (Universitas Pattimura, Indonesia)
Jean Apituley (Universitas Pattimura, Indonesia)
Sahrul Rumalean (Universitas Pattimura, Indonesia)

Parameter: Jurnal Matematika, Statistika dan Terapannya
Vol 5 No 1 (2026): Parameter: Jurnal Matematika, Statistika dan Terapannya

Rainfall is one of the main indicators in climate analysis of a region, as its distribution and intensity can indicate seasonal patterns, climate variability, and long-term climate change. Central Maluku Regency exhibits unpredictable rainfall patterns due to its geographical factors. This study aims to forecast rainfall in Central Maluku Regency using the Autoregressive Moving Average (ARMA) model approach. The data used were obtained from the Central Statistics Agency (BPS), consisting of rainfall data from January 2019 to October 2024. The analysis process includes stationarity tests on variance and mean using the ADF test, model identification, parameter estimation and significance testing, as well as residual diagnostic testing. The results indicate that the ARMA(1,0) model is the best-fitting model, with an accuracy measured by a MAPE of 43,24%, which is considered reasonably accurate. These findings provide important information to support climate data-based decision-making in Central Maluku Regency.