Proyeksi Produksi Beras Provinsi Lampung Menggunakan Model Seasonal Autoregressive Integrated Moving Average (SARIMA)

Abstract

Food security is a strategic issue in Indonesia, including in Lampung Province, which is one of the country’s main rice production centers. Rice self-sufficiency is stated in the 2025–2029 Rencana Pembangunan Jangka Panjang (RPJM). To achieve food sovereignty, the Lampung Provincial Government faces several challenges, including population growth, climate change, land-use conversion, and fluctuations in rice prices. Data from the Badan Pusat Statistik (BPS) show that rice consumption has been increasing annually, yet the existing data have not been fully utilized for food policy formulation. Therefore, an accurate forecasting system is needed to support policy-making. This study employs the Seasonal Autoregressive Integrated Moving Average (SARIMA) model to forecast rice production in Lampung Province, utilizing seasonal sub-round (four-month) data from BPS Lampung. The analysis results indicate that the best model is SARIMA (2,1,0) (0,0,0) [3], with the lowest Akaike Information Criterion (AIC) value. The forecast shows a relatively stable trend compared to previous years, with declines in the early subrounds of 2025 and 2026, followed by increases in the subsequent subrounds. These results can serve as recommendations for local governments in food security planning. However, the SARIMA model has limitations in capturing patterns influenced by external factors such as weather conditions during planting seasons, shifts in planting and harvesting periods, fertilizer shortages, pest and disease outbreaks, and government policies that may affect productivity.