Simulation-based forecasting of temperature dynamics in Indonesia’s new capital city region using seasonal ARIMA

Abstract

Accurate temperature forecasting is essential for supporting climate responsive urban planning and infrastructure development in emerging smart cities. This study investigates the temperature dynamics of Sepaku District, East Kalimantan, Indonesia, the core area of the country’s new capital city (Nusantara), using a Seasonal Autoregressive Integrated Moving Average (SARIMA) model. Daily temperature data collected from January 1, 2022, to March 31, 2024, were analyzed to simulate temporal temperature variations and forecast future trends. The forecasting framework involved time series preprocessing, parameter optimization, and model validation using Root Mean Square Error (RMSE) and Mean Squared Error (MSE). The optimized SARIMA model demonstrated strong predictive capability, achieving an RMSE of 0.93 and an MSE of 1.23, indicating high agreement between observed and predicted temperatures. The results reveal the effectiveness of SARIMA in capturing seasonal temperature fluctuations and providing reliable short term forecasts for the study area. This simulation framework offers a practical tool for climate informed decision making, environmental risk assessment, and sustainable urban development planning in Indonesia’s new capital region.