PREDICTION OF POTENTIAL SEASONAL LAND AND FOREST FIRES (YEAR 2011 – 2022)

• Main Author: Hariz, Fauzan
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/87987
• Institution: Institut Teknologi Bandung

Forest and land fires are frequent disasters in Indonesia, particularly during the dry season, causing severe environmental, health, and economic impacts. Meteorological factors such as temperature, humidity, rainfall, and wind significantly influence the potential for fire occurrence. Seasonal-scale fire prediction can enhance preparedness and support decision-making in disaster mitigation. Therefore, this study aims to evaluate the use of the Seasonal Severity Rating (SSR) in characterizing fire potential and to develop a probabilistic SSR prediction method for Kalimantan Island. This study utilizes ECMWF Reanalysis Version 5 (ERA5) as observational data and the Climate Forecast System Version 2 (CFSv2) as predictive data. The applied method includes the calculation of the Fire Weather Index (FWI), which is further developed into the Daily Severity Rating (DSR), Monthly Severity Rating (MSR), and Seasonal Severity Rating (SSR). Additionally, a Time Lagged Ensemble approach is used to improve prediction accuracy. The model's performance is evaluated by comparing SSR predictions with actual fire occurrences and assessing prediction reliability using the Brier Score (BS). The results indicate that SSR can be used as an indicator of seasonal-scale forest fire potential. The determination of SSR thresholds provides an overview of fire severity levels in Kalimantan, establishing risk categories that can be incorporated into early warning systems. SSR predictions using the CFSv2 model show promising potential in forecasting seasonal fire conditions, although further development is needed to enhance accuracy. Therefore, this study contributes to forest fire mitigation efforts by providing a predictive method that can serve as a foundation for forest and land fire management planning in Indonesia.