FOREST WILDFIRE BURNED AREA SEVERITY PREDICTION USING METEOROLOGICAL AND TERRAIN FEATURES

Abstract

Forest wildfires are among the most destructive natural disasters affecting ecosystems,
biodiversity, and human communities worldwide. The increasing frequency and intensity of
wildfires in recent years have raised significant environmental and socio-economic concerns.
Factors such as rising global temperatures, prolonged drought conditions, deforestation, and
human activities have contributed to the growing risk of wildfire outbreaks. Predicting the
severity of burned areas caused by wildfires is an important step in improving disaster
management strategies and reducing environmental damage. The objective of this project is to
develop a machine learning-based system capable of predicting the burned area severity of forest
wildfires using meteorological and terrain features. The system analyzes historical wildfire data
and identifies patterns between environmental variables and wildfire behaviour. Key
meteorological parameters such as temperature, humidity, wind speed, and rainfall, along with
terrain characteristics such as elevation, slope, and vegetation density, are used as input features
for the predictive model.
Forest wildfires are a major environmental concern, causing significant ecological damage,
economic loss, and threats to human life. Predicting the severity of burned areas in advance is
essential for effective disaster management and mitigation. This project presents a machine
learning-based approach for predicting wildfire burned area severity using meteorological and
terrain features. Key input parameters include temperature, humidity, wind speed, rainfall,
elevation, slope, and vegetation characteristics, which strongly influence fire behavior.