Bushfire Severity Modelling and Future Trend Prediction Across Australia: Integrating Remote Sensing and Machine Learning

TL;DR

This study presents an in-depth analysis of bushfire severity in Australia over the last twelve years, combining remote sensing data and machine learning techniques to predict future fire trends and identifies areas at high risk of future severe bushfires.

Abstract

Bushfire is one of the major natural disasters that cause huge losses to livelihoods and the environment. Understanding and analyzing the severity of bushfires is crucial for effective management and mitigation strategies, helping to prevent the extensive damage and loss caused by these natural disasters. This study presents an in-depth analysis of bushfire severity in Australia over the last twelve years, combining remote sensing data and machine learning techniques to predict future fire trends. By utilizing Landsat imagery and integrating spectral indices like NDVI, NBR, and Burn Index, along with topographical and climatic factors, we developed a robust predictive model using XGBoost. The model achieved high accuracy, 86.13%, demonstrating its effectiveness in predicting fire severity across diverse Australian ecosystems. By analyzing historical trends and integrating factors such as population density and vegetation cover, we identify areas at high risk of future severe bushfires. Additionally, this research identifies key regions at risk, providing data-driven recommendations for targeted firefighting efforts. The findings contribute valuable insights into fire management strategies, enhancing resilience to future fire events in Australia. Also, we propose future work on developing a UAV-based swarm coordination model to enhance fire prediction in real-time and firefighting capabilities in the most vulnerable regions.

Figures (13)

• Figure 1: Map of Australia's Land Cover Types in 2018 on QGIS: This map visualizes using QGIS and visualizes various land cover types across Australia, categorized into Forests, Shrublands, Savannas, Grasslands, Wetlands, Croplands, Urban areas, and Barren lands. The data is derived from the MODIS Land Cover Type Yearly Global 500m dataset. Forest areas are marked in dark green, Shrublands in brown, Savannas in yellow, Grasslands in light green, Wetlands in cyan, Croplands in orange, Urban areas in red, and Barren lands in grey. The map provides a clear overview of the dominant land cover types throughout the country.
• Figure 2: Data extraction workflow for fire severity analysis using multi-source satellite imagery. The process involves combining Landsat data and additional sources such as SRTM, ECMWF/ERA5-Land, and NASA/SMAP to extract topographical, climatic, and spectral indices. Missing bands are handled with zero values, and the calculated indices are used to produce a GeoTIFF output dataset.
• Figure 3: NDVI Map of Australia (2012-2023) on GEE: This map depicts the vegetation health and density across Australia over a period of 12 years. Higher NDVI values, shown in green, indicate areas with dense and healthy vegetation, while lower values, depicted in blue, represent sparse or degraded vegetation cover.
• Figure 4: NBR Map of Australia (2012-2023) on GEE: This map visualizes the burn severity across Australia using the NBR index, where warmer colors (yellow to red) indicate higher burn severity, and cooler colors (white) represent areas with lower or no burn impact over the last decade
• Figure 5: The distribution of major urban centres across Australia (The dataset was acquired from worldpop_australia_2020 and visualized using QGIS), with particular attention paid to Sydney, Melbourne, Brisbane, and Perth. These areas are critical in firefighting resource allocation due to their high population densities and proximity to bushfire-prone regions.