Vegetation and Forest Fire Dynamics in Alberta: A Ground and Remote Sensing Data Analysis

• Main Author: Dastour, Hatef
• Other Authors: Hassan, Quazi K., Achari, Gopal, Ahmed, M. Razu
• Format: Doctoral or Postdoctoral Thesis
• Language: English
• Published: Graduate Studies 2024
• Subjects: Remote Sensing; Machine Learning; Time Series Analysis; Time Series Modeling; Vegetation Greenness; Fire Season; Numerical Simulation; Engineering > Environmental; Artificial Intelligence; Fort McMurray; Slave Lake
• Online Access: https://hdl.handle.net/1880/119087; https://doi.org/10.11575/PRISM/46683

This thesis presents a comprehensive analysis of the interplay between forest fires and vegetation dynamics in Alberta, Canada, under the lens of climate change. By synthesizing data from remote sensing, climate records, and fire databases, the study reveals the intricate relationships between vegetation cover changes and climatic factors throughout 2001–2022. It highlights the significant lead and lag times between the Normalized Difference Vegetation Index (NDVI) and climate variables such as Land Surface Temperature (LST), relative humidity, and precipitation, offering insights into the temporal dynamics of vegetation response to climatic influences. The research further explores the patterns and trends of forest fires, correlating them with interpolated climate data across various subregions. Using trend analysis and anomaly detection methods, the study identifies significant warming and drying trends, alongside variable precipitation changes, which have influenced both human-caused and lightning-induced forest fires. The findings underscore the differential impact of climate variables on fire occurrence and source, with notable patterns emerging in subregions like Athabasca Plain and Central Mixedwood. Building on these insights, the thesis develops a robust forest fire spread model, validated through high-precision simulations of the 2011 Slave Lake and 2016 Fort McMurray wildfires. The model leverages regional physical features, climatic data, and MODIS datasets to offer accurate fire behavior predictions. The phased simulation approach adapts to dynamic factors such as weather conditions and firefighting strategies, enhancing the model's applicability for effective fire management. Ultimately, this thesis aims to bridge the gap between theoretical understanding and practical application, providing valuable contributions to Alberta's forest fire management and community protection strategies. The research paves the way for more informed decision-making in the face of climate change by offering a nuanced ...