Fire risk assessment using geospatial information systems and remote sensing tools

Resumo

We live in a world with numerous risks, both natural and anthropogenic. Among them, forest fires are of increasing concern to society and are on the rise in certain regions. Year after year these events affect thousands of hectares and endanger numerous rural populations, as well as having an impact on the environment. The countries of the Mediterranean region and southern Europe suffer severely from their impacts and consequences. Historically, fire has been used as a vegetation management tool. Studying the causes of its development, particularly its uncontrolled development, is not an easy task, since it there is no single cause but rather various different factors which interact with one another. The main factors involved in forest fires include climate, demographics and land management. Climate change, rural abandonment and depopulation are making it increasingly difficult to manage the areas with the highest fire risk. In recent years, Mediterranean countries have experienced episodes with serious consequences for the population. The forest fires in Portugal and the Spanish region of Galicia in 2017, Greece in 2018, Spain in 2022 and southern Italy in 2023 are clear examples of how these events can seriously affect the civilian population, with 181 deaths and a large number of people injured and evacuated. Particularly large one-off forest fire events known as "mega-fires" are becoming more voracious every year, affecting larger areas in less time. This results in the homogenisation of the landscape in subsequent years if it is not managed, thus providing a future landscape conducive to more fires. This research has two main goals. The first is to address the problem of fires in Galicia, and more specifically in the Galician territory of the Gerês-Xures Transboundary Biosphere Reserve. To that end, potential factors and trends in the territory where fires occur are studied. In both cases, the work is based on fire cartography and statistical and mathematical methods aimed at finding the main parameters of forest fires. The second part of the thesis focuses on fuel modelling in forest ecosystems. This is of the utmost importance as it is extremely helpful to know the state of vegetation susceptible to fire. The tools used range from field sampling (as the basis of the studies) to the use of remote sensing technologies, including satellite images and in-house sensors, that scan the study areas using LiDAR technology. This research has been structured in various international scientific publications in prestigious journals to advance scientific knowledge and improve existing tools for managing a problem that requires solutions beyond mere fire extinguishing. This also ensures greater dissemination of the results in terms of the progress made.