University of the Aegean
Μέθοδοι διαχείρισης της καύσιμης ύλης για τη μείωση της επικινδυνότητας πυρκαγιών σε περιοχές άμεσης προτεραιότητας στη χερσόνησο της Κασσάνδρας Χαλκιδικής
Abstract
dc:descriptionWildfires increasingly affect the natural environment, human lives, and property, due to their increased frequency and their intense destructiveness as an impact of the climate crisis, thus posing great social challenges. The seriousness of the situation is highlighted mainly in the residential and forest flammable vegetation mixing zones, known as Wildland–Urban Interface (WUI), which have been expanding rapidly in recent decades, resulting in a dramatic increase in the number of people and structures affected by wildfires. The accurate knowledge of the spatial distribution of fuel, the assessment of its characteristics and the selection of the appropriate method Fuel management in the appropriate locations is an essential prerequisite for integrated forest fire management and mitigation of their severity. To be effective, forest fire prevention projects must obey an integrated planning. The main objective of this thesis was the design and implementation of a methodology for the spatial optimization of forest fuel treatments in the peninsula of Kassandra, Chalkidiki, with emphasis on both the protection of human values and the sustainable management of forest ecosystems. This methodology aims to assist responsible forest managers in making and supporting their decisions regarding the qualification of specific forest areas for management. In addition, the creation of a framework for evaluating fuel management plans that have already been implemented in the field, regarding their effectiveness in reducing forest fire behavior, is of great interest to the Greek and international scientific community. In recent years, due to the increased incidence of large fires in Greece, there has been a change in the policy approach to dealing with forest fires by strengthening the prevention sector. When fuel reduction measures are accompanied by scientific proof of their effectiveness, they can convince of the need or benefits of investing in these practices. During this process, the main fuel’s management methods were presented and the cost of designing and implementing each method, the revenues that can be generated from them and the degree of effectiveness in reducing the risk of wildfires were evaluated. In this way, decisions about forest management can be made with a more complete understanding of their economic and environmental impacts. The purpose of the economic analysis was to understand whether the proposed actions are cost-effective and whether they have the expected results in reducing the risk of fire in forests. The collection, post-processing of data and their use in the process of modeling fire behavior to evaluate the effectiveness of a ten-year fuel management plan was the subject of study of the first part of the research. Three detailed landscape analysis files (lcp. file) were created for three forestry situations of the Stavronikita forest, Kassandra, Chalkidiki: a) without fuel treatments implementation, b) after the fuel treatment implementation and c) for the year 1969. Field data were collected by sampling within the study area concerning the inventory of basic characteristics of the forest today, while bibliographic data were also used for the forest situation in the past. Subsequently, thousands of spatial simulations of fire behavior were performed with FlamMap's Minimum Travel Time (MTT) algorithm to measure important fire parameters such as burn probability, fire size and conditional flame length for the three silvicultural states. Finally, in the context of the survey, the degree of risk of forest areas of the forest estate was spatially assessed to prioritize the processing of their fuels in the management plan of the next decade. The statistical analysis of the results answered the question whether the variations in factor values are due to the application of fuel treatments, or to a sampling error or to the luck factor. Subsequently, the research focused on its main objective, the development of an innovative method for the spatial optimization of forest fuel treatments, with the study area of the entire peninsula of Kassandra in Chalkidiki. The aim of the survey was to support the decisions of forest managers on where and how management should be implemented. As the available data on the distribution of fuel in the study area were not considered sufficient to achieve high–precision simulation results, it was necessary to retrieve and generate highly accurate topography, vegetation/fuel, and meteorological data. Extensive sampling was carried out in 154 evenly distributed locations within the forests throughout the Kassandra peninsula, where applying the principles and methods of forestry science, tree–metric characteristics (crown height, crown base height) were measured and estimated, while forest composition (dominant species, canopy cover) and natural conditions (e.g., damage from natural disasters) were recorded in detail. Several digital files were also retrieved and updated, such as the subsidized olive oil producing areas of the peninsula (OPEKEPE), all the existing residences inside and outside the Kassandra peninsula of (Hellenic Cadastre), archives of the Forest Map of Chalkidiki, historical records of the dominant wildfires, archives of the road network and forest clusters of Kassandra. Measurements correlated with digital orthophotographs of high spatial resolution (LSO25/2015). Finally, bibliographic data, such as existing forest management plans and fire protection plans, were utilized. Each of the above data yielded accurate descriptions for the land cover category and land use in the entire study area. The aim was to create a local dataset of detailed landscape description called KASSANDRA DATASET, to categorize the different types of vegetation on the Kassandra peninsula into categories that would be suitable for use in fuel models predicting and simulating fire behavior. A grid landscape file (lcp. file) was created, this time for the entire Kassandra peninsula, to provide the required spatial inputs to the FlamMap fire simulation model (elevation, aspect, slope, fuel models, canopy height, canopy base height, canopy cover and canopy bulk density). A total of 10,000 randomly distributed ignitions across the study area were simulated with the Minimum Travel Time (MTT) and Treatment Optimization Model (TOM) algorithms of FlamMap. The aim was to produce two new fire related agents: a) Fire Transmission to Houses (FTH), which calculates the number of buildings exposed to fires starting in a specific location and going to threaten at least one dwelling and b) Fuel Treatment Grid (FTG) to identify areas where fuel management operations slow down the growth rates of a fire. Eight criteria were created based on the experience of the managers, the relevant literature or because they were imposed by Greek and European legislation, covering climatic, technical, topographical, cost, ecological and social dimensions. Gravity factors were assigned for each criterion for two different scenarios a) by applying identical coefficients to all criteria (control scenario) and b) based on the Analytic Hierarchy Process (AHP) pair-by-pair comparison method (realistic scenario). A worst-case weather scenario resembling the prevailing weather conditions was used during a large wildfire in the study area on August 21, 2006. The results produced two final priority maps prioritizing Kassandra's forests for management, one for each scenario, while a quantitative and spatial assessment of the different areas to be managed was carried out. Further analysis assessed the variation in management priority between the main types of forest land cover (coniferous forests, dense Mediterranean scrub, low load scrub and broadleaved forests), as well as the different ranges and distribution of priority values for management between the two scenarios. At the same time, research was conducted on whether and to what extent, for the preparation of spatial optimization studies and evaluation of fuel management plans such as the one developed above, data derived from satellite products can be safely used or the creation of local land cover datasets is required. An attempt was made to compare the primary landscape description dataset, the KASSANDRA DATASET with three widely used European and global land cover products: a) MODIS12Q1 b) the CCI (Climate Change Initiative) of the European Space Agency for Climate Change, and c) the Corine Land Cover of the European program CORINE. The precision measurements used to conduct the comparisons were Overall Accuracy (OA), Producer's Accuracy (PA), User's Accuracy (UA) and Kappa Coefficient (K). The analysis of the results assessed the suitability of satellite products for certain applications, such as specific fire risk assessment studies, which require a high level of detail for specific types of land cover. The methodological framework developed in this PhD thesis has been operationally applied by the local Forest and Fire Service. Specifically, the areas indicated by the final maps as high priority were included in the fire protection plans of the forests of the Kassandra peninsula for immediate implementation through the actions ‘AntiNero I’ and ‘AntiNero II’ in the years 2022 and 2023 respectively. Finally, the simulation results, and specifically the spatial distribution of the FTH factor, were used by the local Fire Brigade for the targeted dispersion of their vehicles and their operational readiness.
Degree
thesis:*- Grantor dc:publisher
- University of the Aegean
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Bachantourian, Margarita
- Μπαχαντουριάν, Μαργαρίτα
Subjects
dc:subject × 10Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/56011