Περίληψη σε άλλη γλώσσα
Wildfires burn annually an area equivalent to half Australia on a global scale. However, the main social, economic and political disaster stems from fires at Wildland-Urban Interface (WUI) zones, where the natural environment and human activities intermingle. This can be attributed partly to the destruction of residences and other infrastructures that exist within the urban zone next to the forest zone and partly to the frequent loss of human lives.Forest fire research has advanced greatly over the past 60 years, prioritizing timely and prompt response to wildfires. Advances in current research have led to computer simulations of weather conditions and simulations of forest fuel behavior under different parameters that affect the initiation or propagation of a wildfire (fire management models). Nevertheless, the main operational models are based on literature data or assumptions, which ignore site-specific parameters in a particular geographic area. The connection between experimental ...
Wildfires burn annually an area equivalent to half Australia on a global scale. However, the main social, economic and political disaster stems from fires at Wildland-Urban Interface (WUI) zones, where the natural environment and human activities intermingle. This can be attributed partly to the destruction of residences and other infrastructures that exist within the urban zone next to the forest zone and partly to the frequent loss of human lives.Forest fire research has advanced greatly over the past 60 years, prioritizing timely and prompt response to wildfires. Advances in current research have led to computer simulations of weather conditions and simulations of forest fuel behavior under different parameters that affect the initiation or propagation of a wildfire (fire management models). Nevertheless, the main operational models are based on literature data or assumptions, which ignore site-specific parameters in a particular geographic area. The connection between experimental data and forest fuel flammability models is still unclear and relies solely on empirical data regarding vegetation types of North America or Australia. Furthermore, recent developments in information and communication technologies, particularly in the fields of Remote Sensing and Geographic Information Systems (GIS), impose a frequent revision and update of all available techniques. To note that, most studies to date focus on aerial photographs or low-resolution satellite images analysis and make use of outdated sets of spatial (vector) data. The current thesis combines raw experimental data of Mediterranean forest materials’ flammability, high resolution satellite imagery, mathematical equations and applied research in a WUI zone of Attica-Greece (Thrakomakedones) in order to propose an innovative semi-empirical method (model) for wildfire management. For this purpose, new methodologies that can assist WUI-area mapping accurately were developed, flammability indices of forest fuels were calculated and fire-hazard risk maps were designed. The specific wildfire management model is proposed both as methodology and as best practice aiming at wildfire prevention or wildfire suppression under any circumstances or in any other spatial context. Further, this model may contribute to the creation of an integrated local or nation-wide fire protection plan, which ensures effective wildfire management. Among the main outcomes of this investigation, the following should be noted.During the early experimental study of forest materials’ combustion and pyrolysis properties, all parameters and methodologies referred to in international literature regarding the quantification of flammability and its components using analytical- or lab-scale techniques (i.e., thermal analysis, limiting oxygen index, calorimetry) were taken into consideration and an extensive number of experiments was carried out. Simultaneously, new and optimized lab-scale techniques (flame spread rate, differential thermal analysis using cubic electrical furnace) were developed. The new lab-scale techniques provide comparative results from a common set of measurements at similar conditions and are advantageous over previous ones, as they do not need to be operated by specialized personnel and can be manufactured at low cost, while requiring inexpensive consumables.Foliar flammability data have been statistically evaluated to design 5 risk indices: Calorimetry Index (CI), Flammability Index FI1, FI2, FI3, and Flammability Index at Inclinations (FII). These indices suggest a common ranking of the Mediterranean forest fuels’ flammability, classifying P. halepensis (aleppo pine), forest litter and E. manipuliflora (heather) as most flammable and C. incanus (pink rockrose) as least flammable.These indices were incorporated into the final WUI area wildfire management model, using site-specific fuel models. For this purpose, two new methodologies were developed, using object-oriented supervised classification of high resolution remote sensing data (IKONOS satellite images on 1:5.000 scale). These models were validated using actual field data and they were proved to be much more accurate (88%) compared to currently available land cover data (e.g. CORINE data) or compared to other unsupervised classification methods (e.g. ISODATA).Additional parameters that influence fire-initiation and fire-propagation (both topographical and human) were evaluated and incorporated into the final model. The processing and combination of these parameters along with geospatial data and forest-fuel models in a GIS, makes a prediction of a fire in space and time possible through static and dynamic fire-hazard risk maps. Based on these maps, the WUI zone of Thrakomakedones can be categorized as a high-risk area. The proposed wildfire management model was validated by accurately identifying fire initiation points and burned area during its a-posteriori application on the dynamic weather conditions of July 7, 2004, according to historical wildfire data (actual fire incident in Thrakomakedones).In conclusion, an innovative research tool for quantitative assessment of wildfire risk was proposed through experimental research and a pilot was implemented for the Thrakomakedones WUI area in Attica. Moreover, new methodologies for modeling forest fuels and quantifying flammability we developed. The flammability indices can be used by all researchers who study Mediterranean ecosystems. The product of this research is readily available to process real-time data as a digital geographic database with multiple layers and accompanies the text of the doctorate thesis in digital format (optical disk). It can be used directly by all actors involved in wildfire of forest management, in order to assist:• wildfire prevention: fire-resistant design, vegetation-risk reduction, fire-risk prediction,• wildfire suppression: wildfire propagation estimation, fleet management, selection and application of fire-extinguishing chemical means, protection of human capital during suppression operations,• restoring burned areas.
περισσότερα
