A Multidisciplinary Decision Support System for Forest Fire Crisis Management

A wildland fire is a serious threat for forest ecosystems in Southern Europe affecting severely and irreversibly regions of significant ecological value as well as human communities. To support decision makers during large-scale forest fire incidents, a multidisciplinary system has been developed that provides rational and quantitative information based on the site-specific circumstances and the possible consequences. The systems architecture consists of several distinct supplementary modules of near real-time satellite monitoring and fire forecast using an integrated framework of satellite Remote Sensing, GIS, and RDBMS technologies equipped with interactive communication capabilities. The system may handle multiple fire ignitions and support decisions regarding dispatching of utilities, equipment, and personnel that would appropriately attack the fire front. The operational system was developed for the region of Penteli Mountain in Attika, Greece, one of the mountain areas in the country most hit by fires. Starting from a real fire incident in August 2000, a scenario is presented to illustrate the effectiveness of the proposed approach.     

Decision Support System for Managing Oil Spill Events

The Mediterranean environment is exposed to various hazards, including oil spills, forest fires, and floods, making the development of a decision support system (DSS) for emergency management an objective of utmost importance. The present work presents a complete DSS for managing marine pollution events caused by oil spills. The system provides all the necessary tools for early detection of oil-spills from satellite images, monitoring of their evolution, estimation of the accident consequences and provision of support to responsible Public Authorities during clean-up operations. The heart of the system is an image processing–geographic information system and other assistant individual software tools that perform oil spill evolution simulation and all other necessary numerical calculations as well as cartographic and reporting tasks related to a specific management of the oil spill event. The cartographic information is derived from the extant general maps representing detailed information concerning several regional environmental and land-cover characteristics as well as financial activities of the application area. Early notification of the authorities with up-to-date accurate information on the position and evolution of the oil spill, combined with the detailed coastal maps, is of paramount importance for emergency assessment and effective clean-up operations that would prevent environmental hazard. An application was developed for the Region of Crete, an area particularly vulnerable to oil spills due to its location, ecological characteristics, and local economic activities.     

Habitat mapping using machine learning-extended kernel-based reclassification of an Ikonos satellite image

The spatial resolution of satellite imagery suitable for earth resources studies has improved from 80 m (Landsat-MSS, launched in 1972) to 0.6 m (QuickBird, launched in 2001). The conventional pixel-based methods developed for medium resolution satellite images are not suitable for classification of very high spatial resolution images, because the spectral responses of particular habitat classes are much more variable. On the other hand, in the original Barnsley–Barr kernel-based reclassification algorithm not only the spectral information of a pixel but also the textural information in the vicinity of the pixel is used when the pixel labeling decision is made. The first step of the kernel reclassification algorithm is to perform an initial classification of the original image. In the second step, the adjacency-event matrices are computed for each pixel according to co-occurrence frequencies of the initial classes in the kernel window. The degree of matching between an adjacency-event matrix corresponding to specific pixel and the set of class-specific template matrices produced during training is the criterion for pixel re-labeling. We extend the original kernel-based reclassification algorithm with a decision tree-based reclassification, simultaneously taking into account the class-specific similarity images, which are a side-product of the original algorithm. The advantage of decision tree-extended approach over the original approach seems to be the ability of the former to consider more input information, thus increasing the Kappa classification accuracy for an Ikonos image of our study area from 0.56 to 0.60, using a nomenclature containing 10 habitat classes.     

Heat wave hazard classification and risk assessment using artificial intelligence fuzzy logic

The average summer temperatures as well as the frequency and intensity of hot days and heat waves are expected to increase due to climate change. Motivated by this consequence, we propose a methodology to evaluate the monthly heat wave hazard and risk and its spatial distribution within large cities. A simple urban climate model with assimilated satellite-derived land surface temperature images was used to generate a historic database of urban air temperature fields. Heat wave hazard was then estimated from the analysis of these hourly air temperatures distributed at a 1-km grid over Athens, Greece, by identifying the areas that are more likely to suffer higher temperatures in the case of a heat wave event. Innovation lies in the artificial intelligence fuzzy logic model that was used to classify the heat waves from mild to extreme by taking into consideration their duration, intensity and time of occurrence. The monthly hazard was subsequently estimated as the cumulative effect from the individual heat waves that occurred at each grid cell during a month. Finally, monthly heat wave risk maps were produced integrating geospatial information on the population vulnerability to heat waves calculated from socio-economic variables.