A Methodological Overview of Risk Mapping Approaches Used in Prevention of Forest Fires From Past To Present

Abstract

Natural or cultural caused forest fires were increased disaster risk in particularly regions that were populated by the human towards natural areas. It was known that the forest fires were mostly caused by the human activities. In addition to direct, stalk fire, shepherd fire, cigarette, picnic fire, sabotage, etc. as an indirect factor, global warming has created ideal conditions for the fire occur. The mapping of forest fire risk offers significant advantages in the stages of prevention, detection and response, which is a part of disaster management. The aim of this study was to reveal the availability of the most widely used forest fire risk mapping techniques from past to present, and to offer suggestions on a suitable fire risk mapping and response system for our country. In this context, we focused on usage pattern, accuracy and data structures of the fire weather index (FWI), which is one of the oldest risk mapping techniques, traditional multi-criteria spatial decision support systems that are not dependent on fire occurrence data, and data-dependent multi-criteria spatial decision support systems, simulation (simulation) based risk assessment systems. As a result, it has been suggested that easily applicable techniques such as FWI and methods suitable for automation-based system creation such as machine learning and deep learning should be integrated in a single interface, supported by remote sensing, and concerned digital data. In addition, it has been determined that there is a need for a system in which damage determinations can be made in advance or instantaneously, and prevention-intervention strategies can be determined, by supporting the regions where disaster risk is detected at local scale with simulation inputs obtained from this platform. It has been determined that the most important problems in such a system are the integration of data from different sources and the development of an artificial intelligence-based automatic action system.