Keywords
forest fire, static risk assessment, frequency-area statistics, power-law distribution, districting problems
Start Date
1-7-2006 12:00 AM
Abstract
The efficiency of fire prevention and fire-fighting organization are strictly related with the identification of the elementary Administrative Units, whose physical characteristics and number/quality of resources affect significantly the occurrence and the dynamics of the considered phenomenon. In fact, wildland fire risk is determined by several factors such as vegetation, local climate condition, topography, socio-economical aspects, and, of course, wildfire fighting available resources. A power law distribution can be used to represent the frequency-area relationship empirical wildfire data. Thus, it is possible to think of characterizing a region by its wildland fire regime, which can be defined by the power law parameters estimated on the basis of the historical data relevant to fire occurrences. In this paper, a methodology for optimal regional partitioning, in connection with wildfire risk characterization, is proposed and discussed in detail. In particular, the paper defines a procedure to identify, on the basis of the available data set, over a wide regional area, a number of zones (regions) characterized by different wildfire regimes. These zones are composed by groups of contiguous elementary territorial units, which have the same characteristics, as regards the wildland fires phenomenon. Such a characterization can be used to identify the static risk distribution over the considered territory. Liguria, a small region (5400 km2) placed on the northwestern coastline of Italy and frequently affected by severe wildland fires occurrences, is the basis for case study relevant to the implementation of the proposed approach.
Optimal Regional Partitioning for Wildfire Risk Characterization
The efficiency of fire prevention and fire-fighting organization are strictly related with the identification of the elementary Administrative Units, whose physical characteristics and number/quality of resources affect significantly the occurrence and the dynamics of the considered phenomenon. In fact, wildland fire risk is determined by several factors such as vegetation, local climate condition, topography, socio-economical aspects, and, of course, wildfire fighting available resources. A power law distribution can be used to represent the frequency-area relationship empirical wildfire data. Thus, it is possible to think of characterizing a region by its wildland fire regime, which can be defined by the power law parameters estimated on the basis of the historical data relevant to fire occurrences. In this paper, a methodology for optimal regional partitioning, in connection with wildfire risk characterization, is proposed and discussed in detail. In particular, the paper defines a procedure to identify, on the basis of the available data set, over a wide regional area, a number of zones (regions) characterized by different wildfire regimes. These zones are composed by groups of contiguous elementary territorial units, which have the same characteristics, as regards the wildland fires phenomenon. Such a characterization can be used to identify the static risk distribution over the considered territory. Liguria, a small region (5400 km2) placed on the northwestern coastline of Italy and frequently affected by severe wildland fires occurrences, is the basis for case study relevant to the implementation of the proposed approach.
