Assessment of risk index for urban vegetation fires of Juiz de Fora, MG, Brazil

Abstract

The impact caused by forest fire, although it is an old subject, but increasingly larger proportions, has stimulated the development of new methods for the prevention and reduction of risks in the environment. The use of a reliable index of risk among the existing preventive measures is fundamental to a more efficient planning of prevention and an effective action of fight. The work involving the forest fire prevention must use all the tools available since it is a complex task and one that shows great variability in the time of occurrence. One such tool is the degree of danger by means of an index that reflects the probability of fire occurrence. Meteorological indices have shown good results in predicting events and contributing to the work of forecasting fire in forests. The aim of this study was to analyze 4 descriptive fire risk indices for the period from 01/01/2006 to 12/31/2012. Each of them was built with 4 hazard classes of risk. The indices used were Formula de Monte Alegre (FMA), Index of Nesterov, Index of Telicyn and P-EVAP According to the results, the rate of P-EVAP performed better in accuracy of forecasts of occurrences within the class of High Risk, while the FMA was the most effective in the forecast of no occurrence in classes of Low Risk and No Risk. The days classified as Medium Risk in Telicyn indices and P - EVAP presented closest to the average of occurrences in the entire series averages. Knowledge of descriptive classes of indices for predicting fire is valuable to communicate with public, help the actions of awareness, prevent and plan accident. The results had shown some similarities with studies of Torres et al. (2009). They used the Skill Score (SS) method and concluded that the logarithmic index of Telicyn showed the best results for the study area.