User Tools

Site Tools


Projecting future burnt area in the EU-Mediterranean countries under IPCC SRES A2/B2 climate change scenarios.


The aim of this work is to use the results of statistical modelling of historical (1985-2004) monthly burnt areas in European Mediterranean countries, as a function of monthly weather data and derived fire danger indexes, and to analyse potential trends under present and future climate conditions.


Meteorological variables were extracted from the ECMWF, and the FWI system components were computed from 1961 to 2004. Monthly averages of the indexes were used as explanatory variables in a stepwise multiple linear regression analysis, to estimate the monthly burnt areas in each of the five most affected Mediterranean countries of Europe. Significant regression equations and satisfactory coefficients of determination were found, although with remarkable differences between countries. Two IPCC SRES climate change scenarios (A2/B2) were simulated using the the regional climate model HIRHAM. The multiple regression models were than applied to the A2/B2 scenarios results to predict the potential burnt areas in each country. The models pointed out tangible changes in the potential burnt area extent for the future scenarios compared to the actual conditions.

Input Data

As input data we use the following datasets:

  • Daily wind, temperature, humidity and precipitation were downloaded from ECMWF .
  • Forest fire data (burnt area and number of fires) were extracted from EFFIS.
  • Nuts3 (province) area is stored in ~/ost4sem/studycase/firerisk/tables/input/nuts3area.asc

Computational Implementation

The modelling procedure includes only the calculation of the regression equation and its graphic for each EU country. The climate change projection is not described:

wiki/firemod.txt · Last modified: 2017/12/05 22:53 (external edit)