• English 
  • Spanish 

Trends in number of fires and burned area and their relationships with climatic variables across regions in Spain during 1974-2008

Conference: ICFBR - International Conference on Fire Behaviour and Risk
Year: 2011
Contribution type: Poster
PDF file: Zavala&al_icfbr2011.pdf
Authors:
Zavala, G., Urbieta, I., Rieiro, I., , , Moreno, J.M.

Understanding past trends in fire occurrence and the relationship between fire and climatic variables is of utmost importance in the Mediterranean countries. In this study we used the Canadian Fire Weather Index (FWI) and its components to determine what relationships exist between fire danger and monthly number of fires and burned area across Spain, and how these relationships have varied in the last 35 years. We analyzed fire statistics (number of fires and burned area, 1974-2008), and the ERA-40 reanalysis database (1974-2001) as input for constructing the FWI. We performed trend analysis and maximum likelihood fire-climate models for the various regions within peninsular Spain and the Balearic Islands. Despite the overall trend of increasing FWI in most of Spain during the last decades, our study shows that different regions have been undergoing different trends. While in some regions an increasing trend in the number of fires was detected, in others the opposite trend was detected. Some regions did not exhibit any pattern in burned area, but in some a decreasing trend was demonstrated. In addition, we found significant relationships between the number of fires and monthly burned area and fire danger indices considering just the meteorological variables. In general, as FWI increased more fires or more burned area occurred. In the most Northern regions, where milder climatic conditions and lower fire danger indices are prevalent, fire activity was highest and so were the correlations between the number of fires and burned area. However, Central, Southern and Eastern regions showed no fires below a threshold (FWI of 5 to 10), but then, fire activity increased up to values of FWI of 40-50. It is remarkable that the relationships varied among regions, some having a saturating-type of relationship while others a more exponential one. The goodness of fit and the variance explained by the models was remarkable for some regions. Most of the Mediterranean regions were the most poorly related to the FWI. Therefore, at least some of the regions with greatest fire activity were not predictable. Overall, our study shows that annual burned area is not increasing in Spain in most regions, and even decreasing in some, despite increased fire danger conditions. Yet, the significant relationships between FWI and burned area indicates that ignitions and fire fighting capacity might be controlling the lack of, or negative, trend in burned area demonstrated for some regions.