|
Abstract
|
Historical fire records and meteorological observations spanning over one century (1894-2010) were assembled in a database to collect long-term fire and weather data in Greece. Positive/negative events of fire occurrence on an annual basis were considered as the years where the annual values of the examined parameters were above (positive values) or below (negative values) the 95% confidence limits around the trend line of the corresponding parameter. To analyse the association of positive/negative events of fire occurrence with meteorological extremes, we proceeded with a cross-tabulation analysis based on a Monte Carlo randomization.
Positive/negative values of total annual precipitation were randomly associated with the corresponding values of burned areas, and significant associations were observed for seasonal precipitation totals (spring and fire season). Fire season precipitation is the dominant factor coinciding with negative values of area burned, while years with high spring precipitation coincide with years with large areas burned. These results demonstrate the dual role of precipitation in controlling a fire's extent through fuel build-up and dryness. Additionally, there is a clear outperformance of precipitation-related variables compared with temperature-related weather revealing that, at least in Greece, total area burned at the national scale is controlled by precipitation totals rather than air temperature.
This analysis improves our understanding of the underlying mechanisms of fire regimes and provides valuable information concerning the development of models relating fire activity to weather parameters, which are essential when facing a changing climate that may be associated with shifts in various aspects of the typical fire regimes of ecosystems. Our results may allow fire managers to more easily incorporate the effect of extreme weather conditions into long-term planning strategies. They contribute to the exploration of fire-climate relationships and may become more important if climate change scenarios are used to predict the occurrence of future extreme weather taking into consideration that climate change is discussed on the basis of changes of extremes rather than changes in means. |