Abstract
A comparative study of lightning-, and people-caused fires is presented for peninsular Spain, for the period 1974-1994. Based on records of fire reports, yearly trends for fires started by the two causes were compared. Fire reports assign each fire to one 10 × 10 km grid-cell within the country. This information, together with data on the cause and date of fire, elevation, size of fire, type of vegetation burned, and meteorological conditions at the time of fire initiation, was incorporated to a raster-based geographic information system for further analysis and mapping. Additional information incorporated to the GIS for each grid-cell was the phytogeographic sector to which it belonged and the main land-use types. The study shows that the number of fires has increased recently and, particularly, that of lightning fires. Annual fire occurrence of the two causes was significantly correlated. People-caused fires were widespread throughout most of the country, whereas lightning-caused fires, although also widely dispersed throughout Spain, were more clustered together in certain areas, mainly in the eastern part of the country and along certain mountain ranges. The difference between the geographic distribution of the fires started by the two causes was statistically significant. Additionally, lightning-caused fires occurred at upper elevations and were more clustered towards the summer than people-caused fires. Furthermore, in those grid-cells where fires of both causes occurred, lightning fires tended to occur at upper elevations, affected more woodlands, produced smaller maximum fire-sizes, and were started under different meteorological conditions than people-caused fi-es. Fire frequencies were small, and fire rotation periods high, in most phytogeographic sectors of the country for fires caused by lightning, not so for fires caused by people. In general, fire temporal-, and geographic-patterns, and fire characteristics of lightning-caused fires were different from those of people-caused fires.