The Atlantic Multidecadal Oscillation (AMO) and the forest fires in France in the period 1980-2014

Abstract

The study examines the connection between the Atlantic Multidecadal Oscillation (AMO) and the forest fires (the annual number of fires, the annual burned area and the average burned area per fire) in France in the period 1980-2014. In order to determine the strength of the correlation connection Pearson correlation coefficient (R) was used. Monthly, seasonal and annual values of AMO were used in calculations, and one year phase shift was performed (the values for the previous year were used). In burned area the highest values of R on the monthly level were recorded for April (-0.474) and January (-0.470), and on the seasonal level for winter (-0.459) and spring (-0.447). These values are statistically significant at the level of p≤0.01. By phase shifting the highest level of correlation was obtained for the autumn (-0.489). In the average burned area per fire on a monthly level the highest value of R was for January (-0.522), and on seasonal for winter (-0.506). By phase shifting the highest value of R was obtained for autumn (-0.522). In the number of fires the highest values were recorded by phase shifting for September (-0.382) and autumn (-0.337). All R values recorded during the study had a negative sign (the correlation is antiphase). In addition, downward trends were determined for all three examined indicators of forest fires in the researched period (1980-2014). Results of the research could be used as a basis for the long-term forecast of the risk of forest fires, and the approach used in the research could be applied for the other areas of the world. However, the more detailed research of the effects of other teleconnections are necessary.