Relationship between sea regions with high thermal variability and wildfires from 1981 TO 2008

Daily OISST version 2 data from 1981 to 2008 are used to detect high SST variability regions in order to explore their relationship with wildfires in some particular regions. A burned area (BA) dataset is obtained for Canada for the period 1982-2006, from LTDR dataset version 2 and LAC burned area record. A selection of high variability cells encountered within the Nino 3.4 region is analyzed with standardized seasonal BA, where a significant correlation at year lag in spring and three year lag in summer and autumn was found. Yet, results showed that SST and Oceanic Nino Index did not cause BA with a three-year lag time in Wiener-Granger sense, whereas annual SST did cause BA in spring, at the 0.95 confidence level. The proportion of fires occurred from 1982 to 2006 after an El Nino event increased at one- and two-year lag.

1. INTRODUCTION

Previous works have studied teleconnections, where at least two different time series data are needed, by analysing particular climate oscillations with lagged changes in some in-land processes and variables [2], like air temperature, precipitation, stream flow, or vegetation activity. In this study, we made lag correlation analyses between both variables and we put into practice a Wiener-Granger causality test in every case to verify the significance and nature of the relationship between them.