Canadian Forest Service Publications

Trend analysis of fire season length and extreme fire weather in North America between 1979-2015. 2017. Jain, P.; Wang, X.; Flannigan, M.D. International Journal of Wildland Fire 26(12): 1009-1020.

Year: 2017

Available from: Great Lakes Forestry Centre

Catalog ID: 38287

Language: English

CFS Availability: PDF (request by e-mail)

Available from the Journal's Web site.
DOI: 10.1071/WF17008

† This site may require a fee.

Plain Language Summary

We have constructed a fire weather climatology over North America from 1979-2015 using data from the North American Regional Reanalysis (NARR) dataset as input to the Canadian Fire Weather Index (FWI) System. We tested the presence of trends in potential fire season length and extreme fire weather using the non-parametric Mann-Kendall test with field significance at each spatial grid point established by two statistical techniques that account for spatial correlation. A randomization test established field significance for the lengthening of fire season, an earlier fire season start and a later fire season end over North America. Applying the false discovery rate procedure also allowed us to identify the spatial locations of these significant trends. The lengthening of fire season (significant for 10.5% of trends) is attributable mainly to a later fire season end (significant for 10.3% of trends) rather than to an earlier fire season start (significant only for 2.0% of trends). There were no significant spatial trends found associated with a shortening of the fire season. Field significance was also established for an increase in annual spread days and in the 99th percentile of FWI over both Canada (1979-2002) and CONUS (1979-2015). The proportion of significant trends corresponding to increased spread days was 1.3% over Canada and 12.2% over CONUS. Similarly, the proportion of significant trends corresponding to increasing 99th percentile of FWI was 5.0% over Canada and 13.9% over CONUS. The proportion of significant trends with decreases in these variables were much lower (equal to or less than 1.5%). The smaller proportion of significant positive trends in these variables over Canada reflects the truncation of the time series to 24 years, necessary because assimilation of precipitation observations over Canada ceased in NARR post 2002.

Date modified: