COS 78-3 - The 1988 Yellowstone fires: A geospatial examination of the impact of historical insect damage on forest fire risk

Wednesday, August 8, 2007: 8:40 AM
Willow Glen III, San Jose Marriott
Heather J. Lynch, Ecology & Evolution, SUNY Stony Brook, Stony Brook, NY and Paul R. Moorcroft, Organismic and Evolutionary Biology Dept., Harvard University, Cambridge, MA
In this study, we examine the historical record of mountain pine beetle (Dendroctonus ponderosae Hopkins) activity within Yellowstone National Park, WY for the 25 year period leading up to the 1988 Yellowstone fires (1963-1986) in order to determine how prior mountain pine beetle activity and resulting tree mortality affected the spatial pattern of the 1988 Yellowstone fires.  To obtain accurate estimates of our model parameters, we used a Markov chain Monte Carlo (MCMC) method to account for the high degree of spatial autocorrelation inherent to forest fires.  Our final model included three statistically significant variables: drought, aspect, and sustained mountain pine beetle activity in the period 1972-1975.  Of the two major mountain pine beetle outbreaks to precede the 1988 fires, the older outbreak (1972-1975) was significantly correlated with the burn pattern, whereas the more recent outbreak (1980-1983) was not.  Although regional drought and high winds were responsible for the large scale of this event, the analysis indicates that mountain pine beetle activity in the mid-1970s increased the odds of burning in 1988 by 11% over unaffected areas.  While relatively small in magnitude, this effect combined with the effects of aspect and spatial variation in drought had a dramatic impact on the resulting spatial pattern of burned and unburned areas in 1988.
Copyright © . All rights reserved.
Banner photo by Flickr user greg westfall.