© 2015 Peter Free
Citation — to study
Sarah J. Harta, Tania Schoennagel, Thomas T. Veblen, and Teresa B. Chapman, Area burned in the western United States is unaffected by recent mountain pine beetle outbreaks, Proceedings of the National Academy of Sciences, DOI: 10.1073/pnas.1424037112 (online before print, 23 March 2015)
Citation — to press release
Jim Scott, Study: Western forests decimated by pine beetles not more likely to burn, University of Colorado at Boulder (23 March 2015)
From the press release:
Western U.S. forests killed by the mountain pine beetle epidemic are no more at risk to burn than healthy Western forests, according to new findings by the University of Colorado Boulder that fly in the face of both public perception and policy.
The CU-Boulder study authors looked at the three peak years of Western wildfires since 2002, using maps produced by federal land management agencies. The researchers superimposed maps of areas burned in the West in 2006, 2007 and 2012 on maps of areas identified as infested by mountain pine beetles.
The area of forests burned during those three years combined were responsible for 46 percent of the total area burned in the West from 2002 to 2013.
“The bottom line is that forests infested by the mountain pine beetle are not more likely to burn at a regional scale,” said CU-Boulder postdoctoral researcher Sarah Hart, lead study author. “We found that alterations in the forest infested by the mountain pine beetle are not as important in fires as overriding drivers like climate and topography.”
© 2015 Jim Scott, Study: Western forests decimated by pine beetles not more likely to burn, University of Colorado at Boulder (23 March 2015)
Surprising? — Maybe not
It seems to me that beetle-killed trees would only increase the probability of forest fire, if they noticeably increased:
(i) overall forest trees dryness — (the whole forest cumulatively becomes drier),
(ii) microclimate dryness — (less evaporation from a significant dead forest fraction reduces atmospheric humidity),
(iii) the dead trees significantly increase the risk of several of them simultaneously catching fire, which then creates a hot spot expansive enough to spread to the rest forest.
None of these factors (however) work in just one direction:
For example, dead trees would take up no water, leaving more for the rest. Presumably, this fraction would now be wetter than they might otherwise be, thereby reducing their chance of going up in flames.
Relatedly, dead trees would evaporate no water, but their now wetter companions might evaporate more, balancing out the micro-atmospheric difference.
Third, at a macro spatial level, the influence of drought and topography might always have been so massive as to cancel out the significance of other factors.
Obviously, the study is not conclusive, given its narrow time span.
One also has to accept that the 54 percent non-overlap (between beetle kill and non-beetle kill areas) proves the team’s conclusion. Mathematically, I am a bit skeptical that essentially a half and half distribution (presumably chance) necessarily proved the reported conclusion:
Although MPB [mountain pine beetle] infestation and fire activity both independently increased in conjunction with recent warming, our results demonstrate that the annual area burned in the western United States has not increased in direct response to bark beetle activity. Therefore, policy discussions should focus on societal adaptation to the effects of recent increases in wildfire activity related to increased drought severity.
© 2015 Sarah J. Harta, Tania Schoennagel, Thomas T. Veblen, and Teresa B. Chapman, Area burned in the western United States is unaffected by recent mountain pine beetle outbreaks, Proceedings of the National Academy of Sciences, DOI: 10.1073/pnas.1424037112 (online before print, 23 March 2015) (at Abstract) (paragraph split)
What would the fire distribution have been like, if noticeably fewer trees had been beetle-killed or killed in different patterns of density?
I do not see how the team could so narrowly have pinned down the possible combinations of drought and varying terrain, so as to eliminate the possibility that beetle kill does increase the chance of fire by some percentage in some locales.
Without having access to the whole article, I do not know whether these considerations were statistically eliminated.
The moral? — Our assumptions about cause and effect are often just weak hypotheses
Hence the value of the scientific method.