April 25, 2007

Torching the Forest Fire Myth

Filed under: Droughts

Since the first time you heard about global warming, you probably learned that among other consequences, wildfires will increase in frequency and area burned in many parts of the world. This prediction is based on the obvious link that increased temperatures will increase evapotranspiration, forests will become drier in the absence of any increase in precipitation, the changes in climate will promote a weakening of the forest ecosystems making them more susceptible to countless stresses, and forests will “burn baby burn.” Back in 1988 when the greenhouse engine was getting into gear (a rather low gear compared to the overdrive gear of today), Yellowstone Park burned as did many other forests in the western states, and the image of forest fires being linked to global warming has lived on ever since. There is not a day that a substantial forest fire is not burning somewhere across the planet, so the popular press will never run out of material on this front.

A recent article has appeared in the Journal of Geophysical Research showing that the burned area from forest fires in Ontario, Canada has increased since 1970, and we are sure the greenhouse crusade will add this to their arsenal of facts supporting their view of global environmental change. However, if one reads this interesting article in its entirety, a completely different picture emerges regarding the recent trend upward in burned area (by the way, on the first page of the article, the authors note “It is also possible that the provincial fire statistics (number and size of fires) were underreported prior to the 1960s”).

A team of scientists from various forest-related organizations throughout Canada conducted the research, and they state in the introduction “Human-induced climate change could lead to an increase in forest fire activity in Ontario, owing to the increased frequency and severity of drought years, increased climatic variability and incidence of extreme climatic events, and increased spring and fall temperatures. Climate change therefore could cause longer fire seasons, with greater fire activity and greater incidence of extreme fire activity years.” The news seems to get worse as they note “Fire has also been recognized as a significant source of greenhouse gas emissions into the atmosphere. Most of this is in the form of carbon dioxide (CO2), but quantities of carbon monoxide, methane, long-chain hydrocarbons, and carbon particulate matter are also emitted. From 1989–1996, forest fires constituted about 2% of Ontario’s total CO2 emissions.”

By now, you must be wondering why we are featuring this article in World Climate Report? Keep reading.

Girardin et al. were interested in extending data on the burned area of Ontario back more than a few hundred years, and they used tree rings to estimate burned areas of the past. Not surprisingly, forest fires produce a recognizable signal in the annual tree rings, and the Canadian team used sophisticated statistical wizardry to go from tree ring patterns to total area burned in a given year. They tested their statistical model from 1917 to 1981, and once satisfied that the model was working with reasonable accuracy, they used to tree rings to extend burned area estimates back to 1781.

The figure below shows the results, and in the authors’ own words, we learn “Episodes of succeeding years of large area burned were estimated approximately at 1790 – 1794, 1803 – 1807, 1818–1822, 1838–1841, 1906–1912, 1920–1921, and 1933–1936.” Furthermore “The tree ring model revealed the year 1804 as the year of most extreme area burned.” It gets even more interesting as they reveal “The sliding window analysis showed higher mean area burned values prior to 1840 and through the 1860s–1880s and the 1910s–1920s. Mean values during the mid-20th century were the lowest in the record. A t-test revealed that the mean area burned in the 1940–1969 period was significantly lower than that of 1910–1939.” They conclude “The reconstruction indicated that the most recent increase in area burned in the province of Ontario (circa 1970–1981) was preceded by the period of lowest fire activity ever estimated for the past 200 years (1940s–1960s).” Finally, they write “while in recent decades (circa 1970–1981) area burned has increased, it remained below the level recorded prior to 1850 and particularly below levels recorded in the 1910s and 1920s”.


Figure 1. (a) Reconstruction of area burned in the province of Ontario for 1782–1981 (thick line). Thin line represents instrumental data (1917-2003). (b) 10-year window polynomial curve (from Girardin et al., 2006).

This recent article is a perfect example of confusion the public must feel regarding important elements of the greenhouse debate. One camp could take the article and claim that numerical models are forecasting an increase in forest fires (actually, no global climate model makes such a direct prediction) and that the evidence from Ontario indeed shows an increase in burned area in recent decades. They would be accurate and consistent with the findings of the Girardin et al. team. We at World Climate Report look at the same article and find evidence that the post World War II period has been unusually quiet in terms of big burns in Ontario, and we would be just as accurate and just as consistent with their recent findings. The figure above shows that the “big burn” periods all occurred prior to World War II.

You decide, but as this essay shows, the deeper you dig into this article, the less evidence there is for any claim that the buildup of greenhouse gases has resulted in an increase in wildfires in Ontario. The beat goes on!

Reference:

Girardin, M. P., J. Tardif, and M. D. Flannigan, 2006. Temporal variability in area burned for the province of Ontario, Canada, during the past 200 years inferred from tree rings, Journal of Geophyical Research, 111, D17108, doi:10.1029/2005JD006815.




No Comments

No comments yet.

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

Powered by WordPress