Yet another piece of bad news is that wildfires are on the increase in the USA. I linked to that story in a blog post about extreme severe events and climate. The article was by Deanna Conners by at earthsky.org .
She offers this graphic, along with a claim that 2015 was (as of mid-September) already in the top six wildfire years, along with the reasonable speculation that 2015 may exceed prior years:
These data trace to “the National Interagency Fire Center in Boise, Idaho, [which] publishes a ton of useful statistics on wildfires that are critical for helping state and federal agencies manage the flames. These records date back to the 1960s.” according to Dr. Conners’ posting.
It’s suggestive of an increase. and compellingly so after adding 2015 to the series as a top-five year.
I also pointed to an odd bit of counter-evidence that I’ve seen trumpeted triumphantly by the occasional climate troll. Unfortunately it is on a legitimate US Forest Service webpage.
Now everybody who has any intuition about environmental data who has looked at this says it doesn’t pass the smell test – it looks very much as if the pre-1960 data which NIFC does not report is of something else than what NIFC reports after 1960. Unfortunately the sourcing of the graph is unclear.
In a recent comment here, Dr. Conners reports that she has worked on tracking the graph down independently. I think it’s at least important enough to quote in full in a top level article here.
Thanks for linking to my blog on wildfires. I spent some time tracking down the original source for the pre-1960 wildfire data that you linked to in the footnote of your blog—the original source for the U.S. Forest Service graph was “America’s Renewable Resources: Historical Trends and Current Challenges,” which was published in 1991 by the non-profit organization Resources for the Future. The original graph is located on page 117 of Chapter 3, and the author (R. A. Sedjo) cites “U.S. Department of Agriculture, Forest Service, 1926–1967, Forest Fire Statistics, various annual issues,” as their data source. The author also cites “U.S. Department of Agriculture, Forest Service, 1968–1989, Wildfire Statistics, annual issues,” as their data source for post-1960 data. Unfortunately, the text of that document did not describe the data in detail.
I did find a useful supplemental document (http://www.fs.fed.us/research/sustain/docs/national-reports/2003/data/documents/Indicator%2015/Indicator%2015.pdf) to the more recent U.S. Forest Service reports/graphs that had this to say about the pre-1960 data (page 19):
“Between 1930 and 1950, in excess of 10 million acres were burned by wildfires annually. Most of the area burned during this period was in the Southeastern United States (South RPA Region) and were primarily incendiary fires. Since 1960, between 2 and 5 million acres were burned annually by wildfires. In recent years, the average area burned has increased, especially in the Rocky Mountain and Pacific RPA Regions. A peak fire year occurred in 1988 when 7.4 million acres burned. This was the year of the extensive fires in the Greater Yellowstone Basin. The range of recent variation (since 1960), in terms of area burned, was exceeded in 2000 when wildfires burned more than 8.4 million acres. This was the largest area burned in more than 40 years.”
So it appears that much of the pre-1960 data were related to incendiary forest fires (per http://www.interfire.org/features/wildfires.asp, an incendiary fire is one that is set intentionally) and not to true wildfires. The post-1960 dataset that I analyzed only contained data for wildfires; the National Interagency Fire Center explicitly separates the wildfire data from the prescribed fire data. Hence, comparisons to earlier data may indeed be akin to comparing apples to oranges, as Magma cautioned earlier in their comment.
So the case isn’t exactly proven in terms of which are the apples and which the oranges, but it’s pretty compelling that it is a graph, so to speak, of apples after 1960 and some other fruit or fruits before 1960. Fortunately, the title of the graph is honest enough. It says “Total acreage burned – all lands” which pretty clearly indicates that it isn’t really a graph of wildfires.
I think that it does match the 1960-2006 NIFC data closely. If so it must be a plot of at least two separate and distinct records. If that in turn is so, how such a shabby thing gets on a federal website and how to get it off is another story, but examined closely, it neither supports nor refutes the claim that current wildfire extent was greatly exceeded or even precedented in the past.