Thursday, December 21, 2017

Fire Fire Burning Bright, How Many Acres Burned Last Night

Nick Stokes has extended the discussion on the number of forest acres burned in the US, which basically started in nothing is happening twitter and various blogs that deny climate change is upon us, with the appearance of the graph to the left trying to disprove concern about the current California blazes.  To be straightforward about it Nick don't believe the left hand side of the figure, and he quotes from the US Historical Statistics table which comments

"The source publication also presents information by regions and States on areas needing protection, areas protected and unprotected, and areas burned on both protected and unprotected forest land by type of ownership, and size of fires on protected areas. No field organizations are available to report fires on unprotected areas and the statistics for these areas are generally the best estimates available."

Eli is not going to exactly defend this either, but he will stick by the point he was trying to make that 1900 Galveston hurricane has damn all to do with deaths caused by hurricanes today especially with improved building codes, weather satellites and more.  Since forest fire fighting in the US really took hold in the middle 1930s when the federal government got serious about it the left hand side of the figures have not very much to do with the right hand side

However, Eli did come up with a way to look at this, by examining the number of acres burnt per fire.  One of the interesting things in the above graph which the Bunny did not comment on at the time was the surge about 1980 and the increased variability after that.  Nick points to the National Interagency Fire Center data which covers the period after 1960.  The data in the 1960 to 1970 period is the same as from the Historical Statistics.  The NIFC table at the bottom states that
The National Interagency Coordination Center at NIFC compiles annual wildland fire statistics for federal and state agencies. This information is provided through Situation Reports, which have been in use for several decades. Prior to 1983, sources of these figures are not known, or cannot be confirmed, and were not derived from the current situation reporting process. As a result the figures above prior to 1983 shouldn’t be compared to later data. 
Which explains that step, but it is not a huge one and it is an increase.  Using the data in the Historical Statistics of the United States one can compare the number of acres burnt per fire (apologies, were this an NSF grant Eli would be ethically and contractually bound to use hectares, but it is not) burnt on protected Federal, State and private lands compared to those burnt on unprotected lands about which Nick and others have great doubts.  In this picture the red line represents the number of acres burnt per fire on unprotected lands vs the blue line which is the number of acres burnt per fire on the protected lands.  The ratio is greater than 5 to 1. 

 For a further internal consistency check one can look at the total number of fires in the protected and unprotected categories bearing in mind that the amount of forested area in the US has essentially remained constant.  The number of fires remains roughly constant at 150 to 200K between 1926 and 1955 after which it declines to about 100K. 

The graph to the right shows that essentially all land is protected by 1970 because there are few fires ther, moreover from the graph immediately above by ~ 1940 burning in protected land had reached either a constant level or was slowly declining.  The National Interagency Fire Center table (see first two figures) shows that from about 1980 the amount of forest burnt has increased and the average size of each fire has increased.

Of course, since western and eastern US forests are very different beasts, we now need to look at data from both sides of the continent.


Fernando Leanme said...

This problem has many variables. For example, the USA forest management system controlled fires too much, this allowed dead wood to pile up on the forest floor. There are more roads built into forested areas, more people, but they carry less glass bottles, etc.

I looked up the 2 meter reanalysis for California and it shows temperature hasn't changed much.

What we need is to set up several control plots, double fenced to stop entry by Homo sapiens, where the air, soil, and forest cover temperature and humidity are carefully.monitored for say 20 years. This study can be paralleled by studies where other plots are managed by planting trees, cleaning the dead wood, setting controlled fires, etc.

jgnfld said...

Nice bit of misdirection. Or perhaps you are unaware that one of the major reasons for the large fires of the past were precisely because lumbering practices "allowed dead wood to pile up on the forest floor". The great pine forests of the old "Great Northwest" were essentially reduced to slashings piles which burned mightily and well.

dhogaza said...

To expand on John Garland's comment, in the Pacific Northwest, at least, runaway slash burns were a very common source of forest fires until they began to be heavily controlled late in the 20th century. Fires cause by chainsaws were also common until spark arresters were developed and made mandatory by law.

So a lot of things have changed that makes comparisons over decades difficult (not to mention the eastern/western forest differences mentioned by Eli, or for that matter pond pine vs. wet westside forests in the PNW, etc).

We do know that climate change has led to increasingly virulent and extensive outbreaks of pine bark beetle that kills trees, and we know dry, dead trees burn exceptionally well, just as a "for instance" regarding climate change and increased fire risk. There are many other changes that favor fire resulting from climate change, too.

Of course, everyone here (except perhaps Fernando) knows that the argument that there were lots of fires in the old days means that climate change can't be contributing to increased fires today is as meaningless as stating that the fact that people died before guns were invented and at times at a much higher rate (the medieval plague years in Europe, for instance) means that guns can't kill people.

Denialists won't be convinced, but I'm still glad that folks like Nick Stokes are hard at work debunking them ...

Steve Bloom said...

Thanks for both of these posts, Eli (and to Nick). Very useful.