Updated Saturday, June 12, AD 2021 UTC
Here I apply Kurz's Number to federal data on acres burned in California, Oregon, Washington and the U.S., to estimate how much CO2 equivalent wildfires here have produced so far this year. I compare those results to federal data on the amount of CO2 that all energy sectors (residential, commercial, industrial, transportation and electric power) in California, Oregon, Washington and the U.S. produce.
These calcs are significant because almost all of the political and academic efforts to reduce CO2 focus on energy-related CO2 only. However, wildfire-produced CO2 can approach or exceed energy-related, non-wildfire CO2 in States with big forest fires. Wildfire prevention and suppression policies for publicly-owned lands are dismal if not criminal.
In this article I consider only CO2 and equivalent from wildfires. As far as I know, no government in the U.S. does any wildfire CO2 or thermal accounting; the CO2 equivalent numbers here are in addition to energy-related CO2 numbers that the federal government publishes.
AC means acres. MT means metric tons. MMT means million metric tons:
The calcs on this page use the following numbers of MMT of energy-related CO2 from all sectors: residential, commercial, industrial, transportation and electric power, for the places that I consider here. The numbers below are from the U.S. Energy Information Administration (EIA), as of May 7, 2021:
State or region | MMT ----------------------+--------- California in 2018 | 362.5 Washington in 2018 | 78.8 Oregon in 2018 | 39.5 The U.S. in 2019 | 5,130 The Pac NW in 2017 | Wash & Ore | 118.3 The Pac Coast in 2017 | The Pac NW & Cal | 480.8
For the calcs below, I use acres burned numbers that change daily, and the above, May 7, 2021, EIA State numbers.
California at the close of 2021-06-08:
Cal (Northern + Southern) fires:
1445 + 1917 = 3362
Cal (Northern + Southern) acres burned:
3613 + 15878 = 19491 AC
Cal average acres per fire:
19491 / 3362 ≈ 6 AC/fire
Cal wildfire-produced CO2 & equivalent:
19491 AC * 68.8 MT/AC = 1340980.8 MT
≈ 1 MMT
1 MMT as a percent of the energy-related CO2 that Cal produced in 2018:
100 * 1.3409808 MMT / 362.5 MMT ≈ 0.3699...%
At the close of 2021-06-08, Cal fires have produced 0.4% of the energy-related CO2 that Cal produced in 2018.
The Pac Northwest at the close of 2021-06-08:
Pac NW (OR and WA) fires: 816
Pac NW acres burned: 15673 AC
Pac NW average acres per fire:
15673 / 816 ≈ 19 AC/fire
Pac NW wildfire-produced CO2 & equivalent:
15673 AC * 68.8 MT/AC = 1078302.4 MT
≈ 1 MMT
1 MMT as a percent of the energy-related CO2 that the Pac NW produced in 2018:
100 * 1.0783024 MMT / 118.3 MMT ≈ 0.9114...%
At the close of 2021-06-08, Pac NW fires have produced 0.9% of the energy-related CO2 that the Pac NW produced in 2018.
The Pac Coast at the close of 2021-06-08:
Pac Coast (Cal and Pac NW) fires:
3362 + 816 = 4178
Pac Coast (Cal and Pac NW) acres burned:
19491 AC + 15673 AC = 35164 AC
Pac Coast average acres per fire:
35164 / 4178 ≈ 8 AC/fire
Pac Coast wildfire-produced CO2 & equivalent:
35164 AC * 68.8 MT/AC = 2419283.2 MT
≈ 2 MMT
2 MMT as a percent of the energy-related CO2 that the Pac Coast produced in 2018:
100 * 2.4192832 MMT / 480.8 MMT ≈ 0.5031...%
At the close of 2021-06-08, Pac Coast fires have produced 0.5% of the energy-related CO2 that the Pac Coast produced in 2018.
The U.S. at the close of 2021-06-08:
U.S. fires: 26833
U.S. acres burned: 833479
U.S. average acres per fire:
833479 / 26833 ≈ 31 AC/fire
U.S. wildfire-produced CO2 & equivalent:
833479 AC * 68.8 MT/AC = 57343355.2 MT
≈ 57 MMT
57 MMT as a percent of the energy-related CO2 that the U.S. produced in 2019:
100 * 57.3433552 MMT / 5130 MMT ≈ 1.1178...%
At the close of 2021-06-08, U.S. fires have produced 1.1% of the energy-related CO2 that the U.S. produced in 2019.
Oregon at the close of 2021-06-08:
Oregon fires: [no ytd data from ODF]
Oregon acres burned: [no ytd data from ODF]
Oregon average acres per fire:
[no ytd data from ODF]
Oregon wildfire-produced CO2 & equivalent:
[no ytd data from ODF]
≈ [no ytd data from ODF]
0 MMT as a percent of the energy-related CO2 that Oregon produced in 2018:
[no ytd data from ODF]
≈ [no ytd data from ODF]
At the close of 2021-06-08, Oregon fires have produced [undef.] of the energy-related CO2 that Oregon produced in 2018.
Updated Sat Jun 12, AD 2021 02:29:53 AM UTC
Tap and hold video to get pause button.
Explication and Sources of the Above Numbers
Except for Oregon, I obtain fires and acres burned numbers from Incident Management Situation Reports (IMSR’s, which I call sitreps). During fire season, The Predictive Services Intelligence Section, a section of The National Interagency Coordination Center (NICC), a branch of The National Interagency Fire Center (NIFC) publishes these sitreps daily. A link to the current sitrep is at the NICC fire stats page.
Predictive Services keeps an archive of sitreps here.
The data in those sitreps are as of the previous day. When using data from the sitrep published on day d, I write, “at the close of” day d-1. For instance, where I wrote, “at the close of 2021-06-08”, the sitrep dated 2021-06-09 provided data for those calcs.
I obtain acres burned numbers only from the Fires and Acres Year-to-Date (by Protection) table in the sitreps. These numbers generally either increase or stay the same over time. They may also decrease, in which case a Coordination Center made them more accurate. This table is a “designated authoritative source”:
Fires and Acres Year to Date is derived directly from Situation Reports submitted by individual units. The ICS-209 program does not provide data for this table. Note: this table represents fires and acres by the protecting agency (not by ownership). ...
The information included in the IMSR comes from Situation and ICS-209 reports, which are typically updated once a day in the evening. Both the Situation Report and the ICS-209 program are designated authoritative sources by federal land management agencies.¹ (from Understanding the IMSR)
There have been large (≥ 100 ac) forest fires in Oregon in 2021. However, as of 2021-06-11, the Oregon Department of Forestry has not published a YTD fire report for 2021.
The U.S. Energy Information Agency (EIA) publishes data for States yearly. The energy-related CO2 data that I use here are from 2018, because 2018 is the most recent year for which State data are available. I compare that to 2020 data on acres burned, translated to wildfire-produced CO2 equivalent.
To get the 2018 energy-related CO2 produced by State, I go to this eia.gov page, and download spreadsheets.
As the EIA updates those data to 2019, 2020, etc., you might try this eia.gov data page and scroll down to State CO2 Emissions.
The sources for the National calcs are the NIFC daily sitrep and this eia.gov page. The energy-related CO2 number for 2019 is in the Overview of CO2 Emissions section, in the explication for Figure 1, and also in the full PDF report.
To calculate CO2 equivalent produced from acres burned I use Werner Kurz’s number: Every hectare (HA) of forest burned produces about 170 metric tons (MT) of carbon dioxide equivalent. The WA-Po published an article mentioning Kurz’s number on May 20, 2016. The National Post re-published that article. Kurz is a Doctor of Forest Ecology, a senior research scientist at the Canadian Forest Service and head of its carbon accounting team.
Since U.S. data are in acres, we need to convert Kurz’s number to MT/AC:
1 hectare (HA) = 2.47105381 acres (AC)
170 MT/HA * 1 HA / 2.47105381 AC = 68.796559 MT/AC
≈ 68.8 MT/AC
Kurz’s number is an estimate. It ends with a zero, which might or might not be a significant digit. One would need to consult Kurz to find out for sure.
Assuming the zero is not significant, Kurz’s number would more correctly be written, 170 ± 5 MT/HA. The second digit, 7, is significant, otherwise, Kurz would probably have put it at either 160 or 180. But the number, 170, could be anywhere from 165 (which rounds up to 170) up to, but not including, 175 (which rounds down to 170).
I calculate MT of CO2 equivalent using three significant digits (68.8 MT/AC), then round the result to the nearest MMT, making it an integer. However,
5 / 2.47105381 ≈ 2.0, and a more scientific expression would include the accuracy: 68.8 ± 2.0 MT/AC. For instance, for the U.S. calculation for Oct. 4, 2020:
7,818,340 AC * 68.8 MT/AC = 537,901,792 MT
the actual product is somewhere between,
7,818,340 AC * 66.8 MT/AC = 522,265,112 MT and
7,818,340 AC * 70.8 MT/AC = 553,538,472 MT.
The range of possible values over two is the accuracy for that calc:
(553,538,472 MT - 522,265,112 MT) / 2 = 15,636,680 MT
≈ 15.6 MMT
In rounding the product to the nearest MMT, I am not losing accuracy, because 538,000,000 is in the middle of that range, and the actual number of MMT is somewhere within that range, exactly where we do not know, given ± 2.0 MT/AC as the accuracy of Kurz’s number. Written more scientifically, my results would look like this:
7,818,340 AC * (68.8 ± 2.0 MT/AC) = 537,901,792 ± 15,636,680 MT
≈ 538 ± 15.6 MMT
I’m giving a general accuracy here, instead of in the results: the accuracy of the MT of CO2 & equivalent numbers is at least ± 2.0 MT/AC. One can more simply calculate the accuracy of those numbers as:
± 100 * 5 / 170 ≈ ± 2.94%
It might be more than that, because there are also uncertainties associated with the acres burned numbers. One is that Kurz’s number is for timber fires. Most of the big, 2020 fires were in timber. But some acres burned were brush, some grass. And as you move east from the Cascades and Sierra Nevadas, there is relatively less timber and more brush and grass. The acres burned numbers that I use here include all fuel types.
Another consideration is hills and mountains. How many actual acres of forest are in a hillside acre of forest?
One acre of mountainous or hilly forest is more than one acre of flat terrain forest. The slope of the hill puts additional acreage in the vertical dimension that is not seen when measuring acreages from aerial or satellite photos, which make everything appear flat.
For an extreme example, consider an imaginary building in New Haven that is 100 × 100 ft2, and 400 feet tall. It is covered with ivy on all sides from bottom to top, and the roof, too. Yale wants to inventory all the ivy in New Haven. So they pay for overhead satellite photos, and hire a grad student to trace out the ivy-covered areas of each photo on a touchscreen. A program calculates and sums the areas traced out.
The grad student dutifully traces out 100 x 100 = 10,000 ft2 of ivy-covered area on that lot, because that is what the satellite saw, and who would dare interpret satellite photos? But that building actually has 4 x 100 x 400 + 10,000 = 170,000 ft2 of ivy. The satellite photo was off by 160,000 ft2. That area of ivy was hidden from the satellite in the vertical dimension. The same kind of error happens when using satellite photos to calculate acres of mountainous or hilly forest.
Consider a rectangle of forested hillside. The length of the rectangle is 660 ft., and is along the steepest right line from a lower to a higher elevation on the hillside. Call this the l-t-h line. The width of the rectangle is 66 ft. and roughly follows the contour lines of the hillside. The actual area of that rectangle is 660 ft * 66 ft = 1 acre. Let θ be the angle of the l-t-h line up from the horizontal, that is, the angle that you would move your neck or eyes up from the horizontal to look at an eye-level marker on a surveyor’s grade rod at the top the l-t-h line, if you were standing at the bottom of it. When a satellite views this same rectangular acre of hillside, it does not look like 660 * 66 ft2. It looks like 660 cos θ * 66 ft2. The missing part, (660 - 660 cos θ) * 66 ft2, is hidden from the satellite in the vertical dimension.
If the grade of the hill is 0%, there is no hill:
θ = 0°. cos θ = 1, and the satellite sees an area measuring 660 * 1 * 66 = 1 acre.
If the grade of the hill is 30%,
θ = arctan(30/100) ≈ 16.700°
cos 16.700° ≈ .958 and the satellite sees an area measuring only 660 * .958 * 66 ≈ .958 acre. But it is actually one acre.
If the grade of the hill is 100%,
θ = arctan(100/100) = 45°
cos 45° ≈ .707 and the satellite sees an area measuring only 660 * .707 * 66 ≈ .707 acre. But it is actually one acre.
If the grade of the hill is infinity percent,
θ = arctan(∞/100) = 90°
cos 90° = 0 and it’s a cliff. Probably there are no trees on it, but even so, the satellite sees an area measuring
660 * 0 * 66 = 0 acre, even though it is still one acre. The entire acre is hidden from the satellite in the vertical.
There are uncertainties in all of the numbers that I use, including the energy-related CO2 numbers. Those are only government estimates, and I don’t know where to find their accuracies.
Kurz’s number is applicable to BC forest fires, and is probably accurate for Oregon and Washington fires, as the ecosystems are similar. His number might be somewhat less accurate for Northern California fires, as the ecosystems are somewhat different. His number might be inaccurate for some Southern California fires, as some ecosystems there are very different. Despite these differences, his is the only number that I know of that relates hectares burned to MT of CO2 equivalent produced. So I use it for all wildfires, regardless of ecosystem or geographic location. The U.S. Government does not that I know of derive Kurz-type numbers for U.S. ecosystems; it has historically neglected to take into account CO2 produced by wildfires.
1: I do not use CalFire numbers, because their numbers are not official until CalFire administrators get together with other Cal wildfire agencies, the State Fire Marshal’s Office and CalStats administrators, and publish them and other stats and charts in the Redbook. That publishing project takes more than nine months.