 |
| Forests spreading fast in the Russian Arctic |
The mighty Revkin has a post up on pop-up boreal forests in the Arctic:
“The speed and magnitude of the observed change is far greater than we expected,” said Prof. Bruce Forbes of the Arctic Center, University of Lapland, corresponding author of the paper. Adds Dr. Marc Macias-Fauria from Oxford University, lead author, “Previously people had thought that the tundra would be colonized by trees from the boreal forest to the south as the Arctic climate warms, a process that could potentially take centuries. But what we’ve found is that the shrubs that are already there are transforming into trees in just a few decades.”Sorting out what this means in term of the climate will take some time, I think. But in the first place, if you care about the future of human civilization, you should wince at another invocation of that near-mantra of modern climate science, We Thought It Would Take Centuries But It Is Happening Now (dot tumblr dot com).
Pet peeve: neither Revkin nor even the article's own press release give the title of the article cited. The press release, but not Revkin, gives us the first author on the paper, and Revkin, but not the press release, gives us the correct issue. Drumroll please: "Eurasian Arctic greening reveals teleconnections and the potential for structurally novel ecosystems." Full text free online. Why wouldn't you link to that? Or at least cite it by name?
OK, decline-of-journalism (and public relations) rant over. Getting down to brass tacks, is this likely to be a positive or a negative feedback to global warming? Revkin doesn't speculate, and the article doesn't say. But we can do a rough back-of-the-envelope calculation to get an idea.
There are all sorts of interesting and dramatic effects that pop-up forests have on the local hydrology, ecology, potential for human use and so on, but the big long-term climate impacts, at first blush, would seem to be a) Forests sequester carbon dioxide, and b) Forests capture a lot of solar radiation that snow-covered tundra reflects back into space. So which effect is bigger?
Locally, the net effect will be warming, because the effect of the increased solar radiation will be local, while the effect of reduced atmospheric carbon dioxide, a well-mixed gas, will be distributed across the entire world. This local warming could be seen either as a bug (even more warming in the most rapidly warming part of the planet, unlocking more permafrost carbon, melting more sea ice, losing more of Greenland's glaciers to the sea) or a feature (warming concentrated in a colder climate where few people live, (relative) cooling over the entire world where everybody lives.) But what is the absolute magnitude of the two effects?
Well, to take a stab at it, suppose we have a large amount of tundra --> forest, enough to cut CO2 by 20ppm.
Some figures:
1ppm CO2 = 2,130,000,000 tons of CO2
Mature tundra sequesters about: 60 tons/acre of CO2
Mature boreal forest sequesters about: 182 tons/acre of CO2
The difference between the net absorption of solar radiation by snowy tundra (94W/m^2) and "snowy" forest (which ends up still having a dark canopy) (445W/m^2) is 351W/m^2. The difference without snow is less, though still favoring the forest's absorption. I decided to ignore it, and concentrate on the snowy season, which I ballparked at half the year (0.5).
1 acre = 4,047 m^2
Surface area of the Earth: 5.1 *10^14
Let's suppose, then, that in the 21st century, that 20ppm proves to be the difference between 600ppm and the more preferable 580ppm. The difference in greenhouse gas forcing is easy to calculate (i.e., even I can do it) and comes from this equation:
[5.35 * ln(600/278)] - [5.35 * ln(580/278)] = about 0.18 W/m^2
That's nothing to sneeze it -- it's sizable. Now we need to know how much area would need to shift to the lower albedo (greater absorption) forest to sequester that much carbon:
[(2,130,000,000 tons * 20)/122 tons per acre] * 4,047 meters per acre = 1.413 * 10^12m^2
(1.413*10^12) * 351Wm^2 (the difference in radiation)/2 (only half the year, when the tundra is snowy) = 2.48 * 10^14W.
Since the CO2 force is expressed in W/m^2 over the entire earth, we need to convert the number above to that convention by dividing by the surface area of the Earth: (2.48*10^14)/(5.1*10^14) = 0.49W/m^2.
About 0.18 W/m^2 (of decreased greenhouse gas warming) < about 0.49W/m^2 (of increased absorption of solar radiation). By my crude calculations (and feel free to point to obvious errors and/or better sources in the comments) the rapid expansion of boreal forest in the Arctic will likely be a net positive feedback to global warming, driving up temperatures further.
"and feel free to point to obvious errors"
ReplyDeleteThe one that strikes me is that the W/m2 numbers were measured at midday in one season (presumably about Julian day 140-160 if I read the paper right). So, a correction is needed to compensate for nighttime and lower-insolation at non-midday times, as well as different seasons. This gets complicated because we're talking high North, so in midwinter, it is night most of the day. I'd suggest limiting the analysis to the two months when there is both snow on the ground, and high sun (so a factor of 6 rather than a factor of 2), and then dividing by 2 again to reflect the fact that even in spring the sun isn't always at high noon.
Another second order climate issue is that trees have different surface roughness than tundra, so will interact with wind... see, eg, papers on the climate effects of giant wind turbine installations (Wang et al 2010 comes to mind: http://www.atmos-chem-phys.net/10/2053/2010/acp-10-2053-2010.pdf).
-MMM
Not really an error, but I was curious about the sensitivity to the choice of 600 ppm. Starting at 650 and going down by increments of 50 ppm I get 0.167, 0.181,0.198,0.218,0.243,0.274.
ReplyDeleteSo stabilizing around 400 ppm the effect from the change in CO2 would be about 50% more effective.
-blueshift
"So stabilizing around 400 ppm the effect from the change in CO2 would be about 50% more effective."
ReplyDeleteTrue, and I thought about it, though I didn't talk about it in the post; you have to chose a baseline from which to subtract 20ppm. The higher the baseline, the less the effect of an absolute -20ppm. I chose 600ppm because on a business-as-usual path, we will certainly cross that line in the 21st century. And I suppose I'm pessimistic as to the prospect (politically pessimistic, not economically or scientifically) of very aggressive mitigation that would keep the final number very much below that number.
I mean, we're at 396ppm right now. Absent a political earthquake (which I hope for and am working for) when the power plants being brought on line today are ready for retirement in 30-40 years, you could be looking at about 500ppm. Carbon-cycle feedbacks will at that point probably take us the rest of the way.
"[T]he W/m2 numbers were measured at midday in one season (presumably about Julian day 140-160 if I read the paper right). So, a correction is needed to compensate for nighttime and lower-insolation at non-midday times, as well as different seasons."
You raise a good point, although I'm not sure you have it entirely right. The paper says "Air temperature and incoming and reflected short- and long wave radiation were measured every 30-seconds, and 10-minute averages were recorded on a data-logger (Campbell Scientific Inc., model CR10X)." The albedo recorded is from 4 hours in the middle of the day. Is the incoming radiation measured that way? I don't see any indication of that in the text.
"I'd suggest limiting the analysis to the two months when there is both snow on the ground, and high sun (so a factor of 6 rather than a factor of 2), and then dividing by 2 again to reflect the fact that even in spring the sun isn't always at high noon."
That seems a little excessive -- you want to ignore 10 months out of the year, despite the fact that the forest always captures more solar radiation than the tundra? I'll dig into it a little more, see if I can come up with better numbers.
You both raise good points. Thanks!