What’s Actually Happening in the Amazon, and How Our Cheap Outrage Is Making It Worse
By now everyone on the social internet has heard of the fires spreading across the Amazon. In a way, this growing awareness is good. There are deep and urgent problems that require our attention. Trouble is, #ActForTheAmazon is rooted in a deeply misinformed view of what’s happening, and most of our would-be advocacy is only serving as fuel for the fires.
To give a sense of the going narrative and where it’s gone wrong, we have this tweet from the President of France:
So, about this:
- That picture is a stock photo that’s at least 16 years old.
- The Amazon basin is on fire every year — to the tune of no fewer than tens of thousands of fires each season. (There hasn’t been a single year with fewer than 100,000 since satellite data collection began in 1998.)
- Most of these fires are set by ranchers and farmers to clear out brush and initiate new grass growth. That is, they’re not the product of climate change (though they are making climate change worse, and warmer temperatures do increase the chance of them turning into wildfires).
- Per NASA, the number of fires year-to-date is actually right about average for the past fifteen years. While August itself has been a bit worse, August numbers happen to have low correlation with annual totals.
- While important to carbon capture, the Amazon rainforest doesn’t produce 20% of the world’s oxygen, nor would it really matter if it did.
Now, this isn’t to say that nothing alarming is happening right now. Quite the opposite! But the real problems here are much messier than the headlines would suggest, and the identity of the bad guys less obvious.
But before we get into all that, let’s turn back to that curious “world’s lungs” bit for a minute.
[2021 EDIT: I’ve written an updated version of this story here. While I’m keeping the original intact as a reference point (with some light cosmetic editing), I’m shifting my corrections policy to the new one.]
Memes vs. Science
While the whole “the Amazon produces 20% of our planet’s oxygen” claim is oft-repeated (some say 28%), I couldn’t find an original source. But we know that it’s unscientific anyway, for two reasons:
- The Amazon makes up only 40% of the world’s tropical forests. As more oxygen surplus comes from algae than all forests combined (claims vary from 60-75%), the math just doesn’t work out.
- While it’s true that trees are pretty good at oxygen turnover, there’s a more basic fact in play here: mature forests are mostly oxygen and CO2 neutral.
Simplifying a bit, trees are just banks that take carbon deposits and give oxygen loans. For as long as the bank is open (i.e., for as long as the tree is alive), it takes more deposits and issues new loans. But trees don’t live forever, and mature forests tend towards net-zero as to trees living and dying.* Birth and growth are balanced by death and decline, and so the cycle goes.
Relevant to our point here, when a tree meets its end its carbon deposits are released via a combination of decay and combustion. Just so, the surplus oxygen it’d loaned out is largely offset by the consumption of termites and all the other little things that feast upon its remains.
So, at a very basic level:
- Carbon that’s locked into a bank is carbon that’s not floating about in the atmosphere
- Carbon in the atmosphere contributes to global warming, which is bad for a lot of ultra-sensitive life-supporting balances
- Planting new trees increases the amount of carbon locked into banks, which is good
- Reducing existing forests is equivalent to those banks opening their vaults, which is bad
(Getting technical, felling a tree doesn’t cause it to release all its carbon. You could use the wood as lumber and it would retain most of the carbon until that lumber itself rots or is burned. But the act of cutting the tree down often causes significant releases once you account for branches being burned or left to rot, and for the soil being trampled and displaced. Plus there’s the problem of degradation, where the remaining trees are weaker, and there’s accounting for the carbon released by the logging, and so on. But for the purpose of this article we’re going to focus on just burning, which releases all the carbon.)
You’ll note that none of this has much do to with oxygen. It isn’t that oxygen is unimportant. It’s just that a slight decrease there is (so far as we know) far less concerning than a corresponding increase in carbon dioxide. We’re in no immediate danger of running out of the former, and mature forests aren’t helping all that much on that front anyway. While we obviously don’t want to lose the trees we do have (as the system needs long-term equilibrium), the current fires in the Amazon are hardly imperiling our oxygen supply.
Now, what does matter in a more urgent sense is the carbon being released — which is something worth looking at in more detail.
(*While all forests tend towards neutrality over long enough timelines, there is evidence that many older forests are still banking net carbon / releasing net oxygen. Some can build incredible density, and are protected from die-off by cooler weather and plenty of rain. But that’s effectively just a buildup for a future carbon release. The best case is that the carbon makes its way to the forest floor and is left to be grown over, turning into a fuel like coal that will mostly hold that carbon until/unless we dig it up and burn it. But the problem is that our current warming is making it less likely that the carbon will get that far before being released. We’re seeing this in Canada now, where our forests are actually carbon-positive at present. The big concern in the Amazon is that continued deforestation and degradation will push the ecosystem over a tipping point, where hotter and drier seasons will continue to thin out what’s left, thus triggering a death cycle that turns the rainforest into savanna. This would have all kinds of catastrophic downstream effects.)
One reason this story has blown up is that a certain presentation of INPE’s (Brazil’s NASA) fire counts make this year look like an outlier, especially if you only compare 2019 YTD to the equivalent period in 2018.
Looking at the most popular chart making the rounds (taken from this BBC article):
INPE’s data goes back to 1998, and can be found here (and here). The site is in Portuguese, but the data is pretty self-explanatory. The first link is where the BBC got their chart from, and the latter shows unique fires per month and per year, with high/low/mean numbers at the bottom. The latter is more useful in my judgment, as 2013 happens to be an unfair cutoff (2012 was a pretty bad year for fires). That said, I still give the BBC credit for at least showing some historical context, as many outlets just reported the year-over-year change vs. 2018, which is quite misleading.
Anyway, looking at the numbers, INPE has this August at 42,061 fires as of the 25th, putting it on pace for a little over 52k. That would be worse than most of the past few Augusts, though quite close to 2012 (51k) and much better than 2010 (90k). But while this trajectory isn’t exactly great, there’s a low correlation between August and yearly totals, so it doesn’t tell us much.
Also, it’s worth noting that, without casting shade at INPE’s efforts, their data contrasts quite a bit with NASA’s. While this is tricky to adjudicate, my subjective bias is that NASA’s records are more useful, in that they have best-in-class everything, along with more political independence.
If we look at GFED’s numbers (based on NASA’s data), the picture going into August looks very undramatic:
And even the most current numbers (updated to August 24th) suggest that we’re still not far off 2016’s pace:
(Note: GFED retroactively updated their August numbers today without explanation. See the below screencap from a few days ago. That bottom number now reads 121,430. But any journalist writing about this prior to today would have seen YTD numbers that put 2019 behind 2016’s pace.)
Anyway, here’s the rub: 2016 ended up with 257k fires vs. 301k for 2017. We’re just getting into dry season, and data through August isn’t all that predictive as to annual totals. And even if August does cross the 100k mark (certainly possible based on the most recent revisions), 2019 YTD will be ~150k, which is still well behind 2010’s 228k. Concerning, yes. Outlier, no.
[Note: Acreage burned is a better metric than total fires, but that data lags significantly. GFED stats roll it into their annual emissions estimates, but those are only current as to 2015.]
Anyway, if the number of current fires isn’t the real issue, then what is?
Under New Management
Brazil got a new president this year, Jair Bolsonaro. While he’s not exactly the Brazilian version of Trump, their policies and general outlook do rhyme. Like Trump, Bolsonaro was elected on a “drain the swamp” sort of platform, where his main selling point was his claim that he’d reform the notorious corruption of his predecessors. Also like Trump, he’s not especially a fan of ceding his country’s economic interests to outside forces (nor to any domestic bureaucracy).
While there’s debate to be had as to whether a regulatory rollback of the policies protecting the rainforest will actually help Brazil economically on a longer timeline (even setting climate effects aside), what is certain is that said rollback is going to have the immediate effect of making ranchers happier and more profitable — who happen to be a major constituency for Bolsonaro.
As context, some data from the folks at Yale:
Cattle ranching is the largest driver of deforestation in every Amazon country, accounting for 80% of current deforestation rates. Amazon Brazil is home to approximately 200 million head of cattle, and is the largest exporter in the world, supplying about one quarter of the global market.
(Note: Those 80% and 25% estimates are mildly out of step with some other estimates I came across. But the first number seems no lower than 70%, and their market share is at least 20%.)
So, what’s been happening for a long time is that Brazilian ranchers clear out a bunch of trees to make room for pasture (using mostly-controlled fires as part of the conversion process), then use this land to raise cattle, some third of which they export, mostly to Asia, which makes them a lot of money compared to most other rural career options.
What seems to be happening now is that Bolsonaro is giving ranchers implicit (if not also explicit) encouragement to take this practice up a notch, as it’s an easy way of goosing cheap revenue while making a base aggrieved by regulation feel like the president is on their side. (Some also argue that he’s trying to displace indigenous folks and generally weaken international control over the rainforest so as to allow new highway and infrastructure projects — but these are harder elements for outsiders like me to quantify.)
Now, so far as this goes, this is indeed bad news. While we don’t need the rainforest for oxygen in any immediate sense, we do need all that carbon to stay in those trees, and for the rainforest to remain rainforest. Continued deforestation is something we really want to avoid. And it would be particularly tragic if we couldn’t, in that this failure would be the clear reversal of a trend.
(Source: Deforestation in the Amazon.)
The big questions:
- How much deforestation will Bolsonaro’s policies actually lead to?
- Is there anything that can be done to reduce said number?
In the first case, I’m skeptical that we have enough public data to know at this point. Like, his intent is clear enough. It’s entirely possible that September is going to see a 2010-esque number of fires, and it’s entirely possible that unusually dry weather will lead to many of those fires spreading. But it’s also possible that those numbers will rise rather modestly, and we won’t know until we know. (While we should be appropriately vigilant in tracking developments, the big thing to be avoided here is attacking Bolsonaro with a narrative that will let him turn to his base and say “fake news”, which is more or less what’s happening right now.)
In the second case, I think we have to take a hard look at the sorts of things the international community can do, particularly in terms of what’s likely to be productive vs. what’s likely to have the opposite effect.
Means of Address
Let’s go back to that Macron tweet again.
For those having a hard time making out the text on mobile, the heart of Bolsonaro’s retort:
The French President’s suggestion that Amazonian issues be discussed at the G7 without the participation of the countries of the region evokes a misplaced colonialist mindset in the 21st century.
Two things here:
- That message is going to sell well at home.
- He’s not entirely wrong.
There’s a complicated tension between national sovereignty and global interdependence. While all of us (not to mention the indigenous peoples living in the rainforest) clearly have a lot to lose with continued Amazonian deforestation, it isn’t likely that taking a “we will impose our will on you and your real estate” approach is going to help.
Setting aside all we’ve learned about combative interventions in recent years (Syria, North Korea, Iran, Iraq, China, etc.), just consider our standing when it comes to telling other countries who to elect, how to use their natural resources, and how to treat their indigenous populations. On which grounds are we expecting them to be persuaded? Especially when we’re the ones importing cheap palm oil and beef jerky and thereby profiting from the very behaviour we’re condemning!
This is a defining problem of the coming era. Developing nations are turning to the playbook we used to reach our own heights, and us telling them “but you can’t do that now” is not going to be especially compelling. We need a better class of argument, and a lot less hypocrisy.
A few things stood out as I was researching this story:
- Brazilian cattle exports are only worth about $7bn USD or so (if you include domestic consumption, it’s closer to $25bn).
- Most Brazilian cattle ranchers are brutally inefficient in how they use their land. But they need investment dollars to scale the best improvements.
- While lots of US food companies are getting pretty good at this whole meat alternatives thing, the industry could use more R&D capital.
This suggests to me that there are few things we could do:
- Install a carbon tax to fund education and infrastructure grants for ranchers, with additional incentives if they manage production without cutting trees or starting uncontrolled fires. (There’s something poetic and just about using the wealth we’ve accumulated from centuries of the same sins to enable others to take a better path to their own economic advancement. In contrast, the existing system is tantamount to pulling up the ladder behind us and offering trinkets and platitudes for those left behind, which isn’t likely to win them over to outcomes we want.)
- Shift more global eaters to lab-grown or plant-based alternatives that don’t suck. For my own part, I’m going to start observing Meatless Mondays. I don’t have the desire or will (yet) to go full vegetarian/vegan, but doing this writeup has compelled me to take at least one meaningful step. (I won’t be religious about it. Some weeks I might swap days. But giving up one day a week feels like the right mix of manageable and meaningful.)
- Be more selective about our news consumption. A lot of the clickshare reporting here has fed a narrative that’s enabling Bolsonsaro’s worst instincts and riling his base. This has led to lots of well-intended people sharing messaging that can only work against our collective interests.
All said, while the global alarm about the fate of the Amazon is reasonable, most of our worry thus far has been ill-directed. If we want to save the Amazon, we have to give positive incentives to a lot of farmers and voters while not pushing them further away.
As ever, creative incentives move the world more than moralizing ever has or can.
(I first published the bones of this story on Quora on August 23rd, 2019. I tend to pull this type of writeup together when I’m dissatisfied with coverage of a trending news story. While I don’t make any money off them, I do offer cash for corrections, just to make sure that any mistakes are caught and corrected quickly. If this is your thing, see me on Twitter or Quora.)