Scientists: “There are multiple, consistent lines of evidence from ground-based studies published in the peer-reviewed literature that Amazon forests are, indeed, very susceptible to drought stress.”

Major amplifying carbon-cycle feedback is not a “myth”

Up to 40% of the Amazonian forests could react drastically to even a slight reduction in precipitation; this means that the tropical vegetation, hydrology and climate system in South America could change very rapidly to another steady state, not necessarily producing gradual changes between the current and the future situation.

This statement in the 2007 IPCC is “basically correct but poorly written, and bizarrely referenced,” as tropical forest researcher Simon Lewis told the BBC in January.

That didn’t stop the anti-science blogosphere from spinning this into another phony “gate,” as ClimateSafety explained in an excellent post, “AmazonGate: how the denial lobby and a dishonest journalist created a fake scandal.”

Recently, the anti-science crowd, from FoxNews to Anthony Watts, has been crowing about a new study that supposedly shows the IPCC paragraph was wrong.  But a major statement by 19 top U.S., U.K., and Brazilian scientists who “conduct research on Amazon forests, climate, and/or fire,” thoroughly debunks that notion:

[Citations for original sources can be found here.]

Scientists’ statement on recent press release on Amazon susceptibility to reductions in rainfall: no Amazon rainforest “myths” have been debunked.

The press release from Boston University describing a recent article in the journal Geophysical Research Letters by BU researchers on the response of Amazon forests to the 2005 drought is misleading and inaccurate. It claims that the study “debunks myths about Amazon rainforests”, which is simply not true. First, there is no myth. Rather, there are multiple, consistent lines of evidence from ground-­”based studies published in the peer-­”reviewed literature that Amazon forests are, indeed, very susceptible to drought stress. Second, nothing is debunked by the new study. The new study contributes to our understanding of interpretations of data retrieved from satellites, but it does not prove or disprove anything about what is really happening on the ground. The BU press release also claims that the new BU paper demonstrates that the IPCC statement about the sensitivity of Amazon forests to small reductions in rainfall is inaccurate, which is also not true. While the IPCC statement could be criticized for citing a review paper rather than original research papers, the main conclusion of the IPCC statement – that Amazonian forests are very susceptible to reductions in rainfall – remains our best understanding of the data available at the time of the IPCC report and also today.

The article published by the BU group (in contrast to the BU press release) makes a scientific contribution to our knowledge of Amazon forests. It presents new analyses of the forest canopy conducted using satellite data from the MODIS sensor. The article challenges the findings of a previous analysis of forest response to the 2005 drought using similar data from the MODIS sensor. This earlier study, published in Science in 2007, concluded that southwestern Amazon forests fared well during the severe drought of 2005, reporting that these forests were greener in 2005, not browner as would be expected if the forests were stressed by drought. The new study found that the forests fared neither better nor worse, as indicated by the color of the canopy as seen from satellite images during the 2005 drought. Scientists are likely to continue to debate the differences in their analyses of the satellite imagery, and the articles in question illustrate the scientific learning process as we explore the potential and the limitations of satellite-­”based measurements to give us information about forest response to drought in the Amazon region.

Forest tree measurements made under the forest canopy following the 2005 drought provide a very different picture of the sensitivity of Amazon rainforests to drought. In tree inventories conducted in 55 long”term forest plots scattered across the Amazon forest, the drought of 2005 was associated with a large surge in tree mortality and no gains in growth. These findings, published in the journal Science in 2009, are consistent with the results of two large-­scale experiments, in which large canopy trees began to die after three years of experimentally reduced rainfall. The forest plot results are also consistent with studies of historical rainfall and soil water storage capacity and with simulation model analyses. These studies, published in some of the best peer-­reviewed science journals, provide several consistent lines of evidence that the forests of the Amazon Basin are susceptible to small reductions in rainfall. We do not know why the drought stress and tree mortality documented in the field studies published in the 2009 Science article and predicted based upon rainfall patterns were not detected in the analyses of satellite images by the Saleska-­ and Samanta-­led teams. It could be that tree deaths, which affect only a portion of the tree canopy, are hard to see in satellite images, especially if this tree death is accompanied by the growth of vines and plants on the forest floor. It could also be that the tree mortality induced by drought was sufficiently  delayed to be invisible in the imagery of 2005. This should be the topic of further research.

Reductions in rainfall can affect Amazon forests by increasing tree mortality, but also by increasing their susceptibility to fire. The initial fire kills trees, increasing the likelihood of subsequent fires for years afterwards in a vicious positive feedback loop. In 2005, more than 2000 km2 of forest caught fire in the tiny state of Acre alone. During the severe drought of 1998, approximately 40,000 km2 of forest caught fire. These are indisputable facts. It is important to remember that these droughts are part of the current Amazon climate regime. If climate change increases the frequency, severity or duration of these episodic droughts, then increased forest fire and tree mortality and reduced river flow are the likely results.

The IPCC must be held accountable for the best scientific information that is available in the peer-­review literature at the time of its writing. The passage in the IPCC that refers to the susceptibility of the Amazon forest to drought cites a World Wildlife Fund review report which, in turn, cites an article in the journal Nature. Ideally, the IPCC should have cited the Nature article as well as several other existing articles in support of its statement, and not a WWF report. The point is, however, that the statement made by the IPCC about the sensitivity of Amazon forests drought was consistent with our knowledge at that time, and has been reinforced by new studies.

You can find the signatories here.

You can read a shorter statement by Dr. Lewis on RealClimate.  There is an interesting back and forth between him and Dr. Samanta on RC (here) :


The press release accompanying the GRL article disputed the following IPCC AR4 (2007) claim –

“Up to 40% of the Amazonian forests could react drastically to even a slight reduction in precipitation; this means that the tropical vegetation, hydrology and climate system in South America could change very rapidly to another steady state, not necessarily producing gradual changes between the current and the future situation (Rowell and Moore, 2000). It is more probable that forests will be replaced by ecosystems that have more resistance to multiple stresses caused by temperature increase, droughts and fires, such as tropical savannas.”

for two reasons: (1) this is presented as the consensus view by quoting Rowell and Moore, 2000. (2) There was more than a slight reduction in precipitation during the third quarter of 2005 and, most of the drought-impacted forest area for which we have uncorrupted satellite greenness data showed no enhanced or reduced greenness levels (third quarter average EVI values) as compared to non-drought years (between 2000 and 2008).

It is only in this context that the material in the press release and the GRL must be understood. We do not dispute any other results related to this theme in these two documents.

Arindam Samanta (on behalf of the authors of the GRL papers).

Dr. Lewis replied as follows:

Dear Arindam,Thanks for the response.

On the IPCC statement, as I have said it is not as well-worded as it ought to be. Strictly, perhaps it can be taken as having one of two different meanings,

1. That the IPCC mean that small reductions in precipitation at any given time cause a drastic response (of which your paper ably shows that for satellite-monitored ‘greenness’ there is no such drastic response, and is an important paper I will certainly cite), or

2. They mean responses of vegetation to mean climate regimes with differing precipitation (of which your paper says little).

It seems clear to me that the sentence is about responses to a shift from one climate regime, the recent past and present day, to another, with less precipitation, in the future (it is the IPCC climate change impacts report after all, and they do say ‘… not necessarily producing gradual changes between the current and the future situation’).

If meaning two of the sentence is taken, then what the IPCC say is reasonable, defensible, basic science: warm lower-rainfall environments tend to be dominated by savanna, while warm higher-rainfall environments tend to be dominated by rainforest, with a threshold amount of rainfall separating which vegetation type one finds. If substantial areas of the Amazon are in a climate regime close to the savanna-rainfall threshold, which diverse evidence suggests they are, then there may be a vegetation shift if rainfall consistently decreases in the future due to climate change.

Your response implies you think meaning one is correct, which is mistaken (logically it can’t hold as a proposition). Had your paper cited the IPCC chapter and the sentence you object to and why — which it doesn’t — the misunderstanding could have likely been addressed at the review stage.

Details aside, it’s the ‘debunking Amazon myths’ headlines, and quotes about putting right ‘muddled understanding’, and, “The way that the WWF report calculated this 40% was totally wrong, while [the new] calculations are by far more reliable and correct,” that are problematic and have unnecessarily confused people. There are no calculations in the WWF report (it’s a review), nor are there any new calculated updates on the IPCC ‘up to 40%’ statement in the Samanta paper, and the ‘muddled understanding’ quote highlights the ‘twin pressures’ facing the Amazon, as logging and climate change, when outright deforestation is certainly the number one current pressure in the context of the quote.

I know the media regularly run out of control (its happened to me several times), but in my view it is critical to try and put things right. Most journalists and bloggers will help put things right once they know there is a problem, but you have to tell them.

With best wishes,


Well, some journalists and bloggers will help put things right.  That’s why I’m posting all of this here.

Let me also quote from Dr. Lewis at RC:

Here is a recent paper that is consistent with the IPCC statement:

This provides a way into lots of the literature. Also, see Lewis, S.L. (2006) Tropical forests and the changing earth system. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences, 361, 195-210. Available from this page:

The first paper Lewis sites is a “Towards quantifying uncertainty in predictions of Amazon ‘dieback’,” by Huntingford et al. which concludes:

Simulations with the Hadley Centre general circulation model (HadCM3), including carbon cycle model and forced by a ‘business-as-usual’ emissions scenario, predict a rapid loss of Amazonian rainforest from the middle of this century onwards….

We find that the loss of Amazonian rainforest is robust across the climate uncertainty explored by perturbed physics simulations covering a wide range of global climate sensitivity.

The “Amazon dieback” paper notes:

Besides acting as a positive feedback on climate, whereby additional carbon dioxide is released back into the atmosphere, the loss of the rainforest in itself would clearly be a significant environmental matter.


I’ll write about the meaning of that feedback in another post, but for now, let me repeat what I wrote in “Science stunner: Vast East Siberian Arctic Shelf methane stores destabilizing and venting“:

It is increasingly clear that if the world strays significantly above 450 ppm atmospheric concentrations of carbon dioxide for any length of time, we will find it unimaginably difficult to stop short of 800 to 1000 ppm.


11 Responses to Scientists: “There are multiple, consistent lines of evidence from ground-based studies published in the peer-reviewed literature that Amazon forests are, indeed, very susceptible to drought stress.”

  1. There was also an important World-Bank funded study with top-level science review panel headed by Tom Lovejoy that found the combination of fire, climate change and deforestation has pushed the Amazon close to an irreversible decline. I covered it for IPS at biodiversity meeting in Paris in Feb – don’t know if anyone else did.

  2. catman306 says:

    Should chain saws should be registered, like automobiles and chain saw operators should be licensed?
    Make chain saws seizable contraband? Maybe.
    But then they’d just use bulldozers.

  3. mike roddy says:

    This is a difficult area, since predicting an ecosystem’s reaction to future warming is based on inference from current studies. As in arctic melting and other biological feedbacks, the future is projected as a series of probability bands. So far, however, effects and feedbacks have all been on the high side of IPCC projections.

    Forest reactions to increased localized microclimate heat and drought stresses are well documented, even with basically constant global temperatures. I observed this as a river guide and outdoorsman in Southern Oregon during my youth. Large acreages of clearcut conifer forest shifted to oak savannah, just as occurred in parts of Greece and China after deforestation. The reason was reduced rainfall and cloud cover due to the absence of tree transpiration.

    A tree planter friend from those days, Michael Mews, told me about units that were replanted six times or more, and the forests still never returned. Major parts of Southern Oregon and Rocky Mountain conifer forests may be well on the way to entirely different ecosystems, which will be far less diverse and lacking CO2 sequestering forests.

  4. Leif says:

    Mike Roddy, #3: A few years ago I came across some research that stated that the mere presence of a forest tended to transport rainfall further inland than would be the case without the forest. The hypothesis was that the increased surface area of the forest would increase evaporation which would condense out again over and over. Without the forest rain fall would be concentrated and run off the land faster with far less going to absorption as well as spread over a smaller area. It all seamed intuitive to me once I read the report.

  5. It’s the feedback loops that worry me most, as negative trends will continue to worsen at a rapid pace once certain conditions are met, and at that point, it becomes extremely difficult to reverse the trends. As this post highlights, it’s a real danger for the Amazon, but this dynamic is present throughout the planet, from forests to oceans.

  6. CM says:

    Typo: “Dr. Leeds”. His name is Lewis. Leeds is his university.

    [JR: Doh! My apologies. And I can’t even blame that on my voice dictation software!]

  7. Andy says:

    It wasn’t bizarrely referenced. It was lazily referenced. Instead of citing the numerous articles referenced by the WWF report, the IPCC cited the report instead.

    [JR: You say tomato…. I personally would have kept the gray literature out of the AR4 entirely.]

  8. Rabid Doomsayer says:

    Fare well the Aspen. SAD is it not?

    One species is getting more coverage than the whole Amazonian ecosystem.
    No ecosystem can survive a multidecadal drought intact.

  9. Eli Rabett says:

    Looking at the list of signers, it is clear that Myron Ebell’s cute little blackmail attempt is bearing bitter fruit

  10. Wit's End says:


    The trees lose their leaves and the nasty vines get more light and climb. So the canopy looks green.

    When I first walked through the forests in California several decades ago, they were dense and dark, the canopy blocked out the sun. Look at this picture of them now:

  11. Raleigh Latham says:

    It’s almost unimaginable, but absolutely certain that every nation on earth will be in a TOTAL WAR fighting climate change not far down the road.