Bishop Hill Nepstad enters the fray
Jun 29, 2010 Climate: WG2 Much fun in the comments to Monbiot's Amazongate posting, with Daniel Nepstad, the activist-scientist behind some of the IPCC claims, having his say:
As the lead scientist on the research that underlies the IPCC statements about the sensitivity of the Amazon forest to reductions in rainfall, and after 25 years studying this question, I can say that the evidence has only grown stronger in support of this statement. I ran an enormous rainfall exclusion experiment in an Amazon forest that identified the rainfall threshold beyond which giant forest trees die quite suddenly (published in the journal Ecology in 2007). During the 2005 Amazon drought, tree mortality spiked up in permanent forest plots across the region (Philips et al. 2009 Science), providing further evidence of the drought threshold. The critics have latched onto two papers that seem to contradict our results, both using the same satellite sensor (MODIS). The forest canopy appears to get a bit greener in some Amazon regions during the dry season. Deeper analyses of the same data have found that these studies probably were seeing leaf-changing episodes and changes in cloudiness (which declines in the dry season) which are not evidence that the forests were not drought stressed. In a recent letter signed by 18 scientists including many of the world's authorities on tropical forest response to climate change, we found the IPCC statement to be sound and the NASA study involving MODIS data to be irrelevent to the IPCC statement. I would be happy to explain the science behind the IPCC statement.
Which is fascinating but still leaves us none the wiser as to where the idea that 40% of the Amazon is at risk from slight changes in rainfall comes from. Several commenters on the thread call Dr Nepstad on this evasion and demand to know in which paper it can be found.
No reply last time I looked.
Reader Comments (44)
"In a recent letter signed by 18 scientists including many of the world's authorities on tropical forest response to climate change, we found the IPCC statement to be sound..."
Is that how it's done these days? A few people sign a letter and from that one can infer that a fact is proved? Or does he mean that 18 people of like mind signed a letter saying they think something is so and from that we should accept that it is so?
I'm a bit of a fusty old fuddy-duddy when it comes to these things, if you want to prove something you have to provide evidence. It's a standard that has stood the test of time. Any evidence chaps? Yes, I hear a squeak, speak up at the back, I can't make out what you're saying. Oh, you said "computer model". Thank you for that contribution. Now, any evidence chaps?
This was the bit that puzzled me
Did this involve an enormous umbrella to keep the rain off the trees and pipe it away somewhere else ? Plastic sheeting ?
I am genuinely intrigued. The word 'experiment' suggests a real-world project to keep rain off some part of the forest. Any ideas ?
Jack
I believe that it was exactly that - a big plastic sheet or something of that ilk.
Yes, it's here - plastic sheeting to cover a hectare of forest for 4 rainy seasons cutting out 60% of the rain:
http://www.whrc.org/resources/publications/pdf/NepstadetalEcol.07.pdf
Yes it's here:
http://www.whrc.org/resources/publications/pdf/NepstadetalEcol.07.pdf
"A severe, four-year drought episode was simulated by excluding 60% of incoming throughfall during each wet season using plastic panels installed in the understory of a 1-ha forest treatment plot, while a 1-ha control plot received normal rainfall. "
Fair play to Nepstad - he did the experiment and gathered the data. Real data.
I'm not sure if his results support the 'tipping point" / "threshold" idea. It's more like "some of the trees started to wilt then some started to die". I wonder what the null hypothesis was: "everything will be hunky-dory if we cut 60% of the rain in a rain forest" ?
The article includes the word "may" 19 times.
(published in the journal Ecology in 2007).
It would be nice if he gave a link and said who peer reviewed the paper. Then again, thats par for a agw commenter in the Guardian
June 29, 2010 | Unregistered CommenterJack Hughes
Thanks for the link.
I reading on the various sites yesterday and I seem to remember a graph (I am sure it was over at WUWT) showing that rainfall in the Amazon region had not changed for a long time. In fact, to my eyes it seemed to increase slightly.
To be honest, when I see the statement like "As the lead scientist on the research that underlies the IPCC", I tend to really check the persons credentials!
Jack Hughes,
A minor point, he did not do 'THE' experiment, he did 'an' experiment, a simulation. What it modelled and whether it had anything to do with reality is open to question and discussion.
The letter by the eminent Amazonian scientists (they do live signing things and publishing them don't they?) is signed by
Daniel Nepstad, Woods Hole Research Center, USA, and Instituto de Pesquisa Ambiental da Amazonia, Brazil; Ane Alencar, Instituto de Pesquisa Ambiental da Amazonia, Brazil, and University of Florida, USA; Greg Asner, Carnegie Institute of Science, USA; Alessandro Baccini, Woods Hole Research Center, USA; Paulo Brando, Instituto de Pesquisa Ambiental da Amazonia, Brazil, & University of Florida, USA; Foster Brown, Woods Hole Research Center, USA & Universidade Federal do Acre, Brazil; Mercedes Bustamante, Universidade de Brasilia, Brazil; Eric Davidson, Woods Hole Research Center, USA; Scott Goetz, Woods Hole Research Center, USA; Richard Houghton, Woods Hole Research, USA; Michael Keller, National Ecological Observatory Network, USA; Simon Lewis, Leeds University, UK; Thomas Lovejoy, Heinz Center for Science, Economics and the Environment, USA; Patrick Meir, University of Edinburgh, UK; Paulo Moutinho, Instituto de Pesquisa Ambiental da Amazonia, Brazil, and Woods Hole Research Center, USA; Carlos Nobre, Centro de Previsao de Tempo e Clima—CPTEC, Brazil; Scott Ollinger, University of New Hampshire, USA; Oliver Phillips, Leeds University, UK; George Woodwell, Woods Hole Research Center, USA.
So 60% exclusion of rainfall equates in someone's 'mind' with a slight reduction of rainfall?
Since this was a rainfall 'exclusion' experiment - not a small 'reduction' in rainfall experiment we can conclude.
Conclusion: No rain = no Amazonian rainforest.
We also know form various reports other conclusions.
Conclusion: Logging = no Amazonian rainforest.
Conclusion: Farming = no Amazonian rainforest.
So where is the data that strongly suggest that slight changes in climate, a small reduction in rainfall, equates to a 40% reduction in rainforest?
It appears that we can conclude there are none because there are no scientific references to hand.
Perhaps Daniel Nepstad should use buckets instead of plastic sheets to simulate rainfall reduction scenarios.
http://www.youtube.com/watch?v=MAfCQ-t7xY0
Anyone know if this 'experiment' has any baring with what's happening in the Amazon at the moment?
Is 60% rain exclusion a slight drop in precipitation?
Can this 'experiment' be applied to the whole of the Amazonian Basin?
The one problem i have with this new 'proof' that the IPCC 'got it right' is that the Last IPCC report came out in 2007. Daniel Nepstad published a paper in 2007 in the journal Ecology.
Is it just me or is this another paper published after the cut off for the IPCC report that is now being used after the date to show that the railway engineer and his pals didn't balls it up again!
Nepstad is dealing in worst case scenarios with his studies. Severe drought, severe logging, severe fire damage, etc. That is why he is claiming 'high' sensitivities.
Cutting the whole forest down is related to a high sensitivity in losing trees .......... well we all kinda expected that.
The IPCC AR4 is completely different, it is claiming that a small reduction in rainfall leads to large losses in rainforest (40% is the figure quoted). There is no science at hand to back this statement, none what-so-ever. All the references have turned out to be either highly misleading or completely bogus.
The thing is Daniel Nepstad is actually supporting the IPCC position when he knows that even his own science doesn't support this - he hasn't done any small reduction in rainfall experiments.
No genuine scientist would stick their necks out on such determinable matters without supporting data.
My suggestion is for Nepstad to return to the rainforest with several thousand plastic buckets - do the rainfall reduction experiments to determine rainforest sensitivity to small changes in rainfall.
@ Jack Hughes
You have a good point about the use of certain words in articles and papers i.e. "may". As an engineer by training and over the years writing many specifications for equipment suppliers, the wording in the spec has specific meanings and this is spelled out at the front of the specification. For example if the word "shall" is used it means the supplier must and will do exactly what that part of the spec requires, whereas if "should" is used it means there is some scope for deviation or alternatives. This definition is clearly stated so everyone knows the requirement.
It seems to me that the scientific fraternity ought to define the uncertainties more clearly and that their written word should reflect these uncertainties. Perhaps their papers should follow an engineering discipline for once ?
As an engineer, covering a 1 ha area with plastic sheeting then extrapolating, does not work for me. I searched the paper for "wind", "convection", "humidity", "currents", "pollination".
Surely you are effecting the natural processes of a rain forest by covering it in plastic panels at the height of the understory. Whilst it may be an "experiment", it is a "quantum" experiment in that you have altered the habitat by reading it.
Extrapolating it to the whole of the rainforest does not work me. But I am an engineer so what do I know?...
MactheKnife: I had the same issue, during the early 80s the new input of engineers didn't seem to have a grasp of the nuances of the English Language, and I used to tell them that "shall" is the future imperative, and "will" is the future hopeful.
I thought he'd run a simulation on his computer, it appears not, he's done a real life experiment, but not into minor changes of precipitation. It's like herding cats trying to get these guys to be accurate in their speech
http://news.mongabay.com/2010/0204-amazongate.html
It is a good picture of the Nepstad experiment.
MacTheKnife: The scientific community is well aware of how words ought to be used. The IPCC, for some reason, tries to slip in its things here and there, under the assumption that non-climate-scientists don't know how to read their reports. By non-climate scientists, I mean the rest of the whole world.
No wonder, they are so surprised that anyone can come asking questions at all. They seem to forget that the rest of humanity has been doing 'science' for hundreds of years. ;)
Perusing the said ecology article one notes that Nepstad states the test area was similar to one third of Amazonian forests then the qualifiers start. Due to the depth of groundwater(100meters) Plant Available Water was representative of only 10% of forests. A further qualifier was that the rainfall simulated was representative of only 5% of Amazon forests. I suppose if one allows a margin for error it probably represents the odd tree or two.
Nepatad comments:
"During the 2005 Amazon drought, tree mortality spiked up in permanent forest plots across the region (Philips et al. 2009 Science), providing further evidence of the drought threshold"
The 2005 drought was considered a once in 100 year event perhaps this explains the spike. Philips published a paper in 2004 relating to the aforementioned forest plots where he found recruitment exceeded mortality. As they say its all in the timing. Philips conclusion was most astonishing, it seems trees thrive on increased sunlight and elevated CO2 levels. I
Regarding those eighteen signatures, six belonged to members of an NGO he co-founded as well as scientists from the WHRC where he alternates employment with the WWF and the Moore foundation (a major sponsor of the WWF) I urge to all to peruse his 2007 WWF publication conservatively titled "Amazons Vicious Cycles". It seems 55% of Brazilian forests will be destroyed within 20 years (17 now) without factoring in AGW.
I recommend Nepstad's post at WUWT its inspiring.
It's all very well showing that a lack of water kills trees (duh!) but the one thing we keep hearing over and over again is that under every model scenario wet areas will get wetter and dry areas will get drier*. Since the Amazon satellite data shows growth then all the pessimists have absolutely nothing to back them up, except perhaps a gut feeling that change is always bad.
*Mind you since the Sahara is shrinking due to extra rainfall that is isn't matching too well with reality.
They should adopt "the data is irrelevant" as their slogan.
JamesG said: "It's all very well showing that a lack of water kills trees (duh!) but the one thing we keep hearing over and over again is that under every model scenario wet areas will get wetter and dry areas will get drier*"
A comment on WUWT pointed out the IPCC also project less rain during dry seasons and more rain during wet ones. The IPCC don't even appear to be saying the rainforest will get drier, it'll just get more polarised between wet and dry. Neptstad's experiment cut off rain all year round so it was bound to suffer - a reduction from 8 or 9 feet of rain a year to nothing. That is a world away from what the IPCC have claimed.
This is the consensus letter Nepsted talks about.
https://docs.google.com/viewer?url=http://www.edf.org/documents/10903_Scientists_speak_no_myths_debunked.pdf
It doesn't remotely say what he pretends. Really all they say is that instead of the Amazon getting greener during the 2005 drought, the new study concluded the canopy was "neither better nor worse". Notably Steve Running, who is responsible for the MODIS interpretations was not on this "consensus" so he obviously disagrees.
So:
1. Rainfall hasn't dropped at all.
2. Amazon got greener over last 30 years by satellite data.
3. During very severe drought the Amazon seems to still get greener, or at least is no better, no worse from satellite data.
4. Nepsted proposes that counting a small area of dead trees and extrapolating hugely over the entire Amazon is somehow better than using satellites to measure the greenery.
5. 18 Scientists sign a letter saying in effect "we don't know". The MODIS man doesn't sign it - (because he does know?)
6. none of which matters anyway, because all the models say the Amazon should get wetter anyway.
Note Nepstad's affiliation. This is not the famous Woodshole Oceanographic, this is more of an advocacy outfit.
Woods Hole Research Center
From Wikipedia, the free encyclopedia
Not to be confused with Woods Hole Oceanographic Institution.
The Woods Hole Research Center addresses pressing environmental issues, including climate change, through scientific and policy initiatives. The Center has projects in the Amazon, the Arctic, Africa, Russia, Alaska, Canada, New England, and the Mid-Atlantic, working in collaboration with a wide variety of partners ranging from NGOs to governments and the United Nations.
The Woods Hole Research Center was established in 1985 in Woods Hole, Massachusetts by George Woodwell. In 2005, Dr. John P. Holdren became the director. Holdren was appointed as President Obama's science advisor in 2009. William Yancey Brown became the Center's third director in February 2010. Currently, the organization has about 55 staff members.
Holdren is one of the great eco-hysterics of all time and an acolyte of Paul Ehrlich, we're all doomed,
so kill'em anyway.
So - there is an 'experiment' which 'proves' something other than the IPCC claim.
Why do these guys simply get a free pass on lying?
This is what happens when 'activism' is valued more highly than honesty. (As evidenced at the WWF, see below):
http://wattsupwiththat.com/2010/01/24/the-scandal-deepens-ipcc-ar4-riddled-with-non-peer-reviewed-wwf-papers/#comment-298859
I suppose practical people can take some solace in the fact that if and when the activists have sunk the world's economies - they will suffer most - as there will be no one to tax - and the practical people will be slightly more adept peasant farmers.
And let's hope the future generations of peasant farmers are not too enamored of the succeeding activist religions - aimed at ensuring the season's crops.
The question is how rainfall is distributed in warm periods. The Amazon may get less precip but other places receive more. My recollection is that during previous warm periods that the Sahara was a savanna.
I will wager that during the next drought period, the WWF and the HOLE institute will despatch their minions to to dig trenches around the trees in order to cut off PAW and ensure that this prediction comes true.
Its all if's and maybe's.
Cutting off 60% of rainfall would kill most things, especially people.
I replicated the experiment. I turned off the watering system for my flower planters and excluded water from them for six weeks. Then I turned the water on for four weeks. That is a 60% reduction -- isn't it? Well, the results were unequivocal: The plants were all dead!
Maybe I should publish it and get Nepstad to review the article. Then maybe the IPCC will give me a $1 million grant.
It seems that Nepstad experiment was indeed large, and no doubt required a huge investment in funds and effort. It should be praised for collecting real data as opposed to virtual, simulated substitutes.
However, Nepstad experiment does not seem to pay any attention to the impact of the experiment itself. The area alteration is considerable and the impact far from negligible. A proper control plot should have been identical but with permeable panels instead of waterproof ones.
An hectare is just slightly larger than a large soccer field. Imagine 5600 persons entering the field to install a panel (or 1 person 5600 times), plus trenches, towers, instrumentation, etc.
We have no information on how this affected plant mortality.
We are told that precipitation ranges from 600 to 3000 mm with an average of 2000 mmm. The average precipitation on the dry plot was 2200,1800,1100,1200 for the first 4 years of the experiment. Thus within natural variability and larger than the natural minimum.
I do not understand how tree and palm mortality increased slightly (fig.4, from about 7% to 9.5% and 2.7% to 3.2%), but mortality of plants within stem class larger than 30 cm in diameter increased 445%. Which plant with a stem diameter larger than 30 cm is not a palm or a tree?
Anyway, with present IPCC models ability to simulate current precipitation we don't need to worry. Global errors are of the order of 80%, regional errors in excess of 400%, so who cares....
Nepstad paper
The Epstad paper does not support the IPCC statement: ""up to 40% of the Amazonian forests could react drastically to even a slight reduction in precipitation"
The paper's abstract says "A severe, four-year drought episode was simulated by excluding 60% of incoming throughfall during each wet season using plastic panels installed in the understory of a 1-ha forest treatment plot, while a 1-ha control plot received normal rainfall. After 3.2 years, the treatment resulted in a 38% increase in mortality rates across all stems >2 cm dbh."
i.e. an enormous, sustained, reduction in rainfall has an effect on a cherry picked (why >2 cm? - did the smaller plants appreciate not being soaked?) plant life.
The Nepstad paper doesn't seem to describe the steps that were taken to insure that the temperature and humidity of the plots were identical, or that the umbrellas, and scientists didn't prevent pollination from occurring (for example). So although it was undoubtedly a large effort - they haven't troubled themselves seriously about anything other than the 'expected' outcome. And of course, if the rainfall were to decline because of a CO2 increase, the experimenters should also increase the CO2 levels - which might compensate for any rainfall effect.
Amusing to see that positive feedback (where has that been mentioned before) must be invoked at the end of the paper to keep the alarm bells clanging.
Additional information on the rigor of the science involved can be found here:
http://en.wikipedia.org/wiki/Trofim_Lysenko
Patagon,
Has Nepstad performed more than one experiment? The one you have described has an artificial reduction in rainfall of 60% sounds remarkably similar to the one described in The Amazon's Vicious Cycles: Drought and Fire in the Greenhouse, with one substantial difference:
Regardless of whether there was one or more tests what is plainly apparent is that both experiments represent a substantial reduction in precipitation.
Just a small note: if you haven't seen a rainforest, it's hard to understand the paper. I can testify from personal experience that it is too far removed from most people's experience for them even to understand quite how thoroughly they don't understand the difference between it and their home environment. There's a tendency to assume that rainforest is just a bit of normal woodland turned up to 11, but that would only be true on a log scale.
I would suggest that unless you saw the experiment, or know about rainforests, you're probably not qualified to nitpick. For instance, the question about how a plant can have a stem over 30cm without being a tree seems laughable if you've ever been in a rainforest - they have *grass* that meets that description. The arguments about 'roofing in a hectare of forest' are probably wrong as well - I read 'restricting 60% of rainfall for x years' as 'restricting 100% of rainfall over 60% of the area', because anything else is ludicrously impractical.
Nepstad's work is reasonably sound, in my view. It's just irrelevant to climate change debates. For once, I don't really feel too bad about his views being twisted like this (and him going along with it). Unlike climate change, Amazonian rainforest loss (through logging, etc) is a genuine and immediate threat, and any publicity that attracts attention is good publicity.
Gareth,
No doubt there was a substantial reduction in precipitation. It is the same experiment, as you see from the references in the paper I link. The problem is that the experiment is a major disturbance, not only in precipitation.
dave,
I know well not one but several tropical forest. I have never seen grass with a 30 cm DIAMETER stem there. I still do not understand the numbers.
geronimo wrote:
"I thought he'd run a simulation on his computer, it appears not, he's done a real life experiment, but not into minor changes of precipitation. It's like herding cats trying to get these guys to be accurate in their speech."
"Herding cats"?!! Nah ... never heard of a "catherd"! More like nailing jello to a wall, methinks. But on second thoughts, why should we expect these guys to be any more accurate in their "speech" than they are in their "science"?!
No doubt they are merely exercising "scientific licence" in order to compel their readers to engage in willing suspension of disbelief (how else to explain the acceptance of their definition of "trick",eh?!)
@dave
'any publicity that attracts attention is good publicity'.
Does that extend to misrepresenting the 'science' and the conclusions drawn? We 'hard' scientists have a word for that. We call it 'lying'.
And I'd love to see a picture of grass with a stem about a foot (30 cm) in diamater!
@dave
'I read 'restricting 60% of rainfall for x years' as 'restricting 100% of rainfall over 60% of the area''
which is probably what he did. And surprise surprise he discovered that if some part of a rain forest gets no rainfall at all for 1400 days, some of the plants that are used to regular soakings die (I've never been there but I assume its called the rain forest for a good reason)
Wow! What a startling conclusion. I guess nobody ever dreamt that result was going to happen. Worth the money to give the guy a grant for 4 years to show something that I can regularly show in my greenhouse for free.
But whatever it does demonstrate, it is not that 'up to 40 per cent of the Amazonian forests could react drastically to only a slight reduction in precipitation'...or anything remotely connected to it. It is a total falsehood to suggest otherwise.
Re Herding Cats
When I started learning science nearly 50 years ago, I was used to writing up my experiment with a bit about purpose, a bit about apparatus (maybe even with a diagram drawn with one of those exciting templates), a piece about the exact method I used, then some stuff about observations, an analysis of my results and errors and usually a conclusion....even if it was 'the experiment did not show what I set out to demonstrate'
This simple way of writing up a lab book was good enough for many of the great scientists of the past and got me through a decent Chemistry degree from a 'good' university.
Is there any philosophical reason why 'climate science' does not use this simple structure for reporting their results? And then we would all know exactly what is going on, what has been shown and what has not.
Dave,
"For once, I don't really feel too bad about his views being twisted like this (and him going along with it). Unlike climate change, Amazonian rainforest loss (through logging, etc) is a genuine and immediate threat, and any publicity that attracts attention is good publicity."
I disagree. If there are sound reasons for doing something, argue with them. Using a distortion of the truth to advance a cause is shortsighted and likely to backfire. It seems to be a general problem with all of these things lighted on as poster children of Global Warming, polar bears, submerging coral islands, Himalayan glaciers, glaciers on Kilimanjaro.
Dave wrote: "Unlike climate change, Amazonian rainforest loss (through logging, etc) is a genuine and immediate threat, and any publicity that attracts attention is good publicity."
I wholeheartedly disagree. There is no benefit whatsoever in misdirecting public attention away from genuine and immediate threats towards something unrelated. At best, the solution will have no effect on the actual problem. At worst, the real genuine and immediate threats will be ignored to tackle an imaginary "genuine and immediate" threat. Consider the paradoxical argument, tangentally promoted by Nepstad, that rainforest can be saved by destroying it to grow more biofuels,
Patagon,
Thanks, my confusion has been resolved. The rainfall was reduced for 60% during the rainy season only, which created an annual reduction of one third.
An example of the plastic panels can be seen on this page.
Dave said:
An appeal to authority! Who needs to be an expert when the IPCC said 'slight' while Nepstad's work which he claims supports it is anything but a slight reduction in rainfall?
Dave said:
Is it really fine to twist words for the greater good? Not only is it dangerous to influence public policy with fibs this kind of behaviour in enviromental science does more honest conservationists a disservice.
If politicians simply abandoned the unsettled and uncertain science and said 'It is rational to reduce our dependence on fossil fuels and improve our enegy security, it will be expensive to begin with but the long term benefits of ample energy surpluses will be worth it. Will you vote for us?' I think they might be surprised.
Dave
Brazil is a developing nation they are entitled to exploit their natural resources or to be compensated for conserving. Brazil has set up a conservation fund an invited the international community to contribute to the fund. To date only one nation has taken up that offer, Denmark $1 billion. It should also be noted that Brazil has made more than 30% of the forests conservation zones and have reduced deforestation by approx. since 2004. Brazil has also legislated that private owners can only clear 20% of their land. The legislation has been backed up by enforcement with penalties levied $175 million. Seems to me thay are doing the right thing and the developed countries don't give a shit.
oops reduced deforestation by approx. 70%
It would appear from Nepstad that the IPCC and himself are not wrong in what they are saying, it is just that people are misinterpreting what they are stating.
According to Nepstad it is wrong to conclude that "40% of the Amazon is at risk from slight changes in rainfall" actually means what is says. It is actually correct to say it, but very wrong to state it.
I hope that clears up the misunderstanding over the science on Amazongate.