Another open thread, and a totally off-topic video:
If you do good things, good things will happen to you.
Support Your Global Climate Blog
Recent Comments
Buy the book
KIDS' LIVES MATTER so let's stop climate change
Another open thread, and a totally off-topic video:
If you do good things, good things will happen to you.
Open Mind 
Ran across a copy of Sussman’s “Climategate: A Veteran Meteorologist Exposes the Global Warming Scam” at the local library. Here’s how it starts, in the Foreword:
“Global warming’s story begins with a diabolical bastard named Karl Marx.”
I skimmed through it and it was full of bogus BS and outright lies. It’s been published by WND press (think WorldNetDaily). I can’t believe this kind of utter crap is out there. Utterly ridiculous.
In the film “Gandhi,” at one point Gandhi say “Martial law only shows how desperate the British are.”
Accusations that global warming began with Marx — only shows how desperate the fake skeptics are.
I just watched a short YT video of Sussman saying, “This is all about God. He created the Earth and takes care of it. Everything else is politics.”
There’s an entire right wing industry of writing scare literature aimed at old people.
When we had to take care of an ancient aunt a few years ago we were absolutely stunned at how much she was inundated with muck that was obviously made up for the purpose of shaking nickels and dimes out of the almost senile. Why the combined attorneys general of the 50 states haven’t banded together to shut them down is a mystery that will only be solved when the Book of Life is opened on Judgment Day.
For an almost completely opposite take on humanity, I just finished reading the transcript from the inquisition of wildlife scientist Charles Monnett this year. If this is how “science audits” are conducted, we need to reign in the inquisitors. The questions and calculations presented by the investigators (who aren’t scientists) were bizarre, weird, and in the words of Dr. Monnett, goofy and silly.
I will admit, that the interview made for entertaining reading, and compelling evidence of just how far the less-intelligent bullies in American society will stoop.
I found this item about the investigation. Interesting that this is not about scientific integrity.
There was no database entry choice for “We did not see dead polar bears”. I get the impression that could be the fuel for the main allegation, i.e., the researchers were supposed to be able to see into the future.
“Somebody is deficient in fifth grade math” and “seven of 11 percent”. lol.
Rather entertaining read…I wonder who did the math analysis, and Dr. Monnett’s question, “Don‟t you wonder why somebody that can‟t
even do math is making these allegations and going through this
stuff?”, was probably a bit of a show-stopper. Seriously though I hope whoever did that pass along that math analysis to Agent May is exposed, not necessarily for any legal procedures, but just so the people who have to work with him/her know just how deficient their analysis and math skills are. Not that it’ll matter as chances are the person doesn’t have a science background so isn’t required to do math–gotta wonder if they got shafted on their mortgage payments though–“7 of 11%–oh that’s a good deal…sign me up!”).
I’ll have to keep following this story and see what happens. I’m a wildlife biologist and I work contracts for both government and NGOs, and this treatment of a colleague infuriates and saddens me…I want to jump in and fight too.
[Response: Maybe it’s time to hold these people accountable. Instead of letting the whole thing drop when it fails to find anything … launch a counterattack. And make a giant stink about it — make them rue the day they thought they could get away with slandering scientists.]
I already decided not to read the transcript, then saw some excerpts and spent the time anyway.
Wow. Just wow. Never knew that Franz Kafka and Douglas Adams had an illegitimate love child.
…and of course Pielke is his usual silly self, claiming scientists are keeping mum about this to cover for Obama. Heck I heard about this a couple of days ago, and I tend to follow these things. Every scientist I’ve heard an opinion from is as angry about this as I am.
Here was another: Ian Thomas of USGS.
The Story of Fired USGS Employee in His Own Words
More HERE.
As a government scientist working for an agency responsible for regulating the oil industry (though not a primary agency like the old MMS, or one nearly as reactionary) I find this absolutely bone-chilling.
I can’t stop thinking about this. Here is a link to the response from Dr. Monnett’s representatives at Public Employees for Environmental Responsibility (PEER): http://www.peer.org/docs/doi/7_28_11_Scientific_Misconduct_Complaint.pdf
Response from BOEMRE (http://www.alaskadispatch.com/article/boemre-director-says-offshore-oil-agency-not-witch-hunt):
“We are limited in what we can say about a pending investigation, but I can assure you that the decision had nothing to do with his scientific work, or anything relating to a five-year old journal article, as advocacy groups and the news media have incorrectly speculated. Nor is this a “witch hunt” to suppress the work of our many scientists and discourage them from speaking the truth. Quite the contrary. In this case, it was the result of new information on a separate subject brought to our attention very recently.”
This would appear to suggest that the heavy handed investigation into the original frivolous complaint has finally manged to dig something up.
IMHO, perhaps the most interesting section of the transcript (starting at page 86, line 21 – quoted in full because it’s buried so deep in a 96 page document):
“But why are people gunning for me (Monnett) constantly? Why is it that I’m telling you and Jeff probably told you that we can’t do science here? Why did I outsource this? You know, why do you think I outsourced it? It’s because I couldn’t do it here anymore. They wouldn’t let me do the right kind of analysis here that had some potential to demonstrate negative effects from what we manage as an agency.
“This study was done for 20-some years, and it was permitted to go forward because it never made a ripple in anything, because it was designed in a way that it had almost no potential to identify any sort of a problem. And part of that has to do with the analysis, and part of it just has to do with the, the questions that were asked.
“And when I started the project, I had to make a basic decision. Am I going to try to ensure that this study goes forward forever and that we keep doing it the way we are, or am I going to try to do some science in here and get some of this stuff out? And I chose the latter, which led almost immediately to me having to outsource the study, and so it’s at the National Marine Mammal Lab, because they’re a trust agency, they have responsibility for these, these resources, they give a damn. And they’re scientists, and if they find something in there, they’ll, they’ll publish it, regardless of what my management thinks.
“And my management have been trying to kill this study for a while, ever since really the polar bear thing came out. That was when they realized that it’s dangerous to take data like this, because if there are changes and, you know, God forbid something that has anything to do with the climate change debate.”
I’m wondering if these”investigators/inquisitors” are people who have filtered through from the Bush Administration.
If I remember correctly Bush and his team fired an extremely large number of people (going down to really low level positions) who worked for the government during the Clinton years and then filled those positions with “folks” who would be far more friendly to the Bush Administration way of doing things.
Remember FEMA and Katrina?
Tamino said… “Maybe it’s time to hold these people accountable. ”
Yes. Yes. Yes.
We need FOIA requests for all documents at BOEMRE and OIG that pertain to Dr Monnett. Let’s see, how many FOIA requests were going to the CRU? Was it something like 65 in one week? Surely we can top that.
[Response: Let that be just the beginning.]
Actually, I think PEER is already doing this. They seem pretty hot about this. According to the Mother Jones article the BOEMRE was formed as a response to the MMR being too cozy with oil and drilling companies. It seems like it was a change of name only.
But what a story hook! Just so it’s shelved with the other fiction. . .
Sadly, no – it’s two entries above in the Dewey system from Weart, two entries below Pielke, Jr. At Amazon, 30 of the 55 reviews give it 5 stars – however, 11 give it one star. I’m not surprised that such bilge exists, it’s just disappointing that it does.
Kevin, it’s enough to cast doubt on valid objections to Arctic prospecting, especially when there’s billions of bucks to be made. Job done well, I’m afraid.
The trouble is that many fake skeptics genuinely believe that stuff. They don’t see scientists pointing out problems with business as usual, they see useful idiots doing the bidding of communists bent on totalitarian rule of the world.
They genuinely believe that scientists are falsifying their results as a way to garner more research money.
Its impossible to argue with such people, because I start from the idea that scientists, while they may occasionally be wrong, are basically honorable people searching for the truth. Mind you, I believe that most of the skeptic heroes, his lordship et al, are being deliberately dishonest.
So if each group believes the others experts are dishonest, how can you have a sensible debate?
The reason it is impossible to argue with such people is that they think it is quite acceptable to misrepresent the truth and make up their own “facts” in support of their position–the “noble lie” syndrome–and thus they assume all of their opponents are doing the exact same thing.
Yes. If you interact with some of these folks, you generally do find that at bottom, they view everything as political–especially ‘truth.’
Me, I generally prefer the same *without* the quotation marks.
They are even more psychotic than that, in that they’re not misrepresenting the truth – they don’t actually know what reality is, and too megalomaniac to believe that knowledge is important.
If I may be permitted to snarkily slightly misquote Ron Suskind, quoting Karl Rove –
scientists live “in what we call the reality-based community,” which he defined as people who “believe that solutions emerge from your judicious study of discernible reality.” … “That’s not the way the world really works anymore,” he continued. “We’re an empire now, and when we act, we create our own reality. And while you’re studying that reality—judiciously, as you will—we’ll act again, creating other new realities, which you can study too, and that’s how things will sort out. We’re history’s actors…and you, all of you, will be left to just study what we do.”
Precisely.
I don’t “believe” a priori that Lord Monckton is being dishonest; I hear what he says and then decide based on the logic and evidence that he’s probably being dishonest. That’s an important distinction.
At any rate, why does anyone still think that a “sensible debate” with the crazies is somehow the only way forward? Yet many people do. Apparently Barack Obama, Michael Tobis, Chris Mooney, and Matt Nisbet — just to name a few — have somehow decided that reaching out to cranks is more important than being truthful, or logical, or reality-based — and as a result they themselves are becoming more and more distanced from fact.
— frank
Why?
Because there’s no option but to make the case, and that’s what they think they need to do to make it. It’s tough, and discouraging in the present climate (pun more or less intentional), but the only real choice we have is tactics. Some have more faith than others in the power of rational discourse.
I’m asking in a totally non-snarky way: what tactics do you think are most effect, frank?
“effective“, of course.
Kevin, to get an idea of the type of tactics I advocate, check out my account of an imaginary debate between Christopher Monckton and John Kerry. (I also have other ideas…)
Mindless “compromise” in the face of an implacable bullshit movement isn’t a “tactic”; it’s just a policy of surrender portrayed as victory. You can’t promote pro-science thinking by caving in to anti-science thinking.
— frank
I have to say that I’m with Frank on this. The denialists are ridicul0us–which makes the only proper response ridicule. Yes, there are deluded fools who listen to them spew their nonsense, but their only credibility results from the credulity of their audience–specifically the desire in the audience to pretend everything is fine.
By all means, we should listen to and engage with the best our opponents have to offer. That is how we learn. Unfortunately, the denialists have nothing to offer but lies and self-delusion. I do not see how we profit by engagind such fools, nor by trying to educate those who are so easily fooled by them.
“. . . nor by those so easily fooled by them.”
I don’t think we can afford to give up on the merely naive, of whom there are a great many–at least in terms of scientific literacy.
And I’m not talking about compromise. I’m talking about every time the crap gets spouted, I say, “Well, that’s not right–solar radiation is not over 50% IR, and here’s where you can read about that. . .” or “Well, you know we weren’t talking about ‘ACO2,’ we were talking about radiation in the atmosphere. . .” or, well, whatever.
It’s tiresome, but I can out-write, out-wit and out-last these yoyos, and I will NOT leave a lie uncontradicted.
Kevin, that only further feeds the false impression that there are two ‘sides’ of ‘equal’ credibility in the ‘debate’.
I’ll put it this way. Which of the following communicates the facts about climate change more clearly? Is it this:
Or this:
To me, it’s crystal clear which communication strategy is more effective.
You can’t out-bullshit them, which is what really matters. Sorry.
— frank
This way of thinking comes from the social science where all knowledge is a social construction. Over time, you ends up with peoples that can even remotely think that such things as fact and reality exist independently of the human interpretation.
Can I ask for some help?
For a few years I have been pursuing UK politicians and other UK policy makers claiming climate change is being underestimated. (I do know I’m not the only one.) Recently I have been pointing out that climate models used in the current IPCC round, AR5, have missing feedbacks and so are underestimating the effects of climate change.
Perhaps the most solid example of a missing feedback is the release of carbon as carbon dioxide from melting permafrost but several other feedbacks seem to be missing:
1.release of permafrost carbon as methane,
2.release of methane from undersea hydrates,
3.the drying of the Amazon etc.
The latest that I’ve heard about is the release of carbon from wildfires, particularly in the Arctic.
What I find particularly frustrating is the omission of these feedbacks from estimates of the probability of “dangerous climate change” but I have hit a temporary stone wall in trying to get a response from Government. (See Open letter to Chris Huhne, http://www.brusselsblog.co.uk/?p=306)
Does anyone have any advice on
What are missing feedbacks?
What are their likely impacts?
Am I wrong?
What else can I do?
Geoff, have you asked James Hnasen? I’m sure his email address can be easily found.
Typo: James Hansen
I have tried to find his email address but failed.
Also I think his excellent “carbon fee” proposal should be supported. See “Hansen’s Carbon Fee”,http://www.morejobs.org.uk/?p=34
Here you go mate.
Thanks
On his professional home page… how quaint
No, I’m afraid that this way of thinking is as old as human kind. It is rather scientific thinking that is the innovation and the aberration. People seem to be equipped with large brains not so that they can better discern the truth, but so they can hide from it more effectively.
So-called philosophers like Paul Feyerabend merely provide a convenient theoretical framework on which those who fear the truth can hang their already existing prejudices–and it doesn’t matter whether they are right or left, physical or social science… Science is the cure for delusion–that is why people resist it so doggedly.
Yes, this is the natural way of thinking. However, it is clear in my mind that is much more present in social science than natural one. Simply because the notion of fact is much harder to dismiss in the natural world.
Science is often in the position of speaking truth to power; it is only tolerated at all because of it’s general usefulness.
Didn’t an official tell him to give the ball back? The mom even thanked the official.
The perception of someone in power finding an idea threatening adds credibility to that idea.
So ironically I suspect more people will believe polar bears are in danger from climate change as a result of reading media stories speculating that Monnett has been silenced.
Amazing. Best news for ages.
“in the letter obtained by Fairfax. ”
I’m sure that if the Murdoch press had the letter, they wouldn’t have published it….
Well, unless they decided to paint Cameron as a member in good standing of the socialist plot to redistribute world wealth.
But even then they’d use ‘creative excerpting’ rather than publishing the letter whole.
I’d be surprised if Australia’s carbon pricing survives the next election; we are seeing big money spent to undermine it – and the depth of the pockets involved are extremely deep. Those interests are backing an Opposition that consistently supports promoters of climate science denial whilst giving only insincere lip service to the existence of the problem. On current polling that opposition would win an election hands down.
Hmmm. Deep pockets and shallow minds. Not a good combo.
I’d think US readers would understand how populist – err popular – a No-New-Taxes policy position can be, and although Australia’s economy isn’t suffering the way the US appears to be the prospect of rising energy costs was never an easy sell. Add a barrage of advertising, shock-jock opinionation and big media editorialising (insisting it’s entirely pain with zero gain and will be completely pointess) and it’s been enough to arouse widespread antipathy to carbon pricing in any form.
I think it thoroughly stunned Big Mining (that has enormous political clout) that a carbon price could get up at all; it probably never would have but for peculiar political circumstances. From the abundance of prime time advertising, it looks like they’ll spend whatever it takes to bring this government down and kill any chance of carbon pricing getting established.
I’m not seeing this anywhere, Geoff, Fairfax or otherwise. A hoax perhaps?
Sorry, Geoff. I found it now. So much for my search skills. I must’ve had my media switched to ‘off’ over the weekend.
For Australian readers wondering when the worm will finally turn down here… some balance from News Ltd (don’t faint)
http://www.couriermail.com.au/news/opinion/tony-abbott-makes-a-habit-of-flipping/story-e6frerdf-1226106158058
Since we do have an open thread, I’d like to invite comment and/or constructive criticism on this brief formulation:
I rather like it, finding it clear, concise and reasonably accurate within the bounds of ‘general readership cappability.’ But what I find to be clear and accurate does not always strike others to be so–unfortunately!
So–whaddaya think? (And yes, that first sentence is awkward, but I can improve the mechanics myself.)
“As a dam built across a river causes a local deepening of the stream, so our atmosphere, thrown as a barrier across the terrestrial rays, produces a local heightening of the temperature at the Earth’s surface.”
Don’t fix what ain’t broke ;-)
Tyndall was blessed not only with great physical intuition, experimental expertise and solid math chops, but with verbal grace and poetic flair. So I agree with you heartily that his description ‘ain’t broke.’
On the other hand, there is a difference between an axe and an adze, and I wanted a verbal ‘tool’ that indicates a bit more specifically how that ‘local heightening’ is produced. Does this formulation supply that need? Or does it mislead or misstate in some way?
I would cut the second sentence “The inverse…”. Rather “Below this level, the air is opaque and heat transport of necessity convective, establishing a known lapse rate connecting temperatures at different heights. In order to restore radiative equilibrium with space, temperatures throughout the atmospheric column must go up.”
What happens radiatively in the lower atmosphere has indeed little explanatory value — eddies in the river upstream from the dam.
By increasing concentration of carbon dioxide you increase the effective radiating height from which thermal radiation escapes without being reabsorbed. However, this layer is cooler. For rate at which radiation escapes the climate system to be brought back into balance with the rate at which radiation enters the climate system this cooler layer must warm up.
By warming up it reduces the rate at which heat is transferred from the layer directly below it (both via radiation and via convection), this causes that layer to warm up until the rate at which heat is transferred is brought back into balance, but as it warms up this reduces the rate at which heat is transferred from the layer below it. Like dominos, each layer closer to the surface must warm until ultimately the surface warms, and thus the warming of the effective radiating layer drags the rest of the atmosphere and ultimately the surface along with it.
The lapse rate is re-established at a level that is roughly the same as before, but with an effective radiating layer that is higher. This implies a warmer surface. With a moist adiabatic lapse rate of 6 km/°C we are talking about a doubling raising the effective radiating layer 0.5 km to give us 3° C.
Kevin,
I would leave out the last little bit about surface radiation and its importance globally. The primary thing here is that the TOA forcing has a more direct impact on the surface temperature than the surface forcing does, particularly when the atmosphere is well-mixed vertically.
This is because the atmosphere (not just the surface) helps balance TOA perturbations by adjusting its OLR, and most terrestrial radiation to space originates in the upper troposphere (owing to its IR opacity). For an atmosphere that is fully opaque to longwave radiation from the surface, all OLR originates in the atmosphere– the atmospheric temperature adjusts to balance the forcing at TOA, and is entirely controlled by the forcing at this level; further, because the atmosphere is well-stirred by convection, upper level temperatures lead to corresponding adjustments of the surface air temperature. The surface budget is balanced by adjustment of the difference between the surface and the overlying air temperature, and in equilibrium both the top and bottom of the atmosphere energy budgets should be satisfied.
The surface fluxes not only include increased ¨back radiation” but changes in surface IR radiation, as well as evaporation and sensible heating fluxes. These fluxes tend to to keep the SST close to the surface air temperature so that at equilibrium, TOA forcing has a primary influence on SST. The surface fluxes eventually bring SST and the atmospheric temperature into equilibrium with forcing at TOA. There are some exceptions– if you threw in an evaporation source in a hot desert, you could cool the surface even with a positive TOA forcing (e.g., increased CO2).
Thanks, GP, thanks Chris.
The sentence you both counsel cutting out is in there because the context involves the G & T ‘atmosphere can’t heat the ground because of the Second Law of Thermodynamics’ meme–one of the obsessions of an ‘opponent,’ whose denialism now encompasses HIV, quantum theory, and the existence of Lagrange points–and I’m probably forgetting a couple of things. I think I agree with you both that its more of a distraction than anything else when placed in a more general context.
I assume that you don’t mind if I use your words–or some paraphrase of them–sometime, GP? I like the formulation.
And thanks for the very substantive response, Chris–there’s a lot in there for me to mull over.
Again, thank you both very much!
Huh. Lagrange point denialism is a new one to me. Are astrophysicists promoting the notion is order to get funding on the orbital gravy train, or are they trying to introduce a commie world government?
Kevin,
This, of course raises the question of why you are arguing with a total loon. You are of course aware of the admonition: “Never argue with an idiot. They will drag you down to your level and beat you with experience.
I mean, Lagrange points? That’s obvious if you know anything about celestial mechanics.
Stu, my hypothesis is that it gives him joy and comfort to think that he is smarter than just about anyone else on the planet–the addictive aspect of Dunning-Kruger.
Ray, you only get dragged down to their level if you allow your ego to make you think that you ‘win’ if you convince them. Otherwise, their ravings allow you the opportunity to calmly disseminate correct information–and even better, your responses allow them the opportunity to make ever more apparent how wacky they actually are. There’s a lot of folks reading who are shrewd enough to notice the difference–or so I fondly believe.
Kevin go right ahead.
Eh, the G&T argument is perhaps better addressed at a more fundamental level, in short sentences of monosyllabic words ;-)
I’m not sure what we gain by engaging with the incurably stupid. We won’t convince them with evidence–the learning curve would have to have a positive slope for that. If anything, our willingness to engage lends a patina of respect to that which should be dismissed with laughter–it’s how we wound up with an entire congress that believes you can fund a gummint with zero taxes.
By all means, if our opponents offer compelling evidence and arguments, we should engage and actively consider what they have to say, but I’m afraid I couldn’t be bothered to piss on the stupid even if they were on fire.
Ray, I know we differ on this. I just don’t think that leaving ‘the Big Lie’ uncontested in a public forum is a good idea–and I’m not above satire, either, though I think that ‘sweet reason’ is the proper default. “Pour encourager les autres,” and all that–meaning that he will never learn, but others can learn from his horrible example.
(As a side note, I don’t think that my loon is actually stupid, exactly–more that his distorted cognition has his intelligence–admittedly, much less than he thinks it is, but then he patently regards himself as a universal genius, incapable of error–constantly spinning its wheels with false premises and perceptions. It’s kind of fascinating, if also sad and creepy.)
GP, your comment was worth a chuckle–out loud. Appreciate that, as well.
Just to advertise in the open thread. I´m putting my (long inactive) wordpress blog to rest. I´m writing laymen-level (but hopefully interesting and thought provoking) articles on general climate issues here; I hope to draw some more commenters and discussion (which could, of course, get into some more technical details), if anyone would like to bookmark or blogroll. I only have one post so far on the issue of locating planets in the habitable zone around stars.
I am just one contributor to this blog, although everyone else comments primarily on weather-related topics.
[Response: I have always found your comments to be especially insightful. I’ll be sure to visit the times-union blog, and I recommend all readers do the same. And feel free to use this place for shameless self-promotion.]
The transcript of the Monnett interview is well worth reading, for some of the reasons mentioned above–the culture of harassment and intimidation of scientists at BOEMRE, the absurdity of the criticisms of the polar bear note–but also because it gives a nice picture of what doing aerial wildlife surveys is like.
Pertinent for readers here is this critique by Dr. Monnett about the analysis that he felt should be done of the Bowhead Whale migration, and the analysis that BOEMRE allowed him to do:
(Pp. 90-91, line numbers redacted.)
Probably should have given the link to the PEER site where you can read the transcript yourself:
Click to access 7_28_11_Monnett-IG_interview_transcript.pdf
Holy crap. Our government is spending money for this?
(you have to read quite a way before it starts getting pathetic – around pp 44-50+; I don’t think I have the stomach for more.)
QotD on p. 46, from Dr. Monnett: “if that’s the kind of thing people are after,
[Special Agent Eric May: “that’s why we’re here”] that’s just silly”
…”god, I’ve got people here who are second-guessing my calculations”; “…”that’s what you do in discussions, you throw stuff out for people to think about”
p.33 “it was just an observation, an anecdotal note, with limited analysis”
p.38 “weren’t recognizing it had potential to be controversial/influential”
Download the PDF at MT’s post (“polar bears”). Amazing.
Notes and page#s from the transcript, up until p. 50, to save others some time
Project was BWASP. Two Special Agents, Lynn Gibson and Eric May.
p. 13 Monnett: no way to document a dead polar bear in our [data collection] system
unusual range expansion
15 flying, tried to hold altitude at 1500 ft; to avoid disturbing animals (permit related)
17 also keep altitude for safety
18 I’ve been in airplane accidents in the past
went below 1k ft once, on seeing dead whale
22 2nd report unauthored since analysis was simplistic, incorrect, misleading
23 saw polar bears in 3 circumstances generally-
* dispersed on ice offshore
* barrier islands along coast, sometimes congregated
* ?
24 redesigned data collection 2005, 2006 to make flexible, able to record the bears better
25-26 dead polar bears?
in 2004 4, maybe 5, over a few days
27 – a typical year see 1 swimming bear; that year, a lot
“we had swimming as a behavioral choice ( in data collection), because whales swim”
up to that point, we had never seen a dead bear to record.
28 the dead bear part is in our books, it’s not in the database
how to recognize it’s dead – something that’s in the water, with its head down, with gurry and stuff streaming off it…
29 the last one we saw was bloated like a beach ball
pillsbury doughbear photos
30 “this is one of those things you always look back on, and you wonder why you didn’t recognize how important it was. to us it was just like weird”
By the time we’d seen 4, we realized something unusual had happened
Before we saw the dead bears, we had a couple days when we saw a lot of bears swimming. and that really got our attention
31 and we had a big discussion..maybe they’r a lot mor aquatic htan you’d think…we saw 10 or so, and some cubs with the bears
about a week later, started seeing the dead ones
Sometime later began to realize it was worth writing up, an important observation
We were looking for quick clean products to write up, that’s how we justify our work/study
32 eg Range extensions, or Mallards eating salmon
33 just an observation, anecdotal note, with limited analysis
What started the m.s? we found we’d seen almost as many bears swimming on 1 day as in history of hte (25+ yr) project
This was 2004, with record sea ice retraction from coast
34 then another just on swimming drowning, possibly from storm, etc; 2 papers -> 2 posters
35 dramatic change in distribution, bears no longer dispersed along coastline- bulk of bears at camden bay, vast majority around the bone pile
35-36 paper(s?) reviewed by wife phd ecologist, cowles, manager (political correctness review), derocher, sterling (world’s #1 polar bear guy)
then 3 anon reviewers
38 region was still taking position it wasn’t of national scope – weren’t recognizing it had potential to be controversial/influential
39 “25%” – only looked at about 10% of the area
44 4 bears seen, 3 on transects
46 “if that’s the kind of thing people are after,: [Special Agent May: “that’s why we’re here”] “that’s just silly”
…
p. 67, Department of Interior, Office of the Inspector General Special Agent Lynn Gibson: “why do you include the fourth bear in there?”
I wonder what the investigators thought of the whole thing. Do they consider this a good use of their time? Do they share the political views and motivations of those higher up who instigated the investigation? It didn’t read like they were hostile but that wouldn’t necessarily be evident in a transcript – it’s more like they didn’t know themselves what they were after – apart from catching Dr Monnett lying about something, preferably something to do with polar bears.
The Charles Monnett story become a bit clearer. Looks like the fishing trip pulled up a couple of tin cans to justify continuing the investigation, although the Justice Department has refused an application to pursue criminal charges. Monnett finally received an official letter.
Suspended Arctic scientist to be questioned over research contracts
Yep. Here more from PEER:
http://www.peer.org/news/news_id.php?row_id=1504
Here is the letter:
Click to access 8_2_11_IG_interview_notice.pdf
Yuck.
Should be a red flag. The judiciary is still the least corrupted part of US government, and the JD knows that.
chriscolose wrote:
Please keep the essays you have there up. They are a treasure.
chriscolose wrote:
I will check it out. Can’t promise that it will send any traffic your way, but I will also add it here. And it might be a good idea to get on Google Plus. There is a growing community. I know of over 70 individuals (familiar names) that are now on it. Take the space out of my name, all lower case at gmail and I will get you an invite.
Thought you might be interested in this, if you haven’t already seen it:
http://moregrumbinescience.blogspot.com/2011/08/how-to-find-climate-normals.html
It’s an analysis of long-term temperatures anomalies. I’d be curious to get your thoughts on methodology and assumptions
[Response: I’ll do a post about it. Maybe tonight.]
While web browsing I stumbled across the Jo Nova website and she is touting the work of Murry Salby as disproving that humans have an effect on global CO2 levels. I get the impression that Murry Salby is actually a credible atmospheric scientist, so I suspect that Jo Nova is misrepresenting the conclusions of his paper and it says nothing of the sort. Am I correct on this?
Unfortunately, he does indeed say what she claims. An example:
“The correspondence to observed changes of Co2 on timescales of a couple of years over the satellite era and to the degree seen even over the 20th century, makes it difficult not to conclude that sources involved in changes of Co2 on short timescales are also involved in its change on long timescales. The popularized view has been that CO2 is driving the bus and climate is along for the ride. The observed behavior reveals just the reverse. Climate is at the wheel, and to a significant degree, CO2 is at the back of the bus. Climate projections rely on an ability to predict CO2; it’s the one thing believed to be known, because of the presumption we control it. Namely, future atmospheric CO2 is determined entirely by human emission. That’s what is specified in climate models which then predict how climate will respond, in so-called climate scenarios.
The observed behavior reveals that, much as we might like it, the real world doesn’t work that way. Net emission includes a substantial contribution from natural sources; if you don’t control CO2, you can’t predict it, and if you can’t predict CO2, you can hardly predict how climate will respond.”
and then:
“Let me make one comment. What you’re really talking about is a 4 degree simulation, right? And that’s based on model runs that were done oh, in the late 90′s and early 2000′s for the last IPCC report. Now the conclusion of that was that the increased temperature that was observed, global temperature, in the 80′s and the 90′s was due to CO2. And that appears to have been a sucker-punch that the IPCC went for because no sooner – the ink hadn’t even dried on the report – that Mother Nature intervened and CO2 after the turn of the century continued to increase, in fact if anything, slightly faster, but global temperature didn’t. If anything, it decreased in the first decade of the 21st century. Now I’m confident the IPCC will come up with any explanation, in fact its already come up with several, but your extrapolation of 4 degrees presumes that you essentially have these model projections and that’s what they’re relying on and those model projections rely on “we know what CO2 is” since we’re responsible for CO2, since the human source is essentially what causes atmospheric CO2 to change, quote unquote, that we can extrapolate that. Well, doesn’t look like that works.”
It’s pretty bad stuff. I’ve never heard of him in climate debates before, but he appears to be “coming out” as a “skeptic”.
“Now I’m confident the IPCC will come up with any explanation, in fact its already come up with several …”
So much focus on the IPCC raises flags. (The UN is behind them, you know …) Is it in the remit of the IPCC to “come up with explanations”? I find that unlikely.
I expect all sorts to emerge from the woodwork in Australia now that a carbon-price policy has been announced there.
Hi Tamino,
There is a podcast at the following URL that I think you would be very well suited to deal with.
[audio src="http://www.thesydneyinstitute.com.au/wp-content/uploads/podcasts/2011/THE_SYDNEY_INSTITUTE_MURRY_SALBY_2_AUGUST_2011.mp3" /]
It is by Dr. Murry Salby
http://www.envsci.mq.edu.au/staff/ms/index.html
Curry talks about it here:
The abstract
“Carbon dioxide is emitted by human activities as well as a host of natural processes. The satellite record, in concert with instrumental observations, is now long enough to have collected a population of climate perturbations, wherein the Earth-atmosphere system was disturbed from equilibrium. Introduced naturally, those perturbations reveal that net global emission of CO2 (combined from all sources, human and natural) is controlled by properties of the general circulation – properties internal to the climate system that regulate emission from natural sources. The strong dependence on internal properties indicates that emission of CO2 from natural sources, which accounts for 96 per cent of its overall emission, plays a major role in observed changes of CO2. Independent of human emission, this contribution to atmospheric carbon dioxide is only marginally predictable and not controllable.
Professor Murry Salby holds the Climate Chair at Macquarie University and has had a lengthy career as a world-recognised researcher and academic in the field of Atmospheric Physics. He has held positions at leading research institutions, including the US National Center for Atmospheric Research, Princeton University, and the University of Colorado, with invited professorships at universities in Europe and Asia. At Macquarie University, Professor Salby uses satellite data and supercomputing to explore issues surrounding changes of global climate and climate variability over Australia. Professor Salby is the author of Fundamentals of Atmospheric Physics, and Physics of the Atmosphere and Climate due out in 2011.”
I really don’t buy his argument but it is something only an expert in statistics could probably completely debunk.
Good lord there’s a lot of stupidity in that thread. It’s quite painful reading the comments.
The observed annual increase in atmospheric CO2 is smaller than the annual human emissions. Conservation of mass therefore requires that the rest of the earth system must be a net CO2 sink right now. When Salby writes “emission of CO2 from natural sources […] plays a major role in observed changes of CO2” the actual role it plays is “being consistently small enough to partially mitigate anthropogenic sources of CO2”.
Well, sort of. After all, water vapor has also been increasing, but not (directly) due to human water vapor emissions.
But even without any heavy stats, you can point out:
– All measurements that we have of CO2 in the Holocene point to a very stable value for the past ~6000 years, suddenly shooting up in the industrial revolution.
– When we look at deglaciation events in the ice core record, there is a lag between rising temperatures and CO2 levels which appears to be due to the time it takes oceans to heat up and degas; hence any control on CO2 levels from the ocean can’t be happening as a response to current temperature changes.
– If Natural CO2 flows dominated, you could see a trend, but it would be surprising if it were such a steady upward trend which just happened to be between zero and the amount expected if 100% of human emissions stayed in the atmosphere.
– I’d go on about isotope ratios, but you have to consider the audience.
It’s a bit like the temperature related arguments in which the current warming is not due to CO2 at all, but some ‘natural factor/cycle’ which just happens to have mimicked the calculated effect of CO2, at the same time as we’ve been adding CO2 to the atmosphere, and chose to manifest now and at no other time in the past few thousand years.
I finally listened to the podcast, but it is difficult for me to tell much about his reasoning without being able to see all the charts and graphs he kept refering to. I suspect that Salby’s argument is similar to the Temperature leads CO2 argument and has the same flaws.
I have a statistical decision problem I don’t know how to approach well so I’m asking for suggestions or pointers to applicable literature. {I use use stats when I need it and have never properly studyed more that the most basic portions.] Here is the problem:
A signal, a positive real number bounded well away from zero, is generated once each 3 seconds. These form the sequence s(0), s(1), …
This sequence contains noise generated by a non-Gaussian process [as the s(i) are always positive] with an extremely “fat” right hand tail. Worse, the noise is not i.i.d., being depend upon the time of day [worse in the daytime than at night.]
The underlying signal generation process is hypothesized to be either
H0: s(i) = p(i) + noise
or
H1: s(i) = q(i) + noise
I now describe p(i) and then q(i), as far as is now known:
(0) Process p(i) is monotone increasing in the sense that a moving average of 100+ s(i) is montone increasing. Further, p(i) appears to be asymptotic (in the moving average sense) to some constant value k independent of the signal index i. The actual value of constant k is not of great interest.
(1) Process q(i) is also montonic increasing in the same moving average sense but continues to increase approximately between linearly and quadratically in the signal index i, something like q(i) = b + c*i^n with n, it seems, between 1 and 2 inclusive. Again, the exact form of the continued increase is not of great interest.
What is of interest is efficiently, using as few signal values as may be to destinguish, with very high confidence (99.9+%) between H0 & H1.
I have a overly crude way to do this, but would like to have something with better statistical foundations. Suggestions and pointers appreciated.
David, could you model the noise as lognormal, with time dependent mean? You could then construct a pair of generalized linear models and see which gives the highest likelihood (or AIC if # parameters is different). This would de-emphasize the noisier daytime measurements
You could also approach the problem from a Bayesian standpoint, with prior over the two models.
Either way, the mean would increase with your signal over time, and eventually the two models have to diverge as p(i ) is bounded, while q(i )is not.
Ray — Thanks for some thoughts. These are helping to relieve my frustration at this demanding problem (which has some serious instrumentaion limitations; something I haven’t faced before).
Yes lognormal or some other one-sided fat tailed distribution with time dependent mean is may well adequately describe the noise, which is so far quite poorly characterized.
I’ve never previously used the so-called generalized linear models but indeed this might work for this crazy problem.
Indeed almost any technique will eventually discover whether one has H0 (p) or H! (q) but the difficulty is doing so rapidly; least power consumption.
What you suggested gives me something to think about. Thanks again.
David,
Likelihood based techniques–including Bayesian and model selection techniques–will likely have the most rapid convergence. I’m sort of looking at somewhat similar problems in my day job.
I noticed an analysis of the HadCRUTv3 data from Lubos Motl, reposted at WUWT, in which he determines the temperature change at each station, with a separate trend for each month. Even as a layman I can see there are MANY things wrong with his analysis, but one thing really stood out. He histogrammed the temperature changes and determined a mean (0.7 degC) and standard deviation (2.36 degC) of the resulting distribution. Conclusion? Let me quote: “The ‘error of the measurement’ of the warming trend is 3 times larger than the result!”
Hmm, ever heard of the standard error of the mean?
There are 58579 entries in the histogram, so the standard error is 2.36/sqrt(58579) = 0.01. That’s what is relevant here, leaving aside whether the 0.7 figure is remotely relevant to reality.
Geoff Beacon,. there are recent reports about permafrost melting in Russia:
http://thinkprogress.org/romm/2011/08/05/288347/warming-shrink-russian-permafrost/
“CO2 after the turn of the century continued to increase, in fact if anything, slightly faster, but global temperature didn’t. If anything, it decreased in the first decade of the 21st century. Now I’m confident the IPCC will come up with any explanation,…”
That’s easy-peasy – we can put it all down to natural cycles of variability and internal forcing.
If Salby believes Pielke is a knowledgeable scientist, then the current warming will cause more outgassing of CO2; unless he believes CO2 is not a greenhouse gas(and if it’s not, why all the fuss over whether its increase is natural or anthropogenic?) this will cause more internal forcing, and more temperature rise, and more albeit diminishing natural CO2 emissions. If he were to actually determine the resultant climate sensitivity, I’m afraid some hereditary lords won’t like the answer.
“Now I’m confident the IPCC will come up with any explanation,…””
In the spirit of full accuracy, I was the one who transcribed that paragraph from the lecture mp3; it should have read “an” explanation, not any. My typo.
I think this paper has the same arguments like Proffesor Salby.
Click to access TomQuirkSourcesandSinksofCO2_FINAL.pdf
Professor, sorry.
Just a quick question: I’m having a discussion with a septic who finds Motl’s interpretation of the spread in raw HadCRUT data as “error of measurement” credible. Can someone explain or point me to a clear discussion on howto interpret the spread in station trends? I’m objecting to Motl’s presentation of the spread as uncertainty (0.7 +/-2.36 degrees) but don’t know quite how to explain this. Thanks!
Motl’s piece is here: http://motls.blogspot.com/2011/07/hadcrut3-30-of-stations-recorded.html
I am not sure if my comment got trough due to a bug (Tamino Pelase remove multi-post if so).
Calculating on month produce a smaller signal to noise ratio than year by a factor up to sqrt(12)=3.46. He should have done their his calculation on daily basis to make his point.
The fourth deadly concert/festival stage collapse this year. I don’t recall this happening before (even when too many people were climbing the light towers during the thunderstorm at Woodstock).
“Unpredictable” said the folks in Indianapolis.
Are they not constructing stages as well, are there just more of them, are summer thunderstorms getting more sever, or is it just (climate noise?)
Seems to me that there are one or two of these a year worldwide.
As far as the unpredictable comment: BS. They were warned by the NWS about a half hour before the storm hit. Whether that was enough to evacuate the people around the stage is the question. They warned the audience but in the game of telephone the severity of the warning seems to have been lost and only some of the audience left.
Yesterday I went to Washington DC participate in the protest against the Keystone XL TransCanada pipeline that will traverse Montana South Dakota and Nebraska. The protest, which is taking place at the White House and Lafayette Square across from it, is now in its second week.
The civil disobedience demonstration has been peaceful and well coordinated. Groups included Bill McKibben’s 350.org, Peaceful Uprising and Rising Tide to name several. Arrests take place everyday with the charge something like failure to obey an officer. Yesterday was the largest turnout I was told with some 140 people arrested. James Hansen of GISS was at the demonstration and was one of those arrested. The crowd of demonstrators were very much inspired by him. Same with Prof McKibben. He was arrested last week.
It is a great effort being put forth albeit highly controlled by the Park Police. Lots of young people with great energy and passion for the environment and solving climate problems. That was very inspiring to be a part of. A good mix of young and old alike joining in this rally.
Meanwhile after some ponderous discussions with the EPA the State Department put their stamp of approval on the pipeline.
Here is a link to James Hansen’s interview with the Climate Blog Solve Climate explaining his decision to join the protest:
http://solveclimatenews.com/news/20110826/james-hansen-nasa-climate-change-scientist-keystone-xl-oil-sands-pipeline-protests-mckibben-white-house
Hey all
I have run into a problem (more or less) with is statistical of nature and I thought I would present an open question here to all those who are interested and hopefully someone (if not Tamino) can provide an answer.
Overview
Lichenometry is a commonly used dating technique in geomorphology. It involves measuring the largest diameter of circular lichens (usually Rhizocarpon Geographicum) on blocks of a surface deposited (like a moraine) by a mechanism (usually ice). It is assumed that the largest lichens on the surface are the lichens which colonized the surface first after being deposited. In order to date a surface it is necessary to know the region’s lichen growth rate which can be calculated through monitoring lichen growth stations.
The earliest approaches in lichenometry (Innes 1985) involved searching moraines for the single largest lichen growing on the surface of boulders. This technique is known as the Largest Lichen (LL) approach. Because microclimate impacts affect the growth of lichens on a surface later studies concluded that it was best to use an average of the largest 5 (sometimes 10) lichens found on a surface (5LL) (Evans et al. 1998). An approach used by McCarroll (1994) involved measuring the 50 largest lichens on boulders of equal size found in certain search radii. The results would then be analyzed and a normal distribution would be assumed because of the large sample size. Several new approaches have been adopted including one termed the size-frequency approach (Bradwell 2004). Recently several authors have begun advocating (Jomelli et al. 2007, Chenet et al. 2010) for the use of extreme value theory in lichenometric dating because the largest lichens are themselves extremes. They have argued that the 5LL uses a mean of extremes which is not valid and the FALL approach assumes normality in itself. It has been demonstrated however that these approaches (5LL, LL and FALL) have relatively accurately (+ or – 20 years) many surfaces that were well dated.
My challenge is understanding the best way for me to analyze lichen data with statistical validity. I have taken 51 measurements of the Largest Lichens on equal sized boulders across moraine surfaces at 20 + locations and I have also searched the search areas for the LL and 5LL on boulders of any size. I chose equal size boulders to alleviate issues identified demonstrating a relationship between boulder size and lichen size.
I could just do a mean of the 5LL like has been demonstrated elsewhere but I want to try something different and see if it is necessary to use extreme value theory or perhaps which might be a better measure. Any thoughts?
Rob | September 3, 2011 at 10:30 pm — What to do may depend upon many aspects of lichen establishment and growth. But assuming great simplicity I see nothing wrong with LL, especially if that method has been established as agreeing with other dating techniques in other locations.
The challenge with LL is the same with measuring anything. There will always be outliers in a population. Identifying those is extremely difficult as you know well yourself so that is why some have used the 5LL approach, particularly with a cut-off threshold whereby the LL is excluded if it is 30% greater than the 2nd largest. The issue of course with that is you’re averaging extreme values which in itself may be an incorrect way of doing things. I’ve more or less just opened this for discussion because I am interested in seeing what could be applied (which distributions and so on) in manners that would be effect, particularly for creating error bounds. As you can imagine the 5 LL approach would have no real way of quantifying error propagation.
History of Climate Change Part 1- next part coming soon
http://blog.timesunion.com/weather/history-of-climate-change-part-1-learning-the-greenhouse-effect/1728/
Rob,
One way to look at this is to ask yourself whether your metric converges to some meaningful number as your sample size increases. Another question is whether it is physically reasonable given the processes you are considering.
The thing about extreme values is that the “extreme” will increase as you increase your sample size. The more lichen you look at, the more likely it is that you’ll see one of the “really big ones”. What is more, it is unclear what that will mean–is it an indication of age or of favorable growing conditions.
The mean and other central tendencies (e.g. mode and median) of an extreme value distribution also may not be particularly meaningful. However, if you were dealing with a bimodal (or multimodal) distribution, there is a chance that these could wind up sampling the highest mode if the sample size were large enough. I think that it is this mode that is most physically meaningful–at least as long as the distributions of first colony/2nd colony… are not so broad that the tail or #2 overlaps the peak of #1, etc. Does that make sense?
If this is the case, what might make sense is to look at the “density” of the samples, perhaps using a rank plot. Where you start seeing the slope of the rank plot decreasing (e.g. points are getting denser) as you move from right to left, that means you are starting to approach the highest mode. If you go too far, you may get to the point where the the left tail of the highest mode overlaps the right tail of the 2nd highest. You might Monte Carlo this with Normal distributions or lognormal just to get some idea of the structures you’re looking at.
Does any of this make sense for your problem?
Hi Ray,
I was having a look at the data a bit and here is a QQ rank plot of one of the sample data sets that I collected.
I’m not sure exactly what you mean by a rank plot though. Had a look online but there was no clear instructions on how to make one. I’d like to be able to do what you suggest though and examine the distribution itself to determine if it is bi/multi-modal.
Because it is for the most part continuous data (although there will be duplicates) it is probably best that I do what I have seen elsewhere which is to assign frequency values to intervals and plot the frequency of the midpoints as a histogram. I have done so here:
(note bins are 9.6 mm bins of which there are 10) (I’m sure there’s a less subjective way to know the appropriate bin sizes).
The challenge is of course trying to figure out which measure is worth excluding from a statistical perspective. We know that we are looking for lichens which have colonized prior to the majority of other lichens (i.e. the earliest to grow) therefore it is assumed that these will have grown for longer and are thus of a larger size. The challenge is really excluding outliers and choosing a statistically realistic (and un-arbitrary) method of select which values will characterize the early lichen growth for each surface. Now for me I could look at that QQ plot and say those two highest values are outliers visually but since it is only a sample of a population (i.e. it is within a search area on a surface) can we still really conclude they are outliers without some statistical rigour? My answer is no of course.
Have you perhaps any suggestions on how I proceed further?
I think it is a really interesting question and one that should be asked particularly when attempting to create confidence intervals for the stats we end up with.
[Response: My guess, just from visual inspection of the first plot, is that the two highest points aren’t outliers, instead they’re indicative of non-normality of the distribution. In fact I get the impression of non-normality even ignoring the two highest values.
Just a thought: try running the Shapiro-Wilk test for normality, then repeat the test with the two highest values omitted.]
Rob, It looks to me as if you could try two hypotheses–a unimodal lognormal, and a bimodal distribution with one mode ~80 and the other ~60. These two models have different #s of parameters, so maybe fit using Max Likelihood and look at AIC. My guess is that the unimodal distribution is probably superior according to AIC.
I agree with Tamino–the data are definitely not normal.
Hi Tamino,
Thanks for the response.
I did the Shapiro-wilks test on the dataset including the two “outliers” (test1) and excluding them (test2).
Test 1: (Including)
W = 0.9061, p-value = 0.0005935
Test 2: (Excluding)
W = 0.9501, p-value = 0.0343
I’m gonna be honest I’m not quite sure how to interpret these results as I am unfamiliar with the test you mentioned except that it is a test for normality.
[Response: The p-values are low enough that we can reject the null hypothesis that the data are distributed normally. So, the distribution is not normal.]
All that being said I am also in the belief that it is likely that the distribution itself exhibits non-normality but the question that I have is how best to deal with it given this circumstance. As I explained above the challenge is determining which of the extreme values can best exemplify early colonization of the surface and which represents outliers due to favorable growth conditions. Undoubtedly there is some mixing amongst the two but there are physical limitations that exist also. As i’ve mentioned before, just averaging the largest 5 lichens to me seems statistically arbitrary for these purposes and it has been shown in the literature that taking the Largest Lichen can give very deceiving results.
Have you any thoughts? Thank you for the time by the way.
[Response: I’ll have to ponder that. In the meantime, if other readers have ideas, don’t be shy.]
It was really hot in Houston last month!
http://blog.chron.com/sciguy/2011/09/august-in-houston-was-a-1-in-10000-year-event/
[Response: There’s also an interesting post by John Nielsen-Gammon here.]
Hi Rob
An interesting problem. A few, not necessarily compatible, thoughts on your problem, which raise as many questions as they answer. I hope some may be useful.
1. As others have said both your q-q plot and histogram show clearly that the distribution is positively skewed and certainly not normal. Possible candidate distributions might be log-normal or Gamma.
2. I’m not sure whether it’s worth trying to fit one of these distributions but If, for example, a gamma distribution is deemed appropriate, I suspect that there are tests out there for outliers from a gamma. An obvious place to try is the book by Vic Barnett & Toby Lewis ‘Outliers in Statistical Data’. Its 3rd edition was published in 1994 – I don’t know if there has been another, and I don’t have access to it.
3.I don’t think that extreme value theory in its usual form is relevant if I understand your data correctly. In standard extreme value theory all your data come from the same distribution and you look at the biggest (or smallest) values observed from that distribution. If I understand you correctly your data (lichen of any size) come from a continuous mixture of distributions with the means of those distributions depending on the age of the lichens. Thus the biggest lichens are (you hope) observations from the
distribution with the largest mean and they are ‘ordinary’ observations not extreme values from that distribution. How should you estimate the mean of this largest distribution? My first reaction would be to take the average of the largest few of your observations, like the 5LL method. Whether you use 5 or some other number depends on how far down your ranking you think the observations could realistically be from the distribution with the greatest mean. In a way you would be trying to remove or downweight outliers at the bottom end of the distribution, though if outliers at the top end were a real possibility you’d want to look for those too.
Ian
Hey all,
Thank you for your interest in this subject. Some very good papers to read on the subject are the following:
Jomelli et al. (2007). Assessment study of lichenometric methods for dating surfaces
Geomorphology. 86: 131-143.
Bradwell, T. (2009). Lichenometric Dating: A commentary, in the light of some recent statistical studies. Geografiska Annaler: Series A, Physical Geography. 91,2: 61-69.
Jomelli et al. (2010). Letter to the Editor. A Response to Bradwell’s commentary on recent statistical studies in lichenometry. Geografiska Annaler: Series A, Physical Geography. 92 A,4: 487-489.
The reasoning given by the GEV (Generalized Extreme Value Theory) group regarding the use of that method is that in effect you are dealing with maxima in a total population (i.e. lichens on a surface). However I am swayed by the argument presented by Professor Jolliffe regarding whether it is appropriate. In effect (if I am reading this correctly) using GEV on the maxima or minima of a distribution of features (such as lichens) with all the same age but varying sizes is appropriate but using it across a continuous distribution of ages such as those found on a typical moraine surface with lichen growth is not optimal. The moraine itself may be dated at a certain age but the size of lichens is perhaps simplistically described as:
Lichen Size= Lichen Age (Lichen Growth Rate) + (Microclimate and topographic conditions, Positive or Negative Effect)
To continue then on this thought it would mean that each age of lichens has its own distribution of lichen sizes which displays variance because of the microclimate conditions. I can see another hole in using EVT in this context also, some lichens die and regrow, therefore you could be modeling 2nd generation lichen growth which is embedded in the results (though on the smaller sized end). Another aside is that outliers may not only be due to microclimate effects but could also be due to boulders having lichens already on them prior to deposition. That is one of the primary reasons it is important for geomorphologic analysis of moraines in relation to glaciers that we try to avoid sampling moraines near rockfalls etc…
On this thought now, we do not know that the distribution represented by the top end of the data is normal therefore wouldn’t summarizing a mean of say 5LL etc… be statistically invalid? There is another question embedded here. What would be a good statistical rationalization for how many of the largest we select for averaging. I think that relates to your final comment, how would one go about identifying outliers at the bottom end of a distribution which is embedded into the dataset itself?
I will have more comments soon once I try to get through some of Dr. Ladbury’s suggestions.
In case any of you want to have a look at the data yourself there is a subset of the data uploaded at the following url:
http://www.skepticalscience.com/pics/Raw_CSV_EB_Data_xs.txt
Hi Rob,
I notice you have attracted the notice of Dr. Jolliffe with your problem. Impressive! Certainly, if you looked at the largest lichen, you’d be looking at the extreme, and GEV would indeed be a way to understand this. The question, however, is what that means physically for the process you are trying to understand. It may be that that the largest is an outlier or in some other way anomalous
If you take the five largest and average them, now you are no longer in the realm of extreme value statistics. I’ve looked at a problem somewhat like this when considering a circuit where critical components are, for example 2-for-one redundant, where the second-worst component drives the system failure rate.
Since you may have multiple, independent distributions and are mainly interested in the highest mode, it is likely that the average of the top n lichen sizes will at least be from this upper distribution for n sufficiently small. The question is how small n must be for this to be true. The Q-Q rank plot and the histogram could give you indications of this–e.g. a slight “bump” in the tail of the “main” distribution (the statistical significance of which you could model a la particle physics). Similarly, if you were to see the “interval” between lichen in the “tail” tighten for several sequential entries, that could indicate another mode (the high one). Your rank plot MIGHT give an indication of this, but you’d have to assess the statistical significance, by, for instance, looking at the AIC for a multi-mode and single-m0de lognormal. Your mean for the highest mode (if the multi-mode fit was better) would then emerge as a fit parameter. Does this make sense?
Hi Rob
I’m sorry that I haven’t looked at the references you gave. A few supplementary comments that may be covered by those references:
1. Your model is only one possibility of expressing the relationship. My intuition would be that something like the following two-equation model might be more appropriate than your suggestion, and there are many other possibilities:
Size = age x (growth rate) + error
Growth rate = (some function of explanatory variables such as micro climate) + error.
This doesn’t look at all standard to me (nor does your model) and it could be a non-trivial research problem to work out how best to fit the model in order to estimate age.
2. The fact that your n largest observations come from a non-normal distribution doesn’t rule out using a mean. The mean is possible measure of the centre of a distribution for most distributions (though for some, the median, for example, may be preferred). If the mean is chosen as a measure of the centre of the distribution, then the sample mean is often the best estimate to use.
3. Ray makes some plausible suggestions for working out how big n should be in taking the mean of the largest n values. They will work if the distribution of the oldest lichens has some separation from those of the younger ones. Your plots look pretty smooth so I expect this may not be the case and you’ll need to use educated guesswork.
Ian
Those of you who aren’t above having a couple of cheap laughs at someone else’s expense might want to check out the discussion thread following this denier column: http://tucsoncitizen.com/wryheat/2011/09/08/a-perspective-on-climate-change-a-primer-for-politicians/#comment-10163
The denier is a Tucson Citizen columnist who likes to tout his engineering background. But he’s so innumerate that he would qualify for a “bag of hammers on steroids” designation.
Don’t waste time reading his column or arguing with him, but do spend a minute or two skimming the discussion thread for a good chuckle at his inability to follow some basic atmospheric CO2 “accounting” arithmetic.
Bag of Hammers indeed. What a stupid, stupid man.
“Both Sides Now”
— Horatio Algeranon’s perversification of Joni Mitchell
Rows and flows of Pachauri hair
And IPCC castles in the air
And alarmist scientists ev’rywhere
I’ve looked at clouds that way.
But now they only block the sun
They change the climate for everyone
So many papers I would have done
But clouds got in my way.
I’ve looked at clouds from both sides now
As feedback and cause, and still somehow
It’s cloud illusions I recall
I really don’t know clouds at all
Manns and Jones and hockey sticks
The human driven temp uptick
And arctic sea ice melting quick
I’ve looked at graphs that way
But now it’s just another show
You leave ’em laughing when you go
And if you care, don’t let them know
Don’t give yourself away
I’ve looked at graphs from both sides now
From Mann and Watts and still somehow
It’s graph illusions I recall
I really don’t know graphs at all
Web awards and feeling proud
To say “Conspiracy” right out loud
Dreams and schemes and circus crowds
I’ve looked at climate that way
But now old colleagues are acting strange
They shake their heads, they say I’ve changed
Well something’s lost, but something’s gained
In blogging every day
I’ve looked at climate from both sides now
From journals and blogs and still somehow
It’s blog illusions I recall
I really don’t know climate at all
I’ve looked at climate from both sides now
From up and down, and still somehow
It’s skeptic illusions I recall
I really don’t know climate at all
Um – Judy Collins.Horatio must be one of those young ‘uns. ;-)
No. Judy had a bigger hit with it, but Joanie wrote it. (And also sang it memorably.)
(Sorry to debunk the “Horatio as young ‘un,” idea, Horatio.)
Ack. “Joni.”
Deech56,
But who wrote the song?
I am thoroughly and gratefully chastened.
I’ve looked at “Both Sides” from both sides now
From Judy and Joni
And still somehow
It’s Judy’s version that I recall
I really don’t know Joni’s at all.
..and Horatio hasn’t been a young un’ since Joni wrote “Woodstock”
Do you guys think there’s any possibility Al Gores new project will become a success? The deniers smear campaign has been quite extraordinary and very, very, successful the past several years.
The smear campaign might have been successful with many people.
But there are many other people frustrated with politics/media/business and-or disappointed with the failures engendered by congress obstructing Obama’s chances that they could gratefully, maybe eagerly, join Gore’s endeavour.
Give it time.
I am sure a whole lot of people support Gore communicating the climate crisis. My concern is they do so in silence. For a project like the Climate Reality Project which entirely relies on people spreading the word silence does no good.
I have, for instance, hardly seen any of the biggest climate blogs (apart from Climate Progress that did a piece months ago) writing about this project. If not even bloggers communicating the same message supports it, then who will?
Gore’s endeavour?
I didn’t know, either, but taking a clue from our friend Mr. Hank Roberts turned this up:
http://climaterealityproject.org/the-event/
A circum-global 24-hour broadcast extravaganza. Daniel is right, we should spread the word. We’ve got 2 days to do it.
(Though, Daniel, why did you leave it to me to dig up a link. . . ? Oh, never mind.)
For anyone interested. Climate Science Legal Defense Fund.
I hope that Tamino does a post to publicise this.
Hey Tamino, if I had any idea who you were or how to verify your identity, I’d be glad to nominate you for the following:
(fwd from the AGU below:)
New Award to Raise Awareness of Climate Change
A new award, the AGU Climate Communication Prize, has been established to recognize excellence in communication on climate change, to highlight the importance of scientific literacy and message clarity, and to foster understanding of and respect for science-based values.
The $25,000 cash award is sponsored by Nature’s Own, a company based in Boulder, Colorado, that specializes in the sale of minerals, fossils and decorative stone specimens. Any AGU member scientist is eligible to receive the AGU Climate Communication Prize. The award will be presented during AGU’s 2011 Fall Meeting in San Francisco, 5-9 December. Nominations will be accepted through 30 September.
Nominate a fellow AGU member today!
[Response: I’m flattered. But I’m not an AGU member.]
Anyone else having trouble getting into RC?
Tamino, you might want to contribute to the conversation between Pielke Sr. and Dikran Marsupial over at Skeptical Science, where Dikran expressed concern (here and then in the next comment here) that Pielke is seeing signals in time periods so short that it is unreasonable given the noise, and Pielke responded by recommending Lucia’s Blackboard discussion of the topic.
Finally, the AGGI (THE NOAA ANNUAL GREENHOUSE GAS INDEX) has been updated, the Ozone Depleting Gas Index not yet and eager to hear. To sum it up, the GHG increases are good for 1.1 Watts / Meter square addition in radiative forcing since 1979 and increased 1.4% in effect over 2009. In excerpts
Oh, god, Judy strikes again:
judithcurry.com/2011/10/18/does-the-aliasing-beast-feed-the-uncertainty-monster/
Judith IS the uncertainty monster…
Judy’s blog is full of the sorts of anti-science posts I hate. You get some expert in some narrow little field thinking he’s overturned all of climate science just because it doesn’t conform to the narrow, little norms of his narrow, little discipline–at least in his narrow, little mind.
This is precisely the sort of thing that makes me wonder whether Judy even understands the basics of climate science. It is not as if the signature of CO2 is subtle, and yet she manages to ignore (or deny) it.
Hope everyone will stop by Real Climate to leave a note of appreciation for Gavin Schmidt’s untiring & patient efforts.
OT except that it has to do with conceptualizing big numbers and it has (almost) nothing to do with Brittany Spears.
In honor of Halloween’s auspicious marker of the day the population of our spaceship crosses 7 billion, the BBC has an illustrative website:
http://www.bbc.co.uk/news/world-15391515
I’m # 2,707,963,284 or so,
or:
#76,047,571,234 or so,
(depending on how you look at it.)
On one of my periodic crawls through the web archives I managed to find the largest remaining block of lost Open Mind Posts, January and February 2010 (links added to the Open Mind Archives at Skeptical Science):
Jan 13, 2010 Models
Jan 19, 2010 Hottest Year
Jan 27, 2010 Post Dispatch
Jan 30, 2010 It’s a slow week
Feb 4, 2010 Skikda
Feb 6, 2010 Gridiron Games
Feb 8, 2010 Combining Stations
Feb 8, 2010 The Real Climate McCarthy
Feb 13, 2010 Prime Meridian
Feb 15, 2010 Dropouts
Feb 15, 2010 Open Thread #18
Feb 15, 2010 Summer and Smoke
Feb 16, 2010 Growthgate
Feb 18, 2010 Cherry Snow
Feb 22, 2010 Statistical Geometry
Feb 22, 2010 Snow
Feb 23, 2010 GHCN: preliminary results
Feb 25, 2010 False Claims Proven False
Feb 25, 2010 Shame
Feb 25, 2010 Interesting Comment
Feb 25, 2010 Show and Tell
Feb 26, 2010 Thanks
Feb 28, 2010 Update
Mar 1, 2010 Replication, not repetition
Mar 5, 2010 Global Update
Mar 5, 2010 Message to Anthony Watts