Hackers & Global Warming

Remove all ads!

The worst aspect about data being lost is there is always a stigma that the data was either altered or flat out fabricated to begin with. Data that goes "missing" is highly suspect.

 

What I mean is that the first time they refused to reveal their data (which was a number of years ago), the entire community should have come down on them. Now, yes, the lack of a proper peer review in the original publications is only a reflection on the publishing journal, but the refusal to release the data to those who asked for it, as I have said before, should have been a huge red flag to everyone in the community. The fact that it was downplayed or ignored is a severely negative reflection on the community as a whole.

Now, of course I'm not saying that every climatologist is to blame. I'm sure plenty of them never even knew that such data had been refused, or even that the requests had been made. Considering the apparent high profile of a few of the requesters in the climate debate, though, I can't believe that only a "handful" of scientists ever knew about it.

 

Well, even if it were a few dozen scientists who knew about it, that would still be a tiny fraction of them. With large studies like this, while there will only be a couple of scientists listed as the main authors, there are likely to be a couple dozen "contributing authors" in the paper. The best indication as to what is going on is how those contributing authors are reacting. If you see them distancing themselves from the research, that would be a clear sign to me that the main authors are being blacklisted.

 

Aldeth, again, you're missing my point. My point is that the reaction, any reaction, should have happened years ago. The first time they refused to give out their data, this stuff should have happened. They obviously weren't blacklisted at the time. This tells me that any reaction now will simply be based on embarassment at being caught, not that it happened in the first place.

Since you also seem to have misinterpreted my comment about the "handful" of scientists, let me illustrate. I'll break the climatology field into a few tiers. The first tier are the core researchers at East Anglia. The second tier are the associated researchers who helped them put together their work, some likely at East Anglia and others spread throughout the world. The third tier are those who "peer-reviewed" the research and/or those who work with the publishers and would have inside information about that process. The fourth tier are those who are not directly related in any way, but are active in the community and in the climate debate. The fifth are those who are more-or-less uninvolved.

With those tiers, the first certianly knew what was going on, and are likely the only ones who know today what actually happened to the data (was it lost, never saved, maliciously deleted, whatever). The second tier may or may not know what happened, depending on how closely they were involved, but probably got drawn into the legal wranglings in one manner or another. They'd at least have enough information to start asking questions. The third tier also know about the refusal to give up data, and may have also been pulled into the legal wranglings, though less certainly. The fourth group are probably the bulk of the climatologists, and would have heard about the legal wranglings and the researchers' refusal to cooperate. The fifth group may also be rather sizable, though probably not nearly as large as the fourth, and probably had no idea what was going on.

Still, that's four tiers that should have at least known that information was being withheld years ago. 2-3 of the four, including the largest, should have raised an alarm. None did.

 

NOG,

Perhaps I'm missing your point because you're not making a very good point. Let me go over some of your last post to try to clarify some stuff for you. I'm not a climatologist, but I imagine that the way data is handled, requirements for publication, etc., are pretty similar regardless of the field.

You do realize that an organization is not under any obligation to give out their original data to anyone who wants it, correct? Research is covered under intellectual property law. What gets published in a scientific journal is protected by copyright law. You can chose to release information in a journal without releasing the original data. The only reason my university can allow people to look at my original notebooks and data that I collected for my thesis is that I authorized them to do so.

I suspect that the main reason that they weren't blacklisted when they refused to release the original data years ago is that they were under no obligation to do so. And there are plenty of very valid reasons for not releasing your original data, the most obvious being that you want to do follow-up experiments on it, and get credit for that subsequent research. If you release your original data for anyone to see, they can perform the follow-up experiments, while not having to do all the hard work you did for them. Not releasing original data is not only allowed, it is quite common.

Why, in a field that has researchers all over the world, are the researchers specifically at East Anglia considered the first tier in the field of climatology? Maybe as it pertains to this particular research that is true, but saying that they represent the first tier of the entire field is grossy overstating their importance.

Oh my... do you really think that's how research works? The only thing that the associated researchers get is an advanced copy of the paper before it goes to peer review to confirm that their contributions were not incorrectly used. Most associated researchers are simple citations in the body of the article, and their "contribution" is a footnote in the bibliography. They are allowed access to the original data, but they aren't allowed to keep it. It's the IP of original researchers.

Oh come on NOG. The second tier doesn't even have the inside information. They would see the original data insofar as it is necessary to do the review, but they do not have the rights to retain that information. And FYI, the people who perform peer reviews do NOT work with or for the publishers. As the name implies, a "peer review" is performed by a fellow researcher. The fourth and fifth tiers are so far removed from this process that they aren't worth commenting on.

The last two groups, who have no driect ability to affect anything, are undoubtedly the largest groups. That aside, even if everyone in the whole world knew they refused to release their original data, it doesn't really matter, because, as I stated, they are under no obligation to do so. The problem is not that they didn't release the data, the problem is that they didn't retain it.

You're casting a very broad net that encompasses far more people than could possibly be held accountable. I'm more than happy to help inform you on how to process works in research, but it can be tedious when you're letting a lot of accusations fly when you admit to being not a fully informed source.

 

BTA,

I attended a public university, and I had to sign a release to allow the university to allow others to view my original data. When I graduated, they asked me to sign a blanket statement that basically said anyone can see it, which I did. I do not think the researchers themselves are under any obligation to release their data for all to see.

As for journal publication, there is absolutely a methods sections detailing what you did, and how you acquired and collected the data. However, it is quite rare for the entire dataset to be published. To use my research paper as an example, the journal article was 12 pages, whereas my thesis was over 150 pages. Most of the unseen 138 pages were spectroscopy data. Scientific journal publications would be the size of muliple copies of Merrian-Webster unabridged dictionaries if they included everything (as it is they are about the size of one copy).

I've also peer reviewed journal articles. While you are allowed to see the original data, they make it quite clear to you that you cannot keep the data, or make photocopies or other reproductions of it. Additional research is the lifeblood of scientists, and releasing original data seriously hurts your ability to get follow-on work.

I'm not saying it was OK for the scientist to not keep the data. All I'm saying is that their initial refusal a couple of years back to volunarily release the data may not have raised too many eyebrows. We now know that they were reluctant to release the original data because of what it would reveal (or as the case may be, not reveal), but that ulterior motive was not plainly evident at the start. If a researcher refuses to release his data, it does not necessarily indicate that he is hiding something.

 

In this case, without the original data and procedures, their results are impossible to replicate, thus impossible to validate, thus the research isn't repeatable, thus it isn't real science.

The problem with that is that they already revealed that their data sources were publicly accessable (for most nations, at least). This means protecting their data is out the window, as anyone could collect that data (or at least a similar set) themselves. This doesn't allow for repeatability, though, because they didn't define which data points were used. Essentially, privacy of data is a non-issue, but repeatability of the experiment still is.

Sorry, I should have explained the reason I was breaking tiers up like that. I was breaking them up by proximity to the situation, not by importance in the field.

That depends on how thoroughly they're involved in the project, doesn't it? As for the tier position, though, I believe it stands simply because the people trying to get the data will probably be talking to every name on the research paper to try and find some way in.

I think you misunderstood the purpose of those tiers again. As above, I put them in this tier because they'd be likely to know about either problems in the original data or the legal wrangling to get that original data. As these people are supposed to have access to the original data itself, they would know if there was no data. As they are major players in the field, they'd likely either be contacted in the legal wrangling or at least hear about it.

I don't know how it works in climatology, or in your field, but in CFD, the peer review is often done at the behest of the journal publishing the research. Of course, CFD is all computational work, with no physical components, so it'd be easier for a random researcher to reproduce the work than in, say, particle physics. Then again, the climatology work is also all computer simulation.

The fifth tier I agree with you on. I included them only for completeness. The fourth tier, I have to disagree with you. While, again, I'm no climatologist, I have to imagine that their community works on a similar level to CFD research, which means people hear about what each other are doing, and what kinds of problems their running into. Add to that the fact that this entire field is central to another community: the AGW debate, and I imagine word travels within this group. While I doubt any outside researchers would have access to the original data, they may well hear if someone is being involved in court proceedings or the like.

This makes no sense to me. It's fine for them to withold the data from everyone else, as long as they hold onto it themselves? What's the point of that? Unless you determine that someone, somewhere has the right to access that data, there's no point holding onto it in the first place. Add to that the fact that privacy of the data isn't a concern and I fail to see any reason to withold it in the first place.

I'm assuming it works similarly to CFD. You are obviously assuming it works similarly to your field. Since neither actually are climatology (the field of discussion) and the two are apparently that different from each other, I think it's safe to assume neither is guaranteed to be an accurate representation of the climatology field.

If they are being employed by someone, then odds are their data isn't their own. If they are state employees, and their data is work-product, then it doesn't belong to them. If they are outside contractors, then their contract probably specifies the status of their data, as well as who owns it. I believe, however, that they are state employees. Unless their research was somehow exempted from FOI requests, such as it being secret (yeah, right), they had a legal obligation to release the data.

 

I've covered this already. In order to replicate research, you don't need the original data, you need to know what they did. This is detailed in the "Method" section of the research paper, and the "Method" is published in every scientific journal. If you wanted to replicate my research, you don't need to know every data point I collected, just how I went about collecting my data points. You'd do what I did, publish your composite results just like I did, and see if your composite results matched my composite results.

I covered this also. They would also know that the researchers aren't required to give original data. Everything necessary to repeat their experiments was in the published article. Further, as I also stated, while these groups would have access to the original data when doing their reviews and signing off on their contributions, they would have either returned the original data, or erased it if it was electronic. At least, that's what you're supposed to do if following protocol.

Because after a certain point in time, copyrights and the like expire, and then anyone can access it. Also, once no further research on a subject is planned, most researchers usually do release all their records. Just because most researchers are reluctant to give up their data when they plan on doing follow-on research, it doesn't mean that they NEVER are willing to give up any of their research. After they have completed their follow-on work, they are willing to release it.

Let me try to explain by example. You perform Experiment A, and based on those results, you plan on doing experiments B and C. After performing experiments B and C, you see that B is a dead end, but experiment C, leads to experiment D and E. At this point, most researchers would release their records in totality from experiments A and B, because they are done with them. They would not release the full data of experiment C, because they plan further work on it.

I know nothing of CFD, so I'm not willing to put forth any assumptions on what they do in that field. However, climatology is a subfield of ecology. (Note, in the scientific sense, "ecology" does not study things like going green and other similar items. Ecology in the scientific sense is totally different than how it is used in the vernacular.) Ecology is a subfield of biology. The three hard sciences - biology, chemistry, and physics - all use standardized techniques in their research papers, and those three fields and their subfields are distinct from research papers you'd see in the social sciences. So my research in biochemistry uses similar reporting techniques as all hard sciences do, including climatology. CFD is not a subset of the three hard sciences, and I am not nearly as confident that the techniques are similar.

It's true that most researchers are employees of private companies. However, companies have just as much financial incentive to not release the full data as the researchers themselves. And this is legally covered as well. They refer to this knowledge as "trade secrets". To use a more mundane example, Pepsi is allowed to withhold information of exactly how it makes Pepsi from the public because their formula is a trade secret. They list all the ingredients on the outside of the can, and this is similar to what would be considered the composite results.

 

I'm sure that other columnists more skilled than this hack have made the point, but she makes it quite well. Bear in mind that I am opposed to the Green agenda, but I also consider myself to be an honest man, and I do respect the opinions of others even if I think those opinions, if crystallized into public policy, would spell disaster for millions of people worldwide.

 

Climate research, beyond weather forecasts, is far less a domain of private industry when compared to biology, chemistry and the like. It is as far as I understand it still pretty firmly in the hand of the international universities, because there is little profit in it. It is largely still research on fundamentals.

I think ecological research can by it's very nature only be lab tested and replicated to a certain extent. So an ecologist predicts that a certain factory and it's waste water will very probably annihilate all the fish in the river? Well, let's fact check this claim, build the factory and see whether he was right? If all the fish perish, we'll have proof and he will be vindicated and we can all laud his brilliant foresight. Except that there is no fish left. Well, that's the price to pay for certainty and not making any rash moves that could cost money, private money! Ecological research is by it's very nature predictive and to an extent speculative. That doesn't mean that the results, only because they are very difficult to be lab tested, are wrong or unreliable or unscientific. They just do have, must have, a different methodology.

I find it interesting that there are in fact industries taking climate research very seriously, and do not share the manufacturing or carbon industry's general revulsion against climate research data (i.e. being forced to invest in no-profit measures like cutting emissions).

I read about a large coal power plant in the US, iirc in Ohio, that emitted so much sulphur dioxide that the company eventually bought up all the ground in the neighbouring town because it had become a pain to live there. The townspeople must have been almost literally been breathing acid on foggy days, and with all the vapour from the cooling water that must have been rather often.

I am familiar with power plant climate. Where I was born there are several large coal power plants with neighbouring villages. They tackled their sulphur problem by building in filters, which was rather expensive, but they were required to do so by law. Amazingly, the smoke from the chimneys is white-ish nowadays, not brown, and people now can dry their laundry in their gardens. To the relief of all the ladies in the neighbourhood, their nylon stockings won't ... dissolve any more as well.

The industry taking climate research seriously is the (long-term thinking) re-insurance business, the insurances that insure insurance companies against risks like having to pay for freak accidents like 9/11, or hurricanes or rising sea levels. These are sober people, and I see their keen interest as something that suggests the matter ought to be taken serious.

 

You would think that information would be contained in the Method. Moreover, in the case of raw or homogenized data was used, that's a big enough piece of information that it should be included in the paper as well. If that's all people were asking for, I'd question why it wasn't in the paper to begin with.

 

This heavily depends on the type of research going on. Between the interpolation used and the adjustments (both legitimate and not) made to the data, using the original data is much more important in a computationally based system like this or CFD than in most sciences. It's sort of like running the same algebraic equation, but with different values for the variables. You can't confirm that the original person did their math right unless you use the same values for the same variables, even if you know exactly what equations were used. In physics, doing the same thing will (or at least should) get the same results, so knowing what data points were used is unimportant. Similarly in chemistry. In CFD or any kind of statistical survey-based research, though, knowing the original data is the starting point of the research. Without it, you really can't reproduce the work. You may be able to design and run your own project, but to expect that to confirm the work of another project is ridiculous.

But that's no justification for requiring them to hold onto the data. That just makes it polite and a good idea. The only justifications for saying someone should hold onto the data is either to allow repeat esperiments or continued research.

Actually, my guess is that CFD would be a subset of physics. Either that or computer logic. Like most engineering fields, it's a pretty good mix of the two. Anyway, the kind of work being done in climatology is drastically different from the work being done in general biology, or even general ecology. It's statistics, interpolation, and computer simulation, which is actually exactly what CFD is, though CFD has a much smaller statistics base.

Ah, but East Anglia University isn't a private company. It's a public, state-run, tax-funded university. It'd be like if your local wastewater collection facility (assuming it's state-run over there) refused to release their flow data, even under FOI request.

I'll agree that the incident in itself doesn't destroy climatology. As I'm arguing, however, I think the implications of the lead-up hurt it substantially.

True. Ideally, they should be matched with past data and, as new data comes in, they should be compared to that as well. The typical scientific experiment is out the window, however the simulations and statistical analyses should still be repeatable. These forms of work, however, are highly dependant on the original data. One single point difference in one data point out of thousands can change the results drastically.

Eh, I'd still be careful in those foggy days. There's a lot more coming out of those coal-fired plants than sulphur dioxide. Not all of it is regulated and even what is is only regulated within certain parameters (meaning they can pump a certain amount out). On top of that, in most localities, the enforcement for those regulations is weak. They poison the air of millions for a few days and maybe have to pay a few hundred thousand dollars. It's probably less than it would cost to have prevented the release in the first place. Add to that politics and you can get some real nightmare scenarios. There's one plant near Richmond that should have been shut down a decade ago or so. They aren't even trying to comply any more. Because they feed a substantial part of DC, however, and only polute Richmond and the surrounding areas, the big-wigs in DC told the not-so-big-wigs in the Virginia State government to eat it and let the plant stay open.

 

Who's requiring who to do what, here? University researchers are asked to sign a release to have their original data released for a reason -- they may work for a University, but their work is still their intellectual property. Doing independent research at a University is not the same thing as working for Dow Chemicals, nor should it be.

It isn't like that at all, and I addressed this above. East Anglia University isn't a private company -- if it were, the original data really would have been theirs to do with as they please. Most, if not all, universities do not retain ownership of the intellectual property of its researchers. If a University were to suddenly start doing that, their researchers would likely start applying at different universities, since a substantial portion of a professor or researcher's income comes from publishing their work. It's a lot harder to publish a book based on your research when your research doesn't belong to you.

 

I don't know about other fields, but in CFD, the research belongs to the university, unless it was done under a grant. If it's done through a grant, and the university is simply providing the grant (which can include access to facilities), then the researcher retains ownership. The grant was a gift after all. If the research is done specifically for the University, or the State, or anyone else, with them paying for it, then the results belong to whoever payed for it. Contracts may specify publication rights, as well as joint ownership, but the University owns the data, at least in part. If it's a public university, and not secret research, then that means the data is subject to FOIA requests.

 

I concede that I don't know all the laws as they apply to research. What I do know is that when I did my research at a public university, they asked for my permission before they retained it as an open source that anyone could access. This certainly implies that when you do research at a university, there is at least a level of co-ownership between the researcher and the university. Additionally, I didn't have to ask the unversity permission the have my research published. If the university had exclusive rights to the research, I would have had to do so.

In BTA's response to me he reminded me that some of the information that was being requested seemed like it should have been contained in the original research paper. While I maintain that you don't need a point-by-point analysis of the data to re-perform the work, it appears that the original researchers were even reluctant to give out basic information such as what weather stations were used, which you certainly would need to perform similar research.

On a somewhat ancillary note, it is very rare for a second group of researchers to perform identical research to that already performed. There is a certain level of trust in the scientific community (that was evidently expoited by some of the researchers), and it is considered a waste of resources to do something that already had been done and published.

Most universities require all graduate and post-grad students to perform original research. Even if you aren't receiving a scholarship or grant money, you are at a minimum utilizing their labs and materials, and since original research is also one of the standards by which scientific journals accept material to be published, they at least want the possibility that what you're working on gets published nationally (because the name of the University appears next to your name in the article).

That, however, does not excuse the researchers from not revealing what weather stations were used in their research. A new study may have wished to sample stations that the original researchers did NOT use, and they'd need the list for that too.

 

It proves that science is just as political as anything else. The debates on the major issues are now completely adversarial. It's the same as anything else: "If you are on the other side, why would I do anything to help you?"

I suppose it has always been that way to a certain extent, but now it is total: "If you are not with us, you are against us." I'm a bit surprised that people see the scientific community as different in some way from other major institutions. Everything, and I mean everything, can become a political talking point these days.