3 ms·
They didn't conspire to prevent it from occurring. Some data was lost, and some bad data was destroyed. In fact, the data that was destroyed does not seem to be
by KirinDave 17y ago
They didn't conspire to prevent it from occurring. Some data was lost, and some bad data was destroyed. In fact, the data that was destroyed does not seem to be part of the data that was lost.
And Climate Science didn't need as much transparency years ago before the AGW controversy. Subsequent events rendered old data to be of more value, but we take for granted in late 2009 how inexpensive disks are. They were not nearly so cheap even last year, let alone years ago. However, the volume of data produced by climate science instruments is enormous. It's very unsurprising that some was lost.
But I and others keep pointing out: We know how to reconstruct that data from earlier published data (back before any of these emails were written, before any conspiracy you allege existed) and published code. Why is this unacceptable? It seems to me the reason is because you don't like that it's science butting it's rude head into the libertarian economic religion, pointing out resources are finite and liberating whole geologic ages worth of carbon back into the atmosphere can have consequences. Perhaps I am assuming too much, but that is the opinion I keep butting heads with on news.ycomb, which is libertarian-free-market-only-conservative-type rich to an unusual degree, at least for someone who lives in the Capital Of Liberalism.
- yummyfajitas 17y agoHmm, let me make sure I've got this straight. Some CRU folks conspired to hide data from skeptics. Other data was lost, but not deleted. We know this because the people conspiring to hide the data told us so. As for how to reconstruct the data, please explain how. Nearly all statistical data analysis methods in common use are highly lossy [1], which makes reversing them to get back the original data impossible. It's ridiculously unlikely that the entire CRU used only reversible analysis methods. Regarding your "but you are a libertarian, so hah!", I'll mostly ignore it, except to point out that while I am a libertarian on most things, I tend to support regulations which internalize externalities. If I believed the scientific case for AGW, I'd support a carbon tax. [1] E.g., moving averages hide the noise, or finding the mean/sdev discardes everything besides the mean/sdev.
- KirinDave 17y agoSome CRU scientists "conspired" to not release raw segments of data without processing. They didn't want to give it to "skeptics" because these "skeptics" were anomaly hunting the data to try and make an unscientific argument sound scientific. Other raw data of questionable value was not kept. While certainly not the best practices, it's understandable. As for reversing, we can produce a reasonable approximation of the original data because the "shady operation" was linear interpolation and the adding of a small corrective constant. (Cue dramatic musical hit). Of course we cannot recover the exact values, but we can reproduce likely approximations of the data. This is not deep stat here, it's interpolation and corrections. Go look at the code if you don't believe me. The article you're commenting on goes into P.S. I am not addressing the substance of your argument with the claim, "You are libertarian." I am addressing all your points as directly as I can. I'm offering the reason to other readers of this thread as to why I think you're persisting in this argument. This event would be a greater problem if there weren't multiple lines of evidence reinforcing the conclusion.
- yummyfajitas 17y agoIt's true that in this one specific instance, we can subtract yearlyadj to get back the original data (now that we have the code). I was referring to the raw data in general, not this specific case. If they did nothing more add known vectors to the data, we can reconstruct the original data once we have access to their code (which they also didn't release). The minute they take a moving average, or use a best fit curve, this becomes impossible. You also still haven't explained why you believe climate science deserves lower levels of care and transparency than econometrics.
- KirinDave 17y agoWell, this one specific instance is the subject of our conversation and what some folks are calling the "smoking gun" of a conspiracy. So naturally I thought that's what they're talking about. I am not so concerned with addressing how to recover the lost data, all I was concerned with addressing was that this statistical "trick" was nothing of consequence and not irreversible. People here (including you) have made much ado about changing the data, and I wanted to address that the "changed data" is not so changed at all. You can see what they're doing and the emails explain why. Finally, you asked why I believe that climate science deserves lower levels of care and transparency? I assume fom here on out we're not talking about the lost raw data (which was, from all claims, an accidental oversight) and instead focus on the withheld line of evidence. I didn't say that. For one, it certainly requires more care because–unlike econometrics–Climate Science is actual science. And I don't think care hasn't been exercised here. I think what you want is transparency on a level above and beyond what science normally requires. While ditching the raw data is regrettable and a mistake (probably on the order of NASA losing the original moon landing data), it shouldn't really change anything. Now, if the CRU was the lone voice trying ardently to make a case that AGW was happening and they lost the raw data to the sole line of evidence, that'd be problematic for the case of AGW. But CRU and other institutions have multiple lines of data that have been clearly published and explained that are in accord with the lost batch of raw data. So my opinion, which I try to base off the majority view of science, is unswayed. If the CRU folks had "hidden the decline" from public discourse and corrected for it in their data, that'd also be an ethics problem. But they didn't hide the existence of the observed temp decline, it's mentioned in published papers (as has been mentioned in the hn commentary to this story). P.S. I don't think the CRU folks have been particularly careless or particularly opaque compared to, say, physicists or astronomers or biologists. And I am not convinced that econometrics is a science. I think it's a blanket term for mathematical modeling, which is part of science. Perhaps there are charlatans confusing the issue, but I've seen some pretty absurd things written under the name of econometrics.