I've repeatedlyargued that we need to start building as many modern nuclear fission plants as possible. Preferably pebble bed reactors, using breeder reactors and reprocessing techniques to turn the waste into useful fuel.
And as I've explained on my homepage, I think that cap-and-trade will make coal less profitable, and nuclear power more profitable. It's a very capitalistic approach to the problem of climate change.
One of the catastrophic outcomes of climate change are large sea level rises due to ice melt in the polar regions. Presumably there are models that predict how this could occur with global warming. So the question is, do these data agree with these models?
The last article I read in Science compared model prediction of sea level rise, and found that observations showed the sea levels rising even faster than the models predicted. Perhaps this was just short-term weather, though: more recent measurements may indicate agreement with the models.
... the fact that the current increase in the CO2 level is 35x faster than at any other time in the last half-million years? That the last time the climate changed this abruptly wasn't very pleasant? Biospheres can adapt to gradual changes, but abrupt changes can be catastrophic.
Higher average global temperatures imply higher upper ocean temperatures, which imply a higher water vapor pressure. Thus more water vapor will evaporate into the atmosphere. Yes, Roderick 2007 showed that wind speed had a stronger affect on the evaporation rate than changes in temperature, but I doubt that affects the expected theoretical equilibrium vapor pressure from basic thermodynamics. When that more humid air is carried across a tall mountain range, its temperature decreases and the water precipitates.
As far as I understand it, Antarctica as a whole is warming more quickly than climatologists expected. Antarctica should be warming more slowly mainly because currently most of the land mass is in the northern hemisphere. The fact that Antarctica is warming at all is a little troubling.
You're right. Glaciers melt all the time for reasons unconnected to emissions of fossil fuels. However, the current warming is atypical in many respects (which I've linked in another comment in this article.) Glacier melt isn't- by itself- proof of the anthropogenic origin of abrupt climate change.
Exactly. I've described my research into Greenland's ice sheets. My most recent estimates show that Greenland as a whole is losing ~100 Gtons of ice every year, but my advisor believes my estimate is too low by a factor of 2. As you say, northern Greenland is gaining mass, but southestern Greenland is losing much more mass. Climate change is a very serious problem, and I'm really annoyed that health care is currently distracting the Senate from an issue that affects the future of the entire human race.
I was actually thinking of that guy in Colorado (Peilke?) who has long argued that global atmospheric heat content is what we should be talking about.
Looked around for Pielke's work mentioning heat content and found this. Is that a good reference? I agree that internal energy of the Earth is a more robust and useful variable than temperature, but I'd go one step further. That is, a much more useful variable would be the internal energy of the atmosphere and ocean combined. That would eliminate the spurious temperature swings associated with ENSO events that seem to mislead many people. This heat transfer between the atmosphere and oceans wouldn't distort such a metric.
You are correct that heat is only one form of internal energy, although physicists have a slightly different take on the nature of heat than chemists, so I don't agree with your characterization of heat as strictly a type of energy transfer....
Actually, I'm a physicist too. Never was that good at chemistry. I still think heat is a form of energy transfer, not a state variable. But I'll drop this argument because (like your point) it doesn't seem particularly interesting or relevant.
So yes, by all means be pedantic and talk about "atmospheric internal energy". That is a physically meaningful quality, whereas neither you nor anyone else has suggested why taking any kind of average of dry-bulb temperatures is in any way physically interesting. And if it is not physically interesting, it is not climatologically interesting.... All I can say is that I still don't understand what anyone thinks they are doing with global average temperature, but whatever it is, it isn't physics.
The internal energy of the atmosphere is a weighted mean of temperatures, where the weightings reflect differing heat capacities. A global average temperature cannot be used to determine the internal energy of the atmosphere because it isn't properly weighted (as I believe you're saying.) But as I've said, even an unweighted average improves the signal-to-noise ratio of temperature trends. More measurements improve the statistics in the same way multi-model ensembles improve climate predictions compared to single-model runs. The global temperature isn't intended as a formal variable, it's simply an easy-to-measure diagnostic of the global climate.
You're proposing that the "Urban Heat Island" effect is responsible for increasing temperatures. I've discussed this at length on the climate change article on my homepage. Multiple studies have shown that the UHI isn't responsible for recent temperature increases. Abrupt climate change is responsible, which is caused by our increasing the CO2 level ~30% above the value it's had for the last ~650,000 years. (These numbers taken from the EPICA ice core analysis, again already discussed at length.)
I wish someone would tell me how you compute the mean temperature of a composite substance like the atmosphere. Global atmospheric heat content is meaningful. Global mean temperature is not. Unless someone would care to explain how you actually compute it in a physically meaningful way?
This sounds similar to the arguments presented in a 2007 paper that's widely considered to be some kind of joke.
Perhaps you mean that different substances have different heat capacities. That's only a problem if you want to determine the equilibrium temperature, and even that's just a weighted average. But even an unweighted average improves the signal-to-noise ratio of temperature measurements, which is why climatologists routinely speak of global mean temperatures.
And to be really pedantic, "heat content" isn't physically meaningful either. Heat is a type of energy transfer across a thermodynamic system boundary. Systems don't store heat, they store internal energy, which is also measured in Joules but can be transferred as heat or work. (Yes, this distinction is irrelevant. That's my point.)
Incidentally, whenever I get some free time I plan to copy one of your older comments to the climate change article on my homepage and answer it. That's because compared to most other people arguing against abrupt climate change, you seem significantly more scientifically literate. I'd email you when this happens, but I don't know how to get in touch with you other than comments like this one.
+10 Insightful. I'm planning something like this myself, using open source textbooks, problem sets, and lectures freely available on Youtube. Or perhaps someone can suggest a better video website for long (~50 minute) lectures with high resolution for the equations recorded from an overhead projector- my tool of choice in the classroom. Any other suggestions for a free online physics course? (I can be contacted using the email address on the "about" page accessible from my homepage.)
Also, as you can tell I've never used threads before (I think fork counts as a separate process, not a thread). One reason I wrote this code was to reassure me that the program hasn't stalled when the calculation takes weeks to finish. It's more reassuring to me that the print statements are actually in the same thread (nay, the exact same loops!) as the calculation itself. If the timer was in a different thread, I'd have to think carefully about how to guarantee that a crash in the calculation thread was reflected correctly in the print statement from the notification thread...
I just learned how to use fork()... so I'm fairly new to parallel processing. If the print statement was in a different thread, how would it access the completion percentage of the main calculation? Also, every time the notification prints the local variables are saved to disk ala autosave. I wonder if it's possible to do that in a "different thread."
But honestly, I get 80% of the way there encounter a few interesting problems, overcome them, get 90% of the way there, encounter some pain in the ass problem and stop. After all, the code technically DOES most of the interesting things I want it to do, just not without a small manual tweak here and there. Am I the only one that gets bored like this?
I'm the same way; short attention span and insufficiently masochistic to bother with the finishing touches. But I do enjoy removing the need for manual tweaks. I once spent a ridiculous amount of time automating a system to notify me that a calculation has, for instance, "12 days, 7 hours" remaining. It's easy to write a function that does this but requires frequent manual tweaking, but a fully automated version was surprisingly involved (and still unfinished.) Eventually, an autosave/autorecover functionality was built into this function which is guaranteed to run at a user-specified interval to avoid slowing down the calculation with too-frequent saves but still protects against hardware/software glitches. I like this kind of proactively lazy coding, and often consider my job to be that of building a science factory that is mostly (and hopefully intelligently) automated.
Yes, but with an eigenvalue >10x smaller than the principal component from the real data. Thus the red noise fit isn't statistically significant, but the real data fit's is. That's the point of the Monte Carlo analysis done by Mann et. al. which was mentioned in the first link I gave.
Graduate education should include mandatory classroom instruction, and a heavier emphasis on giving presentations. I regularly suffer through my colleagues' miserable presentations at conferences, so I strongly believe that scientists need to develop better communication skills.
Re-orient science classes so they emphasise curiosity and skepticism rather than rote memorization. I've previously complained about the sad state of science education in high school and general collegiate physics courses. Some people still believe that the seasons are caused by the Earth moving farther towards/away from the Sun, and that sinks/bathtubs drain differently in different hemispheres. Maybe if science classes actually taught people how to think like scientists, these silly myths wouldn't be as widespread. Maybe people would even be interested in science in general if they didn't see the subject as a bunch of equations to be mechanically applied.
The scientific community has a tendency to ignore bizarre claims because they don't want to give credibility to people who believe in things like creationism, electric universe, climate-change-denialism, moon-landing-hoaxers, relativity-deniers, etc. This isn't very productive, because some people apparently get the impression that scientists dismiss these fringe views because of a massive conspiracy of suppression. I think it's a better idea to slowly and patiently explain why these examples of pseudoscience simply aren't consistent with the available evidence. I'm trying to do that on my homepage, but there's only one of me versus a horde of pseudoscientists...
Science journals need to be made open source, like PLoS ONE and ACP. Maybe the general public's science illiteracy is partially based on the fact that crackpots publish their "research" freely on the internet (which is why the internet is now a tarpit of scientific misinformation), whereas scientists publish articles in peer-reviewed journals that can't be accessed by anyone outside of a major university.
As you can tell, I think this article touches on a very serious problem. Sagan said it best:
"We have designed our civilization based on science and technology and at the same time arranged things so that almost no one understands anything at all about science and technology. This is a clear prescription for disaster."-- Carl Sagan
I'm referring to this quote: "I've now done some stuff with random series rather than the MBH proxy series. This has the advantage of allowing you to create as many proxies as you like. I'll hive that off to a separate page: here. What that appears to demonstrate is that M&M are right about one thing: it often does lead to a 'hockey stick' shape in random data. But the problem is that the variance-explained of the PC1 done this way is tiny: the first eigenvalue is about 0.03. Whereas when you run it on real data the first eigenvalue is about 0.55 (back to 1000) or 0.38 (back to 1400). Which means the two problems are very different."
In the other link, the eigenvalues are supposed to be accessible via a link, but I can't get figure 1 to display. Again, don't know if this is just me. Regardless, they're saying much the same thing. The eigenvalues of the MM fit to red noise aren't statistically significant.
But the real point is that the same answer emerges from more straight-forward analyses that don't rely on PCA at all (which avoids all these issues). In fact, as I've mentioned in my article, multiple independent analyses have been performed, all of which agree that the hockeystick shape is accurate.
Oops. The 35x faster link referred to CO2, not temperature. My bad.
Chapter 3 of the 4th IPCC report says temperatures in the last ~30 years have increased faster than at any point in the last ~1000 years, a rate which is steadily increasing.
This isn't nearly as impressive as the anomaly in the CO2 record compared to the last several million years of proxy data. But Meehl 2004 shows that the warming since ~1970 is primarily caused by anthropogenic emissions, and they used models that are consistent with a climate sensitivity having a maximum likelihood value of 2.9C, with a 95% confidence that it's less than 4.9C but greater than 1.7C.
Just to be clear, I want to stress that when I said the case for negative feedback was very weak, I didn't mean the long-term feedback that acts on slow, natural forcings. It's clear that the climate must be relatively stable (i.e. negative feedback) with respect to slow, natural forcings otherwise the Earth wouldn't have been hospitable for the evolution of life. But the rapid, unprecedented speed of CO2 increase is likely to involve a different mix of positive/negative feedback effects than the ones that stabilized the climate before our arrival.
However, its a question whether the climate reacts to warming by positive feedback, and if so how strongly, or by negative feedback.
There's a debate about how much positive feedback exists, but the case for negative feedback is very weak. For example, events such as Heinrich and Dansgaard-Oeschger events are the best examples of abrupt climate change in the paleoclimate record. These ancient events are worrying because they show the climate has a propensity to shift quickly from one state to another, given even small forcings. That requires positive feedback.
Also, the estimated magnitude of the Milankovitch cycles and other forcings are insufficient to account for the temperature variations observed in ice cores from Vostok and EPICA. This requires positive feedback. In fact, the estimates of positive feedback are too small to bridge the gap.
The decisive evidence for feedback would be if the climate were now genuinely warming faster than or differently from ever before.
Approximately 35x faster, which isn't surprising because of the unprecedented (in the last 2 million years) CO2 levels. Also, the warming is happening after the CO2 increase, which makes this warming qualitatively different from all previous deglaciations.
And this is where the question of the refusal of the climate science community to reveal their data becomes important.
Proxy data are available, Wahl and Ammann have made their code available, the CMIP3 database makes model output public for researchers to perform comparisons, etc. I've previously complained about the (widespread) tendency of scientists to keep their data private to wring every last discovery out of it before making it public. It's worrying, but not a problem unique to climatology. Nor ar all climatologists so hesitant to release their code and data. I publish all my code under the GPLv3, for instance.
As a scientist looking at the global warming debate...
Just curious: what degree in which field of science? I've described my research here, for quid pro quo.
... it always marvels me that no-one ever talks about the effect that water vapor has on global temperatures, given that water vapor has a heat capacity of 1.8 kJ/kgK, while CO2 has a heat capacity of only 0.8 kJ/kgK, and also given that the concentration of water vapor on average is composes 10X more of the atmosphere than CO2, meaning that the total impact of CO2 is about 20 fold less than water vapor, which is itself highly variable in concentration. Why doesn't anyone ever complain about the deleterious effects of water vapor in the atmosphere? Why aren't we moving to ban hydrogen vehicles that put out huge amounts of water vapor?
I've talked about water vapor in depth, repeatedly. As I've explained, water vapor is a feedback, not a forcing. It's not dangerous because it doesn't remain in the atmosphere long and isn't well-mixed to the top of the atmosphere. That's why legitimate peer-reviewed journal articles don't "complain about the deleterious effects of water vapor in the atmosphere."
What this discovery points out is that, well, maybe we don't really have a handle on this global warming thing, and that we shouldn't cut off the arms and legs of our civilization with an environmentally friendly electric chainsaw before we have a full grasp of what is going on here.
First, this discovery isn't related to abrupt climate change in any significant manner. Second, the goal of the legislation in the Senate is to jumpstart a new industrial revolution. No chainsaw involved.
Therefore, if you want to decrease humanities carbon emissions without murdering people, or subjecting them to poverty, you need only remove all regulations limiting the implementation of nuclear power around the world.
Murderous hyperbole aside, this isn't far from the mark. I'd say we have to make small tamper-proof nuclear reactors like SSTAR available to developing nations, but keep the reprocessing and enrichment technologies tightly restricted. Actinide poisoning of the fuel (to make weaponization more difficult than simply starting a clandestine enrichment program) is also probably a good idea.
I can't get that graph to load (but my net connection has been flaky lately so it could be my fault.) At any rate, it sounds like a claim that MM have made: that sending "red noise" into the MBH98 program results in a hockeystick. The main problem is that the extracted trend explains very little variance relative to the trend extracted from real data. Here's a 4-part primer on PCA to help people understand the basics.
Do you have any comment on the link I gave regarding the Nature correction?
I read some of it, and their complaints sound very similar to what other scientists go through when trying to get their research published. Peer-review is often an unpleasant process because it's based on confrontation, but this is true for everyone. In this particular case, I think Nature was right to reject their article based on the mountain of evidence against their claims.
It doesn't show a weakness in the process, it shows that computation power isn't infinite. Redoing all the calculations without the benefit of PCA requires use of a large cluster for a long time. This was done (in point 5) and shows that any PCA errors were negligible. Scientists aren't evil monsters engaged in a massive conspiracy. Really. We're ordinary people, just like you.
Obviously you didn't bother to read anything in those links. As I've said, many studies have examined the PCA centering issues, and found no significant differences when using different methodologies. I'm baffled as to why you find it necessary to tell me to google the Wegman Report, when I discussed it at length in the first link of my previous comment. It's perfectly normal for laymen to be confused about these issues (they ARE complicated). But it's ludicrous to suggest that the scientific community as a whole is somehow unaware of these issues or engaged in a massive conspiracy to suppress them.
Nevermind. I don't feel like repeating myself ad nauseum. It's not possible to reason someone out of a position that he didn't reason himself into in the first place.
Slashdot Mirror
I've repeatedly argued that we need to start building as many modern nuclear fission plants as possible. Preferably pebble bed reactors, using breeder reactors and reprocessing techniques to turn the waste into useful fuel.
And as I've explained on my homepage, I think that cap-and-trade will make coal less profitable, and nuclear power more profitable. It's a very capitalistic approach to the problem of climate change.
The last article I read in Science compared model prediction of sea level rise, and found that observations showed the sea levels rising even faster than the models predicted. Perhaps this was just short-term weather, though: more recent measurements may indicate agreement with the models.
... the fact that the current increase in the CO2 level is 35x faster than at any other time in the last half-million years? That the last time the climate changed this abruptly wasn't very pleasant? Biospheres can adapt to gradual changes, but abrupt changes can be catastrophic.
Higher average global temperatures imply higher upper ocean temperatures, which imply a higher water vapor pressure. Thus more water vapor will evaporate into the atmosphere. Yes, Roderick 2007 showed that wind speed had a stronger affect on the evaporation rate than changes in temperature, but I doubt that affects the expected theoretical equilibrium vapor pressure from basic thermodynamics. When that more humid air is carried across a tall mountain range, its temperature decreases and the water precipitates.
As far as I understand it, Antarctica as a whole is warming more quickly than climatologists expected. Antarctica should be warming more slowly mainly because currently most of the land mass is in the northern hemisphere. The fact that Antarctica is warming at all is a little troubling.
You're right. Glaciers melt all the time for reasons unconnected to emissions of fossil fuels. However, the current warming is atypical in many respects (which I've linked in another comment in this article.) Glacier melt isn't- by itself- proof of the anthropogenic origin of abrupt climate change.
Exactly. I've described my research into Greenland's ice sheets. My most recent estimates show that Greenland as a whole is losing ~100 Gtons of ice every year, but my advisor believes my estimate is too low by a factor of 2. As you say, northern Greenland is gaining mass, but southestern Greenland is losing much more mass. Climate change is a very serious problem, and I'm really annoyed that health care is currently distracting the Senate from an issue that affects the future of the entire human race.
Looked around for Pielke's work mentioning heat content and found this. Is that a good reference? I agree that internal energy of the Earth is a more robust and useful variable than temperature, but I'd go one step further. That is, a much more useful variable would be the internal energy of the atmosphere and ocean combined. That would eliminate the spurious temperature swings associated with ENSO events that seem to mislead many people. This heat transfer between the atmosphere and oceans wouldn't distort such a metric.
Actually, I'm a physicist too. Never was that good at chemistry. I still think heat is a form of energy transfer, not a state variable. But I'll drop this argument because (like your point) it doesn't seem particularly interesting or relevant.
The internal energy of the atmosphere is a weighted mean of temperatures, where the weightings reflect differing heat capacities. A global average temperature cannot be used to determine the internal energy of the atmosphere because it isn't properly weighted (as I believe you're saying.) But as I've said, even an unweighted average improves the signal-to-noise ratio of temperature trends. More measurements improve the statistics in the same way multi-model ensembles improve climate predictions compared to single-model runs. The global temperature isn't intended as a formal variable, it's simply an easy-to-measure diagnostic of the global climate.
You're proposing that the "Urban Heat Island" effect is responsible for increasing temperatures. I've discussed this at length on the climate change article on my homepage. Multiple studies have shown that the UHI isn't responsible for recent temperature increases. Abrupt climate change is responsible, which is caused by our increasing the CO2 level ~30% above the value it's had for the last ~650,000 years. (These numbers taken from the EPICA ice core analysis, again already discussed at length.)
This sounds similar to the arguments presented in a 2007 paper that's widely considered to be some kind of joke.
Perhaps you mean that different substances have different heat capacities. That's only a problem if you want to determine the equilibrium temperature, and even that's just a weighted average. But even an unweighted average improves the signal-to-noise ratio of temperature measurements, which is why climatologists routinely speak of global mean temperatures.
And to be really pedantic, "heat content" isn't physically meaningful either. Heat is a type of energy transfer across a thermodynamic system boundary. Systems don't store heat, they store internal energy, which is also measured in Joules but can be transferred as heat or work. (Yes, this distinction is irrelevant. That's my point.)
Incidentally, whenever I get some free time I plan to copy one of your older comments to the climate change article on my homepage and answer it. That's because compared to most other people arguing against abrupt climate change, you seem significantly more scientifically literate. I'd email you when this happens, but I don't know how to get in touch with you other than comments like this one.
+10 Insightful. I'm planning something like this myself, using open source textbooks, problem sets, and lectures freely available on Youtube. Or perhaps someone can suggest a better video website for long (~50 minute) lectures with high resolution for the equations recorded from an overhead projector- my tool of choice in the classroom. Any other suggestions for a free online physics course? (I can be contacted using the email address on the "about" page accessible from my homepage.)
Also, as you can tell I've never used threads before (I think fork counts as a separate process, not a thread). One reason I wrote this code was to reassure me that the program hasn't stalled when the calculation takes weeks to finish. It's more reassuring to me that the print statements are actually in the same thread (nay, the exact same loops!) as the calculation itself. If the timer was in a different thread, I'd have to think carefully about how to guarantee that a crash in the calculation thread was reflected correctly in the print statement from the notification thread...
I just learned how to use fork()... so I'm fairly new to parallel processing. If the print statement was in a different thread, how would it access the completion percentage of the main calculation? Also, every time the notification prints the local variables are saved to disk ala autosave. I wonder if it's possible to do that in a "different thread."
I'm the same way; short attention span and insufficiently masochistic to bother with the finishing touches. But I do enjoy removing the need for manual tweaks. I once spent a ridiculous amount of time automating a system to notify me that a calculation has, for instance, "12 days, 7 hours" remaining. It's easy to write a function that does this but requires frequent manual tweaking, but a fully automated version was surprisingly involved (and still unfinished.) Eventually, an autosave/autorecover functionality was built into this function which is guaranteed to run at a user-specified interval to avoid slowing down the calculation with too-frequent saves but still protects against hardware/software glitches. I like this kind of proactively lazy coding, and often consider my job to be that of building a science factory that is mostly (and hopefully intelligently) automated.
Yes, but with an eigenvalue >10x smaller than the principal component from the real data. Thus the red noise fit isn't statistically significant, but the real data fit's is. That's the point of the Monte Carlo analysis done by Mann et. al. which was mentioned in the first link I gave.
As you can tell, I think this article touches on a very serious problem. Sagan said it best:
I'm referring to this quote: "I've now done some stuff with random series rather than the MBH proxy series. This has the advantage of allowing you to create as many proxies as you like. I'll hive that off to a separate page: here. What that appears to demonstrate is that M&M are right about one thing: it often does lead to a 'hockey stick' shape in random data. But the problem is that the variance-explained of the PC1 done this way is tiny: the first eigenvalue is about 0.03. Whereas when you run it on real data the first eigenvalue is about 0.55 (back to 1000) or 0.38 (back to 1400). Which means the two problems are very different."
In the other link, the eigenvalues are supposed to be accessible via a link, but I can't get figure 1 to display. Again, don't know if this is just me. Regardless, they're saying much the same thing. The eigenvalues of the MM fit to red noise aren't statistically significant.
But the real point is that the same answer emerges from more straight-forward analyses that don't rely on PCA at all (which avoids all these issues). In fact, as I've mentioned in my article, multiple independent analyses have been performed, all of which agree that the hockeystick shape is accurate.
Oops. The 35x faster link referred to CO2, not temperature. My bad.
Chapter 3 of the 4th IPCC report says temperatures in the last ~30 years have increased faster than at any point in the last ~1000 years, a rate which is steadily increasing.
This isn't nearly as impressive as the anomaly in the CO2 record compared to the last several million years of proxy data. But Meehl 2004 shows that the warming since ~1970 is primarily caused by anthropogenic emissions, and they used models that are consistent with a climate sensitivity having a maximum likelihood value of 2.9C, with a 95% confidence that it's less than 4.9C but greater than 1.7C.
Just to be clear, I want to stress that when I said the case for negative feedback was very weak, I didn't mean the long-term feedback that acts on slow, natural forcings. It's clear that the climate must be relatively stable (i.e. negative feedback) with respect to slow, natural forcings otherwise the Earth wouldn't have been hospitable for the evolution of life. But the rapid, unprecedented speed of CO2 increase is likely to involve a different mix of positive/negative feedback effects than the ones that stabilized the climate before our arrival.
There's a debate about how much positive feedback exists, but the case for negative feedback is very weak. For example, events such as Heinrich and Dansgaard-Oeschger events are the best examples of abrupt climate change in the paleoclimate record. These ancient events are worrying because they show the climate has a propensity to shift quickly from one state to another, given even small forcings. That requires positive feedback.
Also, the estimated magnitude of the Milankovitch cycles and other forcings are insufficient to account for the temperature variations observed in ice cores from Vostok and EPICA. This requires positive feedback. In fact, the estimates of positive feedback are too small to bridge the gap.
Approximately 35x faster, which isn't surprising because of the unprecedented (in the last 2 million years) CO2 levels. Also, the warming is happening after the CO2 increase, which makes this warming qualitatively different from all previous deglaciations.
Proxy data are available, Wahl and Ammann have made their code available, the CMIP3 database makes model output public for researchers to perform comparisons, etc. I've previously complained about the (widespread) tendency of scientists to keep their data private to wring every last discovery out of it before making it public. It's worrying, but not a problem unique to climatology. Nor ar all climatologists so hesitant to release their code and data. I publish all my code under the GPLv3, for instance.
See Meehl 2004 for a primary source, but I've also recently discussed a very similar issue.
Just curious: what degree in which field of science? I've described my research here, for quid pro quo.
I've talked about water vapor in depth, repeatedly. As I've explained, water vapor is a feedback, not a forcing. It's not dangerous because it doesn't remain in the atmosphere long and isn't well-mixed to the top of the atmosphere. That's why legitimate peer-reviewed journal articles don't "complain about the deleterious effects of water vapor in the atmosphere."
First, this discovery isn't related to abrupt climate change in any significant manner. Second, the goal of the legislation in the Senate is to jumpstart a new industrial revolution. No chainsaw involved.
Murderous hyperbole aside, this isn't far from the mark. I'd say we have to make small tamper-proof nuclear reactors like SSTAR available to developing nations, but keep the reprocessing and enrichment technologies tightly restricted. Actinide poisoning of the fuel (to make weaponization more difficult than simply starting a clandestine enrichment program) is also probably a good idea.
I can't get that graph to load (but my net connection has been flaky lately so it could be my fault.) At any rate, it sounds like a claim that MM have made: that sending "red noise" into the MBH98 program results in a hockeystick. The main problem is that the extracted trend explains very little variance relative to the trend extracted from real data. Here's a 4-part primer on PCA to help people understand the basics.
I read some of it, and their complaints sound very similar to what other scientists go through when trying to get their research published. Peer-review is often an unpleasant process because it's based on confrontation, but this is true for everyone. In this particular case, I think Nature was right to reject their article based on the mountain of evidence against their claims.
Oops. I should have said point 5 in part 2. Sorry for any confusion that typo caused.
It doesn't show a weakness in the process, it shows that computation power isn't infinite. Redoing all the calculations without the benefit of PCA requires use of a large cluster for a long time. This was done (in point 5) and shows that any PCA errors were negligible. Scientists aren't evil monsters engaged in a massive conspiracy. Really. We're ordinary people, just like you.
Obviously you didn't bother to read anything in those links. As I've said, many studies have examined the PCA centering issues, and found no significant differences when using different methodologies. I'm baffled as to why you find it necessary to tell me to google the Wegman Report, when I discussed it at length in the first link of my previous comment. It's perfectly normal for laymen to be confused about these issues (they ARE complicated). But it's ludicrous to suggest that the scientific community as a whole is somehow unaware of these issues or engaged in a massive conspiracy to suppress them.
Nevermind. I don't feel like repeating myself ad nauseum. It's not possible to reason someone out of a position that he didn't reason himself into in the first place.