Sunday, November 23, 2008
Trusting the Model Over the Measurements?
.
This may shock some readers, but I really do try to keep abreast of the work of the "real climate scientists" and not get trapped in a corner listening to various skeptics argue whether Al Gore is evil or just crazy. Since one of the latest snarky things is for skeptics to point out that the world hasn't warmed in the last ten years (despite large increases in CO2 concentrations), I was interested to read the latest post at RealClimate which took on this (alleged) canard.
Anyway, the writer says:
(Note that the figure above comes from one of the comments in the RealClimate thread, and note that the source is GISS, run by Hansen. I.e. I'm not taking the graph above from the Michael Crichton estate.)
I really am trying to not jump to conclusions here. I know some of Free Advice's readers are on "the other side" so to speak, so by all means, please explain if I've missed something here.
However, it seems to me that what the writer above is saying, is that we shouldn't take seriously the skeptic claim that "global warming has stopped" (for the last ten years), because (a) only one set of measurements shows this, and (b) such a halt in warming doesn't fit our models.
Now point (a) is fine, and point (b) isn't terrible, subject to a caveat. And yet, the caveat the writer gives is NOT, "Of course, these are just models, and in the long-run the data must trump the models." I admit I don't really understand what the stated caveat is, but I'm pretty sure it's not pointing out the obvious danger in ignoring temperature readings (over the course of ten years) because they don't match what your model predicts.
Like I said, I am trying to be fair. I suppose one could argue that I have done an analogous thing in my cynicism regarding the BLS's inflation figures. But even there, I have not been saying, "I think inflation is higher than the BLS is admitting, because look at the money supply and I believe Milton's theory!" Rather, I have been saying, "I think inflation is higher than the BLS is admitting, because no matter how I manipulate the government's own gas price data, I can't get the 'seasonally adjusted' fall that the BLS reports (back in April), and because the BLS' own yr/yr increases suggest far more inflation than their seasonally adjusted, annualized rates."
And you can be darn sure that if ten years go by and gold has never broken $1000,* I will admit that I was waaaay the heck wrong in my forecasts.
By the way, in case you are entirely new to all this stuff: The main reason the graph above annoys the proponents of the theory of manmade global warming is that 1998 was unusually warm. I.e. there was a huge spike from 1997 to 1998, and so it's natural that temperatures fell immediately after it. Gavin Schmidt I think gives a better first pass at the "why hasn't the globe warmed in ten years?!" attack than the guy I linked to in this post. But the guy above amazed me for how casually he relies on "model-generated data." In my line of work, I think that would be a contradiction in terms. You've got the data, and you've got your model. The model is supposed to fit the data, or predict the future data, but not generate it.
* The one out I grant myself is if the government takes over the gold market again, and declares an official price. Then my predictions would of course refer to the black market price of gold.
This may shock some readers, but I really do try to keep abreast of the work of the "real climate scientists" and not get trapped in a corner listening to various skeptics argue whether Al Gore is evil or just crazy. Since one of the latest snarky things is for skeptics to point out that the world hasn't warmed in the last ten years (despite large increases in CO2 concentrations), I was interested to read the latest post at RealClimate which took on this (alleged) canard.
Anyway, the writer says:
Confusion has continued regarding trends in global temperatures. The misconception 'the global warming has stopped' still lives on in some minds. We have already discussed why this argument is flawed. So why have we failed to convince ;-) ?
The confused argument hinges on one data set - the HadCRUT 3V - which is only one of several estimates, and it is the global temperature record that exhibits the least change over the last decade. Other temperature analyses suggest greater change (warming). Thus, one could argue that the HadCRUT 3V represents the lower estimate, if a warming could be defined for such a short interval.
A comparison with other temperature analyses, such as the NASA/GISS...reveals differences. We can also compare with model-generated data (re-analyses), keeping in mind that one must be very careful with these data since they are not appropriate for studying long-term climate change (they give a misrepresentation of trends - at least on a local scale). Nevertheless, information from independent data suggest an increase in global mean temperatures even over the last decade.
(Note that the figure above comes from one of the comments in the RealClimate thread, and note that the source is GISS, run by Hansen. I.e. I'm not taking the graph above from the Michael Crichton estate.)
I really am trying to not jump to conclusions here. I know some of Free Advice's readers are on "the other side" so to speak, so by all means, please explain if I've missed something here.
However, it seems to me that what the writer above is saying, is that we shouldn't take seriously the skeptic claim that "global warming has stopped" (for the last ten years), because (a) only one set of measurements shows this, and (b) such a halt in warming doesn't fit our models.
Now point (a) is fine, and point (b) isn't terrible, subject to a caveat. And yet, the caveat the writer gives is NOT, "Of course, these are just models, and in the long-run the data must trump the models." I admit I don't really understand what the stated caveat is, but I'm pretty sure it's not pointing out the obvious danger in ignoring temperature readings (over the course of ten years) because they don't match what your model predicts.
Like I said, I am trying to be fair. I suppose one could argue that I have done an analogous thing in my cynicism regarding the BLS's inflation figures. But even there, I have not been saying, "I think inflation is higher than the BLS is admitting, because look at the money supply and I believe Milton's theory!" Rather, I have been saying, "I think inflation is higher than the BLS is admitting, because no matter how I manipulate the government's own gas price data, I can't get the 'seasonally adjusted' fall that the BLS reports (back in April), and because the BLS' own yr/yr increases suggest far more inflation than their seasonally adjusted, annualized rates."
And you can be darn sure that if ten years go by and gold has never broken $1000,* I will admit that I was waaaay the heck wrong in my forecasts.
By the way, in case you are entirely new to all this stuff: The main reason the graph above annoys the proponents of the theory of manmade global warming is that 1998 was unusually warm. I.e. there was a huge spike from 1997 to 1998, and so it's natural that temperatures fell immediately after it. Gavin Schmidt I think gives a better first pass at the "why hasn't the globe warmed in ten years?!" attack than the guy I linked to in this post. But the guy above amazed me for how casually he relies on "model-generated data." In my line of work, I think that would be a contradiction in terms. You've got the data, and you've got your model. The model is supposed to fit the data, or predict the future data, but not generate it.
* The one out I grant myself is if the government takes over the gold market again, and declares an official price. Then my predictions would of course refer to the black market price of gold.
Comments:
Glad to see I'm not the only free marketeer who takes Global warming seriously.
Of course, I dislike the politicization of the scientific issues, which tends to obscure scientific truths.
The environment is complicated and difficult to model, so while that doesn't mean global warming isn't an issue, it doesn't mean we have to throw away capitalism to deal with it.
Of course, I dislike the politicization of the scientific issues, which tends to obscure scientific truths.
The environment is complicated and difficult to model, so while that doesn't mean global warming isn't an issue, it doesn't mean we have to throw away capitalism to deal with it.
Zach, two things:
(a) I significantly altered this post after I first put it up, to add a graph and some other stuff. So you may want to reread it. (It all happened within ten minutes, but I notice that you posted a comment before I was happy with it.)
(b) I think some others will soon "correct" your view that I take global warming seriously. 3, 2, 1...
(a) I significantly altered this post after I first put it up, to add a graph and some other stuff. So you may want to reread it. (It all happened within ten minutes, but I notice that you posted a comment before I was happy with it.)
(b) I think some others will soon "correct" your view that I take global warming seriously. 3, 2, 1...
The most intelligent answer to the question : "Is the earths temperature rising?" as of yet is probably "It depends" (on which period of time you are measuring over). What I find most interesting is that if you look at historical temperature measurements, the current fluctuations represent a change so small that it's ludicrous to think that we can draw conclusions from it. That being said, I will not disregard the possibility that carbon dioxide emissions caused it. I will simply state that there is no conclusive (or even close-to-conclusive) evidence pointing in any direction, everything is just based on loose theories, which makes it utterly stupid to jump to rash decisisons that may cause huge damages to an already suffering world economy.
//hpx
//hpx
# posted by 
: November 24, 2008 at 1:32 AM
: November 24, 2008 at 1:32 AM
I haven't had time to really read through the Real Climate post, but the intro gets me fired up. The allegedly confused argument does not "hinge on one data set--HadCRUT 3V."
Lucia shows here using data, http://rankexploits.com/musings/2008/ipcc-2ccentury-remains-in-low-confidence-region-through-oct/ that using HadCRUT and GISSTemp that there is currently low confidence (< 10%) that there is a 2C/century increase in temperature. That's the real argument.
Some skeptics of climate crisis say claim there hasn't been warming for at least a decade, but most of these guys usually refer to RSS or UAH satellite data, not HadCRUT.
Unsurprisingly Real Climate doesn't actually link to anyone making this argument, so we can't read the actual argument from the mouths of these alleged skeptics. Instead, they set up a straw man to knock down.
I'm not getting into the substance of the Real Climate blog post, just the alleged claim they are trying to refute.
Pielke Jr. has a good post on the falsification/validation of climate models. http://sciencepolicy.colorado.edu/prometheus/how-to-make-two-decades-of-cooling-consistent-with-warming-4413
In that post Pielke Jr. takes Real Climate to task for essentially arguing that anything from -10/C and +13/C (per century) would be consistent with the models. As Pielke Jr. writes, "This sort of reasoning turns climate model falsification into a rather meaningless exercise."
So when the author of the Real Climate post writes, "We have already discussed why this argument is flawed. So why have we failed to convince ;-)" they have failed to convince because they don't admit what the actual data are currently saying.
Lucia shows here using data, http://rankexploits.com/musings/2008/ipcc-2ccentury-remains-in-low-confidence-region-through-oct/ that using HadCRUT and GISSTemp that there is currently low confidence (< 10%) that there is a 2C/century increase in temperature. That's the real argument.
Some skeptics of climate crisis say claim there hasn't been warming for at least a decade, but most of these guys usually refer to RSS or UAH satellite data, not HadCRUT.
Unsurprisingly Real Climate doesn't actually link to anyone making this argument, so we can't read the actual argument from the mouths of these alleged skeptics. Instead, they set up a straw man to knock down.
I'm not getting into the substance of the Real Climate blog post, just the alleged claim they are trying to refute.
Pielke Jr. has a good post on the falsification/validation of climate models. http://sciencepolicy.colorado.edu/prometheus/how-to-make-two-decades-of-cooling-consistent-with-warming-4413
In that post Pielke Jr. takes Real Climate to task for essentially arguing that anything from -10/C and +13/C (per century) would be consistent with the models. As Pielke Jr. writes, "This sort of reasoning turns climate model falsification into a rather meaningless exercise."
So when the author of the Real Climate post writes, "We have already discussed why this argument is flawed. So why have we failed to convince ;-)" they have failed to convince because they don't admit what the actual data are currently saying.
One more thing, as I was closing tabs in my browser and preparing to do actual work, I looked again at the graphic Gavin Schmidt posted here: http://www.realclimate.org/index.php/archives/2008/01/uncertainty-noise-and-the-art-of-model-data-comparison/
He argues " However, as most of our readers will realise, these comparisons are flawed since they basically compare long term climate change to short term weather variability.
This becomes immediately clear when looking at the following graph." What struck me as I looked at the graph again is that the 8-year downward temperature trends were associated with large volcano events--El Chichon and Mt. Pinatubo. So where's the large volcanic event that caused the current downward trend? Apparently there hasn't been one. So what's the explanation? Maybe pollution is the answer.
Of course Schmidt doesn't point out this flaw in his argument. He only says, "it should be clear that short term comparisons are misguided." That's not what his graph shows. His graph shows that downward departures from the long term trend were associated with large events.
He argues " However, as most of our readers will realise, these comparisons are flawed since they basically compare long term climate change to short term weather variability.
This becomes immediately clear when looking at the following graph." What struck me as I looked at the graph again is that the 8-year downward temperature trends were associated with large volcano events--El Chichon and Mt. Pinatubo. So where's the large volcanic event that caused the current downward trend? Apparently there hasn't been one. So what's the explanation? Maybe pollution is the answer.
Of course Schmidt doesn't point out this flaw in his argument. He only says, "it should be clear that short term comparisons are misguided." That's not what his graph shows. His graph shows that downward departures from the long term trend were associated with large events.
as jpx83 says it's only part of the problem
if warming
is co2 increase the culprit or a consequence?
and the same old nagging question
what caused the other warmings in the past?
ps :
I live on the bottom of a deep glacial valley
if warming
is co2 increase the culprit or a consequence?
and the same old nagging question
what caused the other warmings in the past?
ps :
I live on the bottom of a deep glacial valley
I don't know what causes global warming or if it's even taking place. The question really is if it is taking place how bad of an impact will it make. There have been speculation that the world could go through a global warming and be prosperous. Those areas with poor growing seasons may actually be better off. Others say it will be horrific.
We also must ask ourselves, "Can we even control for global climate change?" There is a great program that came on this past weekend about how the earth was formed. It obviously went through different cycles were it went from a big frozen ball to an extremely hot place. What it shows is that the world as we know it was formed because of multiple factors, astroids, methane gas, solar winds, even the possibility of Earth's own magnetic field, etc, etc.
Also, the American Physical Society, I believe this is correct, recently changed it stance to global warming. They stated that there is no conclusive proof that global warming is a man-made cause. This was not a denial of global warming, merely stating that the current science out there is not fact.
The other problem I have is that the models can't replicate what has happened in the past. If the models do replicate the theory correctly then it is not fact and I would rather not act on something that isn't fact. Science demands replication in order for something to be fact. Einstein may have been famous after he gave his presentation on his Theory of Relativity, but it was not a scientific fact until 1922 when a solar eclipse could verify that he his theory was indeed correct. That was years after he had presented the math and years after two different outcomes were inconclusive. Until it's proven as fact I say we do nothing.
We also must ask ourselves, "Can we even control for global climate change?" There is a great program that came on this past weekend about how the earth was formed. It obviously went through different cycles were it went from a big frozen ball to an extremely hot place. What it shows is that the world as we know it was formed because of multiple factors, astroids, methane gas, solar winds, even the possibility of Earth's own magnetic field, etc, etc.
Also, the American Physical Society, I believe this is correct, recently changed it stance to global warming. They stated that there is no conclusive proof that global warming is a man-made cause. This was not a denial of global warming, merely stating that the current science out there is not fact.
The other problem I have is that the models can't replicate what has happened in the past. If the models do replicate the theory correctly then it is not fact and I would rather not act on something that isn't fact. Science demands replication in order for something to be fact. Einstein may have been famous after he gave his presentation on his Theory of Relativity, but it was not a scientific fact until 1922 when a solar eclipse could verify that he his theory was indeed correct. That was years after he had presented the math and years after two different outcomes were inconclusive. Until it's proven as fact I say we do nothing.
# posted by : November 24, 2008 at 9:58 AM
: November 24, 2008 at 9:58 AM
Bob
Have you seen Austrian economist Edwin Dolan's writings on global warming (PDF) and environmental issues in general ?? (...shameless plug here for my Amazon review of his book)
Have you seen Austrian economist Edwin Dolan's writings on global warming (PDF) and environmental issues in general ?? (...shameless plug here for my Amazon review of his book)
etw, thanks for mentioning Edwin Dolan's essay, which skewers the facile efforts of many libertarians to dismiss consideration of climate change concerns.
Bob, FWIW here are my comments on Dolan.
Bob, FWIW here are my comments on Dolan.
it seems to me that what the writer above is saying, is that we shouldn't take seriously the skeptic claim that "global warming has stopped" (for the last ten years), because ... (b) such a halt in warming doesn't fit our models.
Bob, you're making a vailiant effort, but it looks like your own predilections are still getting in the way of understanding what the post author is saying.
Perhaps if you take a look again you will see that when the author talks about "model-generated data (re-analyses)" and "re-analysis data" he is referring not to model projections, but to the use of models to make guesstimates (not precise) of what the likely temps were in areas of the globe where actual data is lacking.
I admit I don't really understand what the stated caveat is, but I'm pretty sure it's not pointing out the obvious danger in ignoring temperature readings (over the course of ten years) because they don't match what your model predicts.
How about where the author says "The re-analyses provide a physically consistent description of the atmosphere - suggesting high temperatures in the Arctic - but we can only be sure about this when we actually have been there and made the real measurements (some can be done by satellites too)"?
the guy above amazed me for how casually he relies on "model-generated data." In my line of work, I think that would be a contradiction in terms. You've got the data, and you've got your model. The model is supposed to fit the data, or predict the future data, but not generate it.
I guess you missed his various precautionary statements, where Benestad essentially agrees with you and spells out the limited purpose of his use of "re-analysis" data? Benestad: "it's important to note that the NCEP re-analysis and other re-analyses (e.g. ERA40) are not regarded as being appropriate for trend studies due to changes in observational systems (new satellites coming in etc). Nevertheless, a comparison between the re-analyses and observations can highlight differences, which may suggest where to look for problems."
Bob, you're making a vailiant effort, but it looks like your own predilections are still getting in the way of understanding what the post author is saying.
Perhaps if you take a look again you will see that when the author talks about "model-generated data (re-analyses)" and "re-analysis data" he is referring not to model projections, but to the use of models to make guesstimates (not precise) of what the likely temps were in areas of the globe where actual data is lacking.
I admit I don't really understand what the stated caveat is, but I'm pretty sure it's not pointing out the obvious danger in ignoring temperature readings (over the course of ten years) because they don't match what your model predicts.
How about where the author says "The re-analyses provide a physically consistent description of the atmosphere - suggesting high temperatures in the Arctic - but we can only be sure about this when we actually have been there and made the real measurements (some can be done by satellites too)"?
the guy above amazed me for how casually he relies on "model-generated data." In my line of work, I think that would be a contradiction in terms. You've got the data, and you've got your model. The model is supposed to fit the data, or predict the future data, but not generate it.
I guess you missed his various precautionary statements, where Benestad essentially agrees with you and spells out the limited purpose of his use of "re-analysis" data? Benestad: "it's important to note that the NCEP re-analysis and other re-analyses (e.g. ERA40) are not regarded as being appropriate for trend studies due to changes in observational systems (new satellites coming in etc). Nevertheless, a comparison between the re-analyses and observations can highlight differences, which may suggest where to look for problems."
Daniel, you argue that climate scientists "have failed to convince because they don't admit what the actual data are currently saying," but it seems to me rather clear that they acknowledge that temperature increases have moderated over the past decade.
Further, it seems that they are correct in arguing that (i) they have failed to convince those who already have their minds made up and confuse short-term variations with the long term, (ii) our climate system is complex and GHG releases are just one of many factors at work, along with other natural perturbations and (iii) no mechanisms have been discovered that eliminate the growing and accumulating anthropogenic forcings and feedbacks.
Further, it seems that they are correct in arguing that (i) they have failed to convince those who already have their minds made up and confuse short-term variations with the long term, (ii) our climate system is complex and GHG releases are just one of many factors at work, along with other natural perturbations and (iii) no mechanisms have been discovered that eliminate the growing and accumulating anthropogenic forcings and feedbacks.
I think that Bob will soon confirm that he takes global warming seriously, by responding to comments. 3, 2, 1...
Tom,
I've been in hotels all week. I drove to Texas for the holiday. Sorry but I forgot to clear my schedule with you before leaving town...
I've been in hotels all week. I drove to Texas for the holiday. Sorry but I forgot to clear my schedule with you before leaving town...
This whole "we've got to stop global warming" nonsense is a political scam. It's run by environmentalists who failed to scare us in the '70s when they said that we were entering a new Ice Age, and didn't manage in the '80s to break us when they warned that acid rain would make Europe uninhabitable by the mid-1990s.
Explain how the Earth came out of the last Ice Age if it didn't do so on its own.
It did so on its own.
The Earth warms and cools naturally.
Explain how the Earth came out of the last Ice Age if it didn't do so on its own.
It did so on its own.
The Earth warms and cools naturally.
# posted by : November 28, 2008 at 12:57 AM
: November 28, 2008 at 12:57 AM
Tom,
Sorry about the sarcastic snipe. I hadn't seen this thread in days, so I forgot that I did the 3, 2, 1 thing. I.e. I just skimmed the comments a few days into my trip, and thought you invented the 3, 2, 1 thing because I wasn't answering comments.
Also, I agree your quotations show more nuance in his position, but he still seems to be saying that the earth is probably warmer, since that best fits the model projections. We have one set of actual measurements that show a stable average over the last ten years, and he is arguing that the true average is probably higher.
So again, I grant that that is more subtle than someone saying that a particular thermometer's reading is wrong, but still, he is saying that that system of measurements is probably wrong because it doesn't fit what his model says the true value (which the measurements are sampling in a sense) is.
Sorry about the sarcastic snipe. I hadn't seen this thread in days, so I forgot that I did the 3, 2, 1 thing. I.e. I just skimmed the comments a few days into my trip, and thought you invented the 3, 2, 1 thing because I wasn't answering comments.
Also, I agree your quotations show more nuance in his position, but he still seems to be saying that the earth is probably warmer, since that best fits the model projections. We have one set of actual measurements that show a stable average over the last ten years, and he is arguing that the true average is probably higher.
So again, I grant that that is more subtle than someone saying that a particular thermometer's reading is wrong, but still, he is saying that that system of measurements is probably wrong because it doesn't fit what his model says the true value (which the measurements are sampling in a sense) is.
Bob, it still seems to be you are not understanding what the author is saying.
"the earth is probably warmer, since that best fits the model projections"
"he is saying that that system of measurements is probably wrong because it doesn't fit what his model says the true value (which the measurements are sampling in a sense) is"
He says neither of these things; rather, he is saying that the data gathering system is incomplete/ imprecise/ has holes in it, that the biggest holes are precisely in the areas that we know are warming the greatest (from actual measurements and observations, consistent with the known physics), so the average temperatures produced in various places are underweight.
Post a Comment
"the earth is probably warmer, since that best fits the model projections"
"he is saying that that system of measurements is probably wrong because it doesn't fit what his model says the true value (which the measurements are sampling in a sense) is"
He says neither of these things; rather, he is saying that the data gathering system is incomplete/ imprecise/ has holes in it, that the biggest holes are precisely in the areas that we know are warming the greatest (from actual measurements and observations, consistent with the known physics), so the average temperatures produced in various places are underweight.
Subscribe to Post Comments [Atom]
<< Home
Subscribe to Posts [Atom]