|
Post by icefisher on May 31, 2009 17:58:00 GMT
The question isn't whether any models showed flat 10 year trends or not its why did the models show flat 10 year trends from time to time and is that an explanation for the current trend. I think the answer to that is no. Why? Do you even know why the models show flat 10 year trends? If not how can you answer 'no'? The whole point is that the models tell us that there will be 10 year periods with flat trend that are just caused by internal noise in the climate system. So obviously 10 year flat period in the temperature record does not falsify the models. It's that simple. I can't answer your question unless you identify the model you are talking about. Several, at least, of the GCMs anticipate randomly volcanic events. You seem to be claiming the one you are talking about is just internal noise. . . . so you then on one hand saying its corollated with a La Nina and a solar minimum and on the other hand suggesting it not real and just noise. Which is it? The bottom line here socold you have to take a stand. Its easy to run around in circles and claim one thing one minute and another the next but it just makes you look stupid. I kind of doubt thats the case but appearances. . . .maybe its just noise internal to your brain.
|
|
|
Post by socold on May 31, 2009 18:09:09 GMT
socold: "We don't need an explaination in this case. The question is whether the temperature record is at odds with the model runs. It isn't. The model runs exhibit 10 year flat periods too." Public opinion polls seem to be indicating that your team does have some 'splainin to do, if you want to stifle the Western industrial economies to save the World from catastrophic AGW. Don't you think it is worth the effort? How many of those model runs exhibit 10 year flat periods over the next ninety years? Is one all you can come up with? No the one is actually a 19 year flat period. There will be far more 10 year flat periods. I don't have the individual model run data to point out the specific periods and models.
|
|
|
Post by socold on May 31, 2009 18:10:07 GMT
Why? Do you even know why the models show flat 10 year trends? If not how can you answer 'no'? The whole point is that the models tell us that there will be 10 year periods with flat trend that are just caused by internal noise in the climate system. So obviously 10 year flat period in the temperature record does not falsify the models. It's that simple. I can't answer your question unless you identify the model you are talking about. Several, at least, of the GCMs anticipate randomly volcanic events. You seem to be claiming the one you are talking about is just internal noise. . . . so you then on one hand saying its corollated with a La Nina and a solar minimum and on the other hand suggesting it not real and just noise. Which is it? That is noise. Just because it can be given a name - ENSO - doesn't mean it isn't noise over the trend.
|
|
|
Post by socold on May 31, 2009 18:14:59 GMT
I should have been more clear. It's declining solar minimum which cannot be maintained. At least it isn't being maintained at the moment. Without TSI continually falling there is no increasing negative forcing to offset the increasing positive forcing from GHG. During the Maunder minimum, (little ice age) minimum was extended for at least 80 years socold. But TSI didn't decline for 80 years. In order to cancel out the ever increasing ghg forcing permanently you need an ever increasing negative forcing, not one that drops for 6 years and then stays the same for decades (or even bounces back up) Looks like the correlation falls apart recently www.woodfortrees.org/plot/hadcrut3vgl/from:1850/mean:120/plot/sidc-ssn/from:1850/scale:0.01/mean:120/offset:-0.8
|
|
|
Post by icefisher on May 31, 2009 18:30:02 GMT
I can't answer your question unless you identify the model you are talking about. Several, at least, of the GCMs anticipate randomly volcanic events. You seem to be claiming the one you are talking about is just internal noise. . . . so you then on one hand saying its corollated with a La Nina and a solar minimum and on the other hand suggesting it not real and just noise. Which is it? That is noise. Just because it can be given a name - ENSO - doesn't mean it isn't noise over the trend. The point here Socold is you are rationalizing the failure of the model you are betting on to predict the future because one run suggests maybe an event might happen to bring it home. In poker thats like raising when you need a draw to an inside straight when you know your opponent has 3 aces. No you are not out of the game yet so keep playing. . . .shall we sweeten the pot some?
|
|
|
Post by socold on May 31, 2009 18:41:28 GMT
That is noise. Just because it can be given a name - ENSO - doesn't mean it isn't noise over the trend. The point here Socold is you are rationalizing the failure of the model you are betting on to predict the future because one run suggests maybe an event might happen to bring it home. In poker thats like raising when you need a draw to an inside straight when you know your opponent has 3 aces. No you are not out of the game yet so keep playing. . . .shall we sweeten the pot some? It is being argued that the models are falsified by the recent flat period. But they are not because they show such flat periods will happen. That's it in a nutshell. Poker analogies not necessary.
|
|
|
Post by dmapel on May 31, 2009 19:42:01 GMT
socold: "No the one is actually a 19 year flat period. There will be far more 10 year flat periods. I don't have the individual model run data to point out the specific periods and models. "
OK, we will count the one 19 year flat period as two flat ten year periods. And as you do, we will just assume that there are far more 10 year flat periods in the other models, with huge spikes just before 2100 to make up for all that flatness we can see in the spaghetti.
Good work, and thanks again for not mentioning that pot on the stove foolishness.
|
|
|
Post by nautonnier on May 31, 2009 20:18:47 GMT
socold: "We don't need an explaination in this case. The question is whether the temperature record is at odds with the model runs. It isn't. The model runs exhibit 10 year flat periods too." Public opinion polls seem to be indicating that your team does have some 'splainin to do, if you want to stifle the Western industrial economies to save the World from catastrophic AGW. Don't you think it is worth the effort? How many of those model runs exhibit 10 year flat periods over the next ninety years? Is one all you can come up with? No the one is actually a 19 year flat period. There will be far more 10 year flat periods. I don't have the individual model run data to point out the specific periods and models. So that model is now THE model that is correct is it? It fits where it touches as they say - it has a flat period - it bears NO RELATION to what has actually happened showing a peak in 1998 a drop then steady climb to 2003 then flat then drop in 2007 that has continued. But it has a FLAT PERIOD from 1995 to 2014 ? That's close enough for climatological science obviously. I have no doubt that I could set up a graphics generator that uses random assumptions in a 'climate generator' model that occasionally showed brief periods that matched the climate record. Should I then demand that the governments of the world use that random climate graphic generator as justification to spend trillions of dollars/pounds/yuan against a background of a worldwide financial crisis? So the argument is from you - that climate models do not have to match what happens - the occasional model only has to show some features of what happens - not even at the right time period - and not even consistently - and the ensemble of these erroneous models is then to be trusted when iterated from these random states 80 years into the future? Its no wonder that Al Gore gets on well with climatologists.
|
|
|
Post by icefisher on May 31, 2009 20:44:05 GMT
The point here Socold is you are rationalizing the failure of the model you are betting on to predict the future because one run suggests maybe an event might happen to bring it home. In poker thats like raising when you need a draw to an inside straight when you know your opponent has 3 aces. No you are not out of the game yet so keep playing. . . .shall we sweeten the pot some? It is being argued that the models are falsified by the recent flat period. But they are not because they show such flat periods will happen. That's it in a nutshell. Poker analogies not necessary. You are like that snake oil salesman selling the wood alcohol with a red label saying "Cure All" when somebody pipes up in the audience with "Dang I bought some of that crap in Deadwood last year. . .bout kilt me!" He then drops the red labled bottle in his left pocket reaches in his right pocket and pulls out a green labeled bottle and sez "A few folks find the red labeled product a little strong but this here mild version is just what you need."
|
|
|
Post by socold on May 31, 2009 20:52:55 GMT
No the one is actually a 19 year flat period. There will be far more 10 year flat periods. I don't have the individual model run data to point out the specific periods and models. So that model is now THE model that is correct is it? It fits where it touches as they say - it has a flat period - it bears NO RELATION to what has actually happened showing a peak in 1998 a drop then steady climb to 2003 then flat then drop in 2007 that has continued. But it has a FLAT PERIOD from 1995 to 2014 ? That's close enough for climatological science obviously. Go back and read the posts, it's almost as if you have skipped them. I have made this very clear. In fact I will help by requoting some of them and bolding the significant parts: "Ah so you are saying the models don't match year by year variations in temperature exactly?
Well duh obviously, because the timing of the ENSO-like variation in model runs is not synchoronized with the real world.
What is relevant is that the models show periods of 10 years without temperature rise even when the background trend is 0.15-0.2C/decade. Why? Well because of weather noise, because of situations like we have just had of declining ENSO.
Therefore the argument that 10 years of cooling in the records falsifies the models is itself falsified, because the models can hardly be falsified by something that they actually show.""The decadal variation in the model runs is not expected to line up with the decadal variation in the temperature record.
However we can ask "do the model runs with their overall warming trend show 10 year periods of flat temperature?". It doesn't matter when they show it. It doesn't matter if one run shows a flat trend in 2011-2021 and another shows a flat trend in 2005-2015 for example. All that matters is that they do exhibit flat trends over 10 year periods now and again despite the overall warming trend.
For that reason we know that the models are compatible with 10 year periods of flat temperature. Therefore it cannot be argued that a 10 year flat trend in the actual temperature record falsifies the models."" The whole point is that the models tell us that there will be 10 year periods with flat trend that are just caused by internal noise in the climate system. So obviously 10 year flat period in the temperature record does not falsify the models. It's that simple." Over and over again I seem to be repeating myself, but noone is actually addressing what I say, only a strawman version where it's as if I am claiming a model run shows a super El Nino in 1998. Well here I go again: Noone expects the noise in decadal periods (meaning < 10 year periods) in the model runs, or any particular model run, to match the noise in the actual temperature records. The climate models have ENSO like variation. Ie temperature can spike upwards for a few years and spike downwards for a few years due to nothing but internal chaotic variation in the climate system. Sometimes this is called weather to distinguish it from the longterm trend due to forcing. To reiterate, because it seems I must, an unforced climate model - that is a model with no external forcing (solar, ghg, aerosols) applied will still exhibit rises and falls on decadal (<10 year) timeframes. Ie one year's surface temperature could be 0.4C warmer than a previous year for example and the next could be 0.3C cooler. The real world climate does this too, the biggest element of noise having been spotted and given the name ENSO. Noone can predict it, noone knows when an El Nino or La Nina will form. So the question is what happens when you get an ENSO spike (strong El Nino) followed by a few years of ENSO neutral followed by an ENSO minimum (strong La Nina)? The answer is that this unlikely chain of events (unlikely in the sense that this will not happen in many 10 year periods) will have a cooling effect on the temperature record in that 10 year period. If this cooling effect is greater than a background forced warming trend then the actual temperature trend could be flat or declining. Noone would listen because you used random assumptions. The only reason people are listening to the GCMs is that they are based on physics to do with climate. However noone is listening to GCMs on decadal timescales anyway because that element is not modelled sufficiently (Notice that the models largely agree on the temperature trend for a given scenario over the next 50 years, but for the next 5 years they are very different), so it's a moot point. No the argument from me is that a flat trend over a 10 year period aint enough time to falsify a continuing multi-decadal warming trend. Any 10 year period in time can be unduely influenced by the internal variation. The models show this and we have empirical evidence that this has happened in the real climate recently: The negative trend 1998-2008 relies on starting in 1998 a strong el nino year and ending in 2008 a strong la nina year. This is not some bizzare accident of this particular 10 year period - it is precisely why the trend is negative. This is exactly the kind of low chance 10 year pattern of internal variation that the models are saying would cause a pause in the climb. So no big suprise that the models are not falsified by this happening in the real world.
|
|
|
Post by nautonnier on May 31, 2009 21:29:02 GMT
So that model is now THE model that is correct is it? It fits where it touches as they say - it has a flat period - it bears NO RELATION to what has actually happened showing a peak in 1998 a drop then steady climb to 2003 then flat then drop in 2007 that has continued. But it has a FLAT PERIOD from 1995 to 2014 ? That's close enough for climatological science obviously. Go back and read the posts, it's almost as if you have skipped them. I have made this very clear. In fact I will help by requoting some of them and bolding the significant parts: "Ah so you are saying the models don't match year by year variations in temperature exactly?
Well duh obviously, because the timing of the ENSO-like variation in model runs is not synchoronized with the real world.
What is relevant is that the models show periods of 10 years without temperature rise even when the background trend is 0.15-0.2C/decade. Why? Well because of weather noise, because of situations like we have just had of declining ENSO.
Therefore the argument that 10 years of cooling in the records falsifies the models is itself falsified, because the models can hardly be falsified by something that they actually show.""The decadal variation in the model runs is not expected to line up with the decadal variation in the temperature record.
However we can ask "do the model runs with their overall warming trend show 10 year periods of flat temperature?". It doesn't matter when they show it. It doesn't matter if one run shows a flat trend in 2011-2021 and another shows a flat trend in 2005-2015 for example. All that matters is that they do exhibit flat trends over 10 year periods now and again despite the overall warming trend.
For that reason we know that the models are compatible with 10 year periods of flat temperature. Therefore it cannot be argued that a 10 year flat trend in the actual temperature record falsifies the models."" The whole point is that the models tell us that there will be 10 year periods with flat trend that are just caused by internal noise in the climate system. So obviously 10 year flat period in the temperature record does not falsify the models. It's that simple." Over and over again I seem to be repeating myself, but noone is actually addressing what I say, only a strawman version where it's as if I am claiming a model run shows a super El Nino in 1998. Well here I go again: Noone expects the noise in decadal periods (meaning < 10 year periods) in the model runs, or any particular model run, to match the noise in the actual temperature records. The climate models have ENSO like variation. Ie temperature can spike upwards for a few years and spike downwards for a few years due to nothing but internal chaotic variation in the climate system. Sometimes this is called weather to distinguish it from the longterm trend due to forcing. To reiterate, because it seems I must, an unforced climate model - that is a model with no external forcing (solar, ghg, aerosols) applied will still exhibit rises and falls on decadal (<10 year) timeframes. Ie one year's surface temperature could be 0.4C warmer than a previous year for example and the next could be 0.3C cooler. The real world climate does this too, the biggest element of noise having been spotted and given the name ENSO. Noone can predict it, noone knows when an El Nino or La Nina will form. So the question is what happens when you get an ENSO spike (strong El Nino) followed by a few years of ENSO neutral followed by an ENSO minimum (strong La Nina)? The answer is that this unlikely chain of events (unlikely in the sense that this will not happen in many 10 year periods) will have a cooling effect on the temperature record in that 10 year period. If this cooling effect is greater than a background forced warming trend then the actual temperature trend could be flat or declining. Noone would listen because you used random assumptions. The only reason people are listening to the GCMs is that they are based on physics to do with climate. However noone is listening to GCMs on decadal timescales anyway because that element is not modelled sufficiently (Notice that the models largely agree on the temperature trend for a given scenario over the next 50 years, but for the next 5 years they are very different), so it's a moot point. No the argument from me is that a flat trend over a 10 year period aint enough time to falsify a continuing multi-decadal warming trend. Any 10 year period in time can be unduely influenced by the internal variation. The models show this and we have empirical evidence that this has happened in the real climate recently: The negative trend 1998-2008 relies on starting in 1998 a strong el nino year and ending in 2008 a strong la nina year. This is not some bizzare accident of this particular 10 year period - it is precisely why the trend is negative. This is exactly the kind of low chance 10 year pattern of internal variation that the models are saying would cause a pause in the climb. So no big suprise that the models are not falsified by this happening in the real world. I think that it is you that is reading the objection that you want into what I wrote. take multiple random models - say 30 of them all different but ALL claiming to be accurate models of the 'real world' the modelers that built them would defend them as such. Start them with random start parameters so none of them generate the same 'output climate' model. Now let them all use differing assumptions for minor issues such as convection and clouds - the one assumption that they must ALL have is that CO 2 will cause warming equivalent to 3.7Wm -2 for a doubling of CO 2. Now all the models generate differing climate outputs. When any attempt is made to validate the output SoCold chooses the model that nearly meets that validation test. When another test is used - SoCold chooses ANOTHER model that nearly meets the second validation test. With luck you can find model to match each real world occurrence Of course they show lots of things that Do NOT happen but here for the wrong period at the wrong temperature it has a graphical output that is flat - so thats it proven. The only problem is that NONE of the models match reality or even get close. They occasionally show patterns that have some similarity but don't match - like 19 years flat temperatures to match 4 years flat then a drop- and presumably a giant heated leap to catch up on year 20. Its like a street trader with a pile of broken calculators scrabbling to find the one that shows 2+2=4 and finding one that says 5 and another that says 3 so between them they are right aren't they ? And on this 'science' rests the future economy of the world.
|
|
|
Post by dmapel on May 31, 2009 21:57:40 GMT
nautonnier: "Its like a street trader with a pile of broken calculators scrabbling to find the one that shows 2+2=4 and finding one that says 5 and another that says 3 so between them they are right aren't they ?"
Actually, he has not yet found the one that says 2+2=3. Unless he can point to many other ten year flat periods produced by the multitude of models, he should find another excuse for the failure of ANY of them to approximate reality.
|
|
|
Post by socold on May 31, 2009 22:43:59 GMT
I think that it is you that is reading the objection that you want into what I wrote. take multiple random models - say 30 of them all different but ALL claiming to be accurate models of the 'real world' the modelers that built them would defend them as such. Start them with random start parameters so none of them generate the same 'output climate' model. Now let them all use differing assumptions for minor issues such as convection and clouds - the one assumption that they must ALL have is that CO 2 will cause warming equivalent to 3.7Wm -2 for a doubling of CO 2. Now all the models generate differing climate outputs. When any attempt is made to validate the output SoCold chooses the model that nearly meets that validation test. Wrong, this entire issue is about falsifying the models based on a 10 year temperature trend not validating them. The question is, as I have repeated now many times, whether the recent 10 year flat period in temperature falsifies the models. Because some skeptics are saying the equivalent of "the models don't expect 10 years flat temperature". But that is wrong.
|
|
|
Post by magellan on Jun 1, 2009 4:08:05 GMT
I think that it is you that is reading the objection that you want into what I wrote. take multiple random models - say 30 of them all different but ALL claiming to be accurate models of the 'real world' the modelers that built them would defend them as such. Start them with random start parameters so none of them generate the same 'output climate' model. Now let them all use differing assumptions for minor issues such as convection and clouds - the one assumption that they must ALL have is that CO 2 will cause warming equivalent to 3.7Wm -2 for a doubling of CO 2. Now all the models generate differing climate outputs. When any attempt is made to validate the output SoCold chooses the model that nearly meets that validation test. Wrong, this entire issue is about falsifying the models based on a 10 year temperature trend not validating them. The question is, as I have repeated now many times, whether the recent 10 year flat period in temperature falsifies the models. Because some skeptics are saying the equivalent of "the models don't expect 10 years flat temperature". But that is wrong. Warmologists seem to not understand what a prediction is by definition. As the following demonstrates, even when a model can supposedly "predict" the past, it has nothing to do with predicting the future, which is by definition what a prediction is. www.ucar.edu/news/releases/2006/sunspot.shtmlThe next sunspot cycle will be 30-50% stronger than the last one and begin as much as a year late, according to a breakthrough forecast using a computer model of solar dynamics developed by scientists at the National Center for Atmospheric Research (NCAR) The scientists have confidence in the forecast because, in a series of test runs, the newly developed model simulated the strength of the past eight solar cycles with more than 98% accuracy. What is the reality? www.spaceweather.com/glossary/spotlessdays.htmBy the standard of spotless days, the ongoing solar minimum is the deepest in a century: NASA report. In 2008, no sunspots were observed on 266 of the year's 366 days (73%). To find a year with more blank suns, you have to go all the way back to 1913, which had 311 spotless days (85%) If climate models were relevant predictive tools, we'd all know the name of the model by heart. Should socold and steve comment on the following it would be appreciated. www.itia.ntua.gr/getfile/864/1/documents/2008HSJClimPredictions.pdfAbstract Geographically distributed predictions of future climate, obtained through climate models, are widely used in hydrology and many other disciplines, typically without assessing their reliability. Here we compare the output of various models to temperature and precipitation observations from eight stations with long (over 100 years) records from around the globe. The results show that models perform poorly, even at a climatic (30-year) scale. Thus local model projections cannot be credible, whereas a common argument that models can perform better at larger spatial scales is unsupported.
|
|
|
Post by socold on Jun 1, 2009 6:49:09 GMT
|
|