|
Post by steve on Apr 12, 2010 13:22:35 GMT
The increase in GHGs we are talking about (say a doubling of CO2) changes the radiation balance by roughly 2% - quite a small number. The phrase "lock heat in place" is painting a picture that I don't recognise. So if your modeling error on the first decade of use is 100% and you need better than +/-2% for the model to be useful. . . . . That sort of bald statement suggests you've been in the wrong job for 25 years. Yeah right!
|
|
|
Post by poitsplace on Apr 12, 2010 13:35:27 GMT
As moist air rises it cools adiabatically, it emits and absorbs radiation according to its temperature and constitution. I don't see how this is something that cannot inherently be modelled. The increase in GHGs we are talking about (say a doubling of CO2) changes the radiation balance by roughly 2% - quite a small number. The phrase "lock heat in place" is painting a picture that I don't recognise. I'm saying that if water vapor feedback was strongly positive...heat wouldn't even be able to effectively get out of the tropics into other parts of the world. Now the models say that water vapor's feedbacks will lead to MORE warming...but for every 1C increase in temperature the amount of water vapor (and latent heat) goes up by 5-6% across most of the surface of the earth. This works out to 5-6 watts per degree. In (supposedly) driving up temperatures by 1C, CO2 would have increased the energy carried by latent heat and convection by roughly 6 watts. Even if you (incorrectly) counted water vapor's absorption separately...you STILL come up short on the forcing. I see little reason to believe that water vapor currently provides anything but negative feedback. The weather models veer off of reality at an amazingly high rate. Even with gobs and gobs of current data from all over the world, the probability that they'll be right drops by such a substantial amount every day that it is useless in a week.
|
|
|
Post by icefisher on Apr 12, 2010 15:57:46 GMT
So if your modeling error on the first decade of use is 100% and you need better than +/-2% for the model to be useful. . . . . That sort of bald statement suggests you've been in the wrong job for 25 years. LOL! I laid it all out for you to respond Steve. My point was that if you model is flawed in its construction it doesn't teach you anything. Its OK to to parameterize a model with guesses and miss your predictions; but after the observations are made and you plug in the correct parameters; the model should return the correct answer. No evidence has been submitted that is the case with any GCM. And the inside view of Kevin Trenberth tells me that in fact the models still don't work when the observations are input. If Trenberth could tell us the model had taught us something it would be all over the press what we had learned and it would not be a "travesty" that they can't tell us anything. All that is left is the hope the unknown physical mechanism that are actually changing our climate are just temporary and you don't need a model to tell you that. "Yeah right" is not a statement related to any claim of how they have been useful. I can accept that my view, which is reflected by the public in their view of the importance of AGW, could be wrong. But is "yeah right!" all you can think of in response to that observation?
|
|
|
Post by steve on Apr 12, 2010 16:08:29 GMT
How do you come to that conclusion?
If warming of 1C leads to an increase in absolute humidity of 5-6%, then that increase can be done with a one-off increase in humidity, and thereby a one-off transfer of energy into latent heat.
To determine the extra amount of energy being cycled through the hydrological cycle, you have to look at how quickly the moisture is recycled, eg. by looking at precipitation. Of course even if precipitation also goes up 5-6% this doesn't help you if most of the increase came from low level stratus and cumulous as it means the extra energy is not released high enough to "bypass the CO2".
I didn't talk about the weekly weather forecast. I talked about the *daily* weather forecast. The daily weather forecast is dependent on getting convection right. The amount of convection on a given day is dependent on the temperature/humidity profile, winds and incoming radiation, and hardly at all dependent on the exact amount of CO2 in the atmosphere.
You still have to remember that many models *do* show an element of a negative feedback due to the sorts of process you describe. But the size of negative feedback you would need is *not* supported by the satellite observations because if anything they show a lower warming trend up high. And what tends to happen is that the negative feedback is inversely correlated with a positive feedback due to the increase in water vapour in the mid-troposphere caused by the stronger convection.
|
|
|
Post by icefisher on Apr 12, 2010 16:19:07 GMT
Of course even if precipitation also goes up 5-6% this doesn't help you if most of the increase came from low level stratus and cumulous as it means the extra energy is not released high enough to "bypass the CO2". Sure its high enough Steve. These low level stratus and cumulous are opaque to IR so the heat might bounce back and forth and maybe lift the clouds an inch or so until it does escape. What you are forgetting is IR from CO2 does not warm anything it merely slows cooling. If the intial response to cooling is massive evaporation and lifting of water into the atmosphere forming thick dark low level clouds it not only blocks incoming but it blocks the IR too from the upper atmosphere.
|
|
|
Post by steve on Apr 12, 2010 16:41:24 GMT
Icefisher, you seem to transfer a lot of experience about financial modelling across to the climate domain. That sort of arguing seems to be the dish of the day, but it's just inappropriate.
Claiming to know when a model is useful or not and then implying that a 2% forcing must be unimportant if there is an error in a model is just tiresomely contradictory, and the energy to argue the point is hard to summon.
The physics of the radiative effects of CO2 are not at the whim of central bankers. The models are broadly correct in most key diagnostics. Invented experiments about greenhouses, or theories about latent heat bypassing CO2 are non-starters. Impacts of PDO and ENSO are all there to be studied with models to demonstrate what impacts they have on the climate.
On the other hand models aren't everything. The key evidence of the amount of warming we should expect comes from the observations.
|
|
|
Post by steve on Apr 12, 2010 16:45:32 GMT
Of course even if precipitation also goes up 5-6% this doesn't help you if most of the increase came from low level stratus and cumulous as it means the extra energy is not released high enough to "bypass the CO2". Sure its high enough Steve. These low level stratus and cumulous are opaque to IR so the heat might bounce back and forth and maybe lift the clouds an inch or so until it does escape. What you are forgetting is IR from CO2 does not warm anything it merely slows cooling. If the intial response to cooling is massive evaporation and lifting of water into the atmosphere forming thick dark low level clouds it not only blocks incoming but it blocks the IR too from the upper atmosphere. Perhaps we should set up IR lasers to bounce the clouds into the stratosphere
|
|
|
Post by icefisher on Apr 12, 2010 18:37:59 GMT
Icefisher, you seem to transfer a lot of experience about financial modelling across to the climate domain. That sort of arguing seems to be the dish of the day, but it's just inappropriate. Claiming to know when a model is useful or not and then implying that a 2% forcing must be unimportant if there is an error in a model is just tiresomely contradictory, and the energy to argue the point is hard to summon. The argument was not that 2% is unimportant. The argument was the model error was equal or greater than 2%. Now you can argue that the model is accurate to 2% over the long haul. In which case. . . .you don't need to do any more work on it. You just wait a generation (or however long you want to define the line between climate and weather) and run it again. To give you an example you have a home behind a hill and on the other side of the hill is a beautiful seaview. An architect comes with a beautiful two-story model and tells you if you build a 2nd story you will see the great view and increase the value of your home. So you sign the contract and the 2nd story is built and you go up and peer out your new view window and all you see is the hill. The architect comes and tells you that well you can build another story and the view will reveal itself. And you are telling me you would sign the new contract. LOL! The first time you bought the model when what you really needed was a survey before any model was built. The physics of the radiative effects of CO2 are not at the whim of central bankers. The models are broadly correct in most key diagnostics. Invented experiments about greenhouses, or theories about latent heat bypassing CO2 are non-starters. Hey thats a strawman. I pointed out that perhaps some mini-model elements of the climate model might have predictive capabilities it is the aggregation of them that does not have any predictive power. But it is important to note the successful submodels are not validated by the failure of the aggregation, thus you can't argue that the aggregated model had any usefulness for the success of any submodels. What you do is run the submodels when you need them and do the necessary survey work on what really is going on and what the requirements will be for a model that works. Only in La la land do people pay for models before the field work is done. The only other rationale for having built models already was a bunch of over inflated egos figuring they had it figured out and took mother nature to stick them in the eye. And I have no idea where this "whims of bankers" has anything to do with this. I must have missed that paper. Impacts of PDO and ENSO are all there to be studied with models to demonstrate what impacts they have on the climate. Climate models are not useful tools for studying the effect of ENSO on climate. You "study" ENSO effects on climate via observations. If your observations indicate a correlation you build an ENSO model to capture that correlation and predict climate results of ENSOs. It might well be useful to build a model to predict the effects of ENSO on climate. But the usefulness for that only exists on the assumption you haven't polluted the model with so much nonsense or try to build too big of a model that goes beyond your observations (say global vs regional) that it fails to predict anything. If you work for private enterprise you are held accountable for getting that right. On the other hand models aren't everything. The key evidence of the amount of warming we should expect comes from the observations. Golly. . . . really? Brrrrrrr!!! Today's ENSO update shows the assemblage of ENSO models are predicting La Nina conditions by the end of the year.
|
|
|
Post by icefisher on Apr 12, 2010 18:40:34 GMT
Sure its high enough Steve. These low level stratus and cumulous are opaque to IR so the heat might bounce back and forth and maybe lift the clouds an inch or so until it does escape. What you are forgetting is IR from CO2 does not warm anything it merely slows cooling. If the intial response to cooling is massive evaporation and lifting of water into the atmosphere forming thick dark low level clouds it not only blocks incoming but it blocks the IR too from the upper atmosphere. Perhaps we should set up IR lasers to bounce the clouds into the stratosphere Sounds like a good private venture for you to invest in.
|
|
|
Post by scpg02 on Apr 13, 2010 1:02:36 GMT
Perhaps we should set up IR lasers to bounce the clouds into the stratosphere Sounds like a good private venture for you to invest in. Either of you heard of [url-http://www.haarp.alaska.edu/]HARPP[/url]? They use radio waves to dissipate clouds.
|
|
|
Post by steve on Apr 13, 2010 9:24:54 GMT
In essence, you are arguing that since some homes are behind hills, investing in a 2nd storey is *always* a bad idea.
Merely that financial modelling is not a good analogy.
If the climate results you predict were not used to build your model, then that is in part what makes the model useful.
Models use the same physical schemes and parametrizations across the globe, and use similar schemes (albeit with higher resolution) in regional implementations. That is why it is important that the details of the physical processes underlie the simulation. Faulty parametrizations (nonsense) might give you a better ENSO but worse monsoon.
|
|
|
Post by Ratty on Apr 13, 2010 10:08:38 GMT
Models use the same physical schemes and parametrizations across the globe, and use similar schemes (albeit with higher resolution) in regional implementations. That is why it is important that the details of the physical processes underlie the simulation. Faulty parametrizations (nonsense) might give you a better ENSO but worse monsoon. Disclaimer 1: I have had a few glasses of red wine so give me a little latitude .... Disclaimer 2: I am suspicious of the "science" of climate change and its promoters. My background in commercial software development cause the word "parametrizations" wrt climate models to worry me. My lack of knowledge of climate science makes me wonder if I should even comment but ... damn it! The red wine has taken over .... Are not all parametrizations a product of the human mind? Are not human minds affected and influenced by what their owners read / see / hear? (EG, has any recent musician written something truly original?) I think I rest my case here for fear that my typing skills (and the red wine) diminish my argument. Perhaps someone sober can see where I'm heading with this and rescue me ..... Clue: If you begin with an incorrect premise, can you ever be confident of the resultant conclusion?
|
|
|
Post by icefisher on Apr 13, 2010 15:04:06 GMT
In essence, you are arguing that since some homes are behind hills, investing in a 2nd storey is *always* a bad idea. So you were the architect huh? LOL!
|
|
|
Post by steve on Apr 13, 2010 15:26:53 GMT
Models use the same physical schemes and parametrizations across the globe, and use similar schemes (albeit with higher resolution) in regional implementations. That is why it is important that the details of the physical processes underlie the simulation. Faulty parametrizations (nonsense) might give you a better ENSO but worse monsoon. Disclaimer 1: I have had a few glasses of red wine so give me a little latitude .... Disclaimer 2: I am suspicious of the "science" of climate change and its promoters. My background in commercial software development cause the word "parametrizations" wrt climate models to worry me. My lack of knowledge of climate science makes me wonder if I should even comment but ... damn it! The red wine has taken over .... Are not all parametrizations a product of the human mind? Are not human minds affected and influenced by what their owners read / see / hear? (EG, has any recent musician written something truly original?) I think I rest my case here for fear that my typing skills (and the red wine) diminish my argument. Perhaps someone sober can see where I'm heading with this and rescue me ..... Clue: If you begin with an incorrect premise, can you ever be confident of the resultant conclusion? Well to trust a parametrization you have to trust that its developer is neither stupid nor dishonest. If you don't trust climate scientists, you ought to be aware that many parametrizations will have been developed for weather models, not climate models. Parametrizations are not simple bodges. There will be physical assumptions that underlie them, and the assumptions will have to be justified by comparing their outputs with observations. A simple example of a parametrization would be one that models processes that occur on a scale that is lower than the model resolution. The output of the parametrization may be that the grid box has a 50% coverage of cumulous cloud at a certain level. So you don't know where every single cloud is, but the average amount of rain, convection, albedo etc. will be right.
|
|
|
Post by magellan on Apr 13, 2010 16:32:29 GMT
Models use the same physical schemes and parametrizations across the globe, and use similar schemes (albeit with higher resolution) in regional implementations. That is why it is important that the details of the physical processes underlie the simulation. Faulty parametrizations (nonsense) might give you a better ENSO but worse monsoon. Disclaimer 1: I have had a few glasses of red wine so give me a little latitude .... Disclaimer 2: I am suspicious of the "science" of climate change and its promoters. My background in commercial software development cause the word "parametrizations" wrt climate models to worry me. My lack of knowledge of climate science makes me wonder if I should even comment but ... damn it! The red wine has taken over .... Are not all parametrizations a product of the human mind? Are not human minds affected and influenced by what their owners read / see / hear? (EG, has any recent musician written something truly original?) I think I rest my case here for fear that my typing skills (and the red wine) diminish my argument. Perhaps someone sober can see where I'm heading with this and rescue me ..... Clue: If you begin with an incorrect premise, can you ever be confident of the resultant conclusion? Well to trust a parametrization you have to trust that its developer is neither stupid nor dishonest. If you don't trust climate scientists, you ought to be aware that many parametrizations will have been developed for weather models, not climate models. Parametrizations are not simple bodges. There will be physical assumptions that underlie them, and the assumptions will have to be justified by comparing their outputs with observations. A simple example of a parametrization would be one that models processes that occur on a scale that is lower than the model resolution. The output of the parametrization may be that the grid box has a 50% coverage of cumulous cloud at a certain level. So you don't know where every single cloud is, but the average amount of rain, convection, albedo etc. will be right. Well to trust a parametrization you have to trust that its developer is neither stupid nor dishonest. False dilemma. One can simply trust it's developer is just correct. They need neither be stupid or dishonest, but can be wrong. Please dispense with the logical fallacy arguments.
|
|