|
Post by icefisher on Apr 7, 2009 4:14:41 GMT
Gridley. I had to go back and read your original post on the electrical analogy, but I'll give it a shot. In electrical circuits, if two or more resistances are placed in parallel, the total resistance is then the inverse of the sum of the inverses. ie, 1 / (1/r-1 + 1/r-2 +1/r-3....). In the GHG world, this would mean that more IR is passed (not absorbed) than would be due to any one concentration area. (Do I have the analogy correct?). I don't believe this would apply as the GHG does not present a resistance to the passage of photons. If the conditons are right, the molecule may absorb, or remove, a photon from the IR flux leaving the background. It may then re-emit a photon of equal or lesser value in any direction. The Greenhouse theory says that since 50% of those photons will be sent back from whence they came, this will reduce the rate at which energy is lost or add energy to the emitter (earth), making the earth warmer. A warmed molecule of CO2 though emitting in all directions will need will emit sufficient photons to maintain an equilibrium. Photons being emitted back at earth don't count. Lets face it atoms at the surface of the earth at temperature X are emitting Y photons from half its surface. The GHG absorbs Y-Z (Z being the IR the GHG is transparent to) and in turn on half its surface towards space emits Y-Z after obtaining the correct temperature to do so. Trying to account for photons the GHG might be emitting back to earth is a futile exercise that merely makes the problem unnecessarily complex whether or not there is such an effect (keep in mind the whole process is an abstraction anyway and it does not matter cause there is no place for this additional heat to hide in the system). Heat in always equals heat out. I
|
|
|
Post by jimg on Apr 7, 2009 6:07:39 GMT
Icefisher. Are you arguing that the greenhouse effect is non-existent?
Just CO2 GH effect?
|
|
|
Post by icefisher on Apr 7, 2009 8:20:29 GMT
Icefisher. Are you arguing that the greenhouse effect is non-existent? Just CO2 GH effect? Not at all. If there were no GHG to warm the atmosphere it would be a lot colder. The issue here isn't whether there is a GHG effect its how it works thats at issue. One's antennae has to go up when the explanation is the model was built on recent warming. . . .that begs the question as to whether you know what causes the warming. Anybody can build a model on any 10 year period of temperature increases and assume all the temperature increases where caused by your model's chief variable. Its easy enough to see thats what they did, they merely extended the blade of the hockey stick. Physics has little to do with such modeling devices. Anyway the physics is too tough. You have to have a cloud model among other things. Everybody admits that can't be done. Further, I read the IPCC piece on how they measured certainty. They set the bar way below what a typical accountant will allow as risk in providing information to an investor. (by about a factor 6 and thats based upon the IPCC's own assessment of that. . . .) The reason the bar is set so low is to make it higher they would have had to go into the handle of the hockey stick and that wasn't helping their cause. . . .the fraud is palpably visible. Somebody put such a model in front of me for a financial scheme and I would shred it. Thus they have never even been close to being good enough upon which to base policy decisions if you use Wallstreet as precedence. Think about that!!!!
|
|
|
Post by FurryCatHerder on Apr 7, 2009 13:28:40 GMT
Ice,
What you've written above, and the actual physics, have absolutely nothing in common.
(That is a polite way of saying you're completely and totally wrong.)
|
|
|
Post by gridley on Apr 7, 2009 14:38:22 GMT
Further, I read the IPCC piece on how they measured certainty. They set the bar way below what a typical accountant will allow as risk in providing information to an investor. (by about a factor 6 and thats based upon the IPCC's own assessment of that. . . .) The reason the bar is set so low is to make it higher they would have had to go into the handle of the hockey stick and that wasn't helping their cause. . . .the fraud is palpably visible. Somebody put such a model in front of me for a financial scheme and I would shred it. Thus they have never even been close to being good enough upon which to base policy decisions if you use Wallstreet as precedence. Think about that!!!! icefisher, would you elaborate on this point a bit? I've heard things like this before, but the explanation always went over my head. Thanks, Gridley
|
|
|
Post by gridley on Apr 7, 2009 14:43:16 GMT
jimg, I would say "heat flux" rather than IR, but that is essentially the question I'm asking, yes. So while CO2 may not retard photons, it does retard the actual heat flux (I'm trying very hard to avoid starting another iteration of the "CO2 perpetual motion machine" here). Would you accept that as a valid starting point? Gridley. I had to go back and read your original post on the electrical analogy, but I'll give it a shot. In electrical circuits, if two or more resistances are placed in parallel, the total resistance is then the inverse of the sum of the inverses. ie, 1 / (1/r-1 + 1/r-2 +1/r-3....). In the GHG world, this would mean that more IR is passed (not absorbed) than would be due to any one concentration area. (Do I have the analogy correct?). I don't believe this would apply as the GHG does not present a resistance to the passage of photons. If the conditons are right, the molecule may absorb, or remove, a photon from the IR flux leaving the background. It may then re-emit a photon of equal or lesser value in any direction. The Greenhouse theory says that since 50% of those photons will be sent back from whence they came, this will reduce the rate at which energy is lost or add energy to the emitter (earth), making the earth warmer.
|
|
|
Post by jimg on Apr 7, 2009 16:57:12 GMT
gridley: I don't think the heat flux / resistor analogy works.
Since the parallel resistor combination is the inverse of the sum of the inverses, and heat flux is a direct proportion to: Area, thermal conductivity and difference in temperature, or mass flowrate specific heat and delta-T. (with a couple of other formulas for calculating Q.)
In the GHG scenario, the gas, CO2, water vapor, doesn't actually impede the flow of heat by covection. It does impede the flow of radiation via photon capture.
If a photon were to be emitted of the same wavelength and in the same direction, no net change.
But if it is sent back to the surface, then there is a net loss in heat out.
I do however totally agree that climate physics is far more complicated than what is ascribed to CO2. And would also agrue that there is far more natural variability than is allotted in the models.
|
|
|
Post by icefisher on Apr 7, 2009 19:41:43 GMT
Further, I read the IPCC piece on how they measured certainty. They set the bar way below what a typical accountant will allow as risk in providing information to an investor. (by about a factor 6 and thats based upon the IPCC's own assessment of that. . . .) The reason the bar is set so low is to make it higher they would have had to go into the handle of the hockey stick and that wasn't helping their cause. . . .the fraud is palpably visible. Somebody put such a model in front of me for a financial scheme and I would shred it. Thus they have never even been close to being good enough upon which to base policy decisions if you use Wallstreet as precedence. Think about that!!!! icefisher, would you elaborate on this point a bit? I've heard things like this before, but the explanation always went over my head. Thanks, Gridley The systems are so complicated they prevent true modeling. So a proxy model is created. However, proxy models have extremely low predictive ability. Further they can look great for a long time just before they fly apart. I have done modeling for real estate portfolios and derivatives. The fact is real estate pricing is a complicated subject its affected by credit availability, interest rates, economic climates, inflation, absorption rates (sales) all sorts of stuff way to complicated to model. Its easy though to build a model for real estate that looks at various ways to profit in the real estate market. But these are serendipity models where an investor ultimately puts his finger in the air and guesses at some of the parameters that are not modeled. What the models are useful for are to measure risk based upon changes of assumptions. In a real estate model all the variables that are not actually modeled are effectively summed up in price and absorption rate assumptions. Derivative models are another form of proxy model that usually just models returns at various external interest rate levels. . . .(like various CO2 emission rates) loss rates by the underlying securities is ignored because it is "assumed" to be absorbed by residual holders. . . .a portion of the portfolio held by somebody else. (in the CO2 model thats probably the constant sun assumption) The only way to value the residual is to look at the underwriting of the underlying securities. Gads I had years of work doing that during the S&L crisis of a decade earlier! Talking about grunt work! When the CO2 models were built they were received by experts in the same way that the derivative models were received by experts. Historically everything looked to be in order. . . .the essential inductive premise surrounding the model. Then something unexpected interferred. In the derivative case it was a government incentivized lending system that started wholesale purchasing poorly underwritten product. Sellers were highly motivated as it wasn't difficult to sell what you produced. . . .so underwriting standards were skirted by essentially telling borrowers that they would never be held to lying on their loan applications. . . .with the sellers actually filling out the necessary lies for them specifically to qualify. . . .since no verification system was in place. Prosecution is fruitless both because you have two parties each blaming each other and because it became so endemic without challenge. Private securitizations became so popular because the markets became so fluid. Underwriters saw risk but figured they could manage it by selling into the robust international markets. For CO2 what is unexpected came a bit earlier in the cycle, but its a huge red flag! Imagine if whatever has been making it cooler held off for another decade! Believers in the CO2 models at this point in time are nothing but sheep. . . .or its necessary for that next grant with the grantor's being sheep. Bottom line is proxy models are increasingly becoming popular in all sectors. They share the same shortcomings in that the key to these models is their actual performance and the risk they will not perform should always be carefully analyzed. It might not be comforting to know that in these situations there are always winners and losers. In the case of derivative holders the winners were the guys who were net sellers and the losers were the guys who were net buyers. Its a used car sales model essentially. . . .with an adoring public for any old car (being green). I believe its a good thing to become conscious of the risks of CO2 and individually take action to limit ones own footprint. Such a strategy can pursued in a way that one can actually make (or save) money. Thats a nice little private enterprise approach to the problem. Your mileage may vary! The problem really is when governments buy into this sort of stuff and forget hard learned lessons about the role of governments to be skeptical auditors and instead become promoters.
|
|
|
Post by jimg on Apr 8, 2009 3:28:32 GMT
Well said icefisher.
The models work great. Until they don't.
|
|
|
Post by steve on Apr 8, 2009 13:58:37 GMT
Icefisher. Are you arguing that the greenhouse effect is non-existent? Just CO2 GH effect? Not at all. If there were no GHG to warm the atmosphere it would be a lot colder. The issue here isn't whether there is a GHG effect its how it works thats at issue. One's antennae has to go up when the explanation is the model was built on recent warming. . . .that begs the question as to whether you know what causes the warming. Anybody can build a model on any 10 year period of temperature increases and assume all the temperature increases where caused by your model's chief variable. Its easy enough to see thats what they did, they merely extended the blade of the hockey stick. Well my antenna has gone up! The models are not built on recent warming. Predictions of warming due to increased greenhouse gases were made in 1896 and refined in the 1960s and 1970s long before there was a discernable warming trend. It seems clear (also from your further post) that you don't understand the difference between a climate model and the sorts of models you have used. Can you give me a definition of a "proxy model" by the way. Are they some sort of database full of assumptions? Things like this *are* used in climate modelling - I guess the emissions scenarios that the economists came up with might be counted as "proxy models" of the behaviour of human society. The big difference I see with economic models is that climate models are by and large driven by physics, and the physical laws are fixed. If you have an economic model that tells you that inflation-targetting will improve stability such that central banks start to follow such a policy, then somewhere along the line people's behaviour changes to assume such stability and to take more risks. This change in behaviour breaks the assumptions that were used in the economic model rendering it useless.
|
|
|
Post by jimg on Apr 8, 2009 14:57:59 GMT
Proxies are measurements made indirectly, ie, by measuring tree rings and coming up with a temperature graph.
Although the climate models may be made on physical processes, that doesn't mean that assumptions were not made.
ie, clouds, albedo, water vapor effects, steady sun, sun's influence is limited to TSI, NAO, PDO, El-Nino/La Nina...
What has been seen as occuring in the climate that has been attributed to CO2 is much more likely the combined effect of many other forcings/feedbacks that we don't fully understand yet. 5 percent off in magnitude here and there would add up.
But since we don't understand it, it must be CO2.
Like the derivatives models, the climate models work fine until something changes that is outside of what the assumptions were expected to be. Then they fail.
Besides, how many times have we heard in the last 10-20 years, "this was completely unexpected" regarding science that we thought we understood, but clearly did not.
As our technology improves, we are better able to peer into the depths, but I am thoroughly not convinced that we have "arrived" and can therefore map out our destiny accordingly.
|
|
|
Post by sfbmikey on Apr 8, 2009 15:07:27 GMT
steve, are you saying that somebody came up with a model in 1896 that accurately predicted the temperature up till now???
I would like to see that.
yes, a greenhouse gas will always produce warming, and a bird landing on a bridge will always deflect it. from what I have seen, everything before the 70s was of that nature, not "in the period of 1995-2005 the earth will have major warming"
the model problem is a VERY serious one. all of the models I have heard about were constructed in the late 90s or later, all of them were basically trying to match the data we already had, and all of them are overwhelmingly simplistic (indeed, you could make the point that a non-simplistic model is out of our grasp at the moment). Are you saying that the including a few laws of physics make these models accurate??
|
|
|
Post by icefisher on Apr 8, 2009 16:17:36 GMT
Not at all. If there were no GHG to warm the atmosphere it would be a lot colder. The issue here isn't whether there is a GHG effect its how it works thats at issue. One's antennae has to go up when the explanation is the model was built on recent warming. . . .that begs the question as to whether you know what causes the warming. Anybody can build a model on any 10 year period of temperature increases and assume all the temperature increases where caused by your model's chief variable. Its easy enough to see thats what they did, they merely extended the blade of the hockey stick. Well my antenna has gone up! The models are not built on recent warming. Predictions of warming due to increased greenhouse gases were made in 1896 and refined in the 1960s and 1970s long before there was a discernable warming trend. It seems clear (also from your further post) that you don't understand the difference between a climate model and the sorts of models you have used. Can you give me a definition of a "proxy model" by the way. Are they some sort of database full of assumptions? Things like this *are* used in climate modelling - I guess the emissions scenarios that the economists came up with might be counted as "proxy models" of the behaviour of human society. The big difference I see with economic models is that climate models are by and large driven by physics, and the physical laws are fixed. If you have an economic model that tells you that inflation-targetting will improve stability such that central banks start to follow such a policy, then somewhere along the line people's behaviour changes to assume such stability and to take more risks. This change in behaviour breaks the assumptions that were used in the economic model rendering it useless. I would suggest it is you in fact that doesn't understand the difference between a financial model and a climate model in structure and probably also in presentation. The answer is very little. When doing a model to project the future you might include a variable for the democrat party moaning about the economy and you might manifest it in the model as a GDP growth rate from loss of consumer confidence that occurs periodically. A "better" financial long term model will do this as these events are periodic to some degree. It would look like a PDO estimate. . . .but it granted PDOs appear a lot more set in a cycle than democrats moaning about the economy. Did the climate models do this? Well no for the most part. Dr Easterbrook did it though. Now he didn't do that based on physics he did it based upon some knowledge of the cyclical nature of the PDO just like the economic modeler would do it. When it comes to actual physics in the model, financial models have lots of similarities there too. Amortization rates, interest rates, balloon payments, current tax rates, principal are all known stuff, some just at the start others throughout the model. Other things that might affect your model that can be less cyclic might be inflation, future tax rate increases, and prepayments. These are usually afforded a special place in the model where the modeler will make multiple runs based upon different assumptions. Because the import thing about these models is they can alert you to sensitivity or in other words risk. In the climate model this might include cloudiness and precipitation or irradiance, or even maybe CR. For this stuff no physics or cycles might be known. . . .in the case of CR maybe no affect is known. . . .but that doesn't mean there are no estimates of what it might be. . . .so a good model would at least examine it. Some times financial models just leaves stuff out, I already mentioned that regarding typical derivative models and its abundantly clear that climate models do the same thing. Your typical financial guru purchaser usually runs the models with lots of assumptions to get a feel for the investment sensitivity. Its a good way of judging risk. But when the same guru uses the model to sell the investment they might try to argue for the set of assumptions that gets the best price. Often they will even try to enlist their CPA into signing off on it. . . .which the CPA never does if he is focused on staying licensed. . . .though they often pay their CPAs a lot of money to help them develop these models in the first place. Now you tell me what major difference or similarity I haven't acknowledged.
|
|
|
Post by Pooh on Apr 8, 2009 17:54:49 GMT
...When it comes to actual physics in the model, financial models have lots of similarities there too. Amortization rates, interest rates, balloon payments, current tax rates, principal are all known stuff, some just at the start others throughout the model. ... Other things that might affect your model that can be less cyclic might be inflation, future tax rate increases, and prepayments. ... Because the import thing about these models is they can alert you to sensitivity or in other words risk. ... Its a good way of judging risk. In support of icefisher, the key words are sensitivity and risk. If one excludes an important variable or process, the overall results can be misleading. In the latest financial fiasco, I'm fairly sure that few recognized the nature or extent of the risk. The brightest bulbs on the block combined good mortgages with cats, dogs and liar loans, sold them as bonds (or CDOs) and collected their fees. The rating agencies could not or did not recognize the risk in these bonds, but did collect their fees. Still others salved the fears of the wary by selling credit swaps and derivatives, did not reserve assets to cover the risks, but did collect their fees. At the end of the day, no one knew what the income stream would be, nor what the companies were worth. It has been said that the climate models are driven by physics, and the physical laws are fixed. This, however, raises issues about whether all the necessary physics are represented in the models, whether the parameters and initial conditions are correct, etc. It may also be worth noting that we no longer use the Rutherford model of the atom.
|
|
|
Post by icefisher on Apr 8, 2009 18:31:44 GMT
In the latest financial fiasco, I'm fairly sure that few recognized the nature or extent of the risk. The brightest bulbs on the block combined good mortgages with cats, dogs and liar loans, sold them as bonds (or CDOs) and collected their fees. The rating agencies could not or did not recognize the risk in these bonds, but did collect their fees. Still others salved the fears of the wary by selling credit swaps and derivatives, did not reserve assets to cover the risks, but did collect their fees. Brings to mind when I audited derivatives, mbs splits. Principal payment streams on mortgages were separated from interest rate payments. Prepayments and refis made the principal only streams (po's) more valuable, whereas payment per plan made the interest only streams more valuable. What got interesting was what is known as "risk arbitrage". An auditor would typically find these securities selling at premiums on both sides based upon buyer early payment assumptions for POs and holdem scenarios for the IO half. Careful reading of the prospectus would reveal the money all comes from 6% mortgages of which maybe a 1/4 pt was extracted for servicing, another 1/4 point for the securitization and everybody holding the securities to yield something like 8%. OK you have to ask were does the extra 2.5% come from? Boy could bank managers come up with a thousand and one reasons. And they were difficult to audit because they were all sold under the counter so there was no published market for them. And a published market really wouldn't be all that helpful anyway except to reveal the BS between tranches. Its still real estate based which means each one is different. So we almost always had adjustments to these accounts. We had to look up actual prepayment rates and calculate the eventual dollars returned, estimate the risk and match it to current market. But based upon purchase price I always imagined the broker having two prepayment estimates he used to sell the darned things, telling the PO buyer man its going to prepay like a banshee as interest rates go down and the IO buyer that this was loans from stable neighborhoods where people seldom moved and had their loans with their neighborhood S&L where they personally knew the loan officers. After years of that sort of stuff I have gotten a little hardened to the "so and so told me" argument for climate change.
|
|