|
Post by steve on Jul 18, 2009 16:43:03 GMT
I'm not going to overstate my level of knowledge. Then what you should not question then is Steve's area of expertise. I know what Steve does and he does it very well as its what I do as well in a different field. That particular discussion is *not* McIntyre's area of expertise. I asked him on one of my polite interchanges with him and he said he preferred to stick to climate datasets related to temperature proxy and the like, and that he did not really have the time to look deeply into climate models and climate data. Here's a link to another Steve from Toronto on climate models. This guy has previously looked at practices used in NASA space craft engineering. Perhaps he and McIntyre should meet up. www.easterbrook.ca/steve/?cat=10
|
|
|
Post by magellan on Jul 18, 2009 22:12:50 GMT
Then what you should not question then is Steve's area of expertise. I know what Steve does and he does it very well as its what I do as well in a different field. That particular discussion is *not* McIntyre's area of expertise. I asked him on one of my polite interchanges with him and he said he preferred to stick to climate datasets related to temperature proxy and the like, and that he did not really have the time to look deeply into climate models and climate data. Here's a link to another Steve from Toronto on climate models. This guy has previously looked at practices used in NASA space craft engineering. Perhaps he and McIntyre should meet up. www.easterbrook.ca/steve/?cat=10Once a climate model can pass an independent V&V such as NASA's Software Assurance Standard or standards commonly applied by industry, you won't see any more criticism from me. Until then, any claims they are little more than engineering code FUdged Beyond All Recognition is fodder for the sheeple. Glowing accolades of the predictive skills of climate models even in the face of realtime falsification makes such claimants look all the more foolish. NASA-STD-8739.8 w/Change 1 www.goes-r.gov/procurement/flight_documents/NASA-STD-8739-8-1.pdf
|
|
|
Post by icefisher on Jul 18, 2009 22:14:19 GMT
That particular discussion is *not* McIntyre's area of expertise. I asked him on one of my polite interchanges with him and he said he preferred to stick to climate datasets related to temperature proxy and the like, and that he did not really have the time to look deeply into climate models and climate data. You are failing to understand Steve's area of expertise. Its not unusual for somebody to not understand what auditing is all about. What you are suggesting is if a climate modeler murders his wife, a police investigator is an inappropriate person to investigate the crime because he know nothing about climate modeling.
|
|
|
Post by socold on Jul 19, 2009 3:27:14 GMT
That particular discussion is *not* McIntyre's area of expertise. I asked him on one of my polite interchanges with him and he said he preferred to stick to climate datasets related to temperature proxy and the like, and that he did not really have the time to look deeply into climate models and climate data. Here's a link to another Steve from Toronto on climate models. This guy has previously looked at practices used in NASA space craft engineering. Perhaps he and McIntyre should meet up. www.easterbrook.ca/steve/?cat=10Once a climate model can pass an independent V&V such as NASA's Software Assurance Standard or standards commonly applied by industry, you won't see any more criticism from me. Until then, any claims they are little more than engineering code FUdged Beyond All Recognition is fodder for the sheeple. Glowing accolades of the predictive skills of climate models even in the face of realtime falsification makes such claimants look all the more foolish. NASA-STD-8739.8 w/Change 1 www.goes-r.gov/procurement/flight_documents/NASA-STD-8739-8-1.pdfIndependent V&V is effectively already done for climate models by virtue of having different teams globally producing their own implementations of the same software as well as going through plenty of tests by the very nature of what they are.
|
|
|
Post by icefisher on Jul 19, 2009 3:54:02 GMT
Independent V&V is effectively already done for climate models by virtue of having different teams globally producing their own implementations of the same software as well as going through plenty of tests by the very nature of what they are. Sorry Socold but you missed that one by a country mile. Let me show you an analogy. Literally thousands of financial institutions and investors and their auditors bought into the CDS market. Each probably built their own investment/audit models and everyone of them completely missed the mark on the crucial issues. Huh, massive defaults on home mortgages? ? Indeed some were sounding the alarm but nobody was listening. Investors though were just pulling ready to go evaluation model pieces off the shelf like most people tend to do in these investment bank vehicles. I can guarantee you not a single one of those investor models entailed actually opening a loan file and checking the underwriting on the underlying asset. Bottom line Socold, Magellan is not referring to some ad hoc verification procedure but a formal well planned top to bottom deconstruction of the models. Popularity of ideas is not a good measure of their advisability. . . .ask the retired hardcore drug culture participants if you can find any. Claiming popularity is a good V&V procedure just shows how ignorant you are about such things.
|
|
|
Post by northsphinx on Jul 19, 2009 8:12:27 GMT
Radiation balance measurements show CO2 effects in the atmosphere First check this: science.larc.nasa.gov/ceres/press_releases/images/prsi/ceres_olr_20010525_s.pngNice picture that show outgoing radiation into space from earth. By the way, the clouds is very efficient to stop outgoing radiation. And compare with this: cimss.ssec.wisc.edu/iremis/images/MYD11C3.A2006091_emisA_wavelen8.3_filled.pngGlobal Infrared Land Surface Emissivity. Everywhere except in dry desert is the emissivity close to 1. In dry desert is the emissivity as low as 0,7-0,75. If we use the dry desert outgoing radiation with that dry desert emissivity do we get a daily average surface temp of about 305K. Make sense. That tells me that hardly any of outgoing LW radiation is absorbed in a dry atmosphere. It just go directly into cold space. The CO2 is not trapping any of the outgoing LW radiation. This show that water vapor and especially clouds do.
|
|
|
Post by glc on Jul 19, 2009 9:00:19 GMT
That tells me that hardly any of outgoing LW radiation is absorbed in a dry atmosphere. It just go directly into cold space. The CO2 is not trapping any of the outgoing LW radiation. This show that water vapor and especially clouds do.
How do you explain the CO2 funnel in emission spectra?
|
|
|
Post by steve on Jul 19, 2009 10:12:56 GMT
That particular discussion is *not* McIntyre's area of expertise. I asked him on one of my polite interchanges with him and he said he preferred to stick to climate datasets related to temperature proxy and the like, and that he did not really have the time to look deeply into climate models and climate data. Here's a link to another Steve from Toronto on climate models. This guy has previously looked at practices used in NASA space craft engineering. Perhaps he and McIntyre should meet up. www.easterbrook.ca/steve/?cat=10Once a climate model can pass an independent V&V such as NASA's Software Assurance Standard or standards commonly applied by industry, you won't see any more criticism from me. This is a perfect example of a sceptic demanding a ludicrously high standard of proof. NASA software standards are by far the highest in the world. And for good reason. When you have a one off shot to get a multimillion spacecraft right you try to be very very careful indeed. If Mars Climate Orbiters cost $1000 a pop then it wouldn't matter so much if the first few plummeted into the marsosphere due to a metric-imperial foul-up. But by attempt number 10 most of the software bugs would have been ironed out. Most software in the world gets to be run on a rig that is identical or almost identical to its production platform. Data can then be analysed, bugs spotted and ironed out. It's a completely different kettle of fish.
|
|
|
Post by steve on Jul 19, 2009 10:39:19 GMT
That particular discussion is *not* McIntyre's area of expertise. I asked him on one of my polite interchanges with him and he said he preferred to stick to climate datasets related to temperature proxy and the like, and that he did not really have the time to look deeply into climate models and climate data. You are failing to understand Steve's area of expertise. You are failing to understand the topic of conversation. He has not grasped the subject of climate sensitivity, but is happy to hold forth on his opinion about what is required. If you are going to tell me that all of Steve McIntyre's mining reports calculated levels of resource extraction to within a factor of 2 or 3 then I'm going to doubt you. Yet people invest in mines.
|
|
|
Post by steve on Jul 19, 2009 10:44:56 GMT
Radiation balance measurements show CO2 effects in the atmosphere First check this: science.larc.nasa.gov/ceres/press_releases/images/prsi/ceres_olr_20010525_s.pngNice picture that show outgoing radiation into space from earth. By the way, the clouds is very efficient to stop outgoing radiation. And compare with this: cimss.ssec.wisc.edu/iremis/images/MYD11C3.A2006091_emisA_wavelen8.3_filled.pngGlobal Infrared Land Surface Emissivity. Everywhere except in dry desert is the emissivity close to 1. In dry desert is the emissivity as low as 0,7-0,75. If we use the dry desert outgoing radiation with that dry desert emissivity do we get a daily average surface temp of about 305K. Make sense. That tells me that hardly any of outgoing LW radiation is absorbed in a dry atmosphere. It just go directly into cold space. The CO2 is not trapping any of the outgoing LW radiation. This show that water vapor and especially clouds do. This tells you nothing! The spectrum is taken at 8.3 microns which is away from the CO2 spectral lines.
|
|
|
Post by icefisher on Jul 19, 2009 11:40:40 GMT
You are failing to understand Steve's area of expertise. You are failing to understand the topic of conversation. He has not grasped the subject of climate sensitivity, but is happy to hold forth on his opinion about what is required. If you are going to tell me that all of Steve McIntyre's mining reports calculated levels of resource extraction to within a factor of 2 or 3 then I'm going to doubt you. Yet people invest in mines. LOL! Thats because when Dr Hansen started salting the mine. . . .Steve picked up on it instantly! . . . .Plenty of dummies around who didn't.
|
|
|
Post by northsphinx on Jul 19, 2009 12:11:43 GMT
The heat balance glc, the heat balance is the answer.
To have a net heat flux by radiation is a heat difference needed. There is a big one between earth and space. Also inside a gas is it also needed a temperature difference between molecules, much shorter distance, and smaller temperature difference.
We can ourself see every clear night that the atmosphere not absorbing much of outgoing LW radiation. A cloudy night is completely different. What is the big difference between vapor and clouds, and between CO2 and clouds? I think it is because of the heat balance on a micro scale. Carbon dioxide molecule absorbs and emits LW in narrow bands. A heated CO2 molecule will very efficient heat the next one like dominoes. The absorbing is as efficient as the emission. But and here is the big but: The temperatures of these CO2 is not set by radiation but of the ambient atmosphere since the heat capacity of CO2 is lower than air. There will be a very little temperature difference between CO2 molecules that will make a very small heat transfer by radiation possible. If vapor is available will the overlapping band drain CO2 from energy into H20 bands due to the heat capacity. The long distance radiation effect for CO2 will vanish into a micro level "dominoes" effect.drained by vapor. That will make it impossible for CO2 to heat or cool the atmosphere. The CO2 temperature is set by the atmosphere. Not the other way around. Proof of this is that during clear dry night is the atmosphere cooled from the surface. The air itself is not emitting much heat. If it not emits heat will it not either absorbs heat by radiation.
The vapor have higher heat capacity than the air and will be able to change the air temperature. Clouds have even further higher heat capacity and will very efficient heat or cool air. Or prevent radiation. Pure water as rain is very efficient in this due to its high heat capacity.
When the CO2 "dominoes" reach dry air is they not drained by the H2O and will show the air temperature it is in.
|
|
|
Post by socold on Jul 19, 2009 13:06:24 GMT
Bottom line Socold, Magellan is not referring to some ad hoc verification procedure but a formal well planned top to bottom deconstruction of the models. ... Claiming popularity is a good V&V procedure just shows how ignorant you are about such things. Not popularity. Independant replication of results. That is a means of validating it. V&V doesn't mean a code review or a "deconstruction of the models". It means validating the output and verifying the model meets requirements. As the developers of the models are it's users the general requirements of being a GCM are obviously honored. And the models by their very nature are validated with different input tested against output. For example 20th century simulations are a form of validation. This whole "V&V formal" argument also sounds very much like a typical politically correct mantra that sounds like an authoriative argument simply through use of acronym. From a more practical point of view, the kind of view that any software business would take, what exactly is the problem we are trying to solve here with this "V&V" rubber stamp? One potential problem is the possibility of coding errors in the model. Another problem is the possibility of the physics being implemented incorrectly. In both cases this problem is alievated through multiple independant implementations of the same thing by different teams.
|
|
|
Post by icefisher on Jul 19, 2009 18:12:15 GMT
Bottom line Socold, Magellan is not referring to some ad hoc verification procedure but a formal well planned top to bottom deconstruction of the models. ... Claiming popularity is a good V&V procedure just shows how ignorant you are about such things. Not popularity. Independant replication of results. That is a means of validating it. You are confusing the imaginary world of modeling with the real world of experimentation Socold. As I said above in the portion of my statement you erased (proving my point) some were sounding the alarm and few were listening. V&V doesn't mean a code review or a "deconstruction of the models". It means validating the output and verifying the model meets requirements. As the developers of the models are it's users the general requirements of being a GCM are obviously honored. And the models by their very nature are validated with different input tested against output. For example 20th century simulations are a form of validation. The 20th century simulations I have seen that match model output are stuff like Mann salting the mine by changing history to make the 20th Century fit the models and is precisely the sort of stuff that I am talking about. Bottom line is that kind of fraudulent science is common and references from that nonsense gets embedded in a lot of other science and since there is no experimental validations it just stays there. This whole "V&V formal" argument also sounds very much like a typical politically correct mantra that sounds like an authoriative argument simply through use of acronym. From a more practical point of view, the kind of view that any software business would take, what exactly is the problem we are trying to solve here with this "V&V" rubber stamp? You are the one obsessing on the acronym Socold. V&V is not the only program in the world that provides for a high level of QC. Magellan was using that as an example. Financial auditing has a far higher standard for QC than what you propose as adequate; but it is not perfect and devious individuals can fool auditors. But there is no way in God's Creation you could get an auditor to sign off on AGW, devious people or not, considering the controversy over the science that one can see everyday in the press short of a complete line by line, a review for completeness, and probably even a higher level of QC than is represented by the V&V process. Enron and Worldcom would not have happened if there had been advance controversy a mere fraction of what controversy we see over the science on AGW. The reason that is true is because the individual partner auditor will be putting his entire personal wealth, job, and profession on the line and the personal wealth, jobs, and professions of every one of his partners as well. Magellan said V&V is what he wants to see and presumably his only motivation is to save some taxes and see the nation generally prosper. If he had to bet his entire individual future on it he might want to see more done than V&V. One potential problem is the possibility of coding errors in the model. Another problem is the possibility of the physics being implemented incorrectly. In both cases this problem is alievated through multiple independant implementations of the same thing by different teams. You left out the biggy! First nobody is going to do a climate model unless they believe they can adequately deal with external forcings. Stuff like LIA and its recovery; not to speak of poorly understood historic fluctuations of the planet from ice age to warm period represents a deck with a lot of wild cards. The models play on the theory the most aces you can get is 4 because thats all you can see. Obviously something has accounted for past warmings and coolings other than CO2. The only way AGW got as far as it did politically was from the phony science since largely discredited like Mann's Hockey Stick and salting of the mine by playing games with Bristlecone pines and a number of artifices to suggest the warming we have been experiencing is unprecedented and thus must be from a new kind of forcing. Get real Socold, ultimately these fraudsters are apt to do more harm than good. Because now we know the theory has been fraudulently promoted. Its going to be hard to convince people its not all fraudulent if indeed it is not. At any rate there is no excuse for not conducting an exhaustive review of any model this nation decides to put faith in. You are trying to pooh pooh that but all you look like is a guilty man trying to deflect the police. As an auditor I find it really telling that its that sort of defense, not just from you, but the author's of the science upon which the theory is built upon. One would think that they would be falling all over themselves to devulge datasets, algorythms, and computer codes if indeed the world was at risk. But nope! Each one is looking after their own private lot, suggesting either their priorities are screwed up or they really aren't all that concerned. What we get instead is all this gut wrenching appealing going on by a bunch of blogsters who have really no personal stake in anything but bleating the flock song.
|
|
|
Post by socold on Jul 20, 2009 0:49:20 GMT
V&V is not a program. It's simply the two words "Verification and Validation" and how software companies do each is up to them.
What you are demanding is external auditing, not V&V.
I am also saying that external auditing is pointless in this situation because the product is implemented multiple times by multiple independant teams anyway, therefore alleviating the possibility of fraud and bugs. The models are also audited by external testing of the model by climatologists outside the development groups, including standard model tests that all model implementations must undergo.
|
|