No signifigant change in OLR in the wavelengths absorbed by CO2. There is a slight decrease in the amount of OLR in other wavelengths that may or may not be attributable to us. Methane, for instance.
Not at all. For example, a rise in the methane concentration in the atmosphere will result in more OLR being absorbed which one might intuit to result in a net temperature rise. Recent research, however, has shown that rising temperatures result in lower humidity levels in the upper atmosphere which result in a net cooling effect.
http://www.nature.com/news/2010/100128/full/news.2010.42.html
http://www.noaanews.noaa.gov/stories2010/20100128_watervapor.html
http://www.friendsofscience.org/assets/documents/The_Saturated_Greenhouse_Effect.htm
Again, there are simply too many unknowns at this time to take any dire warning or economic suggestion seriously.
Tell me if I understand what you are saying.
You are treating the atmosphere as a very complex, open thermodynamic system. You are trying to determine the inter-related macroscopic quantities of work, heat, free energy (I suppose since we include the biosphere) and perhaps entropy by examining spectroscopic data in the molecular level. Obviosly, you need probability theory to do this, that is, relating the over all effects of a large population of particles by their individual motions or states.
Is that the jist of climate science?
Present models, for example, when run in reverse don't postdict the medieval warm period or the little ice age. Neither do they postdict the levelling off and actual decrease of the temperature in the past 18 years or so. If a program can't even account for what has happened, how much credibility can it have for predicting the future.
Again, how significant is this 'medieval warm period' or 'little ice age' to the big picture of geologic time? As I have said above, the study involves probability math. There is no exact right or wrong here, only a right or wrong within tolerable limits of error.
In my line of work, the exact numeric conclusion is just as important as the variances that accrue from the methodology employed in arriving at that number. In short, to what confidence level may the end-user employ the information I am providing? Is it correct to within 20, 50 or a 100 ppm? What are the possible or most likely outliers in the data set and how is it significant to the result? Are these outliers random errors or do they occur systematically in the data gathering process? etc. etc.
No, but with the various climate gate scandals presently going on, a rather large amount of attention has been focused on the methods of computer modelling used by various climate "authorities) and the findings have been far less than stellar. Information isn't hard to find.
I'd rather talk about this with you than google, if you don't mind.
I don't see how any one is expected to do any science without models. And models are mathematical algorithms, are they not? They are only as accurate as the mathematical assumptions that you put into them, no? And when you are starting from the motion of infinitessimal particles and how they translate to the big picture, it is obviously a problem of probabilities.
Which goes back to my original question -- how are the errors in postdicting significant? Are they merely statistical glitches occuring in a very, very large population over a very, very long time or is there something fundamentally wrong with the model itself?
I started school a very long time ago with a keen interest in bio mechanical robotics (when the field was hardly more than science fiction) but by the time I barely passed calc III by the skin of my teeth despite hours upon hours spent in the math lab, I knew that the math would be my downfall if I were to pursue engineering.
LOL
I'd think you gave up too easily if that is the case. Everyone is going to have to claw their way through higher math. And the most important practical problems aren't neat. They invariably have solutions involving some messy partial differntial.
The good thing about it is that we aren't expected to do all these by hand. We have all sorts of softwares for this.