Someone brought up the subject of Critical Thinking. Some BS about starting with a blank page to remove your biases. To me, thinking is critical, so build a BS detector. So you can figure out if someone is BSing you or if your are BSing yourself. So that brought me right back to the learning equation, Kimoto with a little modification.
The basic equation is dF/dt = 4F/T, where F is energy flux in Wm-2 in this case and T is in degrees Kelvin. It is based on the Stefan-Bolztman relationship for radiant energy from a black body. Energy is supposedly fungible, meaning it can take various forms. So I modified it dF/dT= aF1+bF2+..zFn all over T. That bugs people.
So let's try dF(T)/dT=aF1+bF2+...zFn, multiplying both sides by that pesky T.
In order for this to work, the coefficients a,b,...z, would have to be relationships that make the different energy fluxes proportional to one another. Zoom! That goes right over the head of average reader. What throws most folks off is the 4th power relationship of flux to temperature in the S-B relationship. What temperature? It is on the other side of the equation, we can deal with that when and if we want to. Right now we just have photons, phonons or electrons each doing their own thing. If we are balancing energy, screw the temperature, let's balance energy. Temperature is easy, let the students deal with that later.
So let's do aF1+bF2+...zFn=Fout+S, where Fout is the layer we know what energy is leaving and S is the energy lost along the way, entropy. Kinda looks to me that might simplify things. Some of the energy fluxes might flow more efficiently, so S for that sucker could be small with respect to the others. Some of the energy fluxes might be a bear to figure out, so S for that one may be ?, a question mark. So if we solve for what we know best, the small ones and the question mark ones would be uncertainty until we figure them out.
By setting up the right model, doing a little interpolation between layers, we may be able to hone in on some of those unknowns. Lets' call this redneck logic, how to solve a problem with more unknowns that you care to mess with at one time.
I will leave this post alone for folks to ponder. Put your thinking caps on, this is not your average method of doing bidness :) Next will be layer leap frogging. A tricky way to get rid of most of the satellite measurement filter overlap. I hear that is a little bit of an issue.