- \frac{dU}{dz}=-U+B,
- \frac{dD}{dz}= D-B,
where U is upwelling radiation and D is downwelling radiation and B(T) is upward/downward emission of radiation according Stefan-Boltzmann's law with T temperature depending on z. Subtracting the equations, we have
- \frac{d(D-U)}{dz}=D-2B+U. (0)
We see a balance at a certain level z between U as incoming radiation from below, D as incoming radiation from above and -2B as emission upward and downward. Discretizing z into discrete levels/nodes z_i=i*h with h a mesh size and i=0,1,2..., the balance of upward and downward radiation involves
- B_{i+1}-2B_i+B_{i-1} (1)
with B_i=B(T(z_i)), if we set B_{i+1}=D(z_{i+1}) and B_{i-1}=U(z_{i-1}).
We will now compare (0)-(1) with the following discretised heat equation for a temperature T(z,t) with t a time coordinate:
- dT_i/dt = (T_{i+1}-2T_i+T_{i-1})*\frac{1}{pow(h,2)} (2)
where T_i=T(z_i,t), which is a discretisation of \frac{\partial T}{\partial t}=\frac{\partial^2 T}{\partial x^2}. We know that the heat equation is a stable physical model subject to perturbations p of (2) of the form
- dT_i/dt = (T_{i+1}-2T_i+T_{i-1})*\frac{1}{pow(h,2)}+p. (3)
while perturbations P of (1) would have the form
- B_{i+1}-2B_i+B_{i-1}+P. (4)
We see that introducing a perturbation P of the form (4) in (2) would correspond to a perturbation p=\frac{P}{pow(h,2)} which would be big since h is small. We thus see a fundamental difference as concerns stability between a heat equation of the form (2), which is stable, and a two-stream model of radiative transfer of the form (0).
Note that the idea of two-stream radiative heat transfer with upwelling/outgoing (longwave) radiation OLR and downwelling (long wave) DLR is firmly implanted in the discussion of the GHE through the two-stream Schwarzschild model. In particular, the existence of DLR is supposedly being demonstrated experimentally with the help of a pyrgeometer, which however is a ghost detector.
Inga kommentarer:
Skicka en kommentar