Stratosphere
                               ------------

                            (Diabatic Heating)


(Written Nov 1995)

There is something strange with the stratospheric winds and temperatures.
Three people tried to calculate the diabatic heating as a residual of the
thermodynamic equation and failed [J. Whitaker (solving the chi problem,
sigma coordinates), P. Peng (finding the forcing for a linear model, sigma 
coordinates) and me (estimating the errors in the diabatic heating, pressure 
coordinates)].  We all got results that were garbage in the stratosphere.  
These are my preliminary results.

1) The diabatic heating computed as a residual is unrealistic above
200 mb.  The field looks like noise and I know of no physical mechanisms
to produce heating rates of those magnitudes.

2) The calculation is very unstable; i.e., the results depend on whether
you use analysis or 6-hour forecast to calculate the heating rates 
(above 200 mb).

3) The peak temperature increment at 100 mb showed an equivalent heating 
rate of 3.5 K/day for 9301 (only month examined).  This temperature
increment is very large and centered at 100 mb.

4) For the month examined (9301), the 700 mb diabatic heating from the
model's physical parameterizations is ~ 0.5 K/day too cold (uniform
about the globe).  Since it is difficult to imagine an error in the
wind field producing a uniform heating, the most likely suspect is
the radiation code.

5) Vertical profiles of the long wave and short wave heating profiles
suggest a very simplistic heating code.  Simple is not always good.

6) See OZONE before using the short-wave heating in the stratosphere.

7) The large-scale features in the stratosphere look reasonable but
there is too much small-scale features.  These small-scale features
may be ruining the calculations.


In the troposhere, the situation is different. P. Peng and I both 
came to the conclusion that the diabatic heating computed as a residual
is better than diabatic heating from the model parameterizations.