The file c230Nov02.f are corrections I began compiling together in
November but didnt complete until January 2003. The are a changes
submitted by various users of the LEM and correct some minor bugs or
improvve portability. Here are some of the comments:

 Mods by Alison to:

- allow byte swapping during model run time. Use *DEFINE BSWAP
 in the top level updates (changed by Jon to allow control with an
 update flag)

 Mods by Marc to:

- imporve portability as some machines need the seed for random
  numbers to be at least 4 integers long.

 
 Mods by Adrian to:  
- stop problem with fort.15 being written.

- test CFL and change timestep every NNCFL steps, 
  rather than NN NNCFL added to namelist CNTRL

 Mods by Malcolm to:

- allows a movie starting at time=0

- allows the option of backscatter with SCT=0, which is useful in some
  circumstances

- avoids out-of-bounds indexing for array DIFF_DG

- correct problem of a counter incremented in the wrong place relative
  to an IPASQP.EQ.0

- remove a potential spurious failure caused by previous STATUS='NEW'
  flag *** NOTE this is also protected by recent versions of lemsub

- sets the value of K, previously omitted (which made the values of
  the surface sensible and latent heat fluxes wrong if ISCBCP.EQ.2)

- corrects an omission in the OVRIDE1 namelist  FTSURF,FQSURF,

- stops you selecting backscatter with the anelastic option , as Andy
  suspects the backscatter is not coded properly in that case.

Mods by Glen to:

- improve portability by avoiding an 'unfinalized MPI error'. This was
  fixed by calling the routine GC_EXIT()

--------------------------------------------------
--------Following in older correction sets--------
--------------------------------------------------
These are just comments pulled from the old correction file.

Mod by Jon from spec_cor:

Correction set for Version 2.3 of the LEM
*/ Correction to stop run if spectral file is 
*/ not read correctly

*/These updates correct the 2-D field diagnostics for MPP 2-D runs

*/ For some reason this message passing no longer works under MPI. 
*/ This change introduces same logic as in TIMSER which does work.

*/ This section is from Adrian Lock to correct an error in the 
*/ calculation of Moist Ri no.

*// Mods to change LEM2.3 MOISTRI2 to work in terms of potential 
*// temperature throughout so that in cloud-free conditions neutrality 
*// (ie. RI=0) is now zero theta_v gradient.
*//

*/ Glen Richardson passed the following on to me. 
*/ I havent tested them but they apparently work
*/ NB These correct some bugs in warm rain.