Ray and Astrid,

I have compared two code sets, following your global change Dec solstice runs 
of timegcm1.3 and timegcm1.41. The main structural difference is the consolidation
of lbc calculations into the lbc module (lbc.F, which replaces bndry.F). I think
the best way to compare these is to make controlled runs of both. I will do this
later, but in the meantime, below are summaries of differences in the source
files on the list you gave me yesterday:

Left arrows (<) are timegcm1-2dev7, which consists of:
  $TGCMROOT/timegcm1-2dev7/src + ~roble/timegcm/timegcm1.3.dsdgc/modsrc.dsdnc

Right arrows (>) are timegcm1.41, which consists of:
  $TGCMROOT/timegcm1.41/src + $TGCMROOT/timegcm1.41/src/modsrc + ~roble/timegcm/timegcm1.41d.dsdgc/modsrc.dsdnc

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
mgw.F:
92,94c98,101
< 	raykms1 = 5.0e-7
< 	raykms2 = 3.0e-7
< 	raykts  = 2.0e-7
---
> 	raykms1 = 1.0e-6
> 	raykms2 = 3.0e-6
> 	raykts  = 2.0e-6
388c391,392
<      |  (0.5*(1.+tanh( (rlat-al0)/dlat0))*facnhe
---
> !!   |  (0.5*(1.+tanh( (rlat-al0)/dlat0))*facnhe
>      |  (0.5*(1.+tanh( (rlat-al0)/dlat0))*facnhe*3.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
gw_drag.F (setting tautmp according to latitude) 
(v1.41 code line 1122 should set tautmp = 1.0e-5*exp(...)
1119c1119,1120
<       tautmp = 1.0e-3*exp(-(rlat-1.134)**2/0.2684**2)*facoron
---
>       tautmp = 1.0e-5*exp(-(rlat-1.134)**2/0.2684**2)*facoron
> ;      tautmp = 1.0e-3*exp(-(rlat-1.134)**2/0.2684**2)*facoron
1121c1122,1124
<       tautmp = 1.0e-3*exp(-(rlat+1.134)**2/0.2684**2)*facoros
---
>       tautmp = 1.0e-
> 
> *exp(-(rlat+1.134)**2/0.2684**2)*facoros

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
gw_share.F:
Files are identical.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
bndry.F:
Replaced by lbc.F in v1.41

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
lbc.F:
Replaces bndry.F in v1.41

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
cons.F:
No significant differences (both versions use prndtl = 10.)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
tuvbnd.F is replaced by uvbnd.F in v1.41:

Sub uvbnd (uvbnd.F) calculates lbc of U,V
Sub uvbgrd (uvbnd.F) calculates background lbc for U,V (Astrid did this part)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
dt.F:
Calculation of TN lbc has been moved to lbc.F.
v1.41 introduced joule heating factor joulefac, but it is set to 1.0, so no difference.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
duv.F:
Calculation of UN, VN lbc has been moved to lbc.F.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
filter.F:
No significant difference (periodic points are set in v1.41)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
magfield.F:
v1.41 introduces sunloc_apex (use apex coords to find sun's longitude)
v1.41 sets polar values of dynpot

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
params.F:
v1.41 extension of upper boundary from +5 to +7

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
xray.F:
Files are identical, and neither version calls xray.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
qjchem.F:
No significant difference.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 
radcool.F:
No significant difference.