[MITgcm-devel] [Fwd: Re: ECCO numerics]

[Dimitris Menemenlis]

Patrick, without being familiar with bulk formulae code, here is what I can 
gather.  New code from GFDL is attached.  It comes from
http://data1.gfdl.noaa.gov/nomads/forms/mom4/CORE/code.html

Difference between old code and new code is the following:

> !!$        ch(i) = ch_n10/(1+ch_n10*xx/cd_n10_rt)**2;                !       b (bug)
> !!$        ce(i) = ce_n10/(1+ce_n10*xx/cd_n10_rt)**2;                !       c (bug)
>         ch(i) = ch_n10/(1+ch_n10*xx/cd_n10_rt)*sqrt(cd(i)/cd_n10) ! 10b (corrected code aug2007)
>         ce(i) = ce_n10/(1+ce_n10*xx/cd_n10_rt)*sqrt(cd(i)/cd_n10) ! 10c (corrected code aug2007)

This seems to correspond to following lines in pkg/exf/exf_bulk_largeyeager04.F

>               rh = rhn/(exf_one + rhn*(ztln-psixh)/karman)
>               re = ren/(exf_one + ren*(ztln-psixh)/karman)

where

>       _RL re                 ! = Ce / sqrt(Cd)     [-]
>       _RL rh                 ! = Ch / sqrt(Cd)     [-]

So I think your code is OK.  Could you confirm.

Man f90 is clean!  D.
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: ncar_ocean_fluxes.2007_08_01.f90
URL: <http://mitgcm.org/pipermail/mitgcm-devel/attachments/20070807/e1e36803/attachment.f90>