[MITgcm-support] temperature equation

[David Wang]

Hello,

I'd like to check each term on the rhs of the model temperature equation 
(dT/dt = rhs). So I got ADVx_TH, ADVy_TH, ADVr_TH, DIFx_TH, DIFy_TH, 
DFrE_TH, DFrI_TH and an additional flux diagnostics that I added to 
account for the KPP non-local transport (defined as nlfx_TH). To make 
sure I get everything, I ran the model for 5 days, dumped the 
temperature field snapshots at the beginning and the end of the 
integration, and also dumped the 5-day average of all flux diagnostics. 
I was expecting the following two numbers to be identical:

1) the temperature difference between two records;
2) (5*86400/drF(k)/rA(i,j)) * ( \
    advx_th(i,j,k)-advx_th(i+1,j,k) + \
    difx_th(i,j,k)-difx_th(i+1,j,k) + \
    advy_th(i,j,k)-advy_th(i,j+1,k) + \
    dify_th(i,j,k)-dify_th(i,j+1,k) + \
    drfe_th(i,j,k+1)-drfe_th(i,j,k) + \
    drfi_th(i,j,k+1)-drfi_th(i,j,k) + \
    nlfx_th(i,j,k+1)-nlfx_th(i,j,k) )

However, I checked several points and found they are not identical. The 
discrepancy can be as large as 10%. So did I get all rhs terms as I 
thought or something went wrong?

The model version is checkpoint58m_post. I use both kpp and gmredi pkg, 
centered 2nd order advection scheme.

-- 
Daiwei (David) Wang	 Graduate Student
Lamont-Doherty Earth Observatory of Columbia University
103B Oceanography
61 Route 9W, Palisades, NY 10964-8000
Voice: (845) 365-8694