[MITgcm-support] balance OBCS correction

[Martin Losch]

Qian,

the best option is alway to balance the flow offline, i.e. before you 
give it to the model. In that way you have full control over the 
balancing period (say, a full year). For that you need to be careful to 
compute the correct fluxes. They have to include the appropriate hFacW 
(for u-flux) and hFacS (for v-Flux). There have been several similar 
questions about this, where people have answered with more details, so 
I'll let you search the archive for this and will not repeat the flux 
computations here.

The OBCSbalance option is for convenience and works as you said every 
time step so that mean seasonal signals are suppressed. OBCS has so far 
no code to balance the flow for longer time scales. How would you do it 
in a general way? You'd have to read the data for a full year and then 
manipulate it, or compute the mean of the current year/season/day and, 
as an approximation, substract it from the next period of interest, 
possibly taking into account pickup files etc. Too complicated, so we 
dropped this project, because (as I said in the beginning) you can do 
this offline.

the maskInC/W/S do not differ from the surface layer of maskC/W/S except 
that the open boundary points and all points beyond the open boundaries 
are masked out (=0). It's OK to use 2D masks, because the velocity 
fields are masked elsewhere by the 3D masks (well, hFacW/S) anyway.

Martin

On 05/13/2016 12:30 PM, 钱钰坤 wrote:
> Hi all,
>
> I've known that the balance OBCS can overwrite the flows given at OBs to
> ensure a zero-mass transport.  In a regional run, is the model-output
> flow at OBs the original (unbalanced) OBCS or corrected (balanced) OBCS?
>   I really want to know how large does such correction affect my simulation.
>
> I wonder if there could be a further improvement for the balance OBCS.
>   The current balance OBCS does correction every time step so that the
> seasonal variation of sea level cannot be reproduced even the original
> OBCS contains such signal.  Generally, a regional run with a
> climatological forcing (annual cycle) will produce net inflow in some
> season/month and net outflow in the remaining, but the sum over the year
> will be zero.  I hope to use a balance OBCS can automatically balance
> the flow during the period of the climatological forcing.
>
> While there seems no such option, I decided to correct the net flow
> myself.  After reading obcs_balance_flow.F, I found that in computing
> the east OB tileArea, maskInW (2D array) is used instead of maskW (3D
> array) (see lines 149-153 in obcs_balance_flow.F for version c65m).  The
> area is for east OB is in vertical-meridional plane so it is strange not
> to use a vertically varying (3D) mask, as there are less ocean grids in
> the deeper ocean.  It is appreciated if someone could tell me the
> difference between maskInW and maskW (I only know that maskInW is the
> interior mask for u point, from the comments in GRID.h).
>
> Thanks in advanced.
>
> ------------------
> Best regards
>
> Yu-Kun Qian (钱钰坤)
> Center for Monsoon and Environment Research
> Department of Atmospheric Sciences
> School of Environmental Science and Engineering
> Sun Yat-sen University
> No. 135 Xingang West Road, Haizhu District
> Guangzhou, 510275, P.R. China
> Tel; 020-84115227
> Email: qianyk at mail3.sysu.edu.cn <mailto:qianyk at mail2.sysu.edu.cn>
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support
>

-- 
Martin Losch
Alfred Wegener Institute for Polar and Marine Research
Postfach 120161, 27515 Bremerhaven, Germany;
Tel./Fax: ++49(0471)4831-1872/1797