[MITgcm-devel] seaice adjoint and EVP

[Martin Losch]

OK, let's stick to email for now. We can talk on Thursday.

See my comment below.
On 21 May 2007, at 04:03, Dimitris Menemenlis wrote:

> Hi Martin,
>
>> 1. Are these figures all with with zMin = 0?
In this case it may be worth turning of individual terms in the rhs  
of the momentum equations
1. dphiSurf/dx and dphiSurf/dy (in seaice_dynsolver)
2. surface wind stress (taux/y=0 in seaice_get_dynforcing)
3. ice-ocean stress (DWATN in seaice_evp)
4. Coriolis
5. stressDivergence
4 and 5 should be zero over open ocean anyway so I do not see how  
these terms can lead to the stripes.

We should get to the bottom of what is causing these stripes. that  
way we can probably understand the noise in the ice fields, too.
>
> Yes, all the figures and results under
> http://ecco2.jpl.nasa.gov/data1/arctic/output/tests/
> (except for the oldtest subdirectory) are with zMin=0.
>
>> 2. Do you have an EVP run that does not blow up at all (regardless  
>> of noise)?
> I have not run any of the zMin=0/SEAICEuseFlooding=.true. tests out  
> for very
> long, but I am almost certain that none of these new integrations  
> will crash,
> including the SEAICE_deltaTevp=60.
> The crashes had to do with snow accumulation and could happen to  
> both LSR or to
> EVP solutions.
That's good news. It mean that we can (in principle) maskRHS flag and  
not worry about the stripes.
>
>> 3. What's the convergence criterion for LSR, and how many  
>> interations do you allow/do? In other words how close is the LSR  
>> solution to VP?
>
>       LSR_ERROR          = 2e-4,
>       SOLV_MAX_ITERS=1500
That's not very much, is it? For an accurate VP solution I would put  
LSR_ERROR = 1e-7 to 1e-13, right?
>
>> c. the same is true for the wind-ice/ocean-ice stress terms which  
>> in involve
>>  averaging perpendicular to the stripes (unless the turning angle  
>> is not
>> equal to zero, in which case there is also averaging in the other  
>> directions,
>> but you don't do that, do you?).
>
> No I use SEAICE_airTurnAngle=SEAICE_waterTurnAngle=0.
Good.
>
>> About question 3 (is it really a VP solution): Could you diagnose  
>> SIsigI and SIsigII (snapshots!!!! I guess one is enough) for all  
>> (or some) solutions and
>>  plot them (plot(SIsigII(:),SIsigI(:),'x')? These should be the  
>> principle components of sigma normalized by the strength/pressure P.
>
> With SEAICE_dumpFreq, SIGMA1, SIGMA2, and SIGMA12 are diagnosed by  
> default for
> the EVP solutions but not for LSR.  Are these the same as SIsigI  
> and SIsigII?
> Figure for SIGMA1, SIGMA2 for EVP solution is here:
> http://ecco2.jpl.nasa.gov/data1/arctic/output/tests/figs/SIGMA2232.ps
> Does it look as expected?
sigma1/2/12 are not the principle stress components. I have added  
diagnostics that are called SIsigI and SIsigII, which is what you  
want. In principle you could computed them yourself (from snapshots):
SIsigI = 0.5*(sigma1 + sqrt(sigma2^2 + 4*sigma12^2)/Press
SIsigII = 0.5*(sigma1 - sqrt(sigma2^2 + 4*sigma12^2)/Press

Press = max(1.e-13,Pstar * HEFF *exp( -20*(1-AREA)));

see seaice_do_diags.F (and seaice_dynsolver.F)

>
>> I am also a little concerned that the LSR and EVP solutions look  
>> so different
>> in the ice-covered area, can that be attributed to that different  
>> boundary
>> conditons? Can you try a run with no slip for the evp solver?
>
> Is LSR no slip by default?  How do you specify no slip for evp solver?
LSR is half slip and that's hardwired. I didn't want to bother this  
the boundary conditions if EVP works, because it's so much simpler to  
do that in EVP. But now I may have to reconsider this decision.
EVP is free slip by default. SEAICE_no_slip = .true. makes it no slip.

Martin