C grid kmtislands tests: Kimmritz et al. 2016 C1 configuration

[JFLemieux73]

./cice.setup -c kmtislands -m daley -e intel -p 1x1 -s gridcd,debug,boxsyme -g gbox80
kmt_type = 'boxislands'

I do not change the initial ice condition (aice=1). I have the same setup as at the end of issue #697. capping=0, avg_zeta, deltaminEVP=2e-9, ndte=2400, eyc=0.09. I want to check if I try as much as possible to get rid of elastic waves if I can reproduce the ref (VP) solution. I look at the sol after 1 day (hist_avg=.false.) Here is the ref sol (uvel):

[JFLemieux73]

It is very stiff...the max vel is around 0.0002 m/s. Here is the B-grid EVP sol:

[JFLemieux73]

And the C-grid EVP sol:

[JFLemieux73]

Both B and C grid EVP sol are noisy. I set Pstar=1e4 and keep ndte=2400 and eyc=0.09. Here is the ref VP sol (uvel):

The B-grid EVP sol:

and finally the C-grid EVP sol:

[JFLemieux73]

Both B and C-grid EVP sol are close to the ref sol. The problem is still very stiff with the max uvel close to 0.0013 m/s.

[JFLemieux73]

Testing this commit: apcraig@012e21b

I have the same settings as above (i.e Pstar=1e4, ndte=2400, eyc=0.09, etc.) . A visual comparison for B,C EVP uvel solutions are the same as above.

[JFLemieux73]

B and C grid results are similar.