[MITgcm-devel] seaice advection scheme and negative SIarea/SIheff

[An T Nguyen]

Hello sea ice gurus,

I'm running a simple [nx,ny,nz] of [16x1x23] domain to test the various 
sea ice advection schemes.  Patrick has helped me by introducing a flag 
to turn off sea ice thermodynamics.  My ice initial condition is 
heff=0.005m, area = 1 at x-location 2.  I used three SEAICEadvScheme of 
2,30,77.  All schemes conserve heff and area, such that the sum(area) = 
1 and sum(heff) = 0.005 for all time steps.  However,  schemes 2 and 33 
have negative SIarea/SIheff at all time steps.  I just grab the last 
time-step (1month) numbers here:

Scheme 77 (DIFF1=0):
SIarea*100:
0.02 0.09 0.37 1.37 4.92 13.2 19.8 20.1 19.1 14.7 6.13 0.22 0.0006 
0.0006 0.001 0.005
SIheff*1000:
0.001 0.005 0.02 0.07 0.25 0.66 1.0 1.0 0.95 0.74 0.31 0.01 0.00 0.00 
0.0001 0.0003

Scheme 30 (DIFF1=0):
SIarea*100:
-0.1 -2.42 -4.01 -1.80 5.36 15.1 22.84 25.0 21.2 13.9 6.49 1.11 -1.46 
-1.57 -0.3 0.6
SIheff * 1000:
-0 -0.1 -0.2 -0.1 0.3 0.8 1.1 1.2 1.1 0.7 0.3 0.1 -0.1 -0.1 -0 0

Scheme 2 (DIFF1=0.002):
SIarea*100
-6.81 -4.83 1.11 8.56 14.9 18.6 19.22 16.84 12.37 7.67 4.75 4.2 4.53 
3.43 -0.08 -4.49
SIheff*1000
-0.34 -0.24 0.55 0.43 0.75 0.93 0.96 0.84 0.62 0.38 0.24 0.21 0.23 0.17 
-.004 -.224

There's no thermodynamics so there shouldn't be any ice 
melting/growing.  Does it make sense to have negative SIarea/SIheff?

In case you have access on ross, the data.seaice & output files are at:
ross:/net/ross/raid1/atnguyen/MITgcm/verification/ice_dynamics_test/run_thermodyn_off/ocean5/
ice_scheme2/STDOUT.0000
ice_scheme30/STDOUT.0000
ice_scheme77/STDOUT.0000

An