river plume problem with WET_DRY

[zhaobiao]

Hi all:
I am running ROMS ICE model. the model run smoothly without tide for 20 years. But after added tidal forcing to open boudary. it reports toubles in ice frazil. so I follow Kate's advices. I turn on #Define WET_DRY and make a test without tide. Fortunately, it work well and never give ice frazil problem. after integrated 8 months. I check the Sea Surface Salinity and found that river plumes is abnormal compared with undefined WET_DRY experiment. the river plumes simulated by experiment with WET_DRY extend enormously than experiment without . see picture below



the vertical integrated V component of experiment with WET_DRY in the estuary can upto 6 m/s. While V component of experiment without WET_DRY is less than 1 m/s. the V componet of experiment with WET_DRY is approximate one order of magnitude greater than experiment without WET_DRY



By the way,I never run ROMS with WET_DRY.the grid file used in WET_DRY experiment is the same as experiment without WET_DRY. th minimum depth in 5m in two experiments
has anyone of you met this kind of problem? i would be nice of you if give some useful guide
thanks in advance

[kate]

Oh no, that looks terrible. Are any of those points ever actually dry?

How recently did you fetch the ice code? The version there now (as of quite recently) is up to date with Hernan's changes, but also includes WET_DRY fixes from the COAWST branch.

[kate]

I'm actually having other problems with WET_DRY. This is still an area of active code changes.

[zhaobiao]

Thanks! Kate. According to your suggestion. I updated all of my cods yestoday. It seems that the model is back to normal now. Although i still don't know what is the problem. .
BTW, I have checked your ana_sponge.h
file. I think the variable visc2_r and visc2_p should be replaced by MIXING(ng)%visc2_r and MIXING(ng)%visc2_p. Maybe i am wrong. at least it is work for my case.

Best wishes!

[zhaobiao]

I'm actually having other problems with WET_DRY. This is still an area of active code changes.

In addition, the subroutine mod_clma.F is different from the old one. if i want to turn on AICLIMATOLOGY, I must turn on AD_SENSITIVITY or IS4DVAR_SENSITIVITY. So i have to remove these lines of code.

[kate]

Thanks, I'll check it out.