strange behaviour of wetting-drying ...

General scientific issues regarding ROMS

Moderators: arango, robertson

Post Reply
Message
Author
turuncu
Posts: 128
Joined: Tue Feb 01, 2005 8:21 pm
Location: Istanbul Technical University (ITU)
Contact:

strange behaviour of wetting-drying ...

#1 Unread post by turuncu »

Hi,

I have a closed domain - Caspian Sea - which the minimum depth is set as 5 meter. In this case, i activated the wetting-drying scheme to have dynamic masking and i set the critical depth as 0.2 meter like most of the applications. I know that the cell assumed as dry when the following criteria valid,

zeta+h <= dcrit+eps

as you already know that eps is a such mall value. In my case some grid points goes dry but i could not see the reason when i look at the zeta. The values for a specific grid point that was set as dry,

zeta = -1.8 meter
h = 5.0 meter (equal to minimum value)


and in this case the drying criteria is not valid.

Anyway, i have only one explanation for it, in my case the output are daily average values (which is used to get the above values) but in a specific time during the day zeta could be less than -4.8 to achieve the wetting-drying criteria but i won't expect to have such a large change during the day interval. Right? I just wonder that is there anybody else had such kind of problem in the past. By the way, i also plan to decrease the output time interval from day to 3 hour to check it.

Regards,

--ufuk

nacholibre
Posts: 81
Joined: Thu Dec 07, 2006 3:14 pm
Location: USGS
Contact:

Re: strange behaviour of wetting-drying ...

#2 Unread post by nacholibre »

Looks like I have barely missed the first anniversary of this question to reply it :shock: !!! Hopefully you moved on and this answer will be helpful for a future reference if not for you. When WET_DRY is defined in ROMS the following condition is applied over the entire grid:

zeta=zeta+(Dcrit-h)*(1-mask_rho)

This might happen if you mask a grid cell that has a positive depth and display the zeta values without applying the mask afterwards. In this case a cell may have a negative value of a surprisingly much larger magnitude than the surrounding cells. But in reality it is masked and it should not have an effect on your results. So in your case if dcrit=0.2 m and you masked a cell where h=2 m, since zeta is most likely set to 0 initially, you actually see zeta=-1.8 m. It stays the same throughout the entire simulation since the cell is actually permanently dry.
Last edited by nacholibre on Tue May 13, 2014 9:55 pm, edited 1 time in total.

jcwarner
Posts: 1172
Joined: Wed Dec 31, 2003 6:16 pm
Location: USGS, USA

Re: strange behaviour of wetting-drying ...

#3 Unread post by jcwarner »

Happy anniversary!
I answered something similar to Kate a few weeks ago.
This line ensures that the water level does not go below the land. If a cell has a landmask = 0, then zeta on that cell gets set to 0 due to multiplication with the land/sea mask. But i can not let the zeta be < h. So i set the water level on land masked cells to = Dcrit-h. I suppose we could change this to:
zeta=MAX(zeta,zeta+(Dcrit-h)*(1-mask_rho))
if mask_rho=1 (ie water), then this line does nothing.
if mask_rho=0 (ie land), then we have
zeta=MAX(0,(Dcrit-h))
and that would solve your output problem. Of course, if you just multiply the output * the landsea mask, then you get your zeros back anyway.
-john

Post Reply