Vertical Viscosity (AKv) result is strange

Report or discuss software problems and other woes

Moderators: arango, robertson

Post Reply
Message
Author
CBian
Posts: 39
Joined: Mon Dec 28, 2009 4:01 pm
Location: Ocean University of China

Vertical Viscosity (AKv) result is strange

#1 Post by CBian » Tue Jan 25, 2011 2:16 pm

Dear Roms users,

I cannot get a good wind-driven current result near the coast, so I output the vertical viscosity (AKv) from model.

The AKv results are very strange. I check some papers, and the AKv should have a bigger value in mixed coastal area than in deep water. However, when I used MY25_MIXING, the surface and bottom AKv = 0. I got the similar results when I used GLS-MIXING. The depth-averaged AKv near coastal area is about 0.001 N/m2, which is far smaller than it should be.

Can anybody give me some suggestion about this? Thanks!
Last edited by CBian on Wed Jan 26, 2011 3:17 pm, edited 1 time in total.

LiuZHQ
Posts: 45
Joined: Mon Apr 06, 2009 5:12 pm
Location: The Hong Kong University of Sci&Tech

Re: Vertical Viscosity (AKv) result is strange

#2 Post by LiuZHQ » Wed Jan 26, 2011 3:43 am

Hi Bian,
I remind you to recheck the "frc_uvstress.cdl" file before generating the forcing netcdf files.
I noticed that there is some defination say "svstr:units = "Newton meter-2 * 1000;" in this file, which indicates that you should time the svstr with 1000 after converting from the wind speed V, or you may get the wrong forcing file.
Just a suggestion, but may not be the exact problem.

ZQ

CBian
Posts: 39
Joined: Mon Dec 28, 2009 4:01 pm
Location: Ocean University of China

Re: Vertical Viscosity (AKv) result is strange

#3 Post by CBian » Wed Jan 26, 2011 11:31 am

Dear Qiang,

Thanks for your suggestion.

I check the frc_uvstress.cdl:


short sustr(sms_time, eta_u, xi_u) ;
sustr:long_name = "surface u-momentum stress" ;
sustr:units = "Newton meter-2" ;
sustr:time = "sms_time" ;
sustr:scale_factor = 0.0005 ;

short svstr(sms_time, eta_v, xi_v) ;
svstr:long_name = "surface v-momentum stress" ;
svstr:units = "Newton meter-2 * 1000" ;
svstr:time = "sms_time" ;
svstr:scale_factor = 0.0005;



This is very very strange, why 'sustr' use different units with 'svstr'? Do you use this CDL to generate wind stress and it works well? Because I use 'svstr' with units "Newton meter-2", the result seems OK. However, in my ROMS output log file, it is:


GET_2DFLD - surface u-momentum stress, t = 360 00:00:00
(Rec=1440, Index=1, File: byecs-frc.nc)
(Tmin= 0.2500 Tmax= 360.0000)

(Min = -1.60126703E-04 Max = 1.01056013E-04)
GET_2DFLD - surface v-momentum stress, t = 360 00:00:00
(Rec=1440, Index=1, File: byecs-frc.nc)
(Tmin= 0.2500 Tmax= 360.0000)

(Min = -2.41676363E-04 Max = 4.41068446E-06)


The surface stress seems 1000 times smaller than the wind-stress in my forcing data. But the wind-induced current result seems OK. Why this happened? I am confused.

=====================================

And for my first doubt, the vertical viscosity (AKv), it seems have no much relation with wind stress. There must be some model-configuration mistake induce the very small AKv value at the well mixed coastal area. Still waiting for someone to give me some suggestions about this.
Last edited by CBian on Wed Jan 26, 2011 3:19 pm, edited 3 times in total.

LiuZHQ
Posts: 45
Joined: Mon Apr 06, 2009 5:12 pm
Location: The Hong Kong University of Sci&Tech

Re: Vertical Viscosity (AKv) result is strange

#4 Post by LiuZHQ » Wed Jan 26, 2011 12:16 pm

Bian,
plz also chang the "sustr:scale_factor = 0.0005 ;" and "svstr:scale_factor = 0.0005 ;" to "sustr:scale_factor = 1 ;" and "svstr:scale_factor = 1 ;" respectively,

"svstr:units = "Newton meter-2 * 1000" ;" to "svstr:units = "Newton meter-2" ;"
and then generate the new wind forcing files.
good luck,
ZQ

zhaoqiang
Posts: 2
Joined: Wed Feb 11, 2009 7:38 pm
Location: Ocean University of China

Re: Vertical Viscosity (AKv) result is strange

#5 Post by zhaoqiang » Thu Jan 27, 2011 5:38 am

I did not read the codes of this part, but I think sustr is divided by the density of water in the model, so you got the small values in your log file. If you do not use CDL file to generate your nc file, you can ignore it.
The value of AKv you got is very small. It depends on the stability of the water. And you need not confuse the value of AKv at the surface or bottom, cause what we really care about is the turbulence in interior region.

CBian
Posts: 39
Joined: Mon Dec 28, 2009 4:01 pm
Location: Ocean University of China

Re: Vertical Viscosity (AKv) result is strange

#6 Post by CBian » Mon Jan 31, 2011 2:00 pm

Dear ZhaoQiang,

Thanks for your suggestion, it's really helpful.

I made a stupid mistake, the vertical viscosity near coast is about 0.01 N/m2 not 0.001 N/m2. This doesn't seem too small.

In my opinion, when there is a very strong northward wind (eg. 10 m/s), the wind-induced current near the coast(depth < 20m) should be also northward. However, in my model results, the wind-induced surface current is northward, but the bottom current is southward. The water is well-mixed. So I wonder the vertical viscosity (0.01 N/m2) is still very small? Or there are other reasons lead to this strange result?

Post Reply