bottom friction coefficient

Discussion of how to use ROMS on different regional and basin scale applications.

Moderators: arango, robertson

Post Reply
Message
Author
azhang
Posts: 20
Joined: Fri May 11, 2007 12:49 pm
Location: CSDL/NOS/NOAA

bottom friction coefficient

#1 Post by azhang » Fri Jun 29, 2007 4:02 pm

I am applying ROMS3.0 (minimum water depth 10cm with wet-dry option) in Tampa Bay tidal simulation. It ran well for 30 day simulation, but the elevation amplitude is too larger than the tidal prediction in the upper bay. Therefore, I tried to increase bottom friction by changing RDRG2 since I use quadratic bottom friction option. The original RDRG2=0.003 is fine, but I increased RDRG2=0.005, then ROMS blew up after a few days run. I tried LOGDRAG as well with Z0B=0.02, ROMS blew up after a few hour simulation. I wonder why bottom friction coefficient is so sensitive to ROMS, and what value of z0b I should use for LOGDAG option. Can someone give me some advice on how to reduce tidal range in shallow waters for ROMS application, in other words, which parameters should I adjust in ROMS *.in control file.
Thanks very much in advance

AJ

jcwarner
Posts: 812
Joined: Wed Dec 31, 2003 6:16 pm
Location: USGS, USA

#2 Post by jcwarner » Fri Jun 29, 2007 5:16 pm

for tidal applications I typically use
#define UV_LOGDRAG
I often work in the ranges of
zob=0.002, dt ~ 60 sec, ndtfast ~20 to 30

-j

azhang
Posts: 20
Joined: Fri May 11, 2007 12:49 pm
Location: CSDL/NOS/NOAA

#3 Post by azhang » Fri Jun 29, 2007 7:21 pm

Hi, John,
Thanks a lot. I tried LOGDRAG option with z0b=0.02 and 0.002, The model blew up at almost same time (about 3 hour simulation). By the way, I use dt=30s, and Ndtfast=30. Will it be helpful to increase vertical layer (currently I use 11 vertical layers) or reduce dt?

AJ

jcwarner
Posts: 812
Joined: Wed Dec 31, 2003 6:16 pm
Location: USGS, USA

#4 Post by jcwarner » Fri Jun 29, 2007 7:39 pm

when the model "blows up" , and I want to really find out why, then i follow a procedure to
1) run model til it blows up
2) restart model, let it run til it blows again
3) repeat step 2, until it blows up in only a few steps.
4) then save every time step and see exactly where/why it is happening.
Is it blowing up at a wet/dry interface, at the open boundary, at some weird grid location, ???
it is difficult to tell these things until you see why and where the instability is developing.

-j

Post Reply