Flow runs to fast in ideal domain

General scientific issues regarding ROMS

Moderators: arango, robertson

Post Reply
Message
Author
phamsymt
Posts: 26
Joined: Fri May 09, 2008 3:21 pm
Location: center for environment

Flow runs to fast in ideal domain

#1 Unread post by phamsymt »

Hello everyone

I did for ideal domain size with length of 200km in y, and 200km in x direction. The resolution is 3km x 3km.

I put advection (source) at the left boundary with value of 0.4 m/s, while value in the inner domain are 0.1 m/s. Direction of current is only from left to right (u=0.4m/s at left boundry, u=0.1 m/s in other grid points, v = 0 m/s).

Temperature and salinity are set constant.

2D timestep is 20 second, and 3D timestep is 120 second.

Unit of x_rho, y_rho,...in grid file are all in "meter"

However, just after 50 number of timesteps of simulation, the flow from source at left boundary reaches right boundary. I calculated that 50 (number of timesteps) x 120 (3D timestep) = 6000 (seconds) ~ 1.67 hours. It means that just after 1.67 hours current with highest value at source of 0.4 m/s can transport 200km (Please see the distribution of velocity at initial time and after 50 number of timestep in attached file). It is so impossible.

In reality, it must take around 6 - 7 days for current with value of 0.4 m/s to transport with distance of 200km


I do not know what is happen in this matter. Could you kindly give any advice

Thank you very much


Sy

User avatar
kate
Posts: 4088
Joined: Wed Jul 02, 2003 5:29 pm
Location: CFOS/UAF, USA

Re: Flow runs to fast in ideal domain

#2 Unread post by kate »

The flow doesn't get from one side to the other based on the advection, but rather based on surface wave speeds. You didn't tell us how deep your domain is - that sets the wave speed.

phamsymt
Posts: 26
Joined: Fri May 09, 2008 3:21 pm
Location: center for environment

Re: Flow runs to fast in ideal domain

#3 Unread post by phamsymt »

Thanks for your comment.

The depth of my domain is 160 meter.

The vertical layers are 30. For simplicity, all layers have the same velocity. "no-slip condition was set.

I did not run with tide and all atmospheric forcing were set zeros.

User avatar
wilkin
Posts: 875
Joined: Mon Apr 28, 2003 5:44 pm
Location: Rutgers University
Contact:

Re: Flow runs to fast in ideal domain

#4 Unread post by wilkin »

As Kate pointed out, the fastest wave in your uniform depth ocean will be the surface gravity wave with speed sqrt(gh) which for 160 m depth is 39.6 m/s. To travel 200 km at this speed takes:

200e3/sqrt(9.81*160) = 5e3 seconds = 1.4 hours.

So that's what's going on in the start of your simulation.
John Wilkin: DMCS Rutgers University
71 Dudley Rd, New Brunswick, NJ 08901-8521, USA. ph: 609-630-0559 jwilkin@rutgers.edu

phamsymt
Posts: 26
Joined: Fri May 09, 2008 3:21 pm
Location: center for environment

Re: Flow runs to fast in ideal domain

#5 Unread post by phamsymt »

It is really nice comment. I reduced depth the surface gravity wave speed decrease a lot.

However, in case I just only want to observe how flow base on the advection transport to other boundary. Could you kindly suggest me how to do that?

Thanks

User avatar
kate
Posts: 4088
Joined: Wed Jul 02, 2003 5:29 pm
Location: CFOS/UAF, USA

Re: Flow runs to fast in ideal domain

#6 Unread post by kate »

How about a passive dye tracer? Or the FLOATS option?

phamsymt
Posts: 26
Joined: Fri May 09, 2008 3:21 pm
Location: center for environment

Re: Flow runs to fast in ideal domain

#7 Unread post by phamsymt »

I set "undef" for all "FLOAT", "T-PASSIVE" and "SEDIMENT". Also I just ran ROMS alone, not coupling with WRF or SWAN.

User avatar
kate
Posts: 4088
Joined: Wed Jul 02, 2003 5:29 pm
Location: CFOS/UAF, USA

Re: Flow runs to fast in ideal domain

#8 Unread post by kate »

FLOATS need to have a floats.in file, but are a good way to find out how fast the flow is going. You should look into it.

phamsymt
Posts: 26
Joined: Fri May 09, 2008 3:21 pm
Location: center for environment

Re: Flow runs to fast in ideal domain

#9 Unread post by phamsymt »

It should be a good comment, since wave only transports energy, not for material as float.

However, it raise me a question that, in order to convince the audience how float is transported in right way, it is had better to put figure of flow together. Although transport of float depend on several other factors, at least it can roughly prove the reliability of float transport (for example, in simplicity it can calculate: distance of float transport = velocity of current * time). However, u & v-velocity in the out-put file ( his.nc ) show value base on the speed of the surface gravity wave. Such speed increases with increasing the depth. For instance, in my case, the depth is 160m, only 1.6 hours speed of wave can reach other boundary with distance of 200km, while float just only go several km far from its original location.

Could you kindly give more advice on that

Thanks

Post Reply