Ocean Modeling Discussion

zhangtianhao

Hi there,

I wonder that why vertical mixing could be parameterized by k-ε model or MY2.5 model. However, horizontal mixing is just computed by setting an eddy viscosity in ocean.in file.

Is ocean modeling insensitive to horizontal viscosity？

Thanks, Hoty

zhangtianhao

For the incoming 2-D forcing, it depends. ROMS will check the dimensions of each field you provide. If they match that of your ROMS grid, it assumes Uwind and Vwind are aligned in the directions of the grid. If not, it assumes it needs to do the interpolation and the rotation to the ROMS grid ...

zhangtianhao

Hi there,
I have a question which is maybe easy for you.
I wonder that Uwind and Vwind in forcing nc file refers to eastward and northward direction in reality or eta and xi direction in model domain.

Thanks!
Hoty

zhangtianhao

kate wrote: ↑Mon Mar 08, 2021 7:24 pm Hz is the height of the vertical grid box, which you could call Hz=∂z/∂σ.
For consistency between the baroclinic and barotropic solutions, you also need D=ΣHz.

Thanks, Kate!

I misunderstood the defination of Hz and now I get it. We can say that Hz is layer thickness in a physical sense.

Hoty

zhangtianhao

In website https://www.myroms.org/wiki/Vertical_S-coordinate, as the picture shown, in red circle 1, it says Hz=∂z/∂σ or Hz=Δz/Δσ. That means Hz is the slope of z-σ curve. However, in red circle 2, it says that the sum of Hz is total water depth D. As for me, it means that Hz is layer thickness (the ...

zhangtianhao

kate wrote: ↑Thu Dec 10, 2020 5:08 pm It is the vertical viscosity used by the model, the sum of all the bits that go into it. The details depend on which vertical mixing scheme you are using.

Kate,

Thanks, I will take a try.

Cheers,

Hoty

zhangtianhao

I believe it is called Akv in the ROMS output. Kate, Thanks for your reply. But I remember that the defination of Akv [Input] is vertical viscosity (ν in the picture below) but not vertical eddy viscosity (Km in the picture below) . I wonder that if Akv [Output] is the sum of background viscosity ...

zhangtianhao

Hi all,

I plan to calculate the value of vertical eddy viscosity —— Km. I wonder if ROMS could output this variable, or I have to calculate it by myself?

Thanks.

Hoty

zhangtianhao

Hi everyone, I have a quesiton: when I activate BULK_FLUX in .h file，ROMS require variables such as Uwind, Vwind, swrad, lwrad, rain..... However, I only have sustr and svstr data (not Uwind and Vwind data). And I still want to input swrad, lwrad, rain, Qair... in model. I woder how to achieve this ( ...

zhangtianhao

Thanks, Kate and Wilkin, To sum up for other users, ROMS can calculate uvbar by uv values in ini_fields.F (if not define PERFACT_START). So it is not necessary to input uvbar in init file. However, Fla, Shc, (and Red?) boundary condition requir for uvbar, so we have to calculate uvbar by uv ...

zhangtianhao

There is a different approach to imposing an effective horizontal acceleration that is uniform throughout the vertical water column. This is to introduce a horizontal gradient in the air pressure (e.g. within ana_pair.h) and impose this force via #define ATM_PRESS. I use this method to impose a ...

zhangtianhao

Hi forum, I met some problems when dealing with boundary conditions: LBC(isFsur) == Clo Che Clo Che ! free-surface LBC(isUbar) == Clo Shc Clo Shc ! 2D U-momentum LBC(isVbar) == Clo Shc Clo Shc ! 2D V-momentum LBC(isUvel) == Clo RadNud Clo RadNud ! 3D U-momentum LBC(isVvel) == Clo RadNud Clo RadNud ...

zhangtianhao

Hi there, When using BODYFORCE, we can apply the surface stress to the whole water column. And BODYFORCE requirs sustr and svstr which can be set in smflux.F. However, as for me, I only want to set a varying horizontal acceration in the whole water column, so I don't have sustr/svstr data but I ...

zhangtianhao

I see your point. Then you might have to code some workaround, like ... # if !defined BODYFORCE ! the original code ... DO i=Istr,Iend cff=om_v(i,j)*on_v(i,j) cff1= svstr(i,j)*cff cff2=-bvstr(i,j)*cff rvfrc(i,j)=rvfrc(i,j)+cff1+cff2 ... #else ! the code you want, whatever that is ... But before you ...

zhangtianhao

cff1= 0.0_r8 cff2= 0.0_r8 rufrc(i,j)=rufrc(i,j)+cff1+cff2 I can't find that code anywhere in my version of rhs3d.F. Sorry, I don't make myself clear, Firstly , it seems that the miximun value of levbfrc is 1 but not 0, Secondly , Line 2170 in rhs3d.F: # ifndef BODYFORCE DO i=Istr,Iend cff=om_v(i,j ...

zhangtianhao

(But I modify sbstr code, I don't want to transform it to bodyforce, it is still used in 2d equation), as shown below: All you need to do to prevent the bottom stress also being applied as a bodyforce is make LEVBFRC = 0 in roms.in ...

zhangtianhao

Thanks for all of your advice, I use BODYFORCE now, and I correct the τ unit. But ROMS still blowed up. In smflx.h , I set sustr=a*h , (a means acceleration, h means depth, unit=m/s2*m), so the sustr has kinetic unit, as shown below: open(55,file='/ubuntu/COAWST/Projects/acceleration/EW01.txt' ...

zhangtianhao

Where ROMS expresses anything in kinematic units (per unit mass), such as converting stress to friction velocity (tau/rho0), or heat flux (W/m2) to degrees meters/s (Hf/(Cp*rho0) in the surface temperature boundary condition, it always uses the constant parameter rho0. Likewise, to calculate Brunt ...

zhangtianhao

Are you trying to change ROMS equation of state? The definition of in situ density anomaly or potential density anomaly is that a 1000 kg m-3 value is subtracted to the density. The in situ density anomaly, as a field, is only used in ROMS in the pressure gradient algorithm where its value is not ...

zhangtianhao

Hi all,

I found that when I set a (rho+1000)*acceleration term in the equation, ROMS showed rho speed trouble, I wonder that how to check this problem.

Thanks

zhangtianhao

The variable "rho" in the model is insitu density anomaly. You add rho0 to it to get full density. Note that rho0 may or may not be 1000. Thanks, Kate, rho0 equals 1025 in my case, but I find in rho_eos.F file: den(i,k)=den1(i,k)*bulk(i,k)*cff den(i,k)=den(i,k)-1000.0_r8 It seems that in this code ...

zhangtianhao

Hi there, I have a problem and maybe it's easy to you: I want to modify the code and use the density of seawater in fourmula, should I use rho(:,:,:)+1000 or just rho? I wonder that if the rho is the density anomaly :D I don't find the defination of in forum. BTW, I want to use the density to ...

zhangtianhao

OK, I think I get it: So the steps would be (following what you said): Using GridBuilder: Add enough depth to your bathymetry so that everything that you expect to get wet is under water and load it into GridBuilder Use the new 0 contour as your coast line (let GridBuilder use topography to set ...

zhangtianhao

rx0 can become very large if the depth changes sign within the unmasked part of the grid - I think the topography might not match up quite with your coastline after smoothing and raising it, This may have caused land to be exposed near the coast in unmasked cells which is not good and may be ...

zhangtianhao

Metrics information for Grid 01: =============================== Minimum X-grid spacing, DXmin = 1.00000000E+17 km Maximum X-grid spacing, DXmax = Infinity km Minimum Y-grid spacing, DYmin = 1.00000000E+17 km Maximum Y-grid spacing, DYmax = Infinity km Minimum Z-grid spacing, DZmin = -5.47066471E+03 ...

zhangtianhao

clear; modis3='C:\Users\xzz\Desktop\ROMS.nc'; ncdisp(modis3); ncid3=netcdf.open(modis3); %1) Enter name of netcdf grid file to be created. grid_file='ROMS_new.nc'; %2) Obtain grid information. LP = 178; MP = 281; L = LP-1; M = MP-1; xi_psi = L; xi_rho = LP; xi_u = L; xi_v = LP; eta_psi = M; eta_rho ...

zhangtianhao

I think the problem is the bathymetric filtering. The positive and negative adjustment filters only work on wet cells and rx0 is only calculated on wet cells - if you do filtering first then alter the mask it may tag previously dry cells as wet - the topography of these previously dry cells has not ...

zhangtianhao

I think the problem is the bathymetric filtering. The positive and negative adjustment filters only work on wet cells and rx0 is only calculated on wet cells - if you do filtering first then alter the mask it may tag previously dry cells as wet - the topography of these previously dry cells has not ...

zhangtianhao

Are you using GSHHG coastlines to generate the mask? Under the Mask Settings menu choose "Use GSHHG Coastlines (More Accurate)" - default is "Use Topography (Faster)" which is very useful for quickly placing a grid and getting the resolution correct. Once you are happy with the basic grid I suggest ...

zhangtianhao

I think an old version of GridBuilder had an issue with resetting depths with land masks to 0m when exporting to a grid nc file regardless of users intent but that was fixed a while ago; you can set the min depth to anything on export. I am working with version 1.2 of GridBuilder and I am able to ...

zhangtianhao

Maybe you should change tke boundary to gra

Hoty

2020-01-17

zhangtianhao

Thanks for suggestions from Kate, johnluick and jcwarner, that's really useful. Firstly, I read this papaer https://darchive.mblwhoilibrary.org/bitstream/handle/1912/6207/1-s2.0-S0098300413001362-main.pdf?sequence=1. I found I misunderstood some parameters. I confused land/sea mask with wer/dry mask ...

zhangtianhao

-changing hmin to zero is basically removing all the 'land' parts of your grid, so that removed your wet dry problem. - an hmin of -100 means that you have a small hill about 100 m above sea level. that might be troubles for roms. so maybe you could set landmask=0 for cells that are above, lets say ...

zhangtianhao

What I do these 2 days: (1)My COAWST version does not need to define limit_bstress. (2)I found that many people's model blowed up because of large T,S and velocities. So I checked T and S , they seemed also fine before blowing up, the same as uvw. (3)My model blowed up with kinetic_enrg, poten_enrg ...

zhangtianhao

Hi there, I met a problem with WET_DRY option . When I use WET_DRY to run a simple coastal ocean case, ROMS blows up quickly (19 steps, 1.9s). I set DCRIT=0.1m. Notice: I use Gridbuilder and set h of land area as negative value (though people usually set all h>0), for example,-15m. And rx0 and rx1 ...

zhangtianhao

With WET_DRY, zeta in the land area is always some DCRIT above the land, nothing to worry about. It should not be changing in time. Have you tried fussing with your timestep? I changed dt from 0.1 to 0.05, and it delays the blow-up time from 1.9s to 44s . I am keeping trying. Here is the complete ...

zhangtianhao

Hi all, Now I can run this case but it blows up after several steps. And it's weird that: usually results become large and model blows up, but this time results suddenly become NaN . So I checked my forcing file, but there are no NaN in it. One thing I noticed: I turn on WET_DRY, and I set depth of ...

zhangtianhao

The Flather condition requires values from outside the domain for both ubar/vbar and zeta. You can provide them via ANA_xxx options or you can specify a file: NBCFILES == 1 ! number of boundary files BRYNAME == /center1/AKWATERS/kshedstrom/Arctic/Files/Arctic4_bdry_1980_SODA3.3.1.nc | /center1 ...

zhangtianhao

Attached is my bry condition

zhangtianhao

So you think you have already remapped the thing to your grid. You have to have the dimensions *exactly* right or else ROMS will decide to remap for you. Some words on grid sizes. If you want powers of two or three in the grid for your decompositions without needing leftovers, you need Lm, Mm to ...

zhangtianhao

So you think you have already remapped the thing to your grid. You have to have the dimensions *exactly* right or else ROMS will decide to remap for you. Some words on grid sizes. If you want powers of two or three in the grid for your decompositions without needing leftovers, you need Lm, Mm to ...

zhangtianhao

at /home/ubuntu/app/COAWST_V3.2/Build/get_data.f90:104 I suspect you are reading an atmospheric field here? The model wouldn't try to interpolate from your grid file, assuming all those fields are already on your grid. So, what atmospheric fields are you reading and are you expecting ROMS to do the ...

zhangtianhao

Hi there, I met some problems when I ran a simple case:Index '282' of dimension 2 of array 'finp' above upper bound of 281. I checked interpolate.f90 and considered ‘finp’ may be related to grid file. Then I check my grid file, but the dimensions of the whole varibales are 281, no 282 appeared(my dim ...

zhangtianhao

kate wrote:If any part of a boundary is open, you have to set the whole boundary to open. The land masking will take care of the rest. Your north can be closed. If you set it to open, no harm done, but you'll be asking ROMS to do extra work of reading fields (or computing analytic fields).

Thanks, Kate.

zhangtianhao

Yes, ROMS is expecting ubar, vbar in the initial and boundary files. It will ensure consistency between baroclinic and barotropic flows during initialization and after every timestep, but the boundary files ought to be generated in such a way that the u and ubar fields are consistent. If you are ...

zhangtianhao

Hi everyone, I have some confusion about my boundary condition. I use Cha for zeta boundary, Fla for 2D-momentum boundary and RADNUD for 3D-momentum boundary.(My region is in South China Sea) 1)I wonder that if I must input the ubar and vbar in the boundary ncfile or in the init ncfile ? 2)And if I ...

zhangtianhao

THETA_S == 8.0d0 ! surface stretching parameter THETA_B == 0.5d0 ! bottom stretching parameter TCLINE == 20.0d0 ! critical depth (m) use something like Gout=get_roms_grid('ocean_his.nc',tidx); this will give you a field Gout.s_rho that you can avg to get the vertical locations of your grid. how did ...

zhangtianhao

I think the result should be like this:
( I draw this picture by the analytical formula)

zhangtianhao

Init file : all values = 0 Forcing file : Uwind = 8m/s Here is my in.file: !============================================================================== ! ! Application title. TITLE = POOL ! C-preprocessing Flag. MyAppCPP = POOL ! Input variable information file name. This file needs to be ...

zhangtianhao

I set a closed boundary pool(50m*50m), △x=1m, △t=0.2s. There is just one forcing condition, a constant Uwind = 8m/s. But my result looks so strange. (I think it should be a smooth parabola) Here is my result.png. It looks unnatural. The x-coordinate is u(m/s) ,and the y-coordinate is s_rho. I just wa ...

Search found 50 matches