Estuary_Test

[mamckeon]

Hello,

I'm a beginner ROMS user and I've been trying to get the estuary_test case going, but I haven't been able to get it to jive either with the wikiROMS page for estuary_test or the Warner et al. 2005 ocean modeling paper. It seems to be working well enough except for a weird overall freshness in the ocean layer that is exacerbated at the western boundary, and seems to be causing a third turbidity maximum. Here's what an average of the last two tidal cycles (of ten total tidal cycles) looks like:

The header file is essentially the estuary.h provided with the ROMS code:

#undef ANA_GRID
#define UV_ADV
#define UV_LOGDRAG
#define TS_U3HADVECTION
#define SALINITY
#define SOLVE3D
#define SPLINES
#define SEDIMENT
#ifdef SEDIMENT
# define SUSPLOAD
#endif
#define AVERAGES
#define NORTHERN_WALL
#define SOUTHERN_WALL
#define EAST_FSGRADIENT
#define EAST_M2CLAMPED
#define EAST_M3GRADIENT
#define EAST_TCLAMPED
#define WEST_FSCHAPMAN
#define WEST_M2CLAMPED
#undef WEST_M2REDUCED
#undef FSOBC_REDUCED
#define WEST_M3GRADIENT
#define WEST_TRADIATION
#define WEST_TNUDGING
#define GLS_MIXING
#undef MY25_MIXING
#if defined GLS_MIXING || defined MY25_MIXING
# define KANTHA_CLAYSON
# undef CANUTO_A
# define N2S2_HORAVG
#endif
#define ANA_INITIAL
#define ANA_SEDIMENT
#define ANA_SMFLUX
#define ANA_STFLUX
#define ANA_BTFLUX
#define ANA_SSFLUX
#define ANA_BSFLUX
#define ANA_SPFLUX
#define ANA_BPFLUX
#define ANA_FSOBC
#define ANA_M2OBC
#define ANA_TOBC

As far as I can tell, ana_tobc is setting the western boundary salinity to 30. Any ideas what I'm doing wrong here?

Thanks!
-Maggie

[jcwarner]

Maggie-
i did an svn update on the rutgers code and ran estuary_test. It ran for 10 days.
here is what i did in matlab:
ncload ocean_his.nc
saltmean=salt(end-47:end,:,3,:);
figure
pcolorjw(squeeze(mean(saltmean,1)))
colorbar

i do not get the same picture of avg salt that you show. I am not sure why your salt at the western obc shows a decrease.
For the sediment, in the paper i ran the est case for 10 tidal cycles, then i restarted the model and ran it for days 10-20 with the sediment. this allowed the est circ to establish itself beofre the sed was applied. Can you try that and see if the double sed peaks go away? (i do get teh double sed peaks, but i would not use them. they are distorted by the initial conditions.) let the salt get correct first, then go with the sediment.

-john

[mamckeon]

Hi,

Thank you for checking back! I'd done something weird setting up the test case, but it is working now. Thanks for the tip about starting the sediment after the estuary circulation has been set up.

Best,
Maggie

[luhputria]

Hi all,

I am already able to compile ROMS estuary case and right now try to plot using Matlab.
But I found the difference between my result and wikiroms result (https://www.myroms.org/wiki/ESTUARY_TEST_CASE). I try to compare my plotting result with a figure in this website.

1. For example in MY25 day 9.9 for salinity figure:
From wikiroms, we got a picture as attched (Estuary_4panel_salt.gif). While my plotting is shown in the picture (11 day 9.9167.jpg)
My question is: How can the example result in wikiroms has salinity in day 9.9375? Because time ended in 241 time step or 10 days (like my example). In step 239, the day is 9.9167 and step 240 is 9.958. So, I'm wondering, how can the example get salinity contour in 9.9375. Even though I am plotting for day 9.9167, the contour of salinity is still different between mine and example result.

2. In sediment case
The result from the example in wikiroms is shown in figure Estuary_4panel_sed.gif, while mine is shown in figure sediment day 10.jpg.
Besides the difference in day (19.9375 in wikiroms example and day 10 in my result), the value of suspended sediment concentration is also different. Suspended sediment concentration in example result is 20-80 while mine is 0.01-0.045. I compiled existing file in estuary test example and similar with the value stated on the website. But the result is different.

Anyone here has the same issue like me? I am a new user and want to know more about ROMS.

Thanks before.
Regards,
Luh

[stevenmiguelfigueroa]

Hey Luh,

Nice results and analysis.

I think the test cases do not always match what you expect right away. Sometimes you need to go into the settings to check/tinker with them.

To control the time stepping and frequency of saving to a history file, you can modify NTIMES, DT, and NHIS in the .in file.

A result I got using Estuary_test was SSC in the range of 0-0.1 kg m^-3 (attached). One reason for the difference could be that in Warner et al. (2005), the units of SSC seem to have been mg L^-1 based on their Fig. 2.

Warner, J.C., Sherwood, C.R., Arango, H.G. and Signell, R.P., 2005. Performance of four turbulence closure models implemented using a generic length scale method. Ocean Modelling, 8(1-2), pp.81-113.

[luhputria]

Hi Steven,

Thanks for your information. I really appreciate it.

Anw, when I modified the estuary test case (domain is same, change into flat seabed, without river flow, constant flow from western edge) and tried to change the lateral boundary condition in sediment input file, I found something interesting.
When I used clamped and different boundary condition (i.e. "Cla in west, radnud in east" or "cla in west, gra in east"), with initial bed thickness 0.1m (similar with estuary test case), after 50days of simulation, almost all the sediment along 30km from western edge is washed away as shown in picture 1.clarad15.jpg. IMHO, this result is not appropriate if I want to see the longterm morphology changing in the seabed, because the flow is not coming with the sediment that settled outside the domain. But this option is really good if we investigate the vertical profiles of sediment concentration without the intervention sediment from outside of the domain.
I got more reasonable result when I choose the boundary condition which is gradient or periodic in western and eastern edge as shown in picture gragra50con.jpg and per15.jpg.
And based on this both graphs, I assume that to model the incoming flow with the sediment, we can choose gradient and periodic lateral boundary condition. And gradient boundary condition is more suitable for flat seabed, while periodic if the seabed is uneven.

If anyone has different idea about my opinion, or if my assumption is incorrect, please don't hesitate to share or correct me.
I tried to learn the physical meaning of lateral boundary condition using this documents: http://www.oceanwave.jp/research/ROMS/roms_obc.pdf and Marchesiello, P., J. C. McWilliams, A. F. Shchepetkin, 2001: Open boundary conditions for long-term integration of regional ocean models, Ocean Modelling, 3, 1-20. This paper discussed the lateral boundary condition in roms input file. I'm wondering whether the same meaning is also working for sediment input file.

Thank you.

Regards,
Luh