Errors and questions implementing a USGS sediment model

sediment
,
1

Hi,
I want to study suspended and bedload sediment transport in the Yucatan peninsula. I am having some doubts to set up my realistic sediment transport case, using the USGS model. After looking at some example cases (e.g. Tidal flat2DV, ESTUARY, DUNE test cases), I am defining in cppdefs “define REGIONAL” and “define SEDIMENT”,

  1. also, is it necessary to define the below?

define ANA_GRID
define ANA_SMFLUX
define ANA_SRFLUX
define ANA_STFLUX
define ANA_SSFLUX

When defining the above, I get the following error:

image
image1920×1080 163 KB

  1. I need to modify the following analytical fields: ana_grid.F / ana_initial.F / analytical.F ?

  2. For my own realistic case I am implementing a bottom with multiple sediment classes. In that case, is it necessary to define “define MIXED_BED”?

I really appreciate your guidance on this.
César.

2

Hi Cesar,

to me it sounds like your CONFIG name key in cppdefs.h doesnt appear in the param.h
Check your cppkey here :

                /* Configuration Name */

define XXXXXX

In the other hand, Mixed bed is not mandatory , it depends of your area and the grain size diameter climatology. It could be used if you want to see some cohesive behavior (flocculation/consolidation)

I hope it helps you

Guillaume

3

Also,

ANA_GRID cppkey should be undefined , but normally if you activate the REGIONAL key , the model does it alone.

++
Guillaume

4

Hi Guillaume,
Thank for your kind reply.

The error was because I was defining “ANA_GRID”; now I understand that it should not be defined in regional configuration.

There are some things I am trying to understand. I am doing some tests with a resolution of 2 km. I know that for bed transport the resolution should be higher.
In “sediment.in” does the order of the sediments have any relevance?
When I do a test with two kinds of sediments (sand and fine sediment), using the following configuration,

2 Sd(1-NST), CSED, SRHO, WSED, ERATE, TAU_CE, TAU_CD, BED_FRAC(1:NLAY)

  0.020    100.0  2650     0.03  0.0005      0.5      0.4    0.2
  0.200    400.0  2650.    20    0.0005      0.5      0.4    0.8

3 BTHK(1:NLAY)
1.0
4 BPOR(1:NLAY)
1.0
5 Hrip
0.
6 Lrip
0.
7 bedload_coeff
1.
8 morph_fac
0.

after one week simulated I get the results below:

image
image921×808 66.5 KB

When I only consider a single sand it is as if the sediment “disappears”. Like this:

image
image1222×1033 67.4 KB

My first thought, is that the sand sedimented quickly, however the erosion and deposition flux results are zero; when I plot it (croco_gui), the resulting figure is white. Which I do not understand, it seems that I am not defining the model outputs for sediment well.

Thanks for your attention. Please let me know your opinion.

César.

5

Hi Gui,

In continuity with the message sent a few days ago. I have been working on building the physics of my sediment model in cppdefs.
My interest is to see bedload transport in the inner shelf due to currents (for the moment). That is why I have defined bedload according to Meyer-Peter Muller. In cppdefs (define REGIONAL) I have set the following:

                  /* USGS Sediment model */

ifdef SEDIMENT
define SUSPLOAD
define BEDLOAD
define ANA_SEDIMENT
undef Z0_BL
undef Z0_RIP
define BEDLOAD_WENO5
undef BEDLOAD_UP5
undef MORPHODYN
undef SPONGE_SED
ifdef BEDLOAD
define BEDLOAD_MPM
endif
endif

Is it ok like this? maybe BEDLOAD_WENO5 should go inside the “ifdef BEDLOAD”?
With this I thought I would see something different than the image I shared earlier, which only had SUSPLOAD and BEDLOAD defined, but there was no change.

I really appreciate your guidance on this issue.

Thank you very much for your attention.
Cesar.

6

Hi cesar
First the order has no relevance…
Some remarks about the namelist sediment.in :

  • the critical shear stress for erosion tau_ce seems very low in your experiment , that implies it s very easy to erode , it should be higher
  • Erosion rate E0 should be between 1e-3 et 1e-5 (as tau_ce ,look at empirical formula)
  • porosity = 1 ? maybe lower this value to 0.4 - 0.7
  • morph_fac should be 1 (that maybe explain your zero flux values)
7

No it seems good , but to be clean :

ifdef SEDIMENT
define SUSPLOAD
define BEDLOAD
define ANA_SEDIMENT
undef Z0_BL
undef Z0_RIP
undef MORPHODYN
undef SPONGE_SED
ifdef BEDLOAD
define BEDLOAD_MPM
define BEDLOAD_WENO5
undef BEDLOAD_UP5
endif
endif

With a different Morph_fac value, yea it should be different I hope

Let me know !

Guillaume

8

Hello Gui,
Thanks for your recommendations. I have been able to see few changes, but I have made progress.

Likewise, I think bed shear stress for erosion is low, however, I don’t find any difference when I increase notably this value, also I don’t notice the difference for higher values of tau_cd. It looks as if the model does not consider those parameters (yes, what I say does not make sense).

I have changed domain. I am testing with two online nested domains. The large domain has ~1500m resolution and the child ~500m. I still can’t notice differences when changing tau_ce and tau_cd, likewise, when I visualize the erosion flow the figure is blank.

About the child model. I guess it is necessary to create a “sediment.in.1” file for the sediment model in the subdomain, is that correct?
The question is because in the output variables of the child only the averages of the sediment classes (kg/m3) are saved; the rest of the variables for the sediment are not saved.

Thanks for your support. I remain attentive to any suggestion.

Regards,
Cesar.