Problem during simulating sediment from point source

[phamsymt]

Hi.
I simulated sand_01 transport from a point source (released only 24 hour) at the surface layer (25 layers with average depth around 70 m), sand_01 concentration of point source is 70 kg/m3, water discharge is 0.01 m3/s horizontal resolution in 6x6 km.
Some other parameters in sediment.in as follow: Sand_sd 50 (1.25 mm), sand_srho (2650 kg/m3), sand_erate (5.0d-5), sand_tau_cd (0.05 N/m2), initial sandmass (0), bed_thickness (0.1 m).

If I set small particle settling velocity (Sand_wsed=0.1 mm/s) the results seem to be ok after 30 days of simulation. However, if I increase Sand_wsed=4 or 9 mm/s) I got a big problem that the "sandmass" of bottom always increase fast even after the point source stop leasing sediment. After 10 days of simulation, the "sandmass" up to value 3000-100000 kg/m3 (this high value seems just to occur at the same grid cell to point source). Moreover, the bed_thickness also increase very high at this grid cell (up to 100m - greater than initial terrain). As a result, It makes the concentration of sand_01 continuously increase a lot. This must be unreal.

In my understand, I think that Sand_wsed=4 mm/s is not so high. The "sandmass" must decrease after stopping the sediment release. Is it right?

And could I switch off the re-suspend of sand from the bottom? if so how can I do that ? and does it effect to mass conservation?

Thank you very much

Phamsymt

[jcwarner]

The values of settling velocity that you are using are reasonable. There must be sediment coming from somewhere for the bed mass to keep increasing. what did you set for the porosity? If you set Erate=0 then there will be no resuspension.
-john

[m.hadfield]

How are you implementing your point source and how do you know it stops after 24 hours?

[phamsymt]

I am so sorry for keep you waiting so long, because I was out of the office.

I implemented the point source as a 1 river source. In order to constrain it stops after first 24 hours, I set the runoff equate 0 after first 24 hours.

In addition, to release sediment at surface I only gave the value at layer of surface ( for example, only layer number 25 (surface layer) have value 50kg/m3 (just first 24 hours), other layers have value 0 kg/m3).

I set the SAND_POROS =0.5, MUD_PROS=0.5 and bed_porosity =0.9.

The sediment in my domain was only fed from the aforementioned river source, not from the initial condition nor boundaries (sand_01, mud_01 =0). However, I still can not figure out why the bed mass kept increasing as shown in the attached pictures (please find the attached figures including bedthickness.JPG and sandmass_bottom.JPG. In those figures, the cross sign describes location of releasing sediment source, some different colorful cells around cross sign show the value of them)

I highly appreciate any comments

phamsymt

[jcwarner]

is the river coming in from a masked area? It needs to be coming in from a masked area. What was the flow rate?
50kg/m3 is rather high. how did you come up with that value? maybe it should be more like 0.5 kg/m3??

[phamsymt]

The flow rate is 0.01 m3/s. In my problem, the sediment concentration is quite high because it is released from mining activity. We may have 3 scenarios of releasing sediment from that with respect to the depth of release (at the surface (70m), middle (30m) or bottom).

The location of sediment release is very far from the land therefore, I could not set my river source from the the mask because if setting the mask, that grid cell become a small "island" or small "cylinder". As a result, it will effect to the simulated dynamic process.

So, in this case if the river source can not suitably apply for my sediment point source, could you kindly suggest me other method for doing that?

Thank you vey much

phamsymt

[jcwarner]

can you just use a TS_PSOURCE and not a UV source?

[m.hadfield]

I have never successfully set up a point source detached from the land mask using the standard ROMS facilities. (If someone can tell me how, I would love to hear it.) For point sources of this type I add code to step3d_t.F. This represents the source of sediment mass, but not the accompanying water volume. There is a report describing such an application at the link below.

http://www.epa.govt.nz/Publications/NIW ... report.pdf

Attached is the copy of step3d_t.F used for one of the simulations described in this report.

NB: Things have moved on since those simulations. They are no longer relevant to the project described.

[phamsymt]

Many thanks for your great suggestion on focusing on TS_PSOURCE & UV_PSOURCE. I did turn off LuvSrc (LuvSrc ==F) before. Now, I turn on it, the results seems to be correct. The "sandmass_01" does not increasing anymore with the range just from 1.0e-004 to 4.0e-004. However, I still can not understand why this significantly effects on the results of "sandmass_01" because even if turning off "LuvSrc" the background momentum still move the sediment away. Could you kindly explain a little bit about that?

In addition, I have the other problem that the simulated "sand_01" has the minus value in some places. I can understand why in the discussion "viewtopic.php?f=19&t=2170". Since ROMS has not controlled that, for analyzing, could I set "final_sand_01=abs(sand_01) or move away sand_01 with minus value?

Thank you very much for any comment

phamsymt