[MITgcm-support] Problem: Extreme Tidal Currents in the Simulation at Narrow Channels.

Thu May 23 02:19:02 EDT 2024

Hi,

I am beginning to understand your problem. As far as I can see, it’s a setup and tuning issue.
You are comparing to observations (I guess) and find that your tidal velocities are too high in a narrow and shallow channel. I don’t know what the real coastlines look like (are they “porous” with lots of small inlets etc?) and how well the topography is represented by your 100m grid. I think the model just does, what it is supposed to do: there’s a volume flux of sum(UVEL*hFacW*drF*dyG), note that in your case because of the non-linear free surface both UVEL and hFacW vary with time, that the model needs to conserve going from one grid cell to the next. If there’s no dissipation of UVEL (and no other surface forcing), then UVEL will increase as hFacW*drF*dyG decrease, i.e the “green” velocities push through the narrowing channel and become “red”.

I don’t know what coastal modellers do in this case (and the MITgcm is not a particular coast model code), but I guess one needs to introduce turbulent friction and an sea bed friction locally. What do you do for vertical mixing? (I’d use the TKE scheme pkg/ggl90)
I would try to (and you have done some of this already)
- add bottom friction and lateral friction:  no_slip_sides=.TRUE., no_slip_bottom=.FALSE., bottomDragQuadratic = 0.001, (or higher), or instead use zRoughBot=0.01,  which may give more accurate drag with varying topography
-  increase “viscAhGridMax” to 0.5 (1 is the maximum for numerical stability)
- use scaled viscosity (viscAhGrid=0.01 or larger instead of viscAh)
- use biharmonic viscosity (viscA4Grid = 0.01)
- make the channel deeper and wider, because you probably cannot take into account bottom and lateral friction of irregular coastlines and topography in an appropriate way

Martin

> On 23. May 2024, at 07:47, Killampalli Rajesh <rajesh70530 at gmail.com> wrote:
> 
> 
> Hi Martin 
>  
> Here, I've included the depths and model results where I'm experiencing high velocities. As you mentioned, I've checked the volume flux that has already been observed, but not at the open boundary; instead, I've checked the estuary, where I experience higher velocities for an entire tidal cycle, with the volume flux remaining constant during the ebb and flood phases. Since the velocities are caused by topography and geometric factors, adjusting the tidal height and velocity in this area is more difficult, yet it is still tried. I lowered the input velocities by the time tidal height are matched with observation in order to correct amplitudes, however the velocities are too much. 
>  
> Kind regards
> K Rajesh
> 
> 
> On Wed, May 15, 2024 at 7:51 PM Killampalli Rajesh <rajesh70530 at gmail.com <mailto:rajesh70530 at gmail.com>> wrote:
>> Dear Martin 
>> 
>> Thank you for your response. I repent for having missed it. My model domain is the Coringa Estuary on India's east coast. where several creeks/channels meet to the estuary is my model area. I'm making a high-resolution tide model with a horizontal resolution of 100m for this domain. Include name list and model code files. Please see links 1 and 2. In link 2 I followed the method to preprocess the baroclinic tidal current where Data from TPXO7.2 was utilized to produce boundary conditions (OBEam, OBEph) amplitude and phase. 
>> 
>> http://wwwcvs.mitgcm.org/viewvc/MITgcm/MITgcm_contrib/tides/?diff_format=s&sortby=log&pathrev=HEAD
>> https://github.com/RajeshKillampalli/MITgcm_tidesetup_02/tree/main
>> 
>> On Tue, May 14, 2024 at 7:02 PM Killampalli Rajesh <rajesh70530 at gmail.com <mailto:rajesh70530 at gmail.com>> wrote:
>>> Dear MITgcm community
>>> 
>>> I have added additional points to the earlier post "Extreme tidal currents in the simulation at narrow channels".Model mesh resolution is 100 metres. I'm attempting to resolve this issue by adjusting the bottomdragliner coefficient, which was previously set to 0.002. When it increases to 0.006, the model blows up and writes NaNs. For 0.005, the model may run, but the velocities are not reduced as expected, and less variance is recognised. Currently, the maximum channel flow was 1.25m/s; after modifying to 0.005, it was 1.21m/s redused. and I experimented with ViscAh from 10 to 100, but no substantial difference was seen. Can someone advise how to solve this?          
>>> 
>>> Best Regards
>>> Killampalli Rajesh
>>> 
>>> On Wed, May 8, 2024 at 6:56 PM Killampalli Rajesh <rajesh70530 at gmail.com <mailto:rajesh70530 at gmail.com>> wrote:
>>>> Dear MITgcm community 
>>>> 
>>>> Thank you for your assistance; I'm modelling tidal circulation in the lagoon, which is connected to the coastal seas.The research region includes lagoon and coastal waterways. The boundary conditions include baroclinic tidal currents at open boundaries in data.obcs, OBEam, and OBEph only on the eastern border, and OBCS_Tides specified in the model. The model results is verified using tidal heights with the tide gauge data, however the tidal currents are exaggerated. In example, the model findings suggest extremely strong tidal currents at the mouth/channel, which contradicts reality.
>>>> 
>>>> Best Regards
>>>> Killampalli Rajesh
>>>> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mitgcm.org/pipermail/mitgcm-support/attachments/20240523/1e33375c/attachment.html>