Coupling DeltaRCM with CEM

[elbeejay]

Have done some preliminary work to couple the CEM model with DeltaRCM through the use of the pymt framework. Mariela Perignon converted the original DeltaRCM model to Python, making it WMT/BMI compliant in the process. Using the BMI variables, this model coupling is achieved with relative ease.

Currently however, the coastline reshaping by the CEM model is not working as we expected. When the models are coupled from initialization in an empty basin, the CEM forces wash away any deposition generated by DeltaRCM. When the CEM model is run on a pre-developed DeltaRCM topography we are seeing 'spotty' reworking along the channels that does not seem realistic. An example of this behavior is attached.

Any suggestions, tips, or pieces of advice are greatly appreciated. As far as I know, this model coupling has not been done before, and it could be a nice combination of reduced-complexity models. If there is someone else I should be contacting about this please let me know.

CouplingDemo.pdf

Jay

[iovereem]

hi Jay,

Catching up....
Thank you for sharing these results, and for pioneering this combination of reduced complexity models.

One first thought I had when looking at your figures, is that CEM may recognize too many interior distributary channels as 'coastline'? Because they would have a depth below sea level and a land-sea boundary?

I have assigned Eric Hutton to this issue too, because he has worked with the CEM model and coupled it to a river model. He used it with distinct river mouths, which may be another issue?

[iovereem]

Andrew Ashton - CEM developer wrote:

My first guess is that CEM isn't evolving much at all (particularly in
one time step) but rather "fixing" an input grid that is quite different
than what the model is expecting. Probably could be tested by changing
the wave height to something small (0m) or really small (0.00001m)

To be honest I have no idea how the model grids are being passed to one
another or how CEM is interpreting a bathy grid into shorelines.

I think most of Katherine Ratliff's work was focused on having
topography on the floodplain. Dealing wish subaqueous bathy is, well,
confusing to CEM.

This seems like more than a coupling problem as right now CEM assumes
that there is a shoreface geometry.

What is the scientific question that you are hoping to answer?

[elbeejay]

Thank you both for getting back to me. It sounds like this model coupling may be more involved than I initially thought. I am going to try infilling the channels prior to passing the elevation grid to CEM to see if that presents a more clear 'coastline' to be evolved. I'll also try reducing the wave height to see what impact that has on the results. I appreciate the suggestions.

Jay

[elbeejay]

As an update/follow-up I have not been actively pursuing this model coupling but I did compile the efforts and code I had put together thus far into a repository. So if anyone else ends up going down this road at least some of the first steps are documented.