stoiver
commented
4 months ago @ufloth the anuga algorithms work with the centroid values of elevation and stage as well as the other quantities. So maybe there is a miss match between your interpolation of the elevation using pts (which does first interpolate to vertices) and then averaging to the centroids.
If you have a DEM raster or tif, you can directly interploate to the centroids with a set_quantity command like
domain.set_quantity('elevation', file = name_of_asc_file.asc, location='centroids')Should be able to also do this with a tif file at least with the latest develop branch of anuga.
A good way to define a thin dyke would be to use a riverwall. This way you can be very explicit with specifying the height of the dyke along a poly line. The riverwall controls the flow across the line (edge of triangles) using the weir equation.
The documentation https://anuga.readthedocs.io/en/latest/examples/notebook_create_domain_with_riverwalls.html shows how to use riverwalls
Hello everybody, i try to model a scenario where a coastal plain is flooded by a rising water level. The plain is is crossed by a humble dike whose crest is partially below the maximum water level. Even if the model runs fine, i had to observe that the elevations of the dike crest in the mesh are below those of the original *.asc DEM-file. The extent of the dike crest was hereby defined by an interior region. Even when i define the spatial resolution of the dike crest with 1 m, the resulting crest elevation is below the actual one. I also tried to modify the alpha value (default 0.1) from 0.01 to 0.4, but that only yielded minor effects.
I would be glad if you could provide any advice to my issue.
Thank you in advance and best regards,
ufloth