A serious bug fix in the cdklm kernel related to the reconstruction of h. The wrong distance from the reference point was used, as well as the wrong north vector
A previously know bug fix in the cdklm kernel, fixing Coriolis force according to north in the adjust slope functions
Now re-using the angle texture and correct north for the Coriolis force in the model error.
Minor bugfix to support open flow-relaxation BC's without supplying BoundaryConditionData
babrodtk
commented
5 years ago
This PR contains