The operators described in the arXiv paper are interesting because they modify the boundary closures to guarantee SBP and WENO. The downside is that the paper does not explicitly present the matrices, like Q, in an easy summary for implementation purposes. I am not sure if they fit into the existing method design and data structures herein. However, I would be curious if the modified boundary closures cure and/or alleviate the "hotspots" we have encountered with the 2D Upwind runs for KHI.
The operators described in the arXiv paper are interesting because they modify the boundary closures to guarantee SBP and WENO. The downside is that the paper does not explicitly present the matrices, like
Q, in an easy summary for implementation purposes. I am not sure if they fit into the existing method design and data structures herein. However, I would be curious if the modified boundary closures cure and/or alleviate the "hotspots" we have encountered with the 2D Upwind runs for KHI.