Daniel Shapero

@victorsndvg that all sounds about right. The sparse matrix classes in SiGMA mostly follow the template pattern, but there is a [wrapper class](https://github.com/danshapero/sigma/blob/master/src/matrix/sparse_matrix_composites.f90) to hide some of the details. In...

Well that's embarrassing, good catch! My first thought is to use the GNU multi-precision library. If you want to fix this I'd gladly accept the PR, or I can do...

Wow, thanks for catching this! Gotta love those sanitizers. Your guess about the cause makes perfect sense to me. There are several other places in that function that use the...

I think your option (d) would be quick to test. Do you want to give it a try? If it works then great, if not then we can explore other...

Thanks for looking into this @HansBrende ! I'll close the issue now that we know this. I'm sure that at the time it made sense to avoid conditionals if there's...

Hi Bryan! The code in that notebook to compute `u_shear` isn't evaluating the basal velocity as such, but rather the projection of the velocity field onto the 1st-degree Legendre polynomial,...

I made a branch with some demonstrations in the hybrid XZ notebook [here](https://github.com/danshapero/icepack/blob/surface-basal-velocity-demo/notebooks/tutorials/06-hybrid-ice-stream-xz.ipynb). The important part is right before the conclusion section. In case you don't want to go digging...

One of the main tests for the hybrid solver is to take the vertical degree to be 0 and checking that it gives the same result as SSA. Given that...

Yes you're both exactly right, I can't believe I missed this -- it's the double counting because of the symmetry of the strain rate tensor. I'll write a test case...

I made a PR with the fix. If either or both of you can look at the test case and tell me whether or not you're convinced that would be...