Ed Doddridge

The expanding documentation (see e.g. #133, #155, #157) is making progress on this issue. As a larger library of images and animations becomes available, it would be good to place...

I've edited your comment to include links to the papers. I think the model contains all the features I'll require to do these experiments, but if not I'll put them...

Following #196 it is now possible to run simulations similar to those presented by [Yang et al. (2016)](http://doi.wiley.com/10.1002/2015JC011296) - namely, simulations of an idealised Beaufort Gyre in which the interface...

I agree that using NetCDF for the output is a highly desirable behaviour. Seems to me we are agreed on this, and now we just need to decide how to...

There's also another option - wrap the Fortran program in python. Does this option have downsides in terms of speed or complicating any future desires to make it run in...

That's good to know. Given that python dependencies should be easier to solve than Fortran dependencies, perhaps a python wrapping is the best option?

If we slightly reframe this issue as "Develop a benchmarking system and profile model performance" then it would give us very useful information as we embark on any rewriting efforts....

I'm a little surprised, but my intuition is that most of the time is spent writing outputs. Those setups are dumping output every 100 time steps, which is unlikely in...

I know you're currently working towards implementing a LAPACK algorithm for #1, but I did a quick test swapping the loop nesting order in `SOR_solver`. As might be expected from...

Bizarre! I'm afraid I don't have any useful insights to offer. To confirm - this also shows that we're using 20% of the flops available, yes? In which case we...