search

bearecinos / smith_glacier

Experiments with Fenics_ice over Smith, Pope, and Kohler Glaciers; three narrow ∼ 10 km wide, interconnected West Antarctic ice streams (Note: this repository documentation is still a work in progress).
1 stars 0 forks source link

Improve interpolation of fenics_ice output for STREAMICE #49

Closed bearecinos closed 1 year ago

bearecinos commented 1 year ago

@jrmaddison @dngoldberg

We discussed several issues with the interpolation of the fenics_ice output into the grid use for the STREAMICE experiment. The interpolation happens in re_grid_fenics_ice_output.py.

Several things need to be improved:

  1. The resolution of the grid given by @dngoldberg
  2. We need to interpolate bed and thickness (H) from BedMachine directly not use fenics ice output.
  3. We need to interpolate velocity observations from the original ITSLIVE data.
  4. We need to use the model output from the L-curve priors with the original ITSLIVE data. So the output of the workflow shown in Fig. 3 h and f (red line)
  5. We need to set to zero all alpha values in floating ice.
  6. We need to interpolate alpha^2 ... not alpha!

If I forgot something let me know.

bearecinos commented 1 year ago

So I've done all but 6 as I remembered that we dont output alpha^2 from the l-curve workflow to reduce the amount of data produced for each inversion. ... I guess we said that having only alpha is ok.