Shared sea ice model performance metrics

[rgrumbine]

A starting point for consideration is at https://github.com/rgrumbine/ice_scoring

This is organized largely by parameter to be examined -- concentration field, ice edge, ice drift being the most developed.

[eclare108213]

@rgrumbine Here is the paper I mentioned in the meeting today:
Urrego-Blanco, J. R., E. C. Hunke, N. M. Urban, N. Jeffery, A. K. Turner, J. R. Langenbrunner, and J. M. Booker (2017), Validation of sea ice models using an uncertainty-based distance metric for multiple model variables, J. Geophys. Res. Oceans, 122, doi:10.1002/2016JC012602.?

[rgrumbine]

Excellent! I've done a quick glance through, and it doesn't look like there would be any difficulty in adapting/adding this measure in to the setup I already have.

Particularly important/useful is considering the Pareto front. Mention of that also puts me in mind of evolutionary methods for multiobjective optimisation.

For ice drift/motion evaluation, I considered 13 different measures in https://polar.ncep.noaa.gov/mmab/papers/tn315/ and arrived at 5 being meaningful and independent. For favorite, error radius. Codes still compute all 13.