In init_interp: If a given type cannot be found: Consider using something other than bare ground

**Bill Sacks < sacks@ucar.edu > - 2017-01-17 19:35:14 -0700**
**Bugzilla Id:** [2405](http://bugz.cgd.ucar.edu/show_bug.cgi?id=2405)
**Bugzilla Depends:** 2410,
**Bugzilla CC:** andre@ucar.edu, dlawren@ucar.edu, oleson@ucar.edu, rfisher@ucar.edu,

In init_interp: Currently, if there is a type (patch or col type) in your new run that was not present in the "template" run, init_interp will use the closest bare ground patch as a template. This most likely occurs when you interpolate from a non-crop run to a crop run, or a crop-without-irrigation run to a crop-with-irrigation run. But in principle, it could occur in other circumstances, too.

This has sometimes caused problems, such as the recent problem where new crop types were given the rootfr of bare ground - which was 0 for all levels. We have fixed these problems as we have discovered them, but it's possible there are other problems lurking now, or that may be added later.

At today's CLM-CMT meeting, we discussed possibly wanting to change this behavior so that, rather than using the bare ground patch as a template in these cases, it uses some vegetated patch. It's not clear what the rules should look like for how to choose the appropriate vegetated patch; this requires more thought.

This could become more important if/when each PFT is put on its own column: In this case, taking column-level values from the bare ground patch would give you very-non-spun-up C & N pools, for example.