Add regression test fv3_ccpp_gsd_sar_25km_debug

[climbfuji]

This PR updates the submodule pointer for FV3 because of a bug fix for coldstarting models using RUC LSM with RAP/HRRR initial conditions (see NCAR/fv3atm#22). The PR also adds a new regression test fv3_ccpp_gsd_sar_25km_debug which would have detected the error.

[climbfuji]

@JeffBeck-NOAA @gsketefian for your information

[JeffBeck-NOAA]

Dom, Thanks again!  Please let us know when the PR is accepted and we can test the code. Happy Holidays! Jeff

…

On Dec 23, 2019, 4:19 PM, at 4:19 PM, Dom Heinzeller ***@***.***> wrote: @JeffBeck-NOAA @gsketefian for your information -- You are receiving this because you were mentioned. Reply to this email directly or view it on GitHub: #19 (comment)

[climbfuji]

Regression testing on hera/intel using rt.conf against DTC-maintained official baseline: all tests pass.

rt_full.log

[climbfuji]

Regression testing using NEMSCompsetRun on hera/intel against EMC baseline. All tests pass except fv3_gfdlmprad_noahmp. This is correct, because the dtc/develop branches contain a bugfix for Noah MP that is not yet included in EMC's code.

NEMSCompsetRun.log

[climbfuji]

Regression testing on hera/intel using rt_ccpp_gsd.conf (first create = CCPP static, then verify = CCPP static and CCPP dynamic). All tests pass, including the newly added fv3_ccpp_gsd_sar_25km_debug test.

rt_ccpp_gsd_create.log
rt_ccpp_gsd_verify.log

[climbfuji]

@llpcarson @grantfirl @ligiabernardet @tanyasmirnova @JeffBeck-NOAA this PR and NCAR/fv3atm#22 are ready to merge. They need to go into dtc/develop in the morning on Thu Dec 26 so that they will make it into the PR from dtc/develop to EMC's branches. Since the changes are straightforward and all tests passed, I will merge them w/o approval if necessary.