Extraction failure for HRDPS due to change in set of variables in dataset

[douglatornell]

Reported by Natasha via email on 24-Aug-2022:

Another issue just popped up with reshapr and the HRDPS dataset. So I was running reshapr for separate years for the hrdps dataset (2015-2020) to extract tair at a certain location. In 2015 and 2016 I got a full year of data, however, when running 2017 and 2018, I only got a partial year, with the missing part of the year as nans. What Michael discovered was that around July 7 2017, and Dec 6 2018, variables were added to the files part way through the year which impacted the reshapr results.
So I just thought I would let you know. We have a work around, which is splitting the year before and after the variable addition, and it works perfectly fine. So no rush on our side.

[douglatornell]

Git annotate in SalishSeaNowcast confirms my suspicion that one cause of this issue was the addition of 3 HRDPS variables for the VHFR FVCOM model. They came in on 6-Dec-2018:

Add GRIB variables for VHFR FVCOM to weather download variables list.

- LHTFL_SFC_0  # upward surface latent heat flux
- RH_TGL_2     # relative humidity at 2m elevation
- PRATE_SFC_0  # precipitation rate at ground level

[douglatornell]

Another cause of this issue was the addition of the HRDPS variable:

TCDC_SFC_0   # total cloud in percent (for parametrization of radiation missing from 2007-2014 GEMLAM)

That variable came into the dataset on 8-Jul-2017.