I ran into a small driver data difference that isn’t in our new modeling and think it is part of the weird ice issues we ran into with Hayley’s simulations. See here (specifically, this line). Issue is that having values for rain and snow on the same day can cause edge-case hiccups.
I looked into whether this impacted GCMs and it does not, so we only need to add this fix to the NLDAS driver munge code. Ideally, we do this before we share the scaled up NLDAS GLMv3 predictions. Once the new NLDAS drivers are built, they should be uploaded to Caldera.
@jread-usgs
I looked into whether this impacted GCMs and it does not, so we only need to add this fix to the NLDAS driver munge code. Ideally, we do this before we share the scaled up NLDAS GLMv3 predictions. Once the new NLDAS drivers are built, they should be uploaded to Caldera.