18.2.1: Some model parameters do not ingest correctly

[jaredwsmith]

To report a non-security related issue, please provide:

• the version of the software with which you are encountering an issue: AWIPS 18.2.1
• environmental information (i.e. Operating System, compiler info, java version, python version, etc.) macOS Big Sur, CentOS 7
• a description of the issue with the steps needed to reproduce it:

It appears that several model fields do not ingest correctly in 18.2.1. Fields I have found with problems so far:

• NAM/GFS vorticity fields (see screenshot)

• RAP13 does not ingest parameters such as Supercell Composite and Significant Tornado at all; HRRR ingests these, but does not display values for these fields

• RAP13 mixed-layer computed CAPE and CIN are not available at all (function correctly with the RAP20 on AllisonHouse)

I've verified these issues both on the Unidata cloud EDEX and AllisonHouse's EDEX offering. I can also confirm that this does not appear to be client-specific; I have reproduced the problems on AllisonHouse's EDEX using the NWS 20.2.3 baseline build on Linux. This is not an exhaustive list of the fields with issues, but these are certainly the most noticeable.

[WxmanJ]

What happens when you try to sample the NAM/GFS vorticity field image, do values display? If values display when sampling, you could try altering the colormap through "Change Colormap" to handle the fractional values (e-5) in min/max. A permanent way to handle this would be through the styleRules or parameters.xml, but if you are using an outside EDEX server, that is something you cannot alter.

I modified this to include Tiffany’s solution. Parameters.xml is a master change, where editing the style rules gives more flexibility with specific NWP. I feel like the HRRR had some crazy large vorticity values in the past compared to other NWP.

[tiffanycmeyer13]

Thank you for bringing these to our attention. We are looking into them.

• I see the same issue you do with vorticity - it's odd because it shows up incorrectly for the ingested heights (1000, 850, 700, 500, 250) but shows correctly for the interpolated heights...We are looking into this.

• I am not able replicate your issue with the HRRR Supercell Composite....

• At Unidata we've only ever provided RAP13 which hasn't ever had mixed-layer computed CAPE/CIN. We don't ingest RAP20 model data.

[tiffanycmeyer13]

Hello, you should be able to render Vorticity plots now. We found an issue with the units. I had to update the database table:
update parameter set unit = '/s' where abbreviation like 'AV';

And then for future use I updated the units from 1/s to /s in /awips2/edex/data/utility/common_static/base/parameter/definition/parameters.xml

<parameter>
<name>Absolute Vorticity</name>
<abbreviation>AV</abbreviation>
<unit>/s</unit>
</parameter>

No restarting of EDEX was necessary, just had to wait for a new model run to come in.

[jaredwsmith]

Thank you, @tiffanycmeyer13! Validated that vorticity is coming in properly over Unidata's cloud EDEX.

Still seeing issues loading the various thunderstorm parameters from the HRRR. Did some troubleshooting and I believe the culprit is an error with the bulk shear products. When I try to load 0-6km bulk shear from the volume browser, I get "NO DATA" across the entire domain except -- interestingly -- the sliver of Southern California you saw the Layer Supercell Composite come through. I tested this on the Unidata CentOS VM. I suspect similar issues are plaguing the RAP ingest as well. A-OK with NAM/GFS.

[jaredwsmith]

Digging further, and it looks like something starts to go wacky with the wind grids as you step up in elevation. Pardon the chaotic image, but loading the wind data from the HRRR as streamlines does a nice job of showing where the data holes are. (Also notice there are no yellow streamlines indicating 6km wind.) It certainly strikes me that whatever is happening with these parameters is due to issues with the wind grids.

[tiffanycmeyer13]

Okay, I've been looking into this in more detail. NOAAPort is where we receive our HRRR data from and they limit the data that they are sending. You can see here the data NOAAPort is sending and there's nothing above 500mb which is why you aren't able to plot any winds above about 5km.

[jaredwsmith]

Sorry for the very late response, @tiffanycmeyer13 :) That makes total sense w/r/t HRRR wind fields and I'm not sure why I didn't consider that before.

Vorticity looks great on all models post-18.2.1-3!

I'm still noticing in 18.2.1-3 that Supercell Composite Parameter and Significant Tornado Parameter, among others, are not being calculated for the RAP13 or NAM12. It is possible that this could be due to CAPE for this model specifically not being selectable by fixed height above ground (FHAG). Mixed-layer CAPE can be computed for every other available model. This worked for the RAP13 and NAM12 in 18.1.1 and versions prior. The issue is reproducible with the latest Unidata AWIPS build against the edex-cloud server.