[PRODUCT] APRFC data products

[Enovotny]

Product Source (URL Preferred)

https://cbt.crohms.org/akgrids/

Product Description

ABRFC QPF, QPE, QTE, QTF

Product Format

• Geotiff
• Grib2
• NetCDF
• Other (Please provide description below)

Format Description

APRFC Forcings for USACE (1).docx

[adamscarberry]

ABRFC QPF, QPE deployed to stable. Thanks to @Enovotny for working on the migrations and processors.

[FHanbali]

QPE and QPF values are suspicious, showing 60+ inches for a one week time window

Suspect mm to inches conversion maybe the issue.

[adamscarberry]

Fix by @msweier deployed to stable. This can be closed if @FHanbali will test the downloaded results again.

[FHanbali]

@msweier I'm still seeing large precip values. I cross checked a download from cumulus for a forecast starting at 09August2024-1200 against original grib files (which have "mm" units). It seems that Cumulus is not converting from "mm" to "inches" properly.

Btw, the DSS download from Cumulus has units of "inches" for the DSS records, but I'm pretty sure the values have not truly been converted to inches. Suggest writing out the records with units of "mm" anyway, that way you wouldn't have to worry about unit conversions.

@adamscarberry I'm curious if all the products in Cumulus are getting unit conversions?

[FHanbali]

Here are the test Cumulus download and original APRFC files
APRFC_QPF_Tests.zip

[FHanbali]

Also wanted to verify, once a fix is complete for this, will the entire archive for this product in Cumulus get corrected please?

[adamscarberry]

We do not convert units as it goes from raw product to geotif (done by async_geoprocessor). Original units should remain.

The fix attempted by @msweier was to change the units for the product in the database which the async_packager uses to create the dss download.

No files should need reprocessed. Once fixed the future downloads should e correct.

I'm not able to look deeper into this issue right now.

[msweier]

@msweier I'm still seeing large precip values. I cross checked a download from cumulus for a forecast starting at 09August2024-1200 against original grib files (which have "mm" units). It seems that Cumulus is not converting from "mm" to "inches" properly.

Btw, the DSS download from Cumulus has units of "inches" for the DSS records, but I'm pretty sure the values have not truly been converted to inches. Suggest writing out the records with units of "mm" anyway, that way you wouldn't have to worry about unit conversions.

@adamscarberry I'm curious if all the products in Cumulus are getting unit conversions?

Thanks for the test data @FHanbali. I'll take a look.