Artifacts in HRRR hour 0 data

[stephent]

Hi all - I've noticed that the hour 0 forecast often has visual artifacts which are not present in subsequent hours. Frequently they appear to be circular or partly circular. Here's an example for the visibility layer with the artifacts marked with arrows:

Contrast this with hour 01 where no such artifacts are visible (as best I can tell):

Is there a known reason for why these appear in hour 00, and are only certain forecast layers affected?

Similar issues are seen in other layers, e.g. low-cloud cover, where the hour 00 data can appear very blocky in contrast to later hours.

[stephent]

Hi @EricJames-NOAA - is this the right place for this type of question? Thanks!

[EricJames-NOAA]

@stephent yes this is a good place for the question! Sorry for my late response. The artifacts you are seeing stem from the use of our non-variational stratiform cloud analysis system used in HRRR. The system "clears" and "builds" clouds based on available observations of cloud hydrometeors in the 3-D volume, mostly using satellite cloud top height observations and METAR ceilometer observations. If METAR observations indicate low clouds or fog, cloud hydrometeors are inserted within a 100 km circle of the observation, unless other observations indicate clear conditions in any of that area.

You can find more details in our paper describing the analysis system:
https://doi.org/10.1175/MWR-D-20-0319.1

[stephent]

Perfect - thank you! I have some reading to do now...