Not sure if this has any effect what so ever, but alot of the Hires models ingest data from PWS networks and other real time data. Thus is from my PWS, recording the pressure wave from the Tonga explosion, would have been around time 12z to 18z data was being ingested.