Problem with total column water vapor field

When I look through total column water vapor field time progression via NCVIEW or any other software that has the same functionality of going through a dataset in time, the time progression looks laggy as in the field abruptly changes. Does anyone notice this?

video sample

Seems like there is a periodic huge jump in between some time steps.

Dear Joongyun,
thank you for your query!
Are there particular times of the day that you see jumps?

Our analysis fields come from 12-hourly 4D-Var assimilation windows,
which contain hourly fields from 22UTC (the evening before) - 09UTC inclusive, and from 10UTC to 21UTC inclusive.
So, in case you see jumps at 10UTC and 22UTC then that’s where the come from.

How serious are the jumps for you?

Ahh, yes. Closely looking at the field, I do see jumps at 10UTC and 22UTC. I was just wondering if this is already known ‘issue’ (or some characteristics?) of some kind of ERA5 dataset. It was concerning since I’d see abrupt changes when I try to make animations of TCWV that includes 10UTC and 22UTC.

So, this ‘jump’ is not actually an issue? Is there any paper or documentation mentioning this?