What is PMS5003 telling me?

Bought a PMS5003 + cable + Part. Mat. Breakout at the end of January. It is sat indoors on my windowsill, hooked up to a Pi Zero W running Raspbian.
It seems to me like the readings are a “randomised” every time I start all.py:

22:00, ~22:55 and ~23:30 was when it was restarted. I don’t understand why there is such a large difference between the readings before 22:00 and after. The fact that this happens whenever all.py is restarted suggests to me that these readings aren’t a true reflection of the environment.

Zooming out to ~30d view, all these discontinuities in the readouts are happening when all.py is restarted:

[nah, you’re not allowed two pictures in one post]

Can anyone help me interpret these data? Do I have a duff PMS5003?

30d overview:

Does anybody else have stats from their PMS5003 that they could share for comparison?

Has anybody else got graphs of their PMS5003 output?

Hi, I take readings every 10 minutes and here are a few example graphs for pm1 and pm2_5


Thanks! Do your stats seem plausible? I.e., can you account for that large spike?

Hi, yes the stats add up due to their location. The top one is currently in the kitchen and you get the spikes when you start cooking etc. I also have gas resistance sensors and co2 sensors and the spikes all tend to line up with the close by activity.

I also have one outside and we are close to a main road and the spikes match up when there is a lot of traffic about. here is a sample from outside.