About the reference system, unfortunately I don’t have any idea.
About the time, we always use UTC timestamps, so if it is not noted then it is UTC. If anything else is used then it has (should have) the timezone attached.
About the dashboards one way would be to try finding the source for the decoders, there, usually, units are described. I’m going also to ping a couple of people that are related with the decoders of these satellites and their dashboards and they may can help you, cc @dl7ndr@somecheese@marcelfr.
The beacon format of all four satellites is very similar.
So I guess if you know the unit(s) of one satellite, you know them for all four.
Unfortunately, there is no unit note left in the decoder files.
As for GRBAlpha @apal would probably have to weigh in. If I remember correctly there were some calibrations and the calculation to human readable values was a bit more complicated than the following.
For the others it is true they are very similar. There is no calibration applied so these values may be strange as a result. Axes should be aligned with the satellite reference frame XYZ.
With CroCube to calculate into nano Tesla the following should work M_nT = M_raw * 100000 / 16384
LasarSat and Veronika need an additional step of subtracting null-field value.
M_nT = (M_raw - 131072) * 100000 / 16384
Regarding Grafana, the CroCube calculation can be applied quite simply. The nested formula for LasarSat and Veronika I am not too sure on the spot. Per my personal philosophy it is not desired to do these calculations in decoders, but rather in post processing. Hint or commentary probably could have been written in the decoder though.
Quick edit and small disclaimer: These magnetometers honestly did not get the attention in engineering they probably deserved, so I would take their actual data with a spoonful of salt. Out of these four satellites the most relevant magnetic measurements would probably come from LasarSat’s Payload Magnetometer.
Another issue about these magnetometers is about their zero points. So, these are prone to offset easily and in space, it is not that easy to do such a calibration. However, they were very useful to derive spin period of the satellite and give constraints of the spin axis w.r.t. the satellite and/or reference frame. Luckily, these do not depend on this zero point, only on the derived best-fit Fourier frequency and the amplitude of these quasi sinusoidal signals. But for such a measurement, you need much higher cadence of magnetometer data readout than what’s in these dashboard.