The real temperate outside is approximately 0°C. I have to be honest - I don’t have a reference thermometer. But outside is cold and sometimes snowing and websites are reporting 0°C - 1°C. But both HA and AG Dashboard are currently reporting higher temperatures, around 4.7°C.
Even a few days ago when HA/AG are reporting 10 °C, there was no way it was 10°C. It felt much colder, around 4-5°C.
It is very possible to have localized temperatures where you are and where your monitor is, that readings are different than what the local news station reports.
We can see the effects of calibration and compensation working here. Do you have even a cheap little portable thermometer to compare with?
I’m not sure that I would be able to accurately discern the difference between 0 and 4 degrees myself personally. Especially when there’s a breeze or clouds and the sun… too many variables.
Maybe your sensor is just in a sunny location?
Thanks - I totally accept your feedback and agree it would make sense to have some kind of reference thermometer. The only other thermometer I have is my indoor Airgradient unit. I’ve just connected it to a USB powerbank and put it outside next to the outdoor monitor and will leave it there for 10 minutes and report back.
It looks like this experiment points out you should request a co2 calibration for your indoor unit anyway 
Other numbers would need more time to settle.
Temp looks within a degree. Could be margin of error from both the sensor or how they are physically positioned. They would probably drift closer if more time is given.
Also I am just some guy on the internet so my opinion is my opinion alone
Ok I ran the experiment for approximately 14 hours. I placed the indoor monitor outside, directly next to the outdoor one. Neither monitor catches any sun (it was evening/overnight anyway).
The grey line is the outdoor monitor, the solid bar graph is the ‘indoor’ monitor. The indoor monitor was consistently reporting a lower temperature than the outside monitor, by at least -1.5 °C. In the end before the power bank ran out, the discrepancy had reached -2.0 °C.
