Hi
I understand it is a known problem with high humidities. It seems particularly bad (no pun intended) in foggy situations.
My system has shown some enormous figures (400-500 on PM10) which are quite unrealistic for a small rural hamlet about 8K from the nearest (small) town.
I have tried inserting a strip of hard polystyrene into the air tube. The idea is that the surface is supposed to be hygroscopic (not absorbing but attracting) so that water particles might condense. It does not appear to work.
I think the Luftdaten team are trying to find an algorithm to compensate.
Incidentally the BME280 temp/humid sensor is so far performing much better than the SHT22 which stuck at 99.90%.
regards
Alan