KAPA GS - Statistical models for the prediction of the daily concentration of particulate matter PM10

The daily concentration of particulate matter (PM10) in Graz of the next day is predicted by methods of multiple linear regression and time series analysis. An adequate model, our first proposal for PM10 with respect to the winter season includes the mean PM10 concentration of the current day, the type of the next day (working day, saturday, sun/-holiday) and three values from the weather forecast of the next day (vertical difference of air temperatures, wind velocity and rain (no, yes)). This approach will also be applied in Bozen and Klagenfurt.