Forecasting particulate pollutant concentrations in a city from meteorological variables and regional weather patterns

A synoptic meteorological classification and its correlation with particulate matter concentrations collected at five sampler points in the city of Valladolid (Spain) have been performed by means of a Q-mode clustering technique. The application of the algorithm to 11 meteorological variables related to the surface level and to the 500-hPa contour has permitted us to identify five synoptic meteorological types which have been split into nine synoptic meteorological classes. Also, the algorithm has been applied to the particulate matter concentrations recorded at each sampler point, and seven pollution levels have been obtained in each of them. The comparison between the days corresponding to the five synoptic meteorological types and the days included in each of the seven pollution clusters of every sampler point has revealed a satisfactory relationship between the particulate matter concentrations and the identified meteorological types at the 500-hPa contour.