Understanding of the behavior of atmospheric pollen concentration, as well as developing predictive models, can greatly help allergic sufferers. The aims of this study were (i) to determine mean multi-year characteristics of temporal and space–time autocorrelation of the pollen counts of Corylus, Alnus, and Betula in Poland, (ii) to create and evaluate Corylus, Alnus, and Betula pollen concentration levels predictions based on previous pollen count values from given sites, and (ii) to develop spatiotemporal predictive models of Corylus, Alnus, and Betula pollen concentration levels, using preprocessed gridded meteorological data. The monitoring of the concentrations of Corylus, Alnus, and Betula pollen in the air was conducted in 11 cities in Poland. Additionally, AGRI4CAST Interpolated Meteorological Data were used as predictor variables. The autocorrelation and cross-correlation functions were used to investigate temporal and spatial patterns. Random forest method was used to predict the high pollen concentration level of Corylus, Alnus, and Betula. The study provided an understanding of the temporal and spatiotemporal autocorrelation of Corylus, Alnus, and Betula pollen counts. The final models also proved to be capable of pollen levels predicting in continuous areas rather than in a single location.
Analysis of Airborne Betula Pollen in Finland; a 31-Year Perspective