Forecasting the Poaceae pollen season in eastern Poland

Abstract

The aim of the present study was to develop forecast models for the grass pollen season by using regression analyses to predict such characteristics as onset, duration and peak pollen concentration (values and timing). The study shows a negative correlation between seasonal pollen index (SPI) and season duration as well as strong correlations between some features of the pollen season and meteorological data. The forecasting models may predict 86–98% of the variation in pollen onset and duration of the pollen season. Less satisfactory results were obtained for the peak date and peak value. The best prediction was obtained for the season duration. The mean minimum temperature of March and cloud cover in the first 10-day period of May were the best variables for forecasting the start of the grass pollen season in Lublin. The rainfall in May was the most important factor to determine season duration. When the developed models are applied for pollen season forecasting, it is possible to predict the season onset and duration with an accuracy of up to one or two days.