Monitoring, modelling and forecasting of the pollen season

Abstract

The section about monitoring covers the development of phenological networks, remote sensing of the season cycle of the vegetation, the emergence of the science of aerobiology and, more specifically, aeropalynology, pollen sampling instruments, pollen counting techniques, applications of aeropalynology in agriculture and the European Pollen Information System. Three data sources are directly related with aeropalynology: phenological observations, pollen counts and remote sensing of the vegetation activity. The main future challenge is the assimilation of these data streams into numerical pollen forecast systems. Over the last decades consistent monitoring efforts of various national networks have created a wealth of pollen concentration time series. These constitute a nearly untouched treasure, which is still to be exploited to investigate questions concerning pollen emission, transport and deposition. New monitoring methods allow measuring the allergen content in pollen. Results from research on the allergen content in pollen are expected to increase the quality of the operational pollen forecasts. In the modelling section the concepts of a variety of process-based phenological models are sketched. Process-based models appear to exhaust the noisy information contained in commonly available observational phenological and pollen data sets. Any additional parameterisations do not to improve model quality substantially. Observation-based models, like regression models, time series models and computational intelligence methods are also briefly described. Numerical pollen forecast systems are especially challenging. The question, which of the models, regression or process-based models is superior, cannot yet be answered.