Seminar: Michael Rzanny

One third of the world’s population suffers from pollen-induced respiratory allergies. For affected individuals, reliable information on local pollen loads and thus the risk of allergy symptoms is essential. However, obtaining accurate pollen forecasts at the local scale remains challenging. This is largely due to limited knowledge of local pollen occurrence, flowering phenology, and pollen release dynamics, as well as the restricted spatial resolution, species coverage, and real-time flowering information available in current pollen forecasting models.The PollenNet project brings together expertise in biology, technology and medicine to provide accurate predictions of local pollen loads for specific species and even the local risk of experiencing allergy symptoms. A key component of the project is the integration of observations collected through the plant identification app Flora Incognita. By submitting records of allergenic plants, app users contribute real-time information on flowering activity that can be used to improve pollen forecasts. Combined with pollen trap measurements and local weather forecasts, these observations will provide the basis for precise and highly resolved predictions of local pollen concentrations and allergy risk.We are currently conducting a Citizen Science project within the app that invites users to contribute images. This approach has proven highly effective in motivating participants to document and photograph the complete flowering cycle of highly allergenic species such as hazel, birch, and various grasses. These images are subsequently used to develop species-specific classifiers for automated flowering-stage recognition. Hundreds of georeferenced observations submitted each day enable both the fine-scale mapping of allergenic plant distributions and the recognition of their current flowering stages at given locations. In this talk, I will introduce the PollenNet project and present some of our first results.