During the pollen season, MeteoSwiss publishes text forecasts and pollen count maps that are updated every six hours. The maps show the concentrations of hazel, alder, birch, grasses and ragweed pollen according to the season (approx. January to end of September). In this way, members of the public can find out the pollen count for anywhere in Switzerland.
Pollen forecasts are an important tool for those suffering from allergies, allowing them to minimise the symptoms of their allergy, plan outdoor activities and take their medication in a more targeted manner. People suffering from allergies can also better coordinate their holidays and holiday destination with the blooming period for the allergenic pollen types.
The forecast maps depict the pollen concentrations of hazel, alder, birch, grasses and ragweed. They show the average pollen concentrations [pollen count per m³] over an area of 4.2 km². The maximum pollen count for the day can be significantly higher, particularly near blooming plants. On the other hand, the local count in the early hours of the morning tends to be lower. This is not a generally applicable rule, however, as the pollen count is largely determined by the weather and the conditions for pollen transport. In general, the map provides information about the emission and/or transport conditions on a particular day. The colours of the pollen categories reflect their relevance to allergy sufferers.
ICON has a horizontal resolution of 2.1 kilometres, allowing for both nationwide and very localised forecasts to be made for Switzerland. This means that the larger Alpine valleys can be realistically represented. For narrow valleys such as the Lauterbrunnental, however, the resolution of 2.1 km is still too coarse for a local pollen forecast. Nevertheless, the map is still helpful for narrow valleys as nearby, wider alpine valleys can provide reference points. It should be borne in mind, however, that plants bloom later as the height above sea level increases.
Calculating pollen transport is very complex and requires a lot of computing power. It is therefore not possible to track every type of pollen. MeteoSwiss focuses on the types of pollen that are most important for allergies. The ICON pollen module calculates the pollen concentrations of hazel, alder, birch, grasses and ragweed. With these five species, the most important species are covered throughout the entire pollen season, as they bloom one after another during the annual cycle. Depending on the progress of the blooming seasons in a particular year, there may be brief overlaps.
The numerical weather forecast model used by MeteoSwiss, ICON, can model the pollen count with the ART pollen module. The pollen module was developed by the Karlsruhe Institute of Technology (KIT) and MeteoSwiss. ICON’s pollen modelling comprises an entire process chain: Based on plant distribution and with the help of details on the weather over recent months, the model calculates the blooming start date for each grid point using an accumulated temperature model. This is followed by the calculation of the pollen emissions, which is greatly dependent on the current weather: High temperatures, dry conditions and a little wind generally favour pollen release. All the pollen that is released is then carried away by the wind and either falls to the ground somewhere, or is washed away by the rain.
The official pollen observation network of MeteoSwiss has been in operation since 2023 and consists of automatic measurement instruments that are able to measure the pollen counts in real time. This has made it possible to correct and improve the ICON pollen forecasts at regular intervals using real-time data.
Pollen forecasts can only be as good as the weather models. Even state-of-the-art weather models such as ICON do not, for example, always forecast precipitation absolutely correctly. Precipitation, however, is very important for the pollen count. This is where the expertise of the pollen forecasters comes into play. Their assessment of the plants’ current readiness to bloom and of the weather model is integrated into the published forecast, which allows certain deviations from the model-based map representation.