This graph shows data of three sample days. Instead of counting all pollen during one day, the pollen were counted every hour. This results often in interesting insights of pollenconcentrations during the day. In this case you can see peaks during morning, afternoon and evening. Mainly based on different weatherconditions.
For most hay fever patients this hay fever season passed with relatively few complaints. The Birch flourished this year quite exuberant, but the weather conditions were such that the number of pollen in the air was less abundant than previous years. Also the concentration of grass pollen remained low until the first half of may due to the lower temperatures and the volatile weather. Unfortunately there is no crystal ball to predict the amount of pollen in the future. How is such a expectancy-model made anyway? A “hay fever expectancy model” resembles an onion and consists of multiple layers of information.
Rain & Pollencount
A rainstorm is nowadays thanks to all sorts of new techniques very well “live and local” to predict. This is also beneficial for predicting pollen concentration, because precipitation is an important part of the model. But there are more factors. It begins with nature itself. What blooms where at the moment? Thanks to pollen calendars is pretty accurately known during which period specific grasses, trees and herbs thrive. With a margin of 1 or 2 weeks is that pretty good to predict. The temperature, however, has a major influence on the development speed of nature. Field observations are therefore used to the predicted flowering stage of trees, grasses and herbs. These observations and the historical and current temperatures are therefore an important part of a hay fever model.
Pollen counts also play an important role in the hay fever model. Thanks to pollen counts it’s pretty well known how much pollen of any kind were in the air. Thereby you can also check the relationships between pollen and other weather aspects such as wind direction and wind speed. Wind from the South-East is generally unfavourable. On the other hand, by persistently calm weather the pollen concentration might get high. This pollen counts are, among others, done by the Elkerliek hospital in Helmond, Netherlands.
Finally, there are the complaints-reports of the hay fever patients themselves. Via the Pollen news apps, people can track their own symptoms in a logbook and transmit the message. These alerts give a pretty good impression of the hay fever situation in Netherlands. These complaints reports play an important role in improving the hay fever expectation model. In a hay fever expectancy model all these factors come together. It shows the current day and a 5-6 forecast for the coming days. The further away, the less accurate the prediction ofcourse. Reason for this is, among other things, the Dutch weather again. Our weather is very changeable and can vary a lot from day to day. Scientists expect that the weather over the next few years by climate change warmer and more inconsistent.
More inconsistent weather means hayfever expectation models should get more precise. Especially because the pollen concentrations in the air not only vary per day, but also per hour. The good news is that with a new method of analysis it is now possible to count pollen concentrations per hour!
Can this type of hourly countings be used to improve the hayfever expectancy model? The answer is a resounding Yes! Currently Pollennews, Hayfeverradar, Rainradar and the Pollen team of the Elkerliek hospital work hard at improving the expectation model. For example, by combining hourly countings and local weather data on an hourly basis.
