No abstract

A range of commercially-available automatic pollen monitors were run in parallel and evaluated for the first time during the 2019 spring season; this includes the Droplet Measurement Technologies WIBS-NEO, Helmut-Hund BAA-500, the Plair Rapid-E, two Swisens Poleno, and two Yamatronics KH-3000 devices. The instruments were run from 19 April to 31 May 2019 and located in Payerne, Switzerland, representative of a semi-rural site on the Swiss plateau. The devices were validated against Hirst-type traps in terms of total pollen counts for daily and sub-daily averages. While the manual measurements cannot be considered a "gold standard" in terms of absolute values, they provide an established reference against which the automatic instruments can be evaluated. Overall, there was considerable spread between instruments compared to the manual observations. The devices showed better performance when daily averages were considered, with three of the seven showing non-significantly different values from the manual measurements. However, when six-hourly averages were considered, only one of the instruments was not significantly different from the Hirst trap average. The largest differences between instruments were evident at low pollen concentrations (< 20 pollen grains/m 3 ), no matter the temporal resolution considered. This is in part, however, to be expected since it is at such low concentrations that the Hirst measurements are most uncertain. It is also important to note that in 2019 many of the instruments tested had only recently been developed. Differences may also have arisen due to their varying abilities to identify specific pollen taxa or because the classification algorithms applied were developed for different pollen taxa and not total pollen, the variable considered in this study Keywords pollen • automatic monitoring • validation • real-time • Hirst 1. Introduction Airborne pollen form just a small fraction of the atmospheric bioaerosol loading ([36]; [18]; [39]), but despite their low number concentrations these particles are of significant importance because of their impact on human health ([38]; [47]; [14]), agriculture and sylviculture ([6]; [20]), as well as on climate through their role in the hydrological cycle ([35]; [10]). Pollen monitoring networks exist in most countries, providing information to a range of end-users from allergy sufferers and their doctors through to researchers. The current standard that is used across these networks is based on manual technology developed in the early 1950s ([17]; [13]; [2]) that is both time-consuming and laborious. Furthermore, this measurement technique suffers from several shortcomings, including the fact that data are delivered at low temporal resolution (usually daily averages) with a delay of between 3-9 days from the time of observation. New technologies developed over the past few years have made it possible to automatically measure pollen at high temporal resolutions and in real-time ([41],[33], [40]). The provision of such observations is dramatically ...

Technologies for monitoring pollen concentrations in real time made substantial advances in the past years and become increasingly available. This opens the possibility to calibrate numerical pollen forecast models in real time and make a significant step forward regarding the quality of pollen forecasts. We present a method to use real-time pollen measurements in numerical pollen forecast models. The main idea is to calibrate model parameterizations and not to assimilate measurements in a nudging sense. This ensures that the positive effect persists throughout the forecast period and does not vanish after a few forecast hours. We propose to adapt in real time both the model phenology scheme and the overall tuning factor that is present in any numerical pollen forecast model. To test this approach we used the numerical pollen forecast model COSMO-ART (COnsortium for Small-scale MOdelling - Aerosols and Reactive Trace gases) on a mesh size of 1.1 km covering the greater Alpine domain. Test runs covered two pollen seasons and included Corylus, Alnus, Betula and Poaceae pollen. Comparison with daily measurements from 13 Swiss pollen stations revealed that the model improvements are large, but fine tuning of the method remains a challenge. We conclude that the presented approach is a first valuable step towards comprehensive real-time calibration of numerical pollen forecast models.