Analyzing the airborne birch and grass pollen monitoring network of Belgium

Abstract

Here, we analyze the Belgian aerobiological surveillance network design that consists of five pollen monitoring stations. Firstly, we focus on the network quality, evaluating how well the interpolated daily observed concentration fields reproduce the spatiotemporal reference data quantified by the Root-Mean-Square-Error (RMSE), from the 2013 to 2021 birch and grass pollen seasons. The reference dataset was derived from the pollen transport model SILAM (System for Integrated modeling of Atmospheric composition) using ECMWF ERA5 meteorology and pollen emission sources specific for Belgium. Secondly, for evaluating the network coverage representativeness, we perform a footprint-based analysis by running SILAM in the backward mode for the same seasons, which indicates the areas contributing to the concentrations observed in each station. We show that the Belgian aerobiological surveillance network performs well with respect to the spatiotemporal reference dataset for airborne birch and grass pollen near the surface. The quality threshold (normalized RMSE) is met in 79% of the birch pollen days and 80% of the grass pollen days over the nine pollen seasons. For the 2015 birch pollen season 100% is reached, for 2020 only 46%. For grass pollen the values range between 63 and 98% for 2021 and 2017, respectively. The footprint analysis shows that on average the coverage of the monitoring stations for birch pollen is quite good with typical large inter-seasonal differences. For grass pollen, the average coverage is larger, and the inter-seasonal variation is much lower. Out of five stations, three crucial monitoring sites for birch and grass pollen observations were identified. The study showed that the Belgian aerobiological monitoring network is of good quality, and its coverage is sufficiently high for airborne birch and grass pollen.