AAAR 37th Annual Conference October 14 - October 18, 2019 Oregon Convention Center Portland, Oregon, USA
Abstract View
Real-time Monitoring and Modelling of Bioaerosols in Dublin, Ireland
Jose Manzano, Eoin McGillicuddy, Gavin Sewell, Paul Dowding, Matt Smith, Roland Sandra-Esteve, Dominique Baisnee, John Sodeau, DAVID O'CONNOR, Technological University Dublin
Abstract Number: 441 Working Group: Bioaerosols
Abstract Primary Biological Aerosol Particles (PBAP) are an omnipresent component of atmospheric aerosols. PBAP consist of an assortment of entities of biological origin with pollen, bacteria and fungal spores among the most studied in the atmosphere.
PBAP has long been associated with health implications such as hayfever, COPD, asthma and aspergillious to name but a few. Allied to this PBAP also have the potential to act as cloud condensation nuclei or ice nuclei in cloud formation and thus have climatic inferences. Hence the need to study the concentration and composition of such particles is of interest to all.
However within Ireland little has been done with regard to the amounts or identity of biological particles in the literature. This work represents first sustained monitoring of PBAP in decades.
The work presented here has looked at both traditional (based on impactation on a filter and subsequent optical analysis) and newer real-time methodologies (utilizing fluorescence) for the monitoring of pollen and fungal spores at a site located in the heart of Dublin, Ireland. The sampling campaign utilized a Hirst volumetric trap, WIBS-4 and Japanese pollen counter to develop a seasonal cycle for the prevalence of pollen and fungal species in the Irish atmosphere. The real-time and traditional instrumentation were also compared and contrasted to evaluate weather real-time instrumentation was comparable to the currently most used technique (Hirst trap) around the world.
The traditional data in tandem with meteorological parameters, phenological data and source mapping were then used to formulate an Irish specific pollen model using several methods (multiple regression, random forest and neural networks).The incorporation of real-time pollen data into the models was then attempted with hope of increasing the accuracy, precision and timeliness of the forecasts.