RESUMO
BACKGROUND: There is evidence that global anthropogenic climate change may be impacting floral phenology and the temporal and spatial characteristics of aero-allergenic pollen. Given the extent of current and future climate uncertainty, there is a need to strengthen predictive pollen forecasts. METHODS: The study aims to use CatBoost (CB) and deep learning (DL) models for predicting the daily total pollen concentration up to 14 days in advance for 23 cities, covering all five continents. The model includes the projected environmental parameters, recent concentrations (1, 2 and 4 weeks), and the past environmental explanatory variables, and their future values. RESULTS: The best pollen forecasts include Mexico City (R2(DL_7) ≈ .7), and Santiago (R2(DL_7) ≈ .8) for the 7th forecast day, respectively; while the weakest pollen forecasts are made for Brisbane (R2(DL_7) ≈ .4) and Seoul (R2(DL_7) ≈ .1) for the 7th forecast day. The global order of the five most important environmental variables in determining the daily total pollen concentrations is, in decreasing order: the past daily total pollen concentration, future 2 m temperature, past 2 m temperature, past soil temperature in 28-100 cm depth, and past soil temperature in 0-7 cm depth. City-related clusters of the most similar distribution of feature importance values of the environmental variables only slightly change on consecutive forecast days for Caxias do Sul, Cape Town, Brisbane, and Mexico City, while they often change for Sydney, Santiago, and Busan. CONCLUSIONS: This new knowledge of the ecological relationships of the most remarkable variables importance for pollen forecast models according to clusters, cities and forecast days is important for developing and improving the accuracy of airborne pollen forecasts.
Assuntos
Alérgenos , Previsões , Pólen , Pólen/imunologia , Previsões/métodos , Humanos , Mudança Climática , Modelos Teóricos , Monitoramento Ambiental/métodosRESUMO
A coordinated observational and modelling campaign targeting biogenic aerosols in the air was performed during spring 2021 at two locations in Northern Europe: Helsinki (Finland) and Siauliai (Lithuania), approximately 500 km from each other in north-south direction. The campaign started on March 1, 2021 in Siauliai (12 March in Helsinki) and continued till mid-May in Siauliai (end of May in Helsinki), thus recording the transition of the atmospheric biogenic aerosols profile from winter to summer. The observations included a variety of samplers working on different principles. The core of the program was based on 2- and 2.4--hourly sampling in Helsinki and Siauliai, respectively, with sticky slides (Hirst 24-h trap in Helsinki, Rapid-E slides in Siauliai). The slides were subsequently processed extracting the DNA from the collected aerosols, which was further sequenced using the 3-rd generation sequencing technology. The core sampling was accompanied with daily and daytime sampling using standard filter collectors. The hourly aerosol concentrations at the Helsinki monitoring site were obtained with a Poleno flow cytometer, which could recognize some of the aerosol types. The sampling campaign was supported by numerical modelling. For every sample, SILAM model was applied to calculate its footprint and to predict anthropogenic and natural aerosol concentrations, at both observation sites. The first results confirmed the feasibility of the DNA collection by the applied techniques: all but one delivered sufficient amount of DNA for the following analysis, in over 40% of the cases sufficient for direct DNA sequencing without the PCR step. A substantial variability of the DNA yield has been noticed, generally not following the diurnal variations of the total-aerosol concentrations, which themselves showed variability not related to daytime. An expected upward trend of the biological material amount towards summer was observed but the day-to-day variability was large. The campaign DNA analysis produced the first high-resolution dataset of bioaerosol composition in the North-European spring. It also highlighted the deficiency of generic DNA databases in applications to atmospheric biota: about 40% of samples were not identified with standard bioinformatic methods.