Forecasting daily total pollen concentrations on a global scale

Basit Öğe Kaydı
dc.authoridPeter, Jonathan/0000-0002-2658-0723
dc.authoridBruffaerts, Nicolas/0000-0001-6310-9140
dc.authoridCoviello, Luca/0000-0002-8544-6540
dc.authoridSaarto, Annika/0000-0002-1568-3495
dc.authoridMakra, Laszlo/0000-0001-7424-8963
dc.authoridBeggs, Paul/0000-0001-9949-1783
dc.authoridPrzedpelska-Wasowicz, Ewa Maria/0000-0002-0481-7953
dc.contributor.authorMakra, Laszlo
dc.contributor.authorCoviello, Luca
dc.contributor.authorGobbi, Andrea
dc.contributor.authorJurman, Giuseppe
dc.contributor.authorFurlanello, Cesare
dc.contributor.authorBrunato, Mauro
dc.contributor.authorZiska, Lewis H.
dc.date.accessioned2025-02-22T14:08:53Z
dc.date.available2025-02-22T14:08:53Z
dc.date.issued2024
dc.departmentDicle Üniversitesien_US
dc.description.abstractBackgroundThere 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.MethodsThe 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.ResultsThe best pollen forecasts include Mexico City (R2(DL_7) approximate to .7), and Santiago (R2(DL_7) approximate to .8) for the 7th forecast day, respectively; while the weakest pollen forecasts are made for Brisbane (R2(DL_7) approximate to .4) and Seoul (R2(DL_7) approximate to .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.ConclusionsThis 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. CatBoost is a preferable model for short-term forecasts, while Deep Learning is for longer ones, but there is no definite answer to what the better model is for every day or city. Past pollen trends are strong indicators of future pollen concentrations. CatBoost can be used to determine the importance of environmental variables in forecasting daily total pollen concentration. Abbreviations: 2mT, 2 m temperature; CB, CatBoost; DL, Deep Learning; DOY, day of the year; ERA5, the fifth generation ECMWF (European Centre for Medium-Range Weather Forecasts) atmospheric reanalysis dataset; pevap, potential evapotranspiration; st, soil temperature.imageen_US
dc.description.sponsorshipEU- COST Action ADOPT [CA18226]en_US
dc.description.sponsorshipWe thank Mr Fernando Samuel Tellez Unzueta, for his help in data collection. The authors would also like to acknowledge Ms Beatrice Dalpedri for her help in statistical hypothesis testing. The authors express their grateful thanks to Brett J. Green for providing pollen data. The study was partly implemented in the frame of the EU-COST Action ADOPT (New approaches in detection of pathogens and aeroallergens), Grant Number CA18226 (EU Framework Program Horizon 2020).en_US
dc.identifier.doi10.1111/all.16227
dc.identifier.endpage2185en_US
dc.identifier.issn0105-4538
dc.identifier.issn1398-9995
dc.identifier.issue8en_US
dc.identifier.pmid38995241en_US
dc.identifier.scopusqualityQ1en_US
dc.identifier.startpage2173en_US
dc.identifier.urihttps://doi.org/10.1111/all.16227
dc.identifier.urihttps://hdl.handle.net/11468/29692
dc.identifier.volume79en_US
dc.identifier.wosWOS:001269195300001en_US
dc.identifier.wosqualityQ1en_US
dc.indekslendigikaynakWeb of Science
dc.indekslendigikaynakPubMed
dc.language.isoenen_US
dc.publisherWileyen_US
dc.relation.ispartofAllergyen_US
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanıen_US
dc.rightsinfo:eu-repo/semantics/closedAccessen_US
dc.snmzKA_WOS_20250222
dc.subjectallergyen_US
dc.subjectartificial intelligenceen_US
dc.subjectenvironmental variablesen_US
dc.subjectfeature importance clusteren_US
dc.subjectpollen forecasten_US
dc.titleForecasting daily total pollen concentrations on a global scaleen_US
dc.typeArticleen_US

Dosyalar

Koleksiyon

WoS İndeksli Yayınlar Koleksiyonu
PubMed İndeksli Yayın Koleksiyonu