Predicting the Poaceae pollen season: six month-ahead forecasting and identification of relevant features

Abstract

In this paper, we approach the problem of predicting the concentrations of Poaceae pollen which define the main pollination season in the city of Madrid. A classification-based approach, based on a computational intelligence model (random forests), is applied to forecast the dates in which risk concentration levels are to be observed. Unlike previous works, the proposal extends the range of forecasting horizons up to 6 months ahead. Furthermore, the proposed model allows to determine the most influential factors for each horizon, making no assumptions about the significance of the weather features. The performace of the proposed model proves it as a successful tool for allergy patients in preventing and minimizing the exposure to risky pollen concentrations and for researchers to gain a deeper insight on the factors driving the pollination season.

Recent developments in monitoring and modelling airborne pollen, a review

“…This method generates a number of regression trees relating to the sample data and then combines them by averaging (Navares & Aznarte 2020). This method has the added capability of determining the most important variables contributing to a model (Zewdie et al 2019a) and as such have been used in several studies to identify the most important variables for improving ML models (Navares & Aznarte 2017;Navares & Aznarte 2020). Aside from this, RFs have been used to develop prediction models for high concentrations of different pollen types, such as Alnus, Betula and Corylus (Nowosad 2016;Nowosad et al 2016) and Poaceae (Navares & Aznarte 2017).…”

Geographical Imputation of Missing Poaceae Pollen Data via Convolutional Neural Networks

“…There is no general consensus about the definition of the pollen season, and hence, season dates might differ according to their definition [11]. This notwithstanding, in Spain, the first symptoms are observed over 25 grains/m 3 [33].…”

“…Examples include regression models [3,4], time series models [5], and process based phenological models [6]. In the last decade, machine learning techniques have been gaining importance due to the success of their applications [4,[7][8][9][10][11][12]. However, these techniques require a significant amount of data, and when dealing with pollen time series, where high concentrations are especially harmful when they are over 25 grains/m 3 [1], the data are incomplete during the full year ( Figure 1).…”

POLLAR: Impact of air POLLution on Asthma and Rhinitis; a European Institute of Innovation and Technology Health (EIT Health) project

“…Many different methods are used to monitor pollen exposure [25–28 ]. Pollen counts can assess the exposure of pollen‐allergic patients [29 ].…”

“…Meteorological data may, in the future, be of interest for predicting the onset of the season, but more data are required [31 ]. Combining several sources using advanced data engineering may also be important but these data are still complex and, in many different areas, not yet available for all pollen species [25–28, 32 ]. Google Trends (GT) is a Web‐based surveillance tool that uses Google to explore the searching trends of specific queries.…”