Predicting daily ragweed pollen concentrations using Computational Intelligence techniques over two heavily polluted areas in Europe

Forecasting daily total pollen concentrations on a global scale

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.

Aerobiological modeling I: A review of predictive models

The present work is the first of two reviews on applied modeling in the field of aerobiology. The aerobiological predictive models for pollen and fungal spores, usually defined as predictive statistical models, will, amongst other objectives, forecast airborne particles' concentration or dynamical behavior of the particles. These models can be classified into Observation Based Models (OBM), Phenological Based Models (PHM), or OTher Models (OTM). The aim of this review is to show, analyze and discuss the different predictive models used in pollen and spore aerobiological studies. The analysis was performed on published electronic scientific articles from 1998 to 2016 related to the type of model, the taxa and the modelled parameters. From a total of 503 studies, 55.5% used OBM (44.8% on pollen and 10.7% on fungal spores), 38.5% PHM (all on pollen) and 6% OTM (5.4% on pollen and 0.6% on fungal spores). OBM have been used with high frequency to forecast concentration. The most frequent model of OBM was linear regression (18.5% out of 503) on pollen and artificial neural networks (4.6%) on fungal spores. In the PHM, the principal use was to characterize the main pollen season (flowering season) based on the model of growth degree days. Finally, OTM have been used to estimate concentrations at unmonitored areas. Olea (14,5%) on pollen and Alternaria (4,8%) on fungal spores were the taxa most frequently modelled. Daily concentration was the most modelled parameter by OBM (25.2%) and season start day by PHM (35.6%). The PHM approaches include greater model diversity and use fewer independent variables than OBM. In addition, PHM show to be easier to apply than OBM; however, the wide range of criteria to define the parameters to use in PHM (e.g.: pollination start day) makes that each model is used with a lesser frequency than other models.

Forecasting Particulate Matter Concentration Using Linear and Non-Linear Approaches for Air Quality Decision Support

Air quality status on the East Coast of Peninsular Malaysia is dominated by Particulate Matter (PM10) throughout the years. Studies have affirmed that PM10 influence human health and the environment. Therefore, precise forecasting algorithms are urgently needed to determine the PM10 status for mitigation plan and early warning purposes. This study investigates the forecasting performance of a linear (Multiple Linear Regression) and two non-linear models (Multi-Layer Perceptron and Radial Basis Function) utilizing meteorological and gaseous pollutants variables as input parameters from the year 2000–2014 at four sites with different surrounding activities of urban, sub-urban and rural areas. Non-linear model (Radial Basis Function) outperforms the linear model with the error reduced by 78.9% (urban), 32.1% (sub-urban) and 39.8% (rural). Association between PM10 and its contributing factors are complex and non-linear in nature, best captured by an Artificial Neural Network, which generates more accurate PM10 compared to the linear model. The results are robust enough for precise next day forecasting of PM10 concentration on the East Coast of Peninsular Malaysia.

Google Trends and pollen concentrations in allergy and airway diseases in France

BACKGROUND

Google Trends (GTs) is a web-based surveillance tool that explores the searching trends of specific queries via Google. This tool proposes to reflect the real-life epidemiology of allergic rhinitis and asthma. However, the validation of GTs against pollen concentrations is missing at the country level.

OBJECTIVES

In the present study, we used GTs (a) to compare the terms related to allergy in France, (b) to assess seasonal variations across the country for 5 years and (c) to compare GTs and pollen concentrations for 2016.

METHODS

Google Trends queries were initially searched to investigate the terms reflecting pollen and allergic diseases. 13- and 5-year GTs were used in France. Then, 5-year GTs were assessed in all metropolitan French regions to assess the seasonality of GTs. Finally, GTs were compared with pollen concentrations (Réseau National de Surveillance en Aerobiology) for 2016 in seven regions (GTs) and corresponding cities (pollen concentrations).

RESULTS

The combination of searches for "allergy" as a disease, "pollen" as a disease cause and "ragweed" as a plant was needed to fully assess the pollen season in France. "Asthma" did not show any seasonality. Using the 5-year GTs, an annual and clear seasonality of queries was found in all regions depending on the predicted pollen exposure for spring and a summer peak but not for winter peaks. The agreement between GT queries and pollen concentrations is usually poor except for spring trees and grasses. Moreover, cypress pollens are insufficiently reported by GTs.

CONCLUSIONS

Google Trends cannot predict the pollen season in France.

Relationship of NDVI and oak (Quercus) pollen including a predictive model in the SW Mediterranean region

Techniques of remote sensing are being used to develop phenological studies. Our goal is to study the correlation among the Normalized Difference Vegetation Index (NDVI) related with oak trees included in three set data polygons (15, 25 and 50 km to aerobiological sampling point as NDVI-15, 25 and 50), and oak (Quercus) daily average pollen counts from 1994 to 2013. The study was developed in the SW Mediterranean region with continuous pollen recording within the mean pollen season of each studied year. These pollen concentrations were compared with NDVI values in the locations containing the vegetation under a study based on two cartographic sources: the Extremadura Forest Map (MFEx) of Spain and the Fifth National Forest Inventory (IFN5) from Portugal. The importance of this work is to propose the relationship among data related in space and time by Spearman and Granger causality tests. 9 out of 20 studied years have shown significant results with the Granger causality test between NDVI and pollen concentration, and in 12 years, significant values were obtained by Spearman test. The distances of influence on the contribution of Quercus pollen to the sampler showed statistically significant results depending on the year. Moreover, a predictive model by using Artificial Neural Network (ANN) was applied with better results in NDVI25 than for NDVI15 or NDVI50. The addition of NDVI25 with the lag of 5 days and some weather parameters in the model was applied with a RMSE of 4.26 (Spearman coefficient r = 0.77) between observed and predicted values. Based on these results, NDVI seems to be a useful parameter to predict airborne pollen.

Applying machine learning to forecast daily Ambrosia pollen using environmental and NEXRAD parameters

Approximately 50 million Americans have allergic diseases. Airborne plant pollen is a significant trigger for several of these allergic diseases. Ambrosia (ragweed) is known for its abundant production of pollen and its potent allergic effect in North America. Hence, estimating and predicting the daily atmospheric concentration of pollen (ragweed pollen in particular) is useful for both people with allergies and for the health professionals who care for them. In this study, we show that a suite of variables including meteorological and land surface parameters, as well as next-generation radar (NEXRAD) measurements together with machine learning can be used to estimate successfully the daily pollen concentration. The supervised machine learning approaches we used included random forests, neural networks, and support vector machines. The performance of the training is independently validated using 10% of the data partitioned using the holdout cross-validation method from the original dataset. The random forests (R= 0.61, R²= 0.37), support vector machines (R= 0.51, R²= 0.26), and neural networks (R= 0.46, R²= 0.21) effectively predicted the daily Ambrosia pollen, where the correlation coefficient (R) and R-squared (R²) values are given in brackets. Three independent approaches-the random forests, correlation coefficients, and interaction information-were employed to rank the relative importance of the available predictors.

Estimating the daily pollen concentration in the atmosphere using machine learning and NEXRAD weather radar data

Millions of people have an allergic reaction to pollen. The impact of pollen allergies is on the rise due to increased pollen levels caused by global warming and the spread of highly invasive weeds. The production, release, and dispersal of pollen depend on the ambient weather conditions. The temperature, rainfall, humidity, cloud cover, and wind are known to affect the amount of pollen in the atmosphere. In the past, various regression techniques have been applied to estimate and forecast the daily pollen concentration in the atmosphere based on the weather conditions. In this research, machine learning methods were applied to the Next Generation Weather Radar (NEXRAD) data to estimate the daily Ambrosia pollen over a 300 km × 300 km region centered on a NEXRAD weather radar. The Neural Network and Random Forest machine learning methods have been employed to develop separate models to estimate Ambrosia pollen over the region. A feasible way of estimating the daily pollen concentration using only the NEXRAD radar data and machine learning methods would lay the foundation to forecast daily pollen at a fine spatial resolution nationally.

Applying Deep Neural Networks and Ensemble Machine Learning Methods to Forecast Airborne Ambrosia Pollen

Allergies to airborne pollen are a significant issue affecting millions of Americans. Consequently, accurately predicting the daily concentration of airborne pollen is of significant public benefit in providing timely alerts. This study presents a method for the robust estimation of the concentration of airborne Ambrosia pollen using a suite of machine learning approaches including deep learning and ensemble learners. Each of these machine learning approaches utilize data from the European Centre for Medium-Range Weather Forecasts (ECMWF) atmospheric weather and land surface reanalysis. The machine learning approaches used for developing a suite of empirical models are deep neural networks, extreme gradient boosting, random forests and Bayesian ridge regression methods for developing our predictive model. The training data included twenty-four years of daily pollen concentration measurements together with ECMWF weather and land surface reanalysis data from 1987 to 2011 is used to develop the machine learning predictive models. The last six years of the dataset from 2012 to 2017 is used to independently test the performance of the machine learning models. The correlation coefficients between the estimated and actual pollen abundance for the independent validation datasets for the deep neural networks, random forest, extreme gradient boosting and Bayesian ridge were 0.82, 0.81, 0.81 and 0.75 respectively, showing that machine learning can be used to effectively forecast the concentrations of airborne pollen.

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

Allergic rhinitis (AR) is impacted by allergens and air pollution but interactions between air pollution, sleep and allergic diseases are insufficiently understood. POLLAR (Impact of air POLLution on sleep, Asthma and Rhinitis) is a project of the European Institute of Innovation and Technology (EIT Health). It will use a freely-existing application for AR monitoring that has been tested in 23 countries (the Allergy Diary, iOS and Android, 17,000 users, TLR8). The Allergy Diary will be combined with a new tool allowing queries on allergen, pollen (TLR2), sleep quality and disorders (TRL2) as well as existing longitudinal and geolocalized pollution data. Machine learning will be used to assess the relationship between air pollution, sleep and AR comparing polluted and non-polluted areas in 6 EU countries. Data generated in 2018 will be confirmed in 2019 and extended by the individual prospective assessment of pollution (portable sensor, TLR7) in AR. Sleep apnea patients will be used as a demonstrator of sleep disorder that can be modulated in terms of symptoms and severity by air pollution and AR. The geographic information system GIS will map the results. Consequences on quality of life (EQ-5D), asthma, school, work and sleep will be monitored and disseminated towards the population. The impacts of POLLAR will be (1) to propose novel care pathways integrating pollution, sleep and patients' literacy, (2) to study sleep consequences of pollution and its impact on frequent chronic diseases, (3) to improve work productivity, (4) to propose the basis for a sentinel network at the EU level for pollution and allergy, (5) to assess the societal implications of the interaction. MASK paper N°32.

Differences in Reporting the Ragweed Pollen Season Using Google Trends across 15 Countries

BACKGROUND

Google Trends (GT) searches trends of specific queries in Google, which potentially reflect the real-life epidemiology of allergic rhinitis. We compared GT terms related to ragweed pollen allergy in American and European Union countries with a known ragweed pollen season. Our aim was to assess seasonality and the terms needed to perform the GT searches and to compare these during the spring and summer pollen seasons.

METHODS

We examined GT queries from January 1, 2011, to January 4, 2017. We included 15 countries with a known ragweed pollen season and used the standard 5-year GT graphs. We used the GT translation for all countries and the untranslated native terms for each country.

RESULTS

The results of "pollen," "ragweed," and "allergy" searches differed between countries, but "ragweed" was clearly identified in 12 of the 15 countries. There was considerable heterogeneity of findings when the GT translation was used. For Croatia, Hungary, Romania, Serbia, and Slovenia, the GT translation was inappropriate. The country patterns of "pollen," "hay fever," and "allergy" differed in 8 of the 11 countries with identified "ragweed" queries during the spring and the summer, indicating that the perception of tree and grass pollen allergy differs from that of ragweed pollen.

CONCLUSIONS

To investigate ragweed pollen allergy using GT, the term "ragweed" as a plant is required and the translation of "ragweed" in the native language needed.

Forecasting methodologies for Ganoderma spore concentration using combined statistical approaches and model evaluations

High concentration levels of Ganoderma spp. spores were observed in Worcester, UK, during 2006-2010. These basidiospores are known to cause sensitization due to the allergen content and their small dimensions. This enables them to penetrate the lower part of the respiratory tract in humans. Establishment of a link between occurring symptoms of sensitization to Ganoderma spp. and other basidiospores is challenging due to lack of information regarding spore concentration in the air. Hence, aerobiological monitoring should be conducted, and if possible extended with the construction of forecast models. Daily mean concentration of allergenic Ganoderma spp. spores in the atmosphere of Worcester was measured using 7-day volumetric spore sampler through five consecutive years. The relationships between the presence of spores in the air and the weather parameters were examined. Forecast models were constructed for Ganoderma spp. spores using advanced statistical techniques, i.e. multivariate regression trees and artificial neural networks. Dew point temperature along with maximum temperature was the most important factor influencing the presence of spores in the air of Worcester. Based on these two major factors and several others of lesser importance, thresholds for certain levels of fungal spore concentration, i.e. low (0-49 s m(-3)), moderate (50-99 s m(-3)), high (100-149 s m(-3)) and very high (150 < n s m(-3)), could be designated. Despite some deviation in results obtained by artificial neural networks, authors have achieved a forecasting model, which was accurate (correlation between observed and predicted values varied from r s = 0.57 to r s = 0.68).

Airborne castanea pollen forecasting model for ecological and allergological implementation

Castanea sativa Miller belongs to the natural vegetation of many European deciduous forests prompting impacts in the forestry, ecology, allergological and chestnut food industry fields. The study of the Castanea flowering represents an important tool for evaluating the ecological conservation of North-Western Spain woodland and the possible changes in the chestnut distribution due to recent climatic change. The Castanea pollen production and dispersal capacity may cause hypersensitivity reactions in the sensitive human population due to the relationship between patients with chestnut pollen allergy and a potential cross reactivity risk with other pollens or plant foods. In addition to Castanea pollen's importance as a pollinosis agent, its study is also essential in North-Western Spain due to the economic impact of the industry around the chestnut tree cultivation and its beekeeping interest. The aim of this research is to develop an Artificial Neural Networks for predict the Castanea pollen concentration in the atmosphere of the North-West Spain area by means a 20years data set. It was detected an increasing trend of the total annual Castanea pollen concentrations in the atmosphere during the study period. The Artificial Neural Networks (ANNs) implemented in this study show a great ability to predict Castanea pollen concentration one, two and three days ahead. The model to predict the Castanea pollen concentration one day ahead shows a high linear correlation coefficient of 0.784 (individual ANN) and 0.738 (multiple ANN). The results obtained improved those obtained by the classical methodology used to predict the airborne pollen concentrations such as time series analysis or other models based on the correlation of pollen levels with meteorological variables.

Ragweed-induced allergic rhinoconjunctivitis: current and emerging treatment options

Ragweed (Ambrosia spp.) is an annually flowering plant whose pollen bears high allergenic potential. Ragweed-induced allergic rhinoconjunctivitis has long been seen as a major immunologic condition in Northern America with high exposure and sensitization rates in the general population. The invasive occurrence of ragweed (A. artemisiifolia) poses an increasing challenge to public health in Europe and Asia as well. Possible explanations for its worldwide spread are climate change and urbanization, as well as pollen transport over long distances by globalized traffic and winds. Due to the increasing disease burden worldwide, and to the lack of a current and comprehensive overview, this study aims to review the current and emerging treatment options for ragweed-induced rhinoconjunctivitis. Sound clinical evidence is present for the symptomatic treatment of ragweed-induced allergic rhinoconjunctivitis with oral third-generation H1-antihistamines and leukotriene antagonists. The topical application of glucocorticoids has also been efficient in randomized controlled clinical trials. Combined approaches employing multiple agents are common. The mainstay of causal treatment to date, especially in Northern America, is subcutaneous immunotherapy with the focus on the major allergen, Amb a 1. Beyond this, growing evidence from several geographical regions documents the benefit of sublingual immunotherapy. Future treatment options promise more specific symptomatic treatment and fewer side effects during causal therapy. Novel antihistamines for symptomatic treatment are aimed at the histamine H3-receptor. New adjuvants with toll-like receptor 4 activity or the application of the monoclonal anti-immunoglobulin E antibody, omalizumab, are supposed to enhance conventional immunotherapy. An approach targeting toll-like receptor 9 by synthetic cytosine phosphate–guanosine oligodeoxynucleotides promises a new treatment paradigm that aims to modulate the immune response, but it has yet to be proven in clinical trials.