AQ-Bench: A Benchmark Dataset for Machine Learning on Global
Air Quality Metrics

Betancourt, Clara; Stomberg, Timo T.; Stadtler, Scarlet; Roscher, Ribana; Schultz, Martin

doi:10.5194/essd-2020-380

Cited by 3 publications

(5 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The prediction method of O3 concentrations based on machine learning mainly relies on the advantage that the machine learning method can effectively capture the hidden nonlinear characteristics in the change in atmospheric composition and can build a prediction model of atmospheric composition through characteristic variables (Mo et al, 2019;Aljanabi et al, 2020;Amato et al, 2020;Betancourt et al, 2021). Machine learning models trained with data from observations or physical models can produce reliable simulations without intensive high-end computing (Ojha et al, 2021).…”

Section: A C C E P T E D Mmentioning

confidence: 99%

A Machine-Learning-Based Classification Method for Meteorological Conditions of Ozone Pollution

Cao

Zhao²,

et al. 2023

Aerosol Air Qual. Res.

View full text Add to dashboard Cite

Ozone pollution is harmful to human health and ecosystem, which occurs in ecosystems and has occurred frequently in China in recent years, especially during the warm seasons.Meteorological conditions are among the important factors affecting the occurrence of ozone pollution. In this study, a classification method for meteorological conditions of ozone pollution levels based on a back propagation (BP) neural network was proposed to reflect the impact of meteorological conditions on the occurrence of ozone pollution. Ozone pollution was divided into three levels according to surface hourly ozone (O3) concentrations and thus into three groups of meteorological conditions. The input physical parameters for the BP neural network were determined by evaluating the relationship between surface O3 concentrations and meteorological parameters and precursors, including relative humidity, temperature, mixing layer height, precipitation, and nitrogen dioxide (NO2) concentrations. The study area focused on 21 cities in Sichuan Province in southwestern China, which was divided into 12 BP classifiers according to the urban geographical location and sample number of each city, and a single BP classifier was trained for 21 cities. The classification results of the trained BP classifiers were verified by comparison to the observations. With 12 individual BP classifiers, the classification accuracy of all 21 cities was more than 60%, of which 18 cities were more than 70%, and 9 cities were more than 80%. With the single BP classifier, the classification accuracy of 20 cities was more than 60%, of which 18 cities were more than 70%, and 14 cities were more than 80%. Overall, the classification performance of the trained single model was better than trained 12 individual models. The classification method can comprehensively reflect the impact of meteorological conditions on the occurrence of ozone pollution.

show abstract

Section: A C C E P T E D Mmentioning

confidence: 99%

A Machine-Learning-Based Classification Method for Meteorological Conditions of Ozone Pollution

Cao

Zhao²,

et al. 2023

Aerosol Air Qual. Res.

View full text Add to dashboard Cite

show abstract

“…The AQ-Bench dataset is available in .csv format at http://doi.org/10.23728/b2share. 30d42b5a87344e82855a486bf2123e9f (Betancourt et al, 2020). To enable a machine learning quick start on the AQ-Bench dataset with reproduction of the baseline experiments, we also provide an introductory Jupyter notebook on https:// gitlab.version.fz-juelich.de/esde/machine-learning/aq-bench (Betancourt et al, 2021).…”

Section: Data and Code Availabilitymentioning

confidence: 99%

“…Our ready-to-use, fully documented dataset is freely available under the DOI https://doi.org/10.23728/b2share. 30d42b5a87344e82855a486bf2123e9f (Betancourt et al, 2020). We also provide our baseline machine learning code at https://gitlab.version.fz-juelich.de/esde/machine-learning/ aq-bench (Betancourt et al, 2021), offering a low-threshold entrance to machine learning in environmental science within a relevant research topic.…”

Section: Introductionmentioning

confidence: 99%

“…It enables them to start with a real-world problem relevant to humans and nature. The dataset and introductory machine learning code are available at https://doi.org/10.23728/b2share.30d42b5a87344e82855a486bf2123e9f (Betancourt et al, 2020) and https://gitlab.version.fz-juelich.de/esde/machine-learning/aq-bench (Betancourt et al, 2021). AQ-Bench thus provides a blueprint for environmental benchmark datasets as well as an example for data re-use according to the FAIR principles.…”

mentioning

confidence: 99%

See 1 more Smart Citation

AQ-Bench: a benchmark dataset for machine learning on global air quality metrics

Betancourt¹,

Stomberg

Roscher

et al. 2021

Earth Syst. Sci. Data

Self Cite

View full text Add to dashboard Cite

Abstract. With the AQ-Bench dataset, we contribute to the recent developments towards shared data usage and machine learning methods in the field of environmental science. The dataset presented here enables researchers to relate global air quality metrics to easy-access metadata and to explore different machine learning methods for obtaining estimates of air quality based on this metadata. AQ-Bench contains a unique collection of aggregated air quality data from the years 2010–2014 and metadata at more than 5500 air quality monitoring stations all over the world, provided by the first Tropospheric Ozone Assessment Report (TOAR). It focuses in particular on metrics of tropospheric ozone, which has a detrimental effect on climate, human morbidity and mortality, as well as crop yields. The purpose of this dataset is to produce estimates of various long-term ozone metrics based on time-independent local site conditions. We combine this task with a suitable evaluation metric. Baseline scores obtained from a linear regression method, a fully connected neural network and random forest are provided for reference and validation. AQ-Bench offers a low-threshold entrance for all machine learners with an interest in environmental science and for atmospheric scientists who are interested in applying machine learning techniques. It enables them to start with a real-world problem relevant to humans and nature. The dataset and introductory machine learning code are available at https://doi.org/10.23728/b2share.30d42b5a87344e82855a486bf2123e9f (Betancourt et al., 2020) and https://gitlab.version.fz-juelich.de/esde/machine-learning/aq-bench (Betancourt et al., 2021). AQ-Bench thus provides a blueprint for environmental benchmark datasets as well as an example for data re-use according to the FAIR principles.

show abstract

“…Such CTMs can be combined with higher resolution dispersion models to provide local air quality levels in street canyons (Gidhagen et al, 2021 ). More recently, statistical methods, such as Machine Learning (ML) (Barré et al, 2021 ; Betancourt et al, 2021 ; Lovrić et al, 2021 ; Nitheesh et al, 2021 ; Rybarczyk and Zalakeviciute, 2021 ), have proven their efficiency and reliability in predicting the concentrations of pollutants in the atmosphere.…”

Section: Introductionmentioning

confidence: 99%

Deep Learning Approach for Assessing Air Quality During COVID-19 Lockdown in Quito

et al. 2022

View full text Add to dashboard Cite

Weather Normalized Models (WNMs) are modeling methods used for assessing air contaminants under a business-as-usual (BAU) assumption. Therefore, WNMs are used to assess the impact of many events on urban pollution. Recently, different approaches have been implemented to develop WNMs and quantify the lockdown effects of COVID-19 on air quality, including Machine Learning (ML). However, more advanced methods, such as Deep Learning (DL), have never been applied for developing WNMs. In this study, we proposed WNMs based on DL algorithms, aiming to test five DL architectures and compare their performances to a recent ML approach, namely Gradient Boosting Machine (GBM). The concentrations of five air pollutants (CO, NO2, PM2.5, SO2, and O3) are studied in the city of Quito, Ecuador. The results show that Long-Short Term Memory (LSTM) and Bidirectional Recurrent Neural Network (BiRNN) outperform the other algorithms and, consequently, are recommended as appropriate WNMs to quantify the effects of the lockdowns on air pollution. Furthermore, examining the variable importance in the LSTM and BiRNN models, we identify that the most relevant temporal and meteorological features for predicting air quality are Hours (time of day), Index (1 is the first collected data and increases by one after each instance), Julian Day (day of the year), Relative Humidity, Wind Speed, and Solar Radiation. During the full lockdown, the concentration of most pollutants has decreased drastically: −48.75%, for CO, −45.76%, for SO2, −42.17%, for PM2.5, and −63.98%, for NO2. The reduction of this latter gas has induced an increase of O3 by +26.54%.

show abstract

AQ-Bench: A Benchmark Dataset for Machine Learning on Global Air Quality Metrics

Cited by 3 publications

References 3 publications

A Machine-Learning-Based Classification Method for Meteorological Conditions of Ozone Pollution

A Machine-Learning-Based Classification Method for Meteorological Conditions of Ozone Pollution

AQ-Bench: a benchmark dataset for machine learning on global air quality metrics

Deep Learning Approach for Assessing Air Quality During COVID-19 Lockdown in Quito

Contact Info

Product

Resources

About