A Machine Learning Approach to Investigate the Surface Ozone Behavior
The concentration of surface ozone (O<sub>3</sub>) strongly depends on environmental and meteorological variables through a series of complex and non-linear functions. This study aims to explore the performances of an advanced machine learning (ML) method, the boosted regression trees (B...
| Главные авторы: | , |
|---|---|
| Формат: | Статья |
| Язык: | English |
| Опубликовано: |
MDPI AG
2020-10-01
|
| Серии: | Atmosphere |
| Предметы: | |
| Online-ссылка: | https://www.mdpi.com/2073-4433/11/11/1173 |
| _version_ | 1827703177587195904 |
|---|---|
| author | Roberta Valentina Gagliardi Claudio Andenna |
| author_facet | Roberta Valentina Gagliardi Claudio Andenna |
| author_sort | Roberta Valentina Gagliardi |
| collection | DOAJ |
| description | The concentration of surface ozone (O<sub>3</sub>) strongly depends on environmental and meteorological variables through a series of complex and non-linear functions. This study aims to explore the performances of an advanced machine learning (ML) method, the boosted regression trees (BRT) technique, in exploring the relationships between surface O<sub>3</sub> and its driving factors, and in predicting the levels of O<sub>3</sub> concentrations. To this end, a BRT model was trained on hourly data of air pollutants and meteorological parameters, acquired, over the 2016–2018 period, in a rural area affected by an anthropic source of air pollutants. The abilities of the BRT model in ranking, visualizing, and predicting the relationship between ground-level O<sub>3</sub> concentrations and its driving factors were analyzed and illustrated. A comparison with a multiple linear regression (MLR) model was performed based on several statistical indicators. The results obtained indicated that the BRT model was able to account for 81% of changes in O<sub>3</sub> concentrations; it slightly outperforms the MLR model in terms of the predictions accuracy and allows a better identification of the main factors influencing O<sub>3</sub> variability on a local scale. This knowledge is expected to be useful in defining effective measures to prevent and/or mitigate the health damages associated with O<sub>3</sub> exposure. |
| first_indexed | 2024-03-10T15:12:44Z |
| format | Article |
| id | doaj.art-2e6f57135f2040d4b0f83a1851c84125 |
| institution | Directory Open Access Journal |
| issn | 2073-4433 |
| language | English |
| last_indexed | 2024-03-10T15:12:44Z |
| publishDate | 2020-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Atmosphere |
| spelling | doaj.art-2e6f57135f2040d4b0f83a1851c841252023-11-20T19:08:46ZengMDPI AGAtmosphere2073-44332020-10-011111117310.3390/atmos11111173A Machine Learning Approach to Investigate the Surface Ozone BehaviorRoberta Valentina Gagliardi0Claudio Andenna1Istituto Superiore di Sanità, Viale Regina Elena 299, 00161 Rome, ItalyINAIL-DIT, Via del Torraccio di Torrenova 7, 00133 Rome, ItalyThe concentration of surface ozone (O<sub>3</sub>) strongly depends on environmental and meteorological variables through a series of complex and non-linear functions. This study aims to explore the performances of an advanced machine learning (ML) method, the boosted regression trees (BRT) technique, in exploring the relationships between surface O<sub>3</sub> and its driving factors, and in predicting the levels of O<sub>3</sub> concentrations. To this end, a BRT model was trained on hourly data of air pollutants and meteorological parameters, acquired, over the 2016–2018 period, in a rural area affected by an anthropic source of air pollutants. The abilities of the BRT model in ranking, visualizing, and predicting the relationship between ground-level O<sub>3</sub> concentrations and its driving factors were analyzed and illustrated. A comparison with a multiple linear regression (MLR) model was performed based on several statistical indicators. The results obtained indicated that the BRT model was able to account for 81% of changes in O<sub>3</sub> concentrations; it slightly outperforms the MLR model in terms of the predictions accuracy and allows a better identification of the main factors influencing O<sub>3</sub> variability on a local scale. This knowledge is expected to be useful in defining effective measures to prevent and/or mitigate the health damages associated with O<sub>3</sub> exposure.https://www.mdpi.com/2073-4433/11/11/1173surface ozonemonthly-daily variationsmachine learningboosted regression treesprecursorsmeteorological parameters |
| spellingShingle | Roberta Valentina Gagliardi Claudio Andenna A Machine Learning Approach to Investigate the Surface Ozone Behavior Atmosphere surface ozone monthly-daily variations machine learning boosted regression trees precursors meteorological parameters |
| title | A Machine Learning Approach to Investigate the Surface Ozone Behavior |
| title_full | A Machine Learning Approach to Investigate the Surface Ozone Behavior |
| title_fullStr | A Machine Learning Approach to Investigate the Surface Ozone Behavior |
| title_full_unstemmed | A Machine Learning Approach to Investigate the Surface Ozone Behavior |
| title_short | A Machine Learning Approach to Investigate the Surface Ozone Behavior |
| title_sort | machine learning approach to investigate the surface ozone behavior |
| topic | surface ozone monthly-daily variations machine learning boosted regression trees precursors meteorological parameters |
| url | https://www.mdpi.com/2073-4433/11/11/1173 |
| work_keys_str_mv | AT robertavalentinagagliardi amachinelearningapproachtoinvestigatethesurfaceozonebehavior AT claudioandenna amachinelearningapproachtoinvestigatethesurfaceozonebehavior AT robertavalentinagagliardi machinelearningapproachtoinvestigatethesurfaceozonebehavior AT claudioandenna machinelearningapproachtoinvestigatethesurfaceozonebehavior |