Local integrated air quality predictions from meteorology (2015 to 2020) with machine and deep learning assisted by data mining
Overall air quality local indices can usefully be established by combining normalised values of common individual pollutant values. This reveals distinctive seasonal trends that are strongly influenced by local meteorological conditions. A newly compiled dataset for 2015 to 2020 covering Dallas Coun...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2022-01-01
|
| Series: | Sustainability Analytics and Modeling |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2667259621000023 |
| _version_ | 1797753894923665408 |
|---|---|
| author | David A. Wood |
| author_facet | David A. Wood |
| author_sort | David A. Wood |
| collection | DOAJ |
| description | Overall air quality local indices can usefully be established by combining normalised values of common individual pollutant values. This reveals distinctive seasonal trends that are strongly influenced by local meteorological conditions. A newly compiled dataset for 2015 to 2020 covering Dallas County (USA), combining six pollutants into a combined local area benchmark (CLAB), is assessed in terms of eleven meteorological variables. It is possible to distinguish the effects of lock-down induced impacts in the CLAB index and some of its component pollutants during 2020. Nine machine learning and three deep learning algorithms are compared in their abilities to predict CLAB from the meteorological variables on supervised and unseen bases. Prediction results for 2019 and 2020 are distinctive for annual and quarterly timeframes. In-depth prediction outlier analysis using a transparent data-matching algorithm provides insight to the few data records for which CLAB is not accurately predicted from ground-level meteorological data. |
| first_indexed | 2024-03-12T17:24:30Z |
| format | Article |
| id | doaj.art-777705961fe0464e94a531dae127e9be |
| institution | Directory Open Access Journal |
| issn | 2667-2596 |
| language | English |
| last_indexed | 2024-03-12T17:24:30Z |
| publishDate | 2022-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Sustainability Analytics and Modeling |
| spelling | doaj.art-777705961fe0464e94a531dae127e9be2023-08-05T05:18:11ZengElsevierSustainability Analytics and Modeling2667-25962022-01-012100002Local integrated air quality predictions from meteorology (2015 to 2020) with machine and deep learning assisted by data miningDavid A. WoodOverall air quality local indices can usefully be established by combining normalised values of common individual pollutant values. This reveals distinctive seasonal trends that are strongly influenced by local meteorological conditions. A newly compiled dataset for 2015 to 2020 covering Dallas County (USA), combining six pollutants into a combined local area benchmark (CLAB), is assessed in terms of eleven meteorological variables. It is possible to distinguish the effects of lock-down induced impacts in the CLAB index and some of its component pollutants during 2020. Nine machine learning and three deep learning algorithms are compared in their abilities to predict CLAB from the meteorological variables on supervised and unseen bases. Prediction results for 2019 and 2020 are distinctive for annual and quarterly timeframes. In-depth prediction outlier analysis using a transparent data-matching algorithm provides insight to the few data records for which CLAB is not accurately predicted from ground-level meteorological data.http://www.sciencedirect.com/science/article/pii/S2667259621000023Air pollution integrated assessmentLocal meteorological influences on air quality2020 COVID-19-related air quality effectsDaily-averaged air quality trendsEffective data mining algorithmsMachine versus deep learning |
| spellingShingle | David A. Wood Local integrated air quality predictions from meteorology (2015 to 2020) with machine and deep learning assisted by data mining Sustainability Analytics and Modeling Air pollution integrated assessment Local meteorological influences on air quality 2020 COVID-19-related air quality effects Daily-averaged air quality trends Effective data mining algorithms Machine versus deep learning |
| title | Local integrated air quality predictions from meteorology (2015 to 2020) with machine and deep learning assisted by data mining |
| title_full | Local integrated air quality predictions from meteorology (2015 to 2020) with machine and deep learning assisted by data mining |
| title_fullStr | Local integrated air quality predictions from meteorology (2015 to 2020) with machine and deep learning assisted by data mining |
| title_full_unstemmed | Local integrated air quality predictions from meteorology (2015 to 2020) with machine and deep learning assisted by data mining |
| title_short | Local integrated air quality predictions from meteorology (2015 to 2020) with machine and deep learning assisted by data mining |
| title_sort | local integrated air quality predictions from meteorology 2015 to 2020 with machine and deep learning assisted by data mining |
| topic | Air pollution integrated assessment Local meteorological influences on air quality 2020 COVID-19-related air quality effects Daily-averaged air quality trends Effective data mining algorithms Machine versus deep learning |
| url | http://www.sciencedirect.com/science/article/pii/S2667259621000023 |
| work_keys_str_mv | AT davidawood localintegratedairqualitypredictionsfrommeteorology2015to2020withmachineanddeeplearningassistedbydatamining |