Preprints
https://doi.org/10.5194/egusphere-2022-396
https://doi.org/10.5194/egusphere-2022-396
 
28 Jul 2022
28 Jul 2022
Status: this preprint is open for discussion.

Predicting peak daily maximum 8-hour ozone, and linkages to emissions and meteorology, in Southern California using machine learning methods

Ziqi Gao1, Yifeng Wang1, Petros Vasilakos1, Cesunica E. Ivey2,a, Khanh Do2,3, and Armistead Goode Russell1 Ziqi Gao et al.
  • 1School of Civil and Environmental Engineering, Georgia Institute of Technology, Atlanta, GA, 30332, USA
  • 2Department of Chemical and Environmental Engineering, University of California, Riverside, Riverside, CA, USA
  • 3Center for Environmental Research and Technology, Riverside, CA, USA
  • anow at: Department of Civil and Environmental Engineering, University of California, Berkeley, CA, USA

Abstract. The growing abundance of data is conducive to using numerical methods to relate air quality, meteorology, and emissions to address which factors impact pollutant concentrations. Often, it is the extreme values that are of interest for health and regulatory purposes (e.g., the National Ambient Air Quality Standard for ozone uses the annual, maximum, daily 4th highest, 8-hour average (MDA8) ozone), though such values are the most challenging to predict using empirical models. We developed four different computational models, including the Generalized Additive Model (GAM), the Multivariate Adaptive Regression Splines, the Random Forest, and the Support Vector Regression, to develop observation-based relationships between the 4th highest MDA8 ozone in the South Coast Air Basin and precursor emissions, meteorological factors, and large-scale climate patterns. All models had similar predictive performance, though the GAM showed a relatively higher R2 value (0.96) with a lower root mean square error and mean bias.

Ziqi Gao et al.

Status: open (until 22 Sep 2022)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse

Ziqi Gao et al.

Data sets

Predicting peak daily maximum 8-hour ozone, and linkages to emissions and meteorology, in Southern California using machine learning methods Ziqi Gao https://doi.org/10.5281/zenodo.6892062

Model code and software

Predicting peak daily maximum 8-hour ozone, and linkages to emissions and meteorology, in Southern California using machine learning methods Ziqi Gao https://doi.org/10.5281/zenodo.6892066

Ziqi Gao et al.

Viewed

Total article views: 139 (including HTML, PDF, and XML)
HTML PDF XML Total Supplement BibTeX EndNote
105 28 6 139 15 2 3
  • HTML: 105
  • PDF: 28
  • XML: 6
  • Total: 139
  • Supplement: 15
  • BibTeX: 2
  • EndNote: 3
Views and downloads (calculated since 28 Jul 2022)
Cumulative views and downloads (calculated since 28 Jul 2022)

Viewed (geographical distribution)

Total article views: 127 (including HTML, PDF, and XML) Thereof 127 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 08 Aug 2022
Download
Short summary
While the national ambient air quality standard of ozone is based on the 3-year average of the 4th highest 8-hour maximum (MDA8) ozone concentrations, these predicted extreme values using numerical methods are always biased low. We built 4 computational models (GAM, MARS, Random Forest & SVR) to predict the 4th highest MDA8 ozone in Southern California using precursor emissions, meteorology, and climatological patterns. All models presented acceptable performance with GAM being the best.