Journal cover Journal topic
Atmospheric Chemistry and Physics An interactive open-access journal of the European Geosciences Union
Journal topic

Journal metrics

Journal metrics

  • IF value: 5.668 IF 5.668
  • IF 5-year value: 6.201 IF 5-year
    6.201
  • CiteScore value: 6.13 CiteScore
    6.13
  • SNIP value: 1.633 SNIP 1.633
  • IPP value: 5.91 IPP 5.91
  • SJR value: 2.938 SJR 2.938
  • Scimago H <br class='hide-on-tablet hide-on-mobile'>index value: 174 Scimago H
    index 174
  • h5-index value: 87 h5-index 87
Discussion papers
https://doi.org/10.5194/acp-2019-753
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/acp-2019-753
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

Submitted as: research article 02 Oct 2019

Submitted as: research article | 02 Oct 2019

Review status
This discussion paper is a preprint. It is a manuscript under review for the journal Atmospheric Chemistry and Physics (ACP).

Improving the prediction of an atmospheric chemistry transport model using gradient boosted regression trees

Peter D. Ivatt1,2 and Mathew J. Evans1,2 Peter D. Ivatt and Mathew J. Evans
  • 1Wolfson Atmospheric Chemistry Laboratories, Department of Chemistry, University of York, York, YO10 5DD, UK
  • 2National Centre for Atmospheric Science, Department of Chemistry, University of York, York, YO10 5DD, UK

Abstract. Predictions from process-based models of environmental systems are biased, due to uncertainties in their inputs and parameterisations, reducing their utility. We develop a predictor for the bias in tropospheric ozone (a key pollutant) calculated by an atmospheric chemistry transport model (GEOS-Chem), based on outputs from the model and observations of ozone from both the surface (EPA, EMEP and GAW) and the ozone-sonde networks. We train a gradient-boosted decision tree algorithm (XGBoost) to predict model bias, with model and observational data for 2010–2015, and then test the approach using the years 2016–2017. We show that the bias-corrected model performs significantly better than the uncorrected model. The root mean square error is reduced from from 16.21 ppb to 7.48 ppb, the normalised mean bias is reduced from 0.28 to −0.04, and the Pearson's R is increased from 0.479 to 0.841. Comparisons with observations from the NASA ATom flights (which were not included in the training) also show improvements but to a smaller extent reducing the RMSE from 12.11 ppb to 10.50 ppb, the NMB from 0.08 to 0.06 and increasing the Pearson's R from 0.761 to 0.792. We attribute the smaller improvements to the lack of routine observational constraints of the remote troposphere. We explore the choice of predictor (bias prediction versus direct prediction) and conclude both may have utility. We show that the method is robust to variations in the volume of training data, with approximately a year of data needed to produce useful performance. Data denial experiments (removing observational sites from the algorithm training) shows that information from one location (for example Europe) can reduce the model bias over other locations (for example North America) which might provide insights into the processes controlling the model bias. We conclude that combining machine learning approaches with process based models may provide a useful tool for improving performance of air quality forecasts or to provide enhanced assessments of the impact of pollutants on human and ecosystem health, and may have utility in other environmental applications.

Peter D. Ivatt and Mathew J. Evans
Interactive discussion
Status: open (until 27 Nov 2019)
Status: open (until 27 Nov 2019)
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
[Subscribe to comment alert] Printer-friendly Version - Printer-friendly version Supplement - Supplement
Peter D. Ivatt and Mathew J. Evans
Peter D. Ivatt and Mathew J. Evans
Viewed  
Total article views: 331 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
252 76 3 331 3 3
  • HTML: 252
  • PDF: 76
  • XML: 3
  • Total: 331
  • BibTeX: 3
  • EndNote: 3
Views and downloads (calculated since 02 Oct 2019)
Cumulative views and downloads (calculated since 02 Oct 2019)
Viewed (geographical distribution)  
Total article views: 285 (including HTML, PDF, and XML) Thereof 280 with geography defined and 5 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Cited  
Saved  
No saved metrics found.
Discussed  
No discussed metrics found.
Latest update: 20 Oct 2019
Publications Copernicus
Download
Short summary
We investigate the potential of using a decision tree algorithm to identify and correct the tropospheric ozone bias in a chemical transport model. We train the algorithm on 2010–2015 ground and column observation data, and test the algorithm on the 2016–2017 data using the ground data as well as independent flight data. We find the algorithm is successfully able to identify and correct the bias, improving the model performance which could improve the performance of air quality forecasts.
We investigate the potential of using a decision tree algorithm to identify and correct the...
Citation