Journal cover Journal topic
Geoscientific Model Development An interactive open-access journal of the European Geosciences Union
Journal topic

Journal metrics

Journal metrics

  • IF value: 4.252 IF 4.252
  • IF 5-year value: 4.890 IF 5-year
  • CiteScore value: 4.49 CiteScore
  • SNIP value: 1.539 SNIP 1.539
  • SJR value: 2.404 SJR 2.404
  • IPP value: 4.28 IPP 4.28
  • h5-index value: 40 h5-index 40
  • Scimago H <br class='hide-on-tablet hide-on-mobile'>index value: 51 Scimago H
    index 51
Discussion papers
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

Development and technical paper 04 Oct 2018

Development and technical paper | 04 Oct 2018

Review status
This discussion paper is a preprint. A revision of the manuscript was accepted for the journal Geoscientific Model Development (GMD).

Application of random forest regression to the calculation of gas-phase chemistry within the GEOS-Chem chemistry model v10

Christoph A. Keller1,2 and Mat J. Evans3,4 Christoph A. Keller and Mat J. Evans
  • 1NASA Global Modeling and Assimilation Office, Goddard Space Flight Center, Greenbelt, MD, USA
  • 2Universities Space Research Association, Columbia, MD, USA
  • 3Wolfson Atmospheric Chemistry Laboratories, Department of Chemistry, University of York, York, YO10 5DD, UK
  • 4National Centre for Atmospheric Sciences, University of York, York, YO10 5DD, UK

Abstract. Atmospheric chemistry models are a central tool to study the impact of chemical constituents on the environment, vegetation and human health. These models are numerically intense, and previous attempts to reduce the numerical cost of chemistry solvers have not delivered transformative change.

We show here the potential of a machine learning (in this case random forest regression) replacement for the gas-phase chemistry in atmospheric chemistry models. Our training data consists of one month (July 2013) of output of chemical conditions together with the model physical state, produced from the GEOS-Chem chemistry model (v10). From this data set we train random forest regression models to predict the concentration of each transported species after the integrator, based on the physical and chemical conditions before the integrator. The choice of prediction type has a strong impact on the skill of the regression model. We find best results from predicting the change in concentration for long-lived species and the absolute concentration for short-lived species. We also find improvements from a simple implementation of chemical families (NOx = NO + NO2).

We then implement the trained random forest predictors back into GEOS-Chem to replace the numerical integrator. The machine learning driven GEOS-Chem model compares well to the standard simulation. For O3, error from using the random forests grow slowly and after 5 days the normalised mean bias (NMB), root mean square error (RMSE) and R2 are 4.2 %, 35 %, 0.9 respectively; after 30 days the errors increase to 13 %, 67 %, 0.75. The biases become largest in remote areas such as the tropical Pacific where errors in the chemistry can accumulate with little balancing influence from emissions or deposition. Over polluted regions the model error is less than 10 % and has significant fidelity in following the time series of the full model. Modelled NOx shows similar features, with the most significant errors occurring in remote locations far from recent emissions. For other species such as inorganic bromine species and short lived nitrogen species errors become large, with NMB, RMSE and R2 reaching >2100 % >400 %, <0.1 respectively.

This proof-of-concept implementation is 85 % slower than the direct integration of the differential equations but optimisation and software engineering would allow substantial increases in speed. We discuss potential improvements in the implementation, some of its advantages from both a software and hardware perspective, its limitations and its applicability to operational air quality activities.

Christoph A. Keller and Mat J. Evans
Interactive discussion
Status: final response (author comments only)
Status: final response (author comments only)
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
[Login for Authors/Topical Editors] [Subscribe to comment alert] Printer-friendly Version - Printer-friendly version Supplement - Supplement
Christoph A. Keller and Mat J. Evans
Christoph A. Keller and Mat J. Evans
Total article views: 618 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
473 136 9 618 12 9
  • HTML: 473
  • PDF: 136
  • XML: 9
  • Total: 618
  • BibTeX: 12
  • EndNote: 9
Views and downloads (calculated since 04 Oct 2018)
Cumulative views and downloads (calculated since 04 Oct 2018)
Viewed (geographical distribution)  
Total article views: 576 (including HTML, PDF, and XML) Thereof 575 with geography defined and 1 with unknown origin.
Country # Views %
  • 1
No saved metrics found.
No discussed metrics found.
Latest update: 20 Feb 2019
Publications Copernicus
Short summary
Atmospheric chemistry models are used to study the impact of air pollutants on the environment. These models are highly complex and require a lot of computing resources. In this study we show that machine learning can be used to simulate atmospheric chemistry with an accuracy that is comparable to the traditional, computationally expensive method. The machine learning model has the potential to be much faster and to provide more detailed air quality forecasts.
Atmospheric chemistry models are used to study the impact of air pollutants on the environment....