20 Oct 2022
20 Oct 2022
Status: this preprint is open for discussion.

Accelerating models for multiphase chemical kinetics through machine learning with polynomial chaos expansion and neural networks

Thomas Berkemeier1, Matteo Krüger1,, Aryeh Feinberg2,3,4,a,, Marcel Müller2,, Ulrich Pöschl1, and Ulrich K. Krieger2 Thomas Berkemeier et al.
  • 1Max Planck Institute for Chemistry, Hahn-Meitner-Weg 1, 55128 Mainz, Germany
  • 2Institute for Atmospheric and Climate Science, ETH Zürich, 8092 Zürich, Switzerland
  • 3Institute of Biogeochemistry and Pollutant Dynamics, ETH Zürich, 8092 Zürich, Switzerland
  • 4Eawag, Swiss Federal Institute of Aquatic Science and Technology, 8600 Dübendorf, Switzerland
  • acurrently at: Institute for Data, Systems, and Society, Massachusetts Institute of Technology, 02142 Cambridge, MA, USA
  • These authors contributed equally to this work.

Abstract. The heterogeneous chemistry of atmospheric aerosols involves multiphase chemical kinetics that can be described by kinetic multi-layer models (KM) explicitly resolving mass transport and chemical reaction. However, KM are computationally too expensive to be used as sub-modules in large-scale atmospheric models, and the computational costs also limit their utility in inverse modelling approaches commonly used to infer aerosol kinetic parameters from laboratory studies. In this study, we show how machine learning methods can generate inexpensive surrogate models based on the kinetic multi-layer model of aerosol surface and bulk chemistry (KM-SUB). We apply and compare two common and openly available methods for the generation of surrogate models, polynomial chaos expansion (PCE) with UQLab and neural networks (NN) through the Python package Keras. We show that the PCE method is well-suited to determine global sensitivity indices of the KM and demonstrate how inverse modelling applications can be enabled or accelerated with NN-suggested sampling. These qualities make them suitable supporting tools for laboratory work in the interpretation of data and design of future experiments. Overall, the KM surrogate models investigated in this study are fast, accurate, and robust, which suggests their applicability as sub-modules in large-scale atmospheric models.

Thomas Berkemeier et al.

Status: open (until 18 Jan 2023)

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

Thomas Berkemeier et al.

Thomas Berkemeier et al.


Total article views: 369 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
288 76 5 369 2 1
  • HTML: 288
  • PDF: 76
  • XML: 5
  • Total: 369
  • BibTeX: 2
  • EndNote: 1
Views and downloads (calculated since 20 Oct 2022)
Cumulative views and downloads (calculated since 20 Oct 2022)

Viewed (geographical distribution)

Total article views: 356 (including HTML, PDF, and XML) Thereof 356 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
Latest update: 29 Nov 2022
Short summary
Kinetic multi-layer models (KM) successfully describe heterogeneous and multiphase atmospheric chemistry. In applications requiring repeated execution, however, these models can be too expensive. We trained machine learning surrogate models on output of the model KM-SUB and achieve high correlations. The surrogate models run orders of magnitudes faster, which suggests potential applicability in global optimization tasks and as sub-modules in large-scale atmospheric models.