Preprints
https://doi.org/10.5194/gmd-2021-191
https://doi.org/10.5194/gmd-2021-191

Submitted as: methods for assessment of models 23 Jul 2021

Submitted as: methods for assessment of models | 23 Jul 2021

Review status: this preprint is currently under review for the journal GMD.

An aerosol classification scheme for global simulations using the K-means machine learning method

Jingmin Li, Johannes Hendricks, Mattia Righi, and Christof G. Beer Jingmin Li et al.
  • Deutsches Zentrum für Luft- und Raumfahrt (DLR), Institut für Physik der Atmosphäre, Oberpfaffenhofen, Germany

Abstract. A machine learning K-means algorithm is applied to data of seven aerosol properties from a global aerosol simulation using EMAC-MADE3. The aim is to partition the aerosol properties across the global atmosphere in specific aerosol regimes. K-means is an unsupervised machine learning method with the advantage that an a priori definition of the aerosol classes is not required. Using K-means, we are able to quantitatively define global aerosol regimes, so-called aerosol clusters, and explain their internal properties as well as their location and extension. This analysis shows that aerosol regimes in the lower troposphere are strongly influenced by emissions. Key drivers of the clusters’ internal properties and spatial distribution are, for instance, pollutants from biomass burning/biogenic sources, mineral dust, anthropogenic pollution, as well as their mixing. Several continental clusters propagate into oceanic regions. The identified oceanic regimes show a higher degree of pollution in the northern hemisphere than over the southern oceans. With increasing altitude, the aerosol regimes propagate from emission-induced clusters in the lower troposphere to roughly zonally distributed regimes in the middle troposphere and in the tropopause region. Notably, three polluted clusters identified over Africa, India and eastern China, cover the whole atmospheric column from the lower troposphere to the tropopause region. A markedly wide application potential of the classification procedure is identified and further aerosol studies are proposed which could benefit from this classification.

Jingmin Li et al.

Status: final response (author comments only)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on gmd-2021-191', Anonymous Referee #1, 30 Aug 2021
    • AC1: 'Reply on RC1', Jingmin Li, 15 Sep 2021
  • RC2: 'Comment on gmd-2021-191', Anonymous Referee #2, 27 Sep 2021

Jingmin Li et al.

Jingmin Li et al.

Viewed

Total article views: 432 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
362 61 9 432 2 2
  • HTML: 362
  • PDF: 61
  • XML: 9
  • Total: 432
  • BibTeX: 2
  • EndNote: 2
Views and downloads (calculated since 23 Jul 2021)
Cumulative views and downloads (calculated since 23 Jul 2021)

Viewed (geographical distribution)

Total article views: 383 (including HTML, PDF, and XML) Thereof 383 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 22 Oct 2021
Download
Short summary
The growing complexity of global aerosol models results in a large number of parameters which describe the aerosol number, size and composition. This makes the analysis, evaluation and interpretation of the model results a challenge. To overcome this difficulty, we apply a machine learning classification method to identify clusters of specific aerosol types in global aerosol simulations. Our results demonstrate the spatial distributions and characteristics of these identified aerosol clusters.