Preprints
https://doi.org/10.48550/arXiv.2406.18171
https://doi.org/10.48550/arXiv.2406.18171
09 Sep 2024
 | 09 Sep 2024
Status: this preprint is open for discussion.

Similarity-Based Analysis of Atmospheric Organic Compounds for Machine Learning Applications

Hilda Sandström and Patrick Rinke

Abstract. The formation of aerosol particles in the atmosphere impacts air quality and climate change, but many of the organic molecules involved remain unknown. Machine learning could aid in identifying these compounds through accelerated analysis of molecular properties and detection characteristics. However, such progress is hindered by the current lack of curated datasets for atmospheric molecules and their associated properties. To tackle this challenge, we propose a similarity analysis that connects atmospheric compounds to existing large molecular datasets used for machine learning development. We find a small overlap between atmospheric and non-atmospheric molecules using standard molecular representations in machine learning applications. The identified out-of-domain character of atmospheric compounds is related to their distinct functional groups and atomic composition. Our investigation underscores the need for collaborative efforts to gather and share more molecular-level atmospheric chemistry data. The presented similarity based analysis can be used for future dataset curation for machine learning development in the atmospheric sciences.

Hilda Sandström and Patrick Rinke

Status: open (until 04 Nov 2024)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2024-2432', Anonymous Referee #1, 19 Sep 2024 reply
  • RC2: 'Comment on egusphere-2024-2432', Anonymous Referee #2, 20 Sep 2024 reply
  • RC3: 'Comment on egusphere-2024-2432', Jonas Elm, 02 Oct 2024 reply
Hilda Sandström and Patrick Rinke

Model code and software

Atmospheric Compound Similarity Analysis Hilda Sandström https://gitlab.com/cest-group/atmospheric_compound_similarity_analysis

Interactive computing environment

Atmospheric Compound Similarity Analysis Hilda Sandström https://gitlab.com/cest-group/atmospheric_compound_similarity_analysis

Hilda Sandström and Patrick Rinke

Viewed

Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.

Total article views: 92 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
92 0 0 92 0 0
  • HTML: 92
  • PDF: 0
  • XML: 0
  • Total: 92
  • BibTeX: 0
  • EndNote: 0
Views and downloads (calculated since 09 Sep 2024)
Cumulative views and downloads (calculated since 09 Sep 2024)

Viewed (geographical distribution)

Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.

Total article views: 87 (including HTML, PDF, and XML) Thereof 87 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 05 Oct 2024
Download
Short summary
Machine learning has the potential to aid the identification organic molecules involved in aerosol formation. Yet, progress is stalled by a lack of curated atmospheric molecular datasets. Here, we compared atmospheric compounds with large molecular datasets used in machine learning and found minimal overlap with similarity algorithms. Our result underlines the need for collaborative efforts to curate atmospheric molecular data to facilitate machine learning model in atmospheric sciences.