Improving the predictions of black carbon (BC) optical properties at various aging stages using a machine-learning-based approach

Romshoo, Baseerat; Patil, Jaikrishna; Michels, Tobias; Müller, Thomas; Kloft, Marius; Pöhlker, Mira

doi:https://doi.org/10.5194/acp-24-8821-2024

Articles | Volume 24, issue 15

https://doi.org/10.5194/acp-24-8821-2024

© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/acp-24-8821-2024

© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 24, issue 15

Research article

|

12 Aug 2024

Research article |

| 12 Aug 2024

Improving the predictions of black carbon (BC) optical properties at various aging stages using a machine-learning-based approach

Baseerat Romshoo, Jaikrishna Patil, Tobias Michels, Thomas Müller, Marius Kloft, and Mira Pöhlker

Supplement

https://doi.org/10.5194/acp-24-8821-2024-supplement

Data sets

Database of physicochemical and optical properties of black carbon fractal aggregates Baseerat Romshoo et al. https://doi.org/10.5281/zenodo.7523058

Model code and software

jaikrishnap/Machine-learning-forprediction-of-BCFAs: Initial release Baseerat Romshoo et al. https://doi.org/10.5281/zenodo.8060206

jaikrishnap/Optical-properties-ofblack-carbon-aggregates: Initial release Baseerat Romshoo et al. https://doi.org/10.5281/zenodo.8071901

Download

Article (4835 KB)
Full-text XML

Short summary

Through the use of our machine-learning-based optical model, realistic BC morphologies can be incorporated into atmospheric science applications that require highly accurate results with minimal computational resources. The results of the study demonstrate that the predictions of single-scattering albedo (ω) and mass absorption cross-section (MAC) were improved over the conventional Mie-based predictions when using the machine learning method.