Beyond binary maps from HCHO∕NO<sub>2</sub>: a deep neural network approach to global daily mapping of net ozone production rates and sensitivities constrained by satellite observations (2005–2023)

Souri, Amir H.; González Abad, Gonzalo; Duncan, Bryan N.; Oman, Luke D.

doi:10.5194/acp-26-809-2026

Articles | Volume 26, issue 1

https://doi.org/10.5194/acp-26-809-2026

© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/acp-26-809-2026

© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 26, issue 1

Research article

|

16 Jan 2026

Research article |

| 16 Jan 2026

Beyond binary maps from HCHO∕NO₂: a deep neural network approach to global daily mapping of net ozone production rates and sensitivities constrained by satellite observations (2005–2023)

Amir H. Souri, Gonzalo González Abad, Bryan N. Duncan, and Luke D. Oman

Supplement

https://doi.org/10.5194/acp-26-809-2026-supplement

Data sets

TROPOMI Level 2 Nitrogen Dioxide total column products, Version 02 Copernicus Sentinel-5P https://doi.org/10.5270/S5P-9bnp8q8

TROPOMI Level 2 Formaldehyde Total Column products, Version 02 Copernicus Sentinel-5P https://doi.org/10.5270/S5P-vg1i7t0

Model code and software

Ozonerates v1.0 Amir Souri and Gonzalo Gonzalez Abad https://doi.org/10.5281/zenodo.15076487

Download

Article (20007 KB)
Full-text XML

Short summary

We create long-term maps of PO₃ magnitudes along with their corresponding sensitivity maps. This is achieved using a deep learning parameterization method that relies on satellite data, atmospheric models, and ground-based remote sensing. Our approach provides more quantitative information than commonly used methods that depend on ratio-based indicators (such as HCHO/NO₂). Additionally, our method considers light and water vapor, making it suitable for applications with GEO satellites.

Beyond binary maps from HCHO∕NO2: a deep neural network approach to global daily mapping of net ozone production rates and sensitivities constrained by satellite observations (2005–2023)

Supplement

Data sets

Model code and software

Beyond binary maps from HCHO∕NO₂: a deep neural network approach to global daily mapping of net ozone production rates and sensitivities constrained by satellite observations (2005–2023)