Preprints
https://doi.org/10.5194/egusphere-2024-3145
https://doi.org/10.5194/egusphere-2024-3145
04 Nov 2024
 | 04 Nov 2024

Hourly surface nitrogen dioxide retrieval from GEMS tropospheric vertical column densities: Benefit of using time-contiguous input features for machine learning models

Janek Gödeke, Andreas Richter, Kezia Lange, Peter Maaß, Hyunkee Hong, Hanlim Lee, and Junsung Park

Abstract. Launched in 2020, the Korean Geostationary Environmental Monitoring Spectrometer (GEMS) is the first geostationary satellite mission for observing trace gas concentrations in the Earth’s atmosphere. Observations are made over Asia. Geostationary orbits allow for hourly measurements, which leads to a much higher temporal resolution compared to daily measurements taken from low Earth orbits, such as by the TROPOspheric Monitoring Instrument (TROPOMI) or Ozone Monitoring Instrument (OMI). This work estimates the hourly concentration of surface NO2 from GEMS tropospheric NO2 vertical column densities (tropospheric NO2 VCDs) and additional meteorological features, which serve as inputs for Random Forests and linear regression models. With several measurements per day, not only the current observations but also those from previous hours can be used as inputs for the machine learning models. We demonstrate that using these time-contiguous inputs leads to reliable improvements regarding all considered performance measures, such as Pearson correlation or Mean Square Error. For Random Forests, the average performance gains are between 4.5 % and 7.5 %, depending on the performance measure. For linear regression models, average performance gains are between 7 % and 15 %. For performance evaluation, spatial cross validation with surface in-situ measurements is used to measure how well the trained models perform at locations where they have not received any training data. In other words, we inspect the models’ ability to generalize to unseen locations. Additionally, we investigate the influence of tropospheric NO2 VCDs on the performance. The region of our study is Korea.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.
Janek Gödeke, Andreas Richter, Kezia Lange, Peter Maaß, Hyunkee Hong, Hanlim Lee, and Junsung Park

Status: final response (author comments only)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2024-3145', Anonymous Referee #1, 25 Nov 2024
  • RC2: 'Comment on egusphere-2024-3145', Anonymous Referee #2, 03 Dec 2024
Janek Gödeke, Andreas Richter, Kezia Lange, Peter Maaß, Hyunkee Hong, Hanlim Lee, and Junsung Park
Janek Gödeke, Andreas Richter, Kezia Lange, Peter Maaß, Hyunkee Hong, Hanlim Lee, and Junsung Park

Viewed

Total article views: 201 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
142 51 8 201 5 4
  • HTML: 142
  • PDF: 51
  • XML: 8
  • Total: 201
  • BibTeX: 5
  • EndNote: 4
Views and downloads (calculated since 04 Nov 2024)
Cumulative views and downloads (calculated since 04 Nov 2024)

Viewed (geographical distribution)

Total article views: 199 (including HTML, PDF, and XML) Thereof 199 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 12 Dec 2024
Download
Short summary
The Korean Geostationary Environmental Monitoring Spectrometer (GEMS) monitors trace gases over Asia, e.g., NO2. GEMS provides hourly data, improving the time-resolution compared to the daily overpasses by other satellites. For the prediction of hourly surface NO2 over Korea from GEMS observations and meteorological data, this study shows that machine learning models benefit from this higher time-resolution. This is achieved by using observations from previous hours as additional inputs.