15 Jul 2021

15 Jul 2021

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

ML-TOMCAT: Machine-Learning-Based Satellite-Corrected Global Stratospheric Ozone Profile Dataset from a Chemical Transport Model

Sandip S. Dhomse1,2, Carlo Arosio3, Wuhu Feng1,4, Alexei Rozanov3, Mark Weber3, and Martyn P. Chipperfield1,2 Sandip S. Dhomse et al.
  • 1School of Earth and Environment, University of Leeds, Leeds, UK
  • 2National Centre for Earth Observation, University of Leeds, Leeds, UK
  • 3Institute for Environmental Physics, University of Bremen, Bremen, Germany
  • 4National Centre for Atmospheric Science, University of Leeds, Leeds, UK

Abstract. High quality stratospheric ozone profile datasets are a key requirement for accurate quantification and attribution of long-term ozone changes. Satellite instruments obtain stratospheric ozone profile measurements over typical mission durations of 5–15 years. Various methodologies have then been applied to merge and homogenise the different satellite data in order to create longer term observation-based ozone profile datasets with minimal data gaps. However, individual satellite instruments use different measurement methods, sampling patterns and retrieval algorithms which complicate the merging of these different datasets. In contrast, atmospheric chemical models can produce chemically consistent long-term ozone simulations based on specified changes in external forcings, but they are subject to the deficiencies associated with incomplete understanding of complex atmospheric processes and uncertain photochemical parameters.

Here, we use chemically self-consistent output from the TOMCAT 3-D chemical transport model (CTM) and a Random-Forest (RF) ensemble learning method to create a merged 42-year (1979–2020) stratospheric ozone profile dataset (ML-TOMCAT V1.0). The underlying CTM simulation was forced by meteorological reanalyses, specified trends in long-lived source gas, solar flux and aerosol variations. The RF is trained using the Stratospheric Water and OzOne Satellite Homogenized (SWOOSH) dataset over the time periods of the Microwave Limb Sounder (MLS) from the Upper Atmosphere Research Satellite (UARS) (1991–1998) and Aura (2005–2016) missions. We find that ML-TOMCAT shows excellent agreement with available independent satellite-based datasets which use pressure as the vertical coordinate (e.g. GOZCARDS, SWOOSH for non-MLS periods) but weaker agreement with the datasets which are height-based (e.g. SAGE–CCI–OMPS, SCIAMACHY-OMPS). We find that at almost all stratospheric levels ML-TOMCAT ozone concentrations are well within uncertainties in the observational datasets. The ML-TOMCAT dataset is thus ideally suited for the evaluation of model ozone profiles from the tropopause to 0.1 hPa. ML-TOMCAT data is freely available via (Dhomse et al., 2021).

Sandip S. Dhomse 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 essd-2021-225', Anonymous Referee #1, 15 Aug 2021
  • RC2: 'Review of the manuscript “ML-TOMCAT: Machine-Learning-Based Satellite-Corrected Global Stratospheric Ozone Profile Dataset from a Chemical Transport Model“ by S.S. Dhomse et al.', Anonymous Referee #2, 18 Aug 2021

Sandip S. Dhomse et al.

Data sets

ML-TOMCAT V1.0: Machine-Learning-Based Satellite-Corrected Global Stratospheric Ozone Profile Dataset Dhomse S. Sandip; Chipperfield M. P.; Wuhu Feng; Carlo Arosio; Mark Weber; Alexei Rozanov

Sandip S. Dhomse et al.


Total article views: 385 (including HTML, PDF, and XML)
HTML PDF XML Total Supplement BibTeX EndNote
282 91 12 385 32 11 10
  • HTML: 282
  • PDF: 91
  • XML: 12
  • Total: 385
  • Supplement: 32
  • BibTeX: 11
  • EndNote: 10
Views and downloads (calculated since 15 Jul 2021)
Cumulative views and downloads (calculated since 15 Jul 2021)

Viewed (geographical distribution)

Total article views: 350 (including HTML, PDF, and XML) Thereof 350 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
Latest update: 14 Oct 2021
Short summary
High quality long-term ozone profile datasets is a key requirement to estimate short and long term ozone variability. Almost all the satellite (and chemical model) datasets show some kind of biases w.r.t. each other. This is because differences in measurement methodologies as well as simplified processes in the models. Here we use satellite datasets and chemical model output to generate 42 years of ozone profile dataset using Random Forest machine learning algorithm that is named as ML-TOMCAT.