Presenting a Concise OMI and TROPOMI NO2 Afternoon Data Record.

Abstract

Satellite instruments like OMI (2004) and TROPOMI (2018) have transformed global monitoring of tropospheric nitrogen dioxide (NO₂). Here we present a new Level 3 (L3) dataset produced within ESA’s CCI ECV project, created by averaging OMI and TROPOMI NO₂ columns spatially and temporally. This compact, user-friendly dataset is suitable for trend analysis, emission estimation, and climate modeling validation. It includes OMI (2004-2021) and TROPOMI (2018-2021) NO₂ data, uncertainty estimates, and additional variables like the averaging kernel, which informs vertical sensitivity to NO₂. The data set can be found at : https://www.temis.nl/airpollution/no2col/cci-no2-omi.php & https://www.temis.nl/airpollution/no2col/cci-no2-tropomi.phpThe L3 dataset consists of OMI and TROPOMI NO2 measurements spatially and temporally averaged in a consistent manner and includes full L3 uncertainty estimates. The uncertainty propagation includes measurement related uncertainties (from L2) as well as a spatial and a temporal representativity component. The data include additional spatiotemporally averaged variables, such as the averaging kernel, which provides relevant information on the vertical sensitivity to NO2. The L3 data have been produced in different resolutions ranging from 0.2 to 2 degrees and have been used for GEOS-Chem model evaluation. We show that the relative L3 uncertainties fall within the 15-20% range in polluted regions, lower than uncertainties in separate level 2 orbit retrievals, and brings tropospheric NO2 columns to within the GCOS ‘goal’ and ‘breakthrough’ requirements. Validation of the L3 against independent MAX-DOAS and PANDORA NO2 columns shows consistency up to 20%.Our aim is to make the L3 dataset as consistent as possible, by minimizing sensor-related differences to obtain a clearer view of NO2 changes over time. While OMI and TROPOMI have similar retrieval algorithms, there is a 15-minute difference in overpass time, the algorithms use different cloud algorithms, and OMI suffers from the ‘row anomaly’ phenomenon that causes a decrease in spatial coverage and hence sampling differences with TROPOMI. We show that by co-sampling (in space and time) OMI and TROPOMI NO2 columns, the consistency between the datasets improves. For example, for Beijing the relative absolute difference of OMI and TROPOMI for winter months in 2019 and 2020 is 14% and 22% drops to 8% and 11% when co-sampling. Furthermore, the row anomaly phenomenon causes reduced coverage in OMI from 2007 onward. We show how long-term trends in tropospheric NO2 columns are affected by the row anomaly and present a recipe to avoid such spurious trends.

Authors 8

1. Isidora Anglou

   Royal Netherlands Meteorological Institute

   Affiliation as printed

   Royal Netherlands Meteorological Institute, R&D Satellite Observations, De Bilt, Netherlands (isidora.anglou@knmi.nl)

2. K. F. Boersma corresponding

   Royal Netherlands Meteorological Institute · Wageningen University & Research

   Affiliation as printed

   Royal Netherlands Meteorological Institute, R&D Satellite Observations, De Bilt, Netherlands (isidora.anglou@knmi.nl)

   Wageningen University, Meteorology and Air Quality group, Wageningen, The Netherlands

3. Isolde Glissenaar corresponding

   Royal Netherlands Meteorological Institute · UiT The Arctic University of Norway

   Affiliation as printed

   Department of Physics and Technology, UiT The Arctic University of Norway, Tromsø, Norway

   Royal Netherlands Meteorological Institute, R&D Satellite Observations, De Bilt, Netherlands (isidora.anglou@knmi.nl)

4. Pieter Rijsdijk corresponding Aachen SRON Netherlands Institute for Space Research

   Royal Netherlands Meteorological Institute · Space Research Organisation Netherlands · Vrije Universiteit Amsterdam

   Affiliation as printed

   Department of Earth Sciences, Vrije Universiteit, Amsterdam, the Netherlands

   Royal Netherlands Meteorological Institute, R&D Satellite Observations, De Bilt, Netherlands (isidora.anglou@knmi.nl)

   SRON Netherlands Institute for Space Research, Leiden, Netherlands

5. Tijl Verhoelst corresponding

   Royal Belgian Institute for Space Aeronomy

   Affiliation as printed

   Belgian Institute for Space Aeronomy (BIRA-IASB), Brussels, Belgium

6. Steven Compernolle corresponding

   Royal Belgian Institute for Space Aeronomy

   Affiliation as printed

   Belgian Institute for Space Aeronomy (BIRA-IASB), Brussels, Belgium

7. Herizo Narivelo

   Centre National de la Recherche Scientifique · Université Toulouse III - Paul Sabatier · Laboratoire d'Aérologie · Université de Toulouse

   Affiliation as printed

   Laboratoire d’Aérologie, Université Paul Sabatier, Université de Toulouse - CNRS, France

8. Henk J. Eskes corresponding

   Royal Netherlands Meteorological Institute

   Affiliation as printed

   Royal Netherlands Meteorological Institute, R&D Satellite Observations, De Bilt, Netherlands (isidora.anglou@knmi.nl)