Comparing TROPOMI and GEMS Observations for the Same Sun-Satellite Geometry

Together with the geostationary imagers over Southeast Asia (GEMS), North America (TEMPO) and Europe (Sentinel 4), TROPOMI and its follow-on low Earth orbit missions will establish the global air quality satellite constellation. The role of the low Earth orbit sensors in this constellation is twofold: firstly to provide the global coverage, including regions that cannot or will not be covered by the geostationary imagers, and secondly to facilitate cross-comparisons of the geostationary imagers, thereby serving as a travelling standard.

With the availability of the GEMS data and the expected release of the TEMPO data, the development of novel methods to intercompare the geostationary and low Earth orbit data is timely. When comparing data from geostationary and low Earth orbit data, one important aspect to overcome is the difference in the Sun-satellite geometry of the observations for an area on Earth. When comparing measured radiances under different geometries, complex corrections are required, for example to deal with the directionality of the surface reflectance. The uncertainties of such corrections may be larger than the expected difference in the radiance between the geostationary and low Earth orbit observations.

As the GEMS field of view includes the sub-satellite point at the equator, there is the unique opportunity for direct comparison when the TROPOMI nadir observations cover this point. In this case, observations of GEMS and TROPOMI with the same the same viewing geometry are available within maximum 30 minutes of each other. The time difference can be accounted for by interpolation, and/or by including a larger geographic area on the Earth in the intercomparison. As the orbital repeat cycle of TROPOMI is 227 orbits, there is an opportunity for direct comparisons approximately every 16 days. Such comparisons can include comparisons of the Level 1B radiances, reflectances, or fitted quantities such as slant column densities for different gases.

In this contribution we will demonstrate the method using case studies of direct comparisons of radiance, reflectances and NO₂ slant columns, as well as comparisons for different seasons.