Keywords:Satellite observations, Validation, Spectroscopy, Air pollution
Satellite observations of NO2, a key atmospheric pollutant, have provided critical information on global distribution of NOx sources and their temporal changes since 1995, and advanced understanding of atmospheric chemistry. A recent remarkable improvement is with their spatial resolution, reaching 7×3.5 km for TROPOMI (Sentinel-5P), launched in 2017. However, there is still a fundamental issue of low bias in the determined tropospheric NO2 column densities (e.g., 25-50% for TROPOMI), when evaluated against ground-based spectroscopic observations. Potential causes would be (1) a-priori vertical profile shapes, (2) co-present aerosols and/or surface albedo, and (3) spatial inhomogeneity of NO2, while (4) uncertainty in the ground-based observations could also be important. At Yokosuka (32.32°N, 139.65°E), located on the south edge of Tokyo-Yokohama metropolitan area, we newly started direct-sun Pandora spectrometer observations in November 2018 and strengthened validation capability, where the scattered-sun MAX-DOAS observations (from a JAMSTEC original instrument) had been solely referenced before. Results during winter 2018/2019 showed TROPOMI (within 0.1° from Yokosuka) had ~40% low bias against MAX-DOAS. The magnitude of difference was similar to the results from longer comparison during Feb-Sep 2018 for TROPOMI and earlier comparisons for OMI (QA4ECV). Pandora (in direct-sun mode) gave values ~24 % less than MAX-DOAS; the cause of difference was likely spatial inhomogeneity, as pollution levels were higher in the northeast for the sight of MAX-DOAS than in south for Pandora. This was consistent with the fact that TROPOMI gave ~18% higher values, when the averaging area was shifted to northeast by 10 km. As such, the importance of factors (3) and (4) was newly evaluated. By considering MAX-DOAS profile shapes in the recalculation of satellite products, we will discuss on factor (1), and the possibility to derive near-surface NO2 concentrations from satellite observations.