Abstract

Validating tropospheric NO2 column density from GEMS version 3.0 and Tropospheric Chemistry Reanalysis version 2 from background to urban conditions over Japan: with a focus on diurnal variations

This study validates tropospheric NO2 vertical column density (TropNO2 VCD) data from the Geostationary Environment Monitoring Spectrometer (GEMS) version 3.0 and the Tropospheric Chemistry Reanalysis version 2 (TCR-2) against Multi-Axis Differential Optical Absorption Spectroscopy (MAX-DOAS) and Tropospheric Monitoring Instrument (TROPOMI) data at three Japanese sites: Yokosuka (urban), Fukue (rural-remote), and Cape Hedo (remote). The GEMS v3.0 dataset showed markedly improved agreement with MAX-DOAS compared to v2.0, especially at Fukue and Cape Hedo remote islands, with significantly reduced normalized mean bias from 142–319% to 8–18%. GEMS v3.0 showed agreement with MAX-DOAS observations at Yokosuka (R = 0.75) and Fukue (R = 0.50). It performed particularly well in autumn and winter. There were positive biases in autumn at Yokosuka (NMB = 3.5%), but negative biases in the other seasons (NMB = −0.4% to −51.2%). The reanalysis TCR-2 and aggregated TROPOMI TropNO2 VCD data (both at 1 × 1° resolution) showed moderate to good correlation at all three sites (R = 0.72 at Yokosuka; R = 0.44 at Fukue; R = 0.57 at Cape Hedo). While the spatial resolution of TCR-2 (1.1 × 1.1°) precluded meaningful comparisons with MAX-DOAS at urban sites such as Yokosuka, TCR-2 exhibited lower biases than TROPOMI at Cape Hedo, which represents a background environment (NMB = 38% vs. 87%). GEMS v3.0 data effectively captured midday NO2 reductions at Cape Hedo, consistent with TCR-2 and MAX-DOAS, marking this as the first study to explore daytime NO2 decreases in remote regions using GEMS. NO2 loss pathways at Cape Hedo were examined using TCR-2 data and box-model simulations. NO2 levels from GEMS exhibited a decline between 10 AM and 3 PM in autumn, aligning with patterns from MAX-DOAS and TCR-2, with NO2 decay rates of 16 ± 10h from GEMS, 21 ± 5h from MAX-DOAS, and 22 ± 20h from TCR-2. The box-model NOx budget analysis revealed that the NO2 + OH reaction was the dominant pathway for NOx loss. The markedly improved agreement of GEMS v3.0 data with MAX-DOAS and TCR-2 data will allow potential exploration of the oxidizing capacity of the atmosphere over the clean marine regions, after further characterization of NOx chemistry.