Interpreting Summertime Hourly Variation of NO₂ Columns with Implications for Geostationary Satellite Applications

Abstract. Accurate representation of the hourly variation of the NO₂ column-to-surface relationship is needed to interpret geostationary constellation observations of tropospheric NO₂ columns. Prior work has revealed inconsistency in the hourly variation in NO₂ columns and surface concentrations. In this study, we use the high-performance configuration of the GEOS-Chem model (GCHP) to interpret the daytime hourly variation in NO₂ total columns and surface concentrations during summer. We use summer-time Pandora sun photometers and aircraft measurements during the Deriving Information on Surface Conditions from Column and Vertically Resolved Observations Relevant to Air Quality (DISCOVER-AQ) field campaign over Maryland, Texas, and Colorado as well as 50 sites (31: contiguous USA, 10: Europe, 9: Asia) from the Pandonia Global Network (PGN). We correct the Pandora columns for 1) hourly variation in the column effective temperature driven by the fractional boundary layer contribution to the total column, and 2) change in local solar time along the line-of-sight of the Pandora instrument. The corrected Pandora observations are increased by about 5–6 × 10¹⁴ molecules cm^-2 at 9 AM and 6 PM across all Pandora sites. We conduct fine resolution (~12 km) simulations over the contiguous US, Europe, and East Asia using the stretched grid capability of GCHP. We also examine the effect of planetary boundary layer height (PBLH) corrections on the total columns. We first evaluate the GCHP simulated absolute NO₂ concentration with Pandora and aircraft observations. We find that fine resolution simulations at 12 km compared with moderate resolution ~55 km reduce the Normalized Bias (NB) versus Pandora total columns (19 % to 10 %) and versus aircraft measurements (25 % to 13 %) over Maryland, Texas, and Colorado. Fine resolution simulations at 12 km compared with moderate resolution at 55 km reduce the NB versus Pandora total columns over the eastern US (17 % to 9 %), western US (22 % to 14 %), Europe (24 % to 15 %), and Asia (29 % to 21 %). We next use the 12 km simulation to examine the hourly variation in the NO₂ column and surface concentrations. We explain the weaker hourly variation in NO₂ columns than at the surface as a function of 1) hourly variation in the column effective temperature, 2) hourly variation in the local solar time along the Pandora line-of-sight and 3) the integral of weakly connected layers; with the lowest 500 m exhibiting greater NO₂ concentrations in morning and evening than midday, while the residual column above 500 m dominates the total column with weaker variability.