Air mass factor sensitivity to a priori profiles in GEMS formaldehyde retrieval during the ASIA-AQ campaign

Abstract

The air mass factor (AMF) plays a crucial role in determining satellite-observed vertical column densities, which are sensitive to a priori vertical profiles of measured species. We investigate the sensitivity of the Geostationary Environment Monitoring Spectrometer (GEMS) HCHO AMFs to the predetermined a priori profiles during the Airborne and Satellite Investigation of Asian Air Quality (ASIA-AQ) campaign and suggest an approach to enhance satellite observation accuracy. Using in-situ DC-8 aircraft observations, we evaluate the operational GEMS a priori profile and multi-model ensemble forecasts. Compared to DC-8, both the GEMS a priori profile and the ensemble model show good agreement (r = 0.96 and 0.95, respectively), but the GEMS a priori profile shows higher HCHO concentrations (NMB = 14%) than the ensemble model (NMB = −17%) especially near the surface. The GEMS a priori profile also shows consistent diurnal variations with DC-8 observations over the Seoul Metropolitan Area (r = 0.71–0.99) but overestimates HCHO concentrations by up to 58% (NMB), mainly due to higher surface concentrations during the morning and midday hours. The ensemble model better reproduces observed vertical profiles with reduced bias (NMBs up to 6%), owing to its high temporal and spatial resolution and the latest meteorological fields and anthropogenic emission inventory. Compared with the operational GEMS AMFs, vertical shape factors derived from the daily ensemble model forecasts increase AMFs by 57% over the Seoul Metropolitan Area in the morning. Validation results with ground-based Pandora observations show that ensemble-based AMFs reduce the positive bias of GEMS relative to Pandora in the morning (NMBs from 25 to 75% to within ±25%). Our sensitivity tests indicate the need for continuous updates of the meteorology and emission inventory to improve AMF calculations for GEMS retrievals. © 2026