Abstract. Organic aerosol (OA) is one of the main components of the global particulate burden and intimately links natural and anthropogenic emissions with air quality and climate. It is challenging to accurately represent OA in global models. Direct quantification of global OA abundance is not possible with current remote sensing technology; however, it may be possible to exploit correlations of OA with remotely observable quantities to infer OA spatiotemporal variability. In particular, formaldehyde (HCHO) and OA share common sources via both primary emissions and secondary production from oxidation of volatile organic compounds (VOCs). We examine OA-HCHO correlations using data from summer-time airborne campaigns investigating biogenic (NASA SEAC4RS and DC3), biomass burning (NASA SEAC4RS) and anthropogenic conditions (NOAA CalNex and NASA KORUS-AQ). In situ OA correlates well with HCHO (r = 0.59–0.97) but the slope and intercept of this relationship vary with chemical regime. For biogenic and anthropogenic regions, the OA-vs-HCHO slope is higher in low NOx conditions, where HCHO yields are lower and aerosol yields are likely higher. The OA-vs-HCHO slope of wild fires is more than 9 times higher than that associated with biogenic and anthropogenic sources. An estimate of near-surface OA is derived by combining observed in situ relationships with HCHO column retrievals from NASA’s Ozone Monitoring Instrument (OMI). We evaluate this OA estimate against OA observations from the US EPA IMPROVE network and simulated OA from the GEOS-Chem global chemical transport model. The OA estimate compares well with IMPROVE data obtained over summer months (e.g. slope = 0.62, r = 0.56 for August 2013), comparable to intensively validated GEOS-Chem performance (e.g. slope = 0.57, r = 0.56) and superior to the correlation with satellite-derived total aerosol extinction (r = 0.41). Improving the detection limit of satellite HCHO and expanding in situ airborne HCHO and OA coverage in future missions will improve the quality and spatiotemporal coverage of this OA estimate, potentially enabling constraints on the global OA distribution.;
