Estimation of secondary organic aerosol from biomass burning using observations of formaldehyde, NO2 and AOD

Wildfires are a major source of atmospheric organic aerosol (OA), emitting primary organic aerosol (POA) and volatile organic compounds that oxidize to form secondary organic aerosol (SOA). This oxidation also produces formaldehyde (HCHO), which is additionally emitted directly from fires alongside nitrogen dioxide (NO₂). Both HCHO and NO₂ are detectable from space, offering the potential to observationally constrain organic aerosol formation during biomass burning. However, this potential remains poorly quantified.

Here, we evaluate whether satellite observations of HCHO and NO₂ can be used to estimate POA and SOA from biomass burning events. We compare simulations from four models with satellite measurements from OMI (HCHO, NO₂) and POLDER (AOD). All models reproduce correctly the observed spatial patterns of HCHO, NO₂ and AOD, but they overestimate trace gas concentrations and slightly underestimate AOD. Despite differences in magnitude, models and observations show linear relationships between HCHO and AOD.

Building on these observed relationships, we develop a satellite-based methodology to estimate POA and SOA with minimal use of model assumptions. The observed HCHO-AOD correlation is combined with satellite-derived mass extinction coefficient to relate observed AOD to organic aerosol. In addition, the relationship between NO₂ and POA fraction, derived from in-situ measurements, is used to separate the two types of organic aerosols. Together, these relationships allow the estimation of POA and SOA from HCHO and NO₂ observations, and sensitivity analysis shows that the method is robust. Application to the Amazon and African savanna indicates that observation-based POA formation is 3.82 and 5.53 times higher, respectively, than modeled values, while SOA formation is higher by factors of 2.4 and 3.5, suggesting model underestimation of organic aerosol production from biomass burning.