UCLA

Biomass burning (BB) has worsened in the Western US over the past several decades, a trend that is expected to continue. BB emits particles and precursors of secondary organic aerosols (SOA) and ozone. Ozone and particles are known to have adverse effects on human health, and BB will have a larger impact on the air quality of the Western US. HONO and HCHO are two of the major sources of radicals in BB plumes that drive the chemistry that forms SOA and ozone.

HONO photlyzes quickly, thus providing an early source of OH radicals in smoke. This photolysis complicates the measurement of HONO emissions, as in-situ observations are made after some photolysis has occurred. Remote sensing (RS) provides one solution to this issue, as it can measure fresher smoke and continuously observe smoke as it travels downwind. Unfortunately, little work has been done to understand the radiative transfer (RT) of sunlight through smoke. In this thesis, a radiative transfer model (RTM) was initialized with observations of smoke from an airborne field campaign to understand the ability of RS to observe trace gases in smoke when data is available to constrain the model and how a lack of data impacts that ability.

Results from the RTM suggest a model uncertainty from observations of BB plumes of 16.5%, and greater than 50% if more averaged assumptions are used. The impact of HONO photolysis can also impact RS retrievals by 15% over the first 2 hours after emission. This motivates better modelling of BB plumes for future satellite observations, which are becoming more prevalent. A parameter called the color index may help to inform these assumptions. Early results of HONO to NO2 emission ratios from satellites agree very well with the observations in this work, and RS observations suggest a high variability of HONO emissions, even by a single fire. Modelling studies of BB plumes must take care to accurately represent HONO emissions. The findings of this thesis motivate further improvements on RT modelling of BB plumes and photolysis rates to better account for RS observations and chemical modelling.