In our study, we compare the molecular-level functionalized organic aerosol composition from three different field sites, a laboratory chamber, and a fully-explicit chemical model to assess temporal variability in organic aerosol composition. Before we dove fully into this study, we started by exploring the data from our first field site, located in a forest in northern Michigan. We made two key observations that shaped the rest of our research. First, we noticed that the bulk aerosol composition remained fairly consistent across all the daytime and nighttime samples we collected. Second, we noticed that comparing individual samples at the molecular-level showed surprisingly little overlap between the composition of these individual samples. In most data in previously published studies, the composition of organic aerosol seems to follow relatively consistent patterns. So, the variability we observed across samples collected at similar times (e.g. day-to-day) or consecutively (e.g. day-to-night) piqued our interest, and motivated us to perform a similar analysis for two additional field sites and to explore possible driving factors of this variability via laboratory chamber experiments and chemical modeling.

We collected ambient aerosol samples at three diverse field sites: (1) in a northern Michigan forest on the PROPHET research tower, which is isolated from large cities and therefore an opportune location to study organic aerosol from biogenic sources; (2) near downtown Atlanta, GA, a large city that is close to large forested areas, which makes it a good place to study a mixture of biogenic and anthropogenic aerosol and is representative of parts of the southeastern U.S.; and (3) in New York City, a good location to study organic aerosol from anthropogenic sources. At each of the three field sites, we noticed similar compositional variability when comparing individual samples to each other.

To validate our ambient observations and further explore this variability and its potential causes, we performed several laboratory chamber experiments using two prominent biogenic volatile organic compound precursors: alpha- and beta-pinene, which your nose would recognize as the characteristic smell of pine trees. These experiments allowed us to study the effect of precursor identity and chemical oxidation conditions in a controlled setting. We also used a chemical model called GECKO-A to simulate the molecular-level gas and aerosol composition in the laboratory chamber and in the ambient atmosphere. Both the laboratory chamber experiments and the modeling results showed similar variability to the variability we observed in the ambient atmosphere.

So, the takeaway is that the composition of functionalized organic aerosol might not always be as consistent as some measurement methods suggest, and measurements from field sites or times of day that look similar in terms of bulk properties may in fact be quite different at the molecular level. We are still exploring the range of implications of these results. Many aerosol properties depend strongly on molecular composition and structure, such as volatility and phase state.  We hypothesize that these aerosol properties might be sensitive to the highly variable molecular-level composition we observed at these three field sites, and might change significantly day-to-day as well. Aerosol properties are linked to aerosol effects on human health and the environment. So, understanding the molecular-level variability of organic aerosol composition is key to understanding how this variability might influence aerosol properties and possible associated impacts on our health and the environment. 

Our paper can be found here.

Written by Jenna C. Ditto & Drew R. Gentner, Chemical & Environmental Engineering, Yale University