The impacts of aerosol loading, composition, and water uptake on aerosol extinction variability in the Baltimore–Washington, D.C. region

摘要

In order to utilize satellite-based aerosol measurements for thedetermination of air quality, the relationship between aerosol opticalproperties (wavelength-dependent, column-integrated extinction measured bysatellites) and mass measurements of aerosol loading (PM used for airquality monitoring) must be understood. This connection varies with manyfactors including those specific to the aerosol type – such as composition,size, and hygroscopicity – and to the surrounding atmosphere, such astemperature, relative humidity (RH), and altitude, all of which can varyspatially and temporally. During the DISCOVER-AQ (Deriving Information onSurface conditions from Column and Vertically Resolved Observations Relevantto Air Quality) project, extensive in situ atmospheric profiling in theBaltimore, MD–Washington, D.C. region was performed during 14flights in July 2011. Identical flight plans and profile locations throughoutthe project provide meaningful statistics for determining the variability inand correlations between aerosol loading, composition, optical properties, andmeteorological conditions.Measured water-soluble aerosol mass was composed primarily of ammoniumsulfate (campaign average of 32 %) and organics (57 %). A distinctdifference in composition was observed, with high-loading days having aproportionally larger percentage of sulfate due to transport from the OhioRiver Valley. This composition shift caused a change in the aerosolwater-uptake potential (hygroscopicity) such that higher relativecontributions of inorganics increased the bulk aerosol hygroscopicity. Thesedays also tended to have higher relative humidity, causing an increase in thewater content of the aerosol. Conversely, low-aerosol-loading days had lowersulfate and higher black carbon contributions, causing lower single-scatteringalbedos (SSAs). The average black carbon concentrations were240 ng m in the lowest 1 km, decreasing to 35 ng m in thefree troposphere (above 3 km).Routine airborne sampling over six locations was used to evaluate therelative contributions of aerosol loading, composition, and relative humidity(the amount of water available for uptake onto aerosols) to variability inmixed-layer aerosol extinction. Aerosol loading (dry extinction) was found tobe the predominant source, accounting for 88 % on average of the measuredspatial variability in ambient extinction, with lesser contributions fromvariability in relative humidity (10 %) and aerosol composition(1.3 %). On average, changes in aerosol loading also caused 82 % ofthe diurnal variability in ambient aerosol extinction. However on days withrelative humidity above 60 %, variability in RH was found to cause up to62 % of the spatial variability and 95 % of the diurnal variabilityin ambient extinction.This work shows that extinction is driven to first order by aerosol massloadings; however, humidity-driven hydration effects play an importantsecondary role. This motivates combined satellite–modeling assimilationproducts that are able to capture these components of the aerosol optical depth (AOD)–PMlink. Conversely, aerosol hygroscopicity and SSA play a minor role in drivingvariations both spatially and throughout the day in aerosol extinction andtherefore AOD. However, changes in aerosol hygroscopicity from day to daywere large and could cause a bias of up to 27 % if not accounted for.Thus it appears that a single daily measurement of aerosol hygroscopicity canbe used for AOD-to-PM conversions over the study region (on the orderof 1400 km). This is complimentary to the results of Chu etal. (2015), who determined that the aerosol vertical distribution from "a single lidar isfeasible to cover the range of 100 km" in the same region.