Cloud water composition during HCCT-2010: Scavenging efficiencies, solute concentrations, and droplet size dependence of inorganic ions and dissolved organic carbon

Abstract

Abstract. Cloud water samples were taken in September/October 2010 at Mt. Schmücke in a rural, forested area in Germany during the Lagrange-type Hill Cap Cloud Thuringia 2010 (HCCT-2010) cloud experiment. Besides bulk collectors, a 3-stage and a 5-stage collector were applied and samples were analysed for inorganic ions (SO42−, NO3−, NH4+, Cl−, Na+, Mg2+, Ca2+, K+), H2O2 (aq), S(IV), and dissolved organic carbon (DOC). Campaign volume-weighted mean concentrations were 191, 142, and 39 μmol L−1 for ammonium, nitrate, and sulfate, respectively, between 4 and 27 μmol L−1 for minor ions, 5.4 μmol L−1 for H2O2 (aq), 1.9 μmol L−1 for S(IV), and 3.9 mgC L−1 for DOC. The concentrations compare well to more recent European cloud water data from similar sites. On a mass basis, organic material (as DOC · 1.8) contributed 20–40 % (event means) to total solute concentrations and was found to have non-negligible impact on cloud water acidity. Relative standard deviations of major ions were 60–66 % for solute concentrations and 52–80 % for cloud water loadings (CWLs). Contrary to some earlier suggestions, the similar variability of solute concentrations and CWLs together with the results of back trajectory analysis and principal component analysis, suggests that concentrations in incoming air masses (i.e. air mass history), rather than cloud liquid water content (LWC) was the main factor controlling bulk solute concentrations at Mt. Schmücke. Droplet effective radius was found to be a somewhat better predictor for cloud water total ionic content (TIC) than LWC, even though no single explanatory variable can fully describe TIC (or solute concentration) variations in a simple functional relation due to the complex processes involved. Bulk concentrations typically agreed within a factor of 2 with co-located measurements of residual particle concentrations sampled by a counterflow virtual impactor (CV) and analysed by an aerosol mass spectrometer (AMS), with the deviations being mainly caused by systematic differences and limitations of the approaches (such as outgassing of dissolved gases during residual particle sampling). Scavenging efficiencies (SEs) of aerosol constituents were 0.56–0.94, 0.79–0.99, 0.71–98, and 0.67–0.92 for SO42−, NO3−, NH4+, and DOC, respectively, when calculated as event means with in-cloud data only. SEs estimated using data from an upwind site were substantially different in many cases, revealing the impact of gas-phase uptake (for volatile constituents) and mass losses across Mt. Schmücke likely due to physical processes such as droplet scavenging by trees and/or entrainment. Drop size-resolved cloud water concentrations of major ions SO42−, NO3−, and NH4+ revealed two main profiles: decreasing concentrations with increasing droplet size and "U"-shapes. In contrast, profiles of typical coarse particle mode minor ions were often increasing with increasing drop size, highlighting the importance of a species' particle concentration size distribution for the development of size-resolved solute concentration patterns. Concentration differences between droplet size classes were typically < 2 for major ions from the 3-stage collector and somewhat more pronounced from the 5-stage collector, while they were much larger for minor ions. Due to a better separation of droplet populations, the 5-stage collector was capable of resolving some features of solute size dependencies not seen in the 3-stage data, especially sharp concentration increases (up to a factor of 5–10) in the smallest droplets for many solutes.