Activation properties of particles as cloud condensation nuclei

Abstract

This PhD dissertation focuses on the activation properties of aerosol particles as cloud condensation nuclei (CCN) to improve the existing knowledge on how aerosol particles evolve in the atmosphere to become effective CCN. To this end, physicochemical properties of atmospheric aerosol particles and their activation properties as CCN have been analyzed using ground-based in-situ techniques at two different environments. To study the activation properties of aerosol particles at the height where clouds might form and asses the influence of anthropogenic influence at remote sites, firstly, the CCN activation properties of aerosol particles are characterized at two different sites in southern Spain: an urban background station in Granada city and a high-altitude mountain station in the Sierra Nevada National Park, with a horizontal separation of 21 km and vertical distance of 1820 m. CCN activity of aerosol particles at the urban environment is driven by primary sources, mainly road traffic. High CCN concentrations occurred during traffic rush hours, although this was also when the fraction of activated particles over total particles was the lowest. This is due to the characteristics of the rush hour aerosol population consisting of ultrafine and less hygroscopic particles. In contrast, the mountain site exhibited larger and more hygroscopic particles, with CCN activity driven by the joint effect of new particle formation (NPF) events and subsequent growth and vertical transport of anthropogenic particles from Granada urban area by orographic buoyant upward flow. This led to the maximum concentrations of CCN and aerosol particles occurring at midday at the mountain site. Clear differences in the diurnal evolution of CCN between NPF event days and nonevent days were observed at the Sierra Nevada station, demonstrating the large contribution of newly formed particles to CCN concentrations after growth. The isolated contribution of NPF to CCN concentration has been estimated to be 175% higher at supersaturation ratio of 0.5% relative to what it would be without NPF events, revealing that CCN concentrations can be highly modified during these events. Also, two empirical models were proposed to parameterize CCN concentrations in terms of aerosol optical or physical parameters. The models could explain measurements successfully at the urban station, whereas at the mountain site both models could not reproduce satisfactorily the observations probably due to the aerosol properties changes caused by upslope transport of urban particles and NPF events. As urban particles were found to affect the CCN activity at the mountain site during summer and autumn, a new field campaign was performed at the urban site during summer to get a deepen insight of the different aerosol sources and processes affecting urban aerosol particles and disentangle its influence on the CCN concentrations. An unsupervised clustering model was used to classify the main aerosol categories and processes occurring in the urban atmosphere and then, the influence of the identified aerosol populations on the CCN properties was analyzed. According to the physical properties of each cluster, its diurnal timing, and additional air quality parameters, the clusters were grouped into five main aerosol categories: nucleation, growth, traffic, aged traffic, and urban background. The results showed that aged traffic and urban background categories are the most efficient CCN sources. By contrast, traffic category was observed as the main aerosol source with the highest frequency of occurrence (32%), however, its impact in the CCN activity is very limited likely due to lower particle mean diameter and hydrophobic chemical composition. Similarly, nucleation and growth categories, associated to NPF events, present high total particle number concentration with large frequency of occurrence (22% and 28%, respectively) but the CCN concentration for these categories is about half of the CCN concentration observed for the aged traffic category. Overall, these results showed the limited direct influence of traffic emissions on the CCN budget, however, when these particles undergo ageing processes, they have a significant influence on the CCN concentrations and may be an important CCN source (aged traffic category showed activation fraction of 0.41 at supersaturation ratio of 0.5%). Thus, urban particles could be transported to other remote environments, where clouds might form, modifying the CCN budget at those sites. Finally, to improve the CCN predictive capability at the high-altitude mountain site and understand the aerosol properties changes at this site, a new field campaign was performed focused on the chemical composition of particles and its relation to the activation properties. A more direct method to calculate CCN based on particle number size distribution measurements and aerosol hygroscopicity was investigated. At this site the sub-micron aerosol mass concentration was constituted of 70% of organic aerosol and, therefore, play a crucial role in defining the overall aerosol hygroscopicity. Different organic aerosol schemes were proposed to assess the organic hygroscopicity influence on CCN prediction. The main organic aerosol sources were identified using the Positive Matrix Factorization method. Results revealed the predominance of secondary organic aerosol with high degree of oxidation in the overall aerosol population. The CCN closure for all organic schemes showed good agreement with observations, with slope and correlation coefficients between predicted and measured CCN concentrations of 1.02-1.40 and 0.89-0.94, respectively, depending on the prediction scheme. However, when the aerosol population is affected by the atmospheric boundary layer (ABL) during morning and midday hours (affected by vertical transport of particles or NPF events), predicted CCN concentrations overestimate the measurements in a wider range (from 0 to 35%). These results evidenced that detailed knowledge of organic sources and organic hygroscopicity are not sufficient to obtain reliable CCN predictions at all atmospheric conditions, especially during ABL influence conditions. Thus, even at mountain environments where the aerosol population shows typically free troposphere conditions, changes on the aerosol properties and mixing state conditions might play a crucial role in CCN predictions.