Abstract.We have investigated the formation and early growth of atmospheric secondary aerosol particles building on atmospheric measurements. The measurements were part of the QUEST 2 campaign which took place in spring 2003 in Hyytiälä (Finland). During the campaign numerous aerosol particle formation events occurred of which 15 were accompanied by gaseous sulphuric acid measurements. Our detailed analysis of these 15 events is focussed on nucleation and early growth (to a diameter of 3 nm) of fresh particles. It revealed that new particle formation seems to be a function of the gaseous sulphuric acid concentration to the power from one to two when the time delay between the sulphuric acid and particle number concentration is taken into account. From the time delay the growth rates of freshly nucleated particles from 1 nm to 3 nm were determined. The mean growth rate was 1.2 nm/h and it was clearly correlated with the gaseous sulphuric acid concentration. We tested two nucleation mechanisms -recently proposed cluster activation and kinetic type nucleation -as possible candidates to explain the observed dependences, and determined experimental nucleation coefficients. We found that some events are dominated by the activation mechanism and some by the kinetic mechanism. Inferred coefficients for the two nucleation mechanisms are the same order of magnitude as chemical re-action coefficients in the gas phase and they correlate with the product of gaseous sulphuric acid and ammonia concentrations. This indicates that besides gaseous sulphuric acid also ammonia has a role in nucleation.

Abstract. The contribution of boundary layer (BL) nucleation events to total particle concentrations on the global scale has been studied by including a new particle formation mechanism in a global aerosol microphysics model. The mechanism is based on an analysis of extensive observations of particle formation in the BL at a continental surface site. It assumes that molecular clusters form at a rate proportional to the gaseous sulfuric acid concentration to the power of 1. The formation rate of 3 nm diameter observable particles is controlled by the cluster formation rate and the existing particle surface area, which acts to scavenge condensable gases and clusters during growth. Modelled sulfuric acid vapour concentrations, particle formation rates, growth rates, coagulation loss rates, peak particle concentrations, and the daily timing of events in the global model agree well with observations made during a 22-day period of March 2003 at the SMEAR II station in Hyytiälä, Finland. The nucleation bursts produce total particle concentrations (>3 nm diameter) often exceeding 10 4 cm −3 , which are sustained for a period of several hours around local midday. The predicted global distribution of particle formation events broadly agrees with what is expected from available observations. Over relatively clean remote continental locations formation events can sustain mean total particle concentrations up to a factor of 8 greater than those resulting from anthropogenic sources of primary organic and black carbon particles. However, in polluted continental regions anthropogenic primary particles dominate particle number and formation events lead to smaller enhancements of up to a factor of 2. Our results therefore suggest that particle concentrations in remote continental regions are dominated by nucleated particles while concentrations in polluted continental regions are dominated by primary particles. The effect of BL particle formation over tropical regions and the Amazon is negligiCorrespondence to: D. V. Spracklen (dominick@env.leeds.ac.uk) ble. These first global particle formation simulations reveal some interesting sensitivities. We show, for example, that significant reductions in primary particle emissions may lead to an increase in total particle concentration because of the coupling between particle surface area and the rate of new particle formation. This result suggests that changes in emissions may have a complicated effect on global and regional aerosol properties. Overall, our results show that new particle formation is a significant component of the aerosol particle number budget.

[1] We use a global aerosol microphysics model to predict the contribution of boundary layer (BL) particle formation to regional and global distributions of cloud condensation nuclei (CCN). Including an observationally derived particle formation scheme, where the formation rate of molecular clusters is proportional to gas-phase sulfuric acid to the power one, improves modeled particle size distribution and total particle number concentration at three continental sites in Europe. Particle formation increases springtime BL global mean CCN (0.2% supersaturation) concentrations by 3 -20% and CCN (1%) by 5 -50%. Uncertainties in particle formation and growth rates must be reduced before the accuracy of these predictions can be improved. These results demonstrate the potential importance of BL particle formation as a global source of CCN.

Abstract. This study investigates the connections between atmospheric sulphuric acid and new particle formation during QUEST III and BACCI/QUEST IV campaigns. The campaigns have been conducted in Heidelberg (2004) and Hyytiälä (2005), the first representing a polluted site surrounded by deciduous forest, and the second a rural site in a boreal forest environment. We have studied the role of sulphuric acid in particle formation and growth by determining 1) the power-law dependencies between sulphuric acid ([H 2 SO 4 ]), and particle concentrations (N 3−6 ) or formation rates at 1 nm and 3 nm (J 1 and J 3 ); 2) the time delays between [H 2 SO 4 ] and N 3−6 or J 3 , and the growth rates for 1-3 nm particles; 3) the empirical nucleation coefficients A and K in relations J 1 =A[H 2 SO 4 ] and J 1 =K[H 2 SO 4 ] 2 , respectively; 4) theoretical predictions for J 1 and J 3 for the days when no significant particle formation is observed, based on the observed sulphuric acid concentrations and condensation sinks. In both environments, N 3−6 or J 3 and [H 2 SO 4 ] were linked via a power-law relation with exponents typically ranging from 1 to 2. The result suggests that the cluster activation theory and kinetic nucleation have the potential to explain the observed particle formation. However, some differences between the sites existed: The nucleation coefficients were about an order of magnitude greater in Heidelberg than in Hyytiälä conditions. The time lags between J 3 and [H 2 SO 4 ] were consistently lower than the corresponding Correspondence to: I. Riipinen (ilona.riipinen@helsinki.fi) delays between N 3−6 and [H 2 SO 4 ]. The exponents in the J 3 ∝[H 2 SO 4 ] n J 3 -connection were consistently higher than or equal to the exponents in the relation N 3−6 ∝[H 2 SO 4 ] n N36 . In the J 1 values, no significant differences were found between the observed rates on particle formation event days and the predictions on non-event days. The J 3 values predicted by the cluster activation or kinetic nucleation hypotheses, on the other hand, were considerably lower on non-event days than the rates observed on particle formation event days. This study provides clear evidence implying that the main process limiting the observable particle formation is the competition between the growth of the freshly formed particles and their loss by scavenging, rather than the initial particle production by nucleation of sulphuric acid. In general, it can be concluded that the simple models based on sulphuric acid concentrations and particle formation by cluster activation or kinetic nucleation can predict the occurence of atmospheric particle formation and growth well, if the particle scavenging is accurately accounted for.

Abstract.A dimensionless theory for new particle formation (NPF) was developed, using an aerosol population balance model incorporating recent developments in nucleation rates and measured particle growth rates. Based on this theoretical analysis, it was shown that a dimensionless parameter L , characterizing the ratio of the particle scavenging loss rate to the particle growth rate, exclusively determined whether or not NPF would occur on a particular day. This parameter determines the probability that a nucleated particle will grow to a detectable size before being lost by coagulation with the pre-existing aerosol. Cluster-cluster coagulation was shown to contribute negligibly to this survival probability under conditions pertinent to the atmosphere. Data acquired during intensive measurement campaigns in Tecamac (MILAGRO), Atlanta (ANARChE), Boulder, and Hyytiälä (QUEST II, QUEST IV, and EUCAARI) were used to test the validity of L as an NPF criterion. Measurements included aerosol size distributions down to 3 nm and gas-phase sulfuric acid concentrations. The model was applied to seventyseven NPF events and nineteen non-events (characterized by growth of pre-existing aerosol without NPF) measured in diverse environments with broad ranges in sulfuric acid concentrations, ultrafine number concentrations, aerosol surface areas, and particle growth rates (nearly two orders of magnitude). Across this diverse data set, a nominal value of L =0.7 was found to determine the boundary for the occurCorrespondence to: C. Kuang (ckuang@bnl.gov) rence of NPF, with NPF occurring when L <0.7 and being suppressed when L >0.7. Moreover, nearly 45% of measured L values associated with NPF fell in the relatively narrow range of 0.1