Mixing state of size-selected submicrometer particles in the Arctic in May and September 2012.

Aerosols have been associated with large uncertainties in estimates of the radiation budget and cloud formation processes in the Arctic. This paper reports the results of a study of in situ measurements of hygroscopicity, fraction of volatile species, mixing state, and off-line morphological and elemental analysis of Aitken and accumulation mode particles in the Arctic (Ny-Ålesund, Svalbard) in May and September 2012. The accumulation mode particles were more abundant in May than in September. This difference was due to more air mass flow from lower latitude continental areas, weaker vertical mixing, and less wet scavenging in May than in September, which may have led to a higher amount of long-range transport aerosols entering the Arctic in the spring. The Aitken mode particles observed intermittently in May were produced by nucleation, absent significant external mixing, whereas the accumulation mode particles displayed significant external mixing. The occurrence of an external mixing state was observed more often in May than in September and more often in accumulation mode particles than in Aitken mode particles, and it was associated more with continental air masses (Siberian) than with other air masses. The external mixing of the accumulation mode particles in May may have been caused by multiple sources (i.e., long-range transport aerosols with aging and marine aerosols). These groups of externally mixed particles were subdivided into different mixing structures (internal mixtures of predominantly sulfates and volatile organics without nonvolatile species and internal mixtures of sulfates and nonvolatile components, such as sea salts, minerals, and soot). The variations in the mixing states and chemical species of the Arctic aerosols in terms of their sizes, air masses, and seasons suggest that the continuous size-dependent measurements observed in this study are useful for obtaining better estimates of the effects of these aerosols on climate change.

Snow cover duration in northern Finland and the influence of key variables through a conceptual framework based on observed variations in snow depth.

Seasonal snow cover duration is the net result from many processes acting on snow fallen on the Earth's surface. Several of these processes feed back into the atmosphere-cryosphere system causing non-linear interactions. The timing of snow retreat is of essential importance, but the duration of snow cover has large spatiotemporal variabilities. However, from a large data set of observed snow depth changes in northern Finland, systematic similar evolutions are identified that allow for a considerable simplification and reduction of the complexity in snow depth changes. Here, a novel conceptual framework is designed based on dividing the season into two main periods (dark and bright period, based on solar irradiance), for which snow depth decrease is parameterized based on three variables, average temperature, incoming shortwave radiation, and light-absorbing particles (LAP) in the snow. The processes are simplified into two linear relations, and a new formulation for concentration enhancement of LAP, which is dependent on snow depth decrease, is given. The results show that the seasonal snow cover duration is shifted by about one day for every 10 mm snow water equivalent of precipitation. This effect is comparable in scale to that of doubling of the amount of LAP concentration in snow. We also found that the combined shift in snow cover duration from interannual variability in ambient temperature and shortwave radiation (warm and bright vs. cold and dark season) is large enough to explain the variability of a couple of weeks for a given precipitation amount in Northern Finland.

Spatiotemporal variability of light-absorbing carbon concentration in a residential area impacted by woodsmoke.

Residential wood combustion (RWC) is responsible for 33% of the total carbon mass emitted in Europe. With the new European targets to increase the use of renewable energy, there is a growing concern that the population exposure to woodsmoke will also increase. This study investigates observed and simulated light-absorbing carbon mass (MLAC) concentrations in a residential neighborhood (Lycksele, Sweden) where RWC is a major air pollution source during winter. The measurement analysis included descriptive statistics, correlation coefficient, coefficient of divergence, linear regression, concentration roses, diurnal pattern, and weekend versus weekday concentration ratios. Hourly RWC and road traffic contributions to MLAC were simulated with a Gaussian dispersion model to assess whether the model was able to mimic the observations. Hourly mean and standard deviation concentrations measured at six sites ranged from 0.58 to 0.74 microg m(-3) and from 0.59 to 0.79 microg m(-3), respectively. The temporal and spatial variability decreased with increasing averaging time. Low-wind periods with relatively high MLAC concentrations correlated more strongly than high-wind periods with low concentrations. On average, the model overestimated the observations by 3- to 5-fold and explained less than 10% of the measured hourly variability at all sites. Large residual concentrations were associated with weak winds and relatively high MLAC loadings. The explanation of the observed variability increased to 31-45% when daily mean concentrations were compared. When the contribution from the boilers within the neighborhood was excluded from the simulations, the model overestimation decreased to 16-71%. When assessing the exposure to light-absorbing carbon particles using this type of model, the authors suggest using a longer averaging period (i.e., daily concentrations) in a larger area with an updated and very detailed emission inventory.

Feasibility of analyzing fine particulate matter in air using solid-phase extraction membranes and dynamic subcritical water extraction.

We have evaluated the feasibility of using Empore solid-phase extraction (SPE) membranes as an alternative to conventional techniques for sampling fine airborne particulate matter (PM), including nanoparticles, utilizing a scanning mobility particle sizer (SMPS) and a condensation particle counter to evaluate their efficiency for trapping fine particles in the 10-800 nm size range. The results demonstrate that the membranes can efficiently trap these particles and can then be conveniently packed into an extraction cell and extracted under matrix solid-phase dispersion (MSPD) conditions. The potential utility of sampling PM using Empore membranes followed by dynamic subcritical water extraction (DSWE) for fast, efficient, class-selective extraction of polycyclic aromatic hydrocarbons (PAHs) associated with the particles, prior to changing the solvent and analysis by GC/MS, was then explored. The performance of the method was tested using National Institute of Standards and Technology (NIST)-certified "urban dust" reference material (SRM 1649a) and real samples collected at a site in central Rome with heavy road traffic. The method appears to provide comparable extraction efficiency to that of conventional techniques and with using GC/MS, detection limits ranged in the few picograms per cubic meter level. Sampling PM by Empore membranes may reduce the risks of losses of semivolatile compounds, while allowing relatively high sampling flow rates and safe sample storage. Moreover, the combination of MSPD with DSWE permits specific fractions of the PM components to be eluted, thereby generating clean extracts and reducing both analysis time and sample manipulation.

Estimates of black carbon and size-resolved particle number emission factors from residential wood burning based on ambient monitoring and model simulations.

In this paper we derive typical emission factors for coarse particulate matter (PM(10)), oxides of nitrogen (NO(x)), black carbon (BC), and number particle size distributions based on a combination of measurements and air quality dispersion modeling. The advantage of this approach is that the emission factors represent integrated emissions from several vehicle types and different types of wood stoves. Normally it is very difficult to estimate the total emissions in cities on the basis of laboratory measurements on single vehicles or stoves because of the large variability in conditions. The measurements were made in Temuco, Chile, between April 18 and June 15, 2005 at two sites. The first one was located in a residential area relatively far from major roads. The second site was located in a busy street in downtown Temuco where wood consumption is low. The measurements support the assumption that the monitoring sites represent the impact of different emission sources, namely traffic and residential wood combustion (RWC). Fitting model results to the available measurements, emission factors were obtained for PM(10) (RWC = 2160 +/- 100 mg/kg; traffic = 610 +/- 51 mg/veh-km), NO(x) (RWC = 800 +/- 100 mg/kg; traffic = 4400 +/- 100 mg/veh-km), BC (RWC = 74 +/- 6 mg/kg; traffic = 60 +/- 3 mg/veh-km) and particle number (N) with size distribution between 25 and 600 nm (N(25-600)) (RWC = 8.9 +/- 1 x 10(14) pt/kg; traffic = 6.7 +/- 0.5 x 10(14) pt/veh-km). The obtained emission factors are comparable to results reported in the literature. The size distribution of the N emission factors for traffic was shown to be different than for RWC. The main difference is that although traffic emissions show a bimodal size distribution with a main mode below 30 nm and a secondary one around 100 nm, RWC emissions show the main mode slightly below 100 nm and a smaller nucleation mode below 50 nm.