Atmospheric Environment (v.37, #19)

Collocated parallel measurements of PM2.5 and PM10 were conducted at 7 sites in Switzerland since January 1998, constituting now one of the longest comparative data sets for PM2.5 and PM10 in Europe. The range of the long-term mean concentrations of PM2.5 was between 7.9 μg/m3 at Chaumont and 24.4 μg/m3 in Lugano. For the sites within the Swiss plateau this range narrows from 15.1 μg/m3 at the rural site of Payerne to 20.8 μg/m3 at the directly traffic exposed site of Bern. The long-term averages of the PM2.5/PM10 ratios of the daily values vary only from 0.75 to 0.76, with the exception of the traffic exposed site of Bern (0.59). The correlation between the daily values of PM2.5 and PM10 at all sites is generally high. For PM10, as well as for PM2.5 the highest concentrations are normally observed during wintertime. An exception is Chaumont (1140-m a.s.l.), which is often positioned above the inversion layer during wintertime and, therefore, has the lowest concentration during wintertime. A minimum of the PM2.5/PM10 ratio is often found during spring, probably due to the influence of relatively coarse biogenic particles. Though the sites have quite different exposition characteristics, the correlation of the daily values of PM2.5 and PM10 between the different sites of the Swiss plateau is very high, indicating a dominant influence of regional meteorology over local events and sources. The findings imply that from the point of view of an efficient use of financial and personal resources, the number of collocated PM2.5 measurements at PM10 sites in a monitoring network can be kept quite limited. The saved resources could rather be used to investigate other particle related parameters providing substantial new information (e.g. on particle sources, formation and effects) like PM1, particle number concentrations, morphology or chemical composition.
Keywords: Fine particles; Parallel measurement; Collocated measurement; PM monitoring; PM spatial distribution;

Size fractionated speciation of sulfate and nitrate in airborne particulates in Beijing, China by Lun Xiaoxiu; Zhang Xiaoshan; Mu Yujing; Nang Anpu; Jiang Guibin (2581-2588).
We report on the speciation of sulfate, nitrate with special attention to their ammonium salts during 1999–2000 in Beijing. Higher concentrations in fine particles were observed for ammonium sulfate and other sulfates during winter. Ammonium sulfate concentration ranged from 2.78 to 23.7 μg m−3 in fine particles, and ranged from 0.15 to 1.39 μg m−3 in coarse particles. Other sulfate concentrations ranged from 1.51 to 11.9 μg m−3 in fine particles, and ranged from 0.10 to 1.00 μg m−3 in coarse particles. Ammonium nitrate was mainly in fine particles with the highest concentration in December (11.9 μg m−3). Other nitrates were mainly distributed in the particles of 2.1–4.7 μm with concentrations ranged between 0.36 and 2.54 μg m−3, and the highest concentration of other nitrates was observed in fine particles in January (9.32 μg m−3). XRD analysis showed that ammonium sulfates existed in different forms in atmospheric aerosols. The results suggested the importance of heterogeneous formations of sulfate and nitrate in Beijing especially in winter.
Keywords: Airborne particulates; Speciation; Sulfate; Nitrate;

Influence of ionic strength on aqueous oxidation of SO2 catalyzed by manganese by Janja Turšič; Irena Grgić; Boštjan Podkrajšek (2589-2595).
We studied the influence of ionic strength as well as type of supporting electrolyte on the rate of oxidation of dissolved SO2 catalyzed by Mn(II) under dark conditions at pH 3. Two electrolytes were chosen; NaCl and NaNO3. NaCl is an important component of atmospheric particles of marine origin and NO3 represents a considerable part of aerosols of continental origin. Laboratory experiments were performed in a system that enables continuous introduction of gas mixture (SO2/air) into synthetic aqueous solution. The solubility of SO2 in aqueous solution increases to some extent with an increase in ionic strength from 0.001 to 0.1 mol l−1 of both supporting electrolytes. The results suggest that the effect of concentration and nature of electrolyte on the catalytic S(IV) autoxidation are not negligible parameters. In NaNO3 medium, the rate constant decreases with increasing ionic strength, and the inhibition appears to be a consequence of the primary salt effect. In NaCl medium, the dependence of the rate constant on the ionic strength is more complex. The reaction rate of the catalyzed reaction is almost constant or increases slightly with the ionic strength in the concentration range of 0.001–0.02 mol l−1, and subsequently a considerable decrease was noticed. It seems that two opposite effects of Cl could be important, the inhibiting and catalytic one.
Keywords: Aqueous S(IV) oxidation; Catalysis; Primary salt effect; Inhibition; Manganese;

Difference between filtering-type bulk and wet-only data sets based on site classification by Masahide Aikawa; Takatoshi Hiraki; Motonori Tamaki; Mitsuru Shoga (2597-2603).
In order to clarify the limitations and the effectiveness of filtering-type bulk sampler, a study on precipitation collected by filtering-type bulk and wet-only samplers was performed. At an urban site, there was a statistically significant difference between the filtering-type bulk and wet-only samples at a 1% significance level in the pH and major chemical components. At a suburban site, pH, NH4 +, and nss-Ca2+ showed a difference at a 1% significance level and NO3 at a 5% significance level, but nss-SO4 2− did not show any difference even at a 5% significance level. At a rural site, although there was a difference in nss-Ca2+ at a 1% significance level and in pH at a 5% significance level, nss-SO4 2−, NO3 , and NH4 + did not show any difference. The attributions of the deposition of gaseous compounds and particulate matter in a filtering-type bulk sampler were estimated to be 32% (nss-SO4 2−), 46% (NO3 ), 36% (NH4 +), and 75% (nss-Ca2+) at an urban site, 14% (nss-SO4 2−), 16% (NO3 ), 27% (NH4 +), and 42% (nss-Ca2+) at a suburban site, and 16% (nss-SO4 2−), 17% (NO3 ), 19% (NH4 +), and 47% (nss-Ca2+) at a rural site. The degree of pollution at a sampling site should be taken into account when using a data set collected with filtering-type bulk samplers.
Keywords: Filtering-type bulk sampler; Wet-only sampler; Acid precipitation; Site classification; Z-tests;

Long-range persistence of acid deposition by Jianlin Zhu; Zhiguo Liu (2605-2613).
Rescaled range analysis and detrended fluctuation analysis are applied to verify the presence of long-range persistence (LRP) of weekly hydrogen ion deposition (WHD) series in National Atmospheric Deposition Program, USA, for the period 1978–2001. Both methods confirm that WHD fluctuations exhibit fractal Brown motion with two different temporal scaling regimes and LRP exists in the series, indicating that the succeeding WHD are not independent of each other. The fluctuations of WHD series behave as a self-affine fractal with self-similarity. WHD caused by pollution (WHDP) shows scaling behavior similar to that of WHD. The deseasonized WHD/WHDP series display uniform scaling regimes with temporal persistence. The dependence duration may persist 6–9 years, implying that the LRP, which is believed to be related to the self-organized critical behavior of atmosphere in the paper, should be considered in the trend prediction of acid deposition as an important factor.
Keywords: Acid rain; Detrended fluctuation analysis; Rescaled range analysis; Self-similarity; Self-organized criticality;

Summer nocturnal ozone maxima in Göteborg, Sweden by Ingegärd Eliasson; Sofia Thorsson; Yvonne Andersson-Sköld (2615-2627).
The magnitude and frequency of nocturnal ozone maxima in a high mid-latitude city (Göteborg, Sweden) has been analysed. Nocturnal ozone maxima have been reported from cities in Europe and North America and can be explained by vertical mixing of high ozone concentrations from higher levels or horizontal transportation from rural areas through local and mesoscale wind systems. Data from four summer months (May–August) in 1994 were used to analyse the relative importance of local- and mesoscale wind systems and vertical mixing in Göteborg during clear and calm weather conditions. Results show that nocturnal ozone maxima frequently occur during these conditions, with a magnitude up to 104 μg m−3. The nocturnal ozone maxima were positively correlated to both situations with a well-developed land breeze and situations with vertical mixing. During the period investigated, in total 17 nights with secondary ozone maximum occurred. The majority of the secondary ozone maxima (80%) appeared early in the night, i.e. an ozone increase within the first 3 h after sunset and sometimes even two peaks occurred. Seven of these occasions can be explained by horizontal advection, eight by vertical mixing and five cannot by certainty be explained to be due to horizontal or vertical transportation only.During the measurement period the Swedish guideline of 80 μg m−3 (for 1-h value) was exceeded 55% of the days (i.e. 68 days, 557 h) and 33% of the nights (i.e. 41 nights, 103 h) in the central parts of Göteborg. The results thus show that in Scandinavia nocturnal ozone concentrations must be considered.
Keywords: Secondary ozone maximum; Urban; City; Local- and mesoscale wind systems; Advection; Vertical mixing;

Diurnal and annual characteristics of particle mass and number concentrations in urban, rural and Arctic environments in Finland by Lauri Laakso; Tareq Hussein; Päivi Aarnio; Mika Komppula; Veijo Hiltunen; Yrjö Viisanen; Markku Kulmala (2629-2641).
In order to be able to detect annual and diurnal variation the nucleation, Aitken and accumulation mode aerosol number concentrations were investigated. In this study particulate matter and aerosol number concentrations based on three years (1999–2001) measurements in four different places, Helsinki, Hyytiälä, Pallas and Värriö in Finland were utilized. Mean PM10 concentration was found to be in urban, urban background, disturbed rural and rural conditions 18.7, 15.3, 10.2 and 6.9  μg m −3 . PM2.5 concentrations in urban, urban background and rural conditions were 9.6, 8.2 and 5.8  μg m −3 . PM1 concentration in rural area were 4.1  μg m −3 . Total number concentrations between 10 and 500  nm in urban, rural and arctic background were 16660, 2110 and about 920  cm −3 , respectively. Annual cycle of different particle sizes was also investigated. Particle mass was found to have maximum during the spring and autumn. Nucleation mode particle concentrations had their maximum both in urban and rural conditions in spring and autumn. For Aitken mode a similar variation was found. Accumulation mode particles were found to have the highest concentrations during summer and the lowest concentrations during winter. In urban conditions local sources drive over natural variability, but a natural cycle can still be seen. Diurnal variation was found to be associated with traffic in urban conditions. In rural conditions, diurnal variation of nucleation mode particles were connected to new particle formation. It was also found, that there is no correlation between ultrafine (<90  nm) and larger (>90  nm) particles.
Keywords: Atmospheric aerosols; Particulate mass; Number concentration;

Within this study denuder laboratory and field experiments as well as direct kinetic laboratory studies were performed in order to quantify an artefact by the aqueous phase formation of nitrous acid from dissolved NO2 and SO2 at wetted denuder walls. A wet annular denuder with a stripping solution of K2CO3 (1 mM, pH=10) with high collection efficiency (>97%) for measurements of acid tropospheric gases (HNO2, HNO3 and HCl) without substantial influence of dispersed particulate matter is used. The samples from the denuder are analysed by ion chromatography. To study the reaction of NO2(aq) with SO3(aq) 2− in the temperature interval between 288 and 328 K for the system NO 2(aq) +SO 3(aq) 2− ⇌ k20,k−20 [NO 2−SO 3]2−, [NO 2−SO 3]2−
NO 2 +SO 3 excimer laser photolysis–long path absorption was applied. The rate constants were measured following the decay of absorbance of the NO2 radical at 400 nm. The rate constant for the adduct formation from NO2(aq) with SO3(aq) 2− is found to be independent of temperature with k 20 (288⩽T⩽328 K)=(1.38±0.2)×10−7  l mol−1  s−1. The following temperature-dependencies for the forward k 22, the reverse reactions k –20, k −22 and for the equilibrium constants K 20 and K 22 were obtained: At pH=10.0: k−20(T)=(3.5±0.6)·106   exp  [−(2440±800)  K/T]  s −1, K20(T)=(1.06±900)  exp  [−(2833±2000)  K/T]  l mol −1. At pH=4.5: k22(T)=(5.3±0.1)×1012   exp  [−(4500±1000)  K/T]  l mol −1   s −1, k−22(T)=(3.7±0.6)×106 exp  [−(3550±800)  K/T]  s −1, K22(T)=(1.5±1.8)×106 exp  [(950±300)  K/T]  l mol −1. These latter data are of interest for tropospheric multiphase modelling.The reaction involves the formation of a long-lived intermediate [NO2–SO3]2− which decays into NO2(aq) and SO3(aq) . To obtain the decay rate constant of this intermediate results at pH=10.0 of the laboratory denuder measurements were fitted to the above mechanism combined to a description of phase transfer resulting in k21=(8.4±0.1)×10−3   s −1   at  T=298  K. The overall HNO2 artefact formation in a wet annular denuder with a solution volume of 10 ml is described by (T=298 K) [HNO 2 ]art k =0.0056[NO 2 ]+(0.0022/ppb)[NO 2 ] [SO 2 ]. The first term accounts for HNO2 formation from NO2 and water alone and the second term quantifies the aqueous phase reaction of dissolved NO2 and SO2. This equation was used for a correction of HNO2 field measurements with the described denuder technique at the IfT-research station Melpitz (Germany) and the results are discussed. Corrected HNO2 measurements were compared with HNO2-DOAS measurements. Generally, the agreement with DOAS and denuder HNO2 measurements is improved after correction of the denuder data according to the above equation. Field measurements are discussed in view of NO2 concentration and meteorological conditions.
Keywords: Nitrous acid; Kinetic investigation; Wet annular denuder; Artefact correction; Field measurements;

In this study a Lagrangain photochemical box model has been used to explore gaseous/heterogeneous sulfur plume chemistry initiated by ship traffic in the MBL. The results indicate that although SO2 ship emission rates have an impact on the concentration levels of SO2, there is no evidence of non-linear loss processes in this chemical system. This finding as well as others have demonstrated that the major loss pathway for SO2 is via heterogeneous loss to sea-salt aerosol. This loss rate, however, can be greatly influenced by the MET condition selected for the MBL. For example, the “stable” MET condition leads to the maximum loss rate, whereas the “unstable” condition gives the minimum value. In the case of H2SO4(g), the “stable” MET condition provides both the highest SO2 and OH concentrations, and hence, the formation rate is also maximized. It was found that increases in aerosol surface area (i.e., heterogeneous loss) as well as shifts to higher latitudes (e.g., lower T and higher O3 levels) tended to increase the rate of loss of SO2 due to enhancements in the rates of heterogeneous chemistry. However, the single largest factor that influenced the rate of SO2 loss was the presence of BL clouds where oxidation via scavenged H2O2 was the dominant process. Quite significant also was the finding that photochemical processes, because of their production of OH, O3, and H2O2, are strongly coupled to SO2 sea-salt heterogeneous reactions, cloud heterogeneous reactions, and gas-phase reactions to form H2SO4. Not surprisingly, therefore, model predictions of the rate of formation of new particles within ship plumes were found to be quite dependent on the intensity of photochemistry.
Keywords: Ship plumes; SO2 oxidation; Heterogeneous reactions; H2SO4 formation; Photochemistry; New particle formation;

Elemental analysis of airborne particulate matter and cooling water in west Texas residences by Helmut Paschold; Wen-Whai Li; Hugo Morales; Nicholas E. Pingitore; Beata Maciejewska (2681-2690).
Ten residences in El Paso, TX were monitored in the summer of 2001 for concurrent indoor and outdoor PM2.5 and PM10 10-min mass concentrations in an attempt to establish the indoor–outdoor PM correlation for typical west Texas residences equipped with evaporative coolers. Simultaneously, the modified tapered element oscillating microbalance instruments were used to collect 48-h PM2.5 and PM10 samples, both indoor and outdoor, for elemental analysis. Water samples were also taken at each home for chemical and PM correlation analysis.The air and water samples were analyzed by inductively coupled plasma–mass spectrometry, with 30 elements identifiable above detection limits in the outdoor PM10 samples and fewer elements in the PM2.5 and indoor PM10. Indoor elemental concentrations in PM10 were found to be approximately 50–70% lower than outdoor concentrations in nine of ten homes, consistent with the PM10 indoor/outdoor (I/O) mass concentrations previously reported. PM2.5 I/O ratio correlations were not as strong as for PM10; however, reduced correlations could be attributed to a pattern of recurring outlier data pairs, consisting of the same three or four elements in all ten homes.Comparison of the elemental concentrations of the evaporative cooler supply water and indoor PM demonstrated little or no correlation, yielding a conclusion that the evaporative cooler was not introducing dissolved solids from the supply water into indoor air. Overall, evaporative cooling appears to provide a cleansing of indoor air.
Keywords: Air pollution; Evaporative cooler; Particulate matter; Indoor–outdoor ratio; Element composition;

Correlations between short-term indoor and outdoor PM concentrations at residences with evaporative coolers by Wen-Whai Li; Helmut Paschold; Hugo Morales; Julian Chianelli (2691-2703).
Airborne particulate matter (PM) was monitored at 10 residences in the El Paso, Texas region in the summer of 2001. Concurrent indoor and outdoor 10-min averaged PM2.5 and PM10 concentrations were recorded for 2 days each to establish the indoor–outdoor PM correlation for typical west Texas residences equipped with evaporative coolers.Indoor PM concentrations stabilize in approximately 10 min in a typical house equipped with evaporative coolers. If the ambient PM concentration remains steady, a 10-min average indoor air sample after the first 10-min period would contain 99% outdoor air and a 1-h average indoor air sample would actually be represented by 95% of the outdoor air. A strong diurnal pattern of PM10 indoor and outdoor was observed in 9 out of the 10 tested houses independent of the possible human activities and other indoor sources at each residence. Consistent with prior regional studies, indoor and outdoor PM10 concentrations at these houses frequently peaked with strong association with each other in the evening hours between 6 and 9 pm. In addition, it is observed that both indoor and outdoor PM10 peaked after the wind speed and wind gust peaked. Indoor PM concentration peaks clearly correlated with documentation of human activities, however, these peaks tended to be of shorter duration due to the high ventilation rates of the evaporative coolers.Evaporative coolers were found to act as PM filters that effectively replace indoor air rapidly creating indoor concentrations approximately 40% of outdoor PM10 and 35% of outdoor PM2.5. Both cooler types, rigid media and aspen pad, appeared to produce similar reduction rates for both PM2.5 and PM10.
Keywords: Air pollution; Evaporative cooler; Particulate matter; Indoor–outdoor ratio; Paso del Norte;

History of tropospheric ozone for the San Bernardino Mountains of Southern California, 1963–1999 by E.Henry Lee; David T. Tingey; William E. Hogsett; John A. Laurence (2705-2717).
Keywords: Ozone trends; Photochemical oxidants; Air pollution; Southern California; San Bernardino Mountains;

A semi-empirical model for predicting the effect of changes in traffic flow patterns on carbon monoxide concentrations by Kim N. Dirks; Murray D. Johns; John E. Hay; Andrew P. Sturman (2719-2724).
A simple semi-empirical model for predicting the effect of changes in traffic flow patterns on carbon monoxide concentrations is presented. The traffic component of the model requires average vehicle emission rate estimates for a range of driving conditions, as well as traffic flow data for the site of interest. The dispersion component of the model is based on a modified empirically optimised box model requiring only wind speed and direction information. The model is evaluated at a suburban site in Hamilton, New Zealand. Despite the simplicity of the model, produces reliable concentration predictions when tested on days with significantly different traffic flow patterns from those days with which the optimum model parameters were evaluated.
Keywords: Carbon monoxide; Urban air quality; Traffic modelling; Vehicle emissions; Emperical modelling;