Mineral dust and tropospheric chemistry (MINATROC 2000) Special Issue,Atmospheric Chemistry and Physics. Editor(s): Y. Balkanski and R. van Dingenen

Aerosol physical properties were measured at the Monte Cimone Observatory (Italy) from 1 June till 6 July 2000. The measurement site is located in the transition zone between the continental boundary layer and the free troposphere (FT), at the border between the Mediterranean area and Central Europe, and is exposed to a variety of air masses. Sub-µm number size distributions, aerosol hygroscopicity near 90% RH, refractory size distribution at 270°C and equivalent black carbon mass were continuously measured. Number size distributions and hygroscopic properties indicate that the site is exposed to aged continental air masses, however during daytime it is also affected by upslope winds. The mixing of this transported polluted boundary layer air masses with relatively clean FT air leads to frequent nucleation events around local noon. Night-time size distributions, including fine and coarse fractions for each air mass episode, have been parameterized by a 3-modal lognormal distribution. Number and volume concentrations in the sub-?m modes are strongly affected by the air mass origin, with highest levels in NW-European air masses, versus very clean, free tropospheric air coming from the N-European sector. During a brief but distinct dust episode, the coarse mode is clearly enhanced. The observed hygroscopic behavior of the aerosol is consistent with the chemical composition described by Putaud et al. (2004), but no closure between known chemical composition and measured hygroscopicity could be made because the hygroscopic properties of the water-soluble organic matter (WSOM) are not known. The data suggest that WSOM is slightly-to-moderately hygroscopic (hygroscopic growth factor GF at 90% relative humidity between 1.05 and 1.51), and that this property may well depend on the air mass origin and history. External mixing of aerosol particles is observed in all air masses through the occurrence of two hygroscopicity modes (average GF of 1.22 and 1.37, respectively). However, the presence of "less" hygroscopic particles has mostly such a low occurrence rate that the average growth factor distribution for each air mass sector actually appears as a single mode. This is not the case for the dust episode, where the external mixing between less hygroscopic and more hygroscopic particles is very prominent, and indicating clearly the occurrence of a dust accumulation mode, extending down to 50 nm particles, along with an anthropogenic pollution mode. The presented physical measurements finally allow us to provide a partitioning of the sub-?m aerosol in four non-overlapping fractions (soluble/volatile, non-soluble/volatile, refractory/non-black carbon, black carbon) which can be associated with separate groups of chemical compounds determined with chemical-analytical techniques (ions, non-water soluble organic matter, dust, elemental carbon). All air masses except the free-tropospheric N-European and Dust episodes show a similar composition within the uncertainty of the data (53%, 37%, 5% and 5% respectively for the four defined fractions). Compared to these sectors, the dust episode shows a clearly enhanced refractory-non-BC fraction (17%), attributed to dust in the accumulation mode, whereas for the very clean N-EUR sector, the total refractory fraction is 25%, of which 13% non-BC and 12% BC.

Abstract. The EU-project MINATROC (MINeral dust And TROpospheric Chemistry) aims at enabling an estimation of the influence of mineral dust, a major, but to date largely ignored component of tropospheric aerosol, on tropospheric oxidant cycles. Within the scope of this project continuous atmospheric measurements of gas-phase HNO3 and SO2 were conducted in June and July 2000 at the CNR WMO station, situated on Monte Cimone (MTC) (44°11' N --10°42' E, 2165 m asl), Italy. African air transporting dust is occasionally advected over the Mediterranean Sea to the site, thus mineral aerosol emitted from Africa will encounter polluted air masses and provide ideal conditions to study their interactions. HNO3 and SO2 were measured with an improved CIMS (chemical ionization mass spectrometry) system for ground-based measurements that was developed and built at MPI-K Heidelberg. Since HNO3 is a very sticky compound special care was paid for the air-sampling and background-measurement system. Complete data sets could be obtained before, during and after major dust intrusions. For the first time these measurements might provide a strong observational indication of efficient uptake of gas-phase HNO3 by atmospheric mineral-dust aerosol particles.

Physical and chemical characterizations of the atmospheric aerosol were carried out at Mt. Cimone (Italy) during the 4 June-4 July 2000 period. Particle size distributions in the size range 6nm-10µm were measured with a differential mobility analyzer (DMA) and an optical particle counter (OPC). Size-segregated aerosol was sampled using a 6-stage low pressure impactor. Aerosol samples were submitted to gravimetric and chemical analyses. Ionic, carbonaceous and refractory components of the aerosol were quantified. We compared the sub- and superµm aerosol mass concentrations determined by gravimetric measurements (mGM), chemical analyses (mmCA), and by converting particle size distribution to aerosol mass concentrations (mmSD). Mean random uncertainties associated with the determination of mmGM, mmCA, and mmSD were assessed. The three estimates of the sub-µm aerosol mass concentration agreed, which shows that within experimental uncertainty, the sub-µm aerosol was composed of the quantified components. The three estimates of the super-µm aerosol mass concentration did not agree, which indicates that random uncertainties and/or possible systematic errors in aerosol sampling, sizing or analyses were not adequately accounted for. Aerosol chemical composition in air masses from different origins showed differences, which were significant in regard to experimental uncertainties. During the Saharan dust advection period, coarse dust and fine anthropogenic particles were externally mixed. No anthropogenic sulfate could be found in the super-µm dust particles. In contrast, nitrate was shifted towards the aerosol super-µm fraction in presence of desert dust.

An intensive measurement campaign was performed in June 2000 at the Mt. Cimone station (44°11' N-10°42' E, 2165 m asl, the highest mountain in the northern Italian Apennines) to study photochemical ozone production in the lower free troposphere. In general, average mixing ratios of important trace gases were not very high (121 ± 20 ppbv CO, 0.284 ± 0.220 ppbv NOx, 1.15 ± 0.8 ppbv NOy, 58 ± 9 ppbv O3), which indicates a small contribution by local pollution. Those trace gas levels are representative of continental background air, which is further supported by the analysis of VOCs (e.g.: C2H6 = (905 ± 200) pptv, C3H8 = (268 ±110) pptv, C2H2 = (201 ± 102) pptv, C5H8 = (111 ± 124) pptv, benzene = (65 ± 33) pptv). Furthermore, significant diurnal variations for a number of trace gases (O3, CO, NOx, NOy, HCHO) indicate the presence of free tropospheric airmasses at nighttime as a consequence of local catabatic winds. Average mid-day peroxy radical concentrations at Mt. Cimone are of the order of 30 pptv. At mean NO concentrations of the order of 40 pptv this gives rise to significant in situ net O3 production of 0.1-0.3 ppbv/hr. The importance of O3 production is supported by correlations between O3, CO, NOz, and HCHO, and between HCHO, CO and NOy.

Its location in the Mediterranean region and its physical characteristics render Mt. Cimone (44°11' N, 10°42' E), the highest peak of the Italian northern Apennines (2165 m asl), particularly suitable to study the transport of air masses from the north African desert area to Europe. During these northward transports 12 dust events were registered in measurements of the aerosol concentration at the station during the period June–December 2000, allowing the study of the impact of mineral dust transports on free tropospheric ozone concentrations, which were also measured at Mt. Cimone. Three-dimensional backward trajectories were used to determine the air mass origin, while TOMS Aerosol Index data for the Mt. Cimone area were used to confirm the presence of absorbing aerosol over the measurement site. A trajectory statistical analysis allowed identifying the main source areas of ozone and aerosols. The analysis of these back trajectories showed that central Europe and north and central Italy are the major pollution source areas for ozone and fine aerosol, whereas the north African desert regions were the most important source areas for coarse aerosol and low ozone concentrations. During dust events, the Mt. Cimone mean volume concentration for coarse particles was 6.18 µm3/cm3 compared to 0.63 µm3/cm3 in dust-free conditions, while the ozone concentrations were 4% to 21% lower than the monthly mean background values. Our observations show that surface ozone concentrations were lower than the background values in air masses coming from north Africa, and when these air masses were also rich in coarse particles, the lowest ozone values were registered. Moreover, preliminary results on the possible impact of the dust events on PM10 and ozone values measured in Italian urban and rural areas showed that during the greater number of the considered dust events, significant PM10 increases and ozone decreases have occurred in the Po valley.