We have made large-scale intensive observations, named the Integrated Measurement Program for Aerosol and Oxidant Chemistry in Tokyo (IMPACT) , at the University of Tokyo, RCAST in Tokyo in 2003 ∼ 2004 to understand key processes which control the abundances of different types of aerosols in urban air. Various kinds of aerosols in fine mode and their precursor gases were measured with high accuracies using the most advanced measurement techniques, including an aerosol mass spectrometer (AMS) . This paper is a summary of the important findings on carbonaceous aerosols (organic aerosols and elemental carbon) . First, we present systematic evaluations of the uncertainties in aerosol measurement. We then discuss the temporal variations and chemical composition of organic aerosols based on the observational data. Water solubility of secondary organic aerosol (SOA) is quantified. Recent results on modeling the distributions of SOA are presented. Finally, the diurnal and seasonal variations of elemental carbon (EC) are analyzed in terms of temporal variations of the EC emissions.
We have conducted daily sampling of atmospheric aerosols since 1986 at Sakai, Osaka and analyzed the ionic constituents (Cl-, NO3-, SO42-, Na+, NH4+, K+, Mg2+, Ca2+) in the samples collected from 1986 to 2004 with ion chromatography. The daily variations of the concentrations show that the particle mass (PM) and ionic constituent concentrations have seasonal variations and yearly trends. In the ionic constituents, Cl- and NO3- exhibited the most significant seasonal variations with low mean concentrations in summer and high mean concentrations in winter and autumn. Yearly mean of PM trends to decrease from 1995 to 2000 because administrative management seems to successfully control the emission of diesel particles. Cl-, NO3-, NH4+, K+ also have a decreasing trend, and among them Cl- is decreased at a high rate since 1997. This decrease of Cl- could be attributed to the reduction of gaseous chlorine emitted mainly from incineration plants. The decrease of NO3- could have been caused by the decreasing trend of NOx from which NO3- is produced by the photo chemical process. On contrary, SO42- has an increasing trend in 1986 ∼ 1995 due to SO2 emission increase in China. From 1995 to 1997, the reduction of sulfur in the diesel fuel contributed a slight decrease in SO42- concentration, but it remains fairly constant in 1998 ∼ 2004.
Emission characteristics of ultrafine particles emitted from a diesel vehicle were measured on the chassis dynamometer using dilution tunnel and high speed particle sizer in transient test cycle. Averaged size distribution for whole test cycle showed a single mode size distribution with a peak of around 50 ∼ 60 nm diameter. Nanoparticles which have a peak diameter about 20 nm were observed only under deceleration condition in case of the tested vehicle. Long term measurement of size distribution and number concentration of ultrafine particles have been performed using Scanning Mobility Particle Sizer to understand the impact of vehicle emission and behavior of the ultrafine particles in the atmosphere at three roadside sites. At roadside, bimodal size distribution with peak diameters of 20 ∼ 30 and 60 ∼ 90 nm was observed in winter. On the other hand unimodal size distribution with peak diameters of 30 ∼ 40 nm was observed in summer. Number concentration is influenced by traffic of heavy duty vehicles, meteorological condition and background atmosphere. The correlation analysis result shows a good correlation between number and NO concentration in the roadside atmosphere.
Elemental and organic carbons (EC and OC) in PM2.5 at roadside, urban, and rural sites were measured from 2002 to 2006 by thermal optical reflectance (TOR) method. Total carbon (TC) were 10 ∼ 40 μg/m3 at roadside sites, whereas 5 ∼ 15 μg/m3 at urban and rural sites. The fractions of TC in PM2.5 (TC/PM2.5) at roadside sites were larger than those at urban and rural sites. EC were 5 ∼ 32 μg/m3 at roadside sites, whereas less than 5 μg/m3 at urban and rural sites. The fractions of EC in TC (EC/TC) at roadside sites were more than 0.5, whereas 0.3 ∼ 0.5 at urban and rural sites, which indicated the influence of vehicle exhaust, especially EC-rich diesel exhaust at roadside sites. OC were about 5 μg/m3, however, those were higher at heavily polluted roadside sites and at a rural site in early winter probably owing to open biomass burning. Part of OC was pyrolized when OC and EC were thermally separated, and the fraction of pyrolized carbon in OC at urban and rural sites was higher than that at roadside sites. The concentration level of EC and OC, TC/PM2.5, and EC/TC showed seasonal changes probably affected by meteorological conditions and seasonal differences in emission/formation. The size distributions of EC and OC were also measured by thermal method. The size distributions of EC had modal diameters at around 0.2 ∼ 0.3 μm at a roadside site, whereas at around 0.5 μm at an urban and a rural site, which reflected the differences in residence time and growth process such as coagulation and condensation.
Air pollutant concentration in an urban area is determined by the reaction-diffusion system, which is usually non-linear. Thus, the concentration often shows complicated tempo-spatial variations. Although such variations were first recognized in the 1980's, there have been few observations since then because of technical insufficiency for the measurement. However, recent development of technology has enabled us to measure the air pollutant concentration which complicatedly varies in time and space. In this study, we adapted a dust sensor for the field measurement. The sensor was originally designed as a built-in device in an air cleaner, and it had sufficiently high time resolution and was compact and inexpensive enough to conduct multipoint measurements of aerosol concentration. The tempo-spatial variations of aerosol concentration were measured at a heavily polluted intersection in Kawasaki City. The observed concentration distributions showed complex patterns being significantly different from those predicted based on a simple decay depending on the distance from the sources. We also proposed a method for correcting the interinstrumental discrepancy, which is a source of serious error in multipoint observations. In addition, some information was obtained on chemical composition of the aerosol particles.
Aerosol particles can be classified according to their electrical mobility. A multi-channel detector has been developed to classify electrically charged airborne particles under the influence of an electric field. The detector is capable of measuring the electrical mobility of particles in the sub-micrometer size range. It is compactly designed as an assembly of two concentrically-aligned cylindrical short electrodes with a fixed clearance. Sheath air and aerosol flows enter from one end, pass through the annulus and exit the other end. An electric field is applied between the inner and outer electrodes. Particles with a given electrical mobility are collected on a designated electrometer ring where electrical signals are measured to obtain size distributions. In this study, a mathematical model was developed to optimize the locations of particle detection sensors and the calculations of particle trajectories were performed for various particle sizes. A prototype detector constructed in the present work gave the particle sizes which are in good agreement with those obtained by a scanning electron microscope. Signal current from the detector was also analyzed to give number concentration of particles. Experimental results agreed well with the theoretical predictions. The proposed model was proven to be useful in designing the detector and the prototype detector showed promising results for aerosol size measurement.
We observed anthropogenic aerosols and dust particles (Kosa) successively from 7 to 9 November 2005 at Cape Hedo, Okinawa with high time resolution apparatuses. An aerosol mass spectrometer with the time resolution of 10 min showed that peaks due to anthropogenic aerosol appeared at 0:00 and 15:00 on November 7. The first peak was not associated with Kosa though the cold front passed. The second one was observed at the observation point just before Kosa arrived, which was measured by Lidar system with 15 min resolution. Kosa event continued after the anthropogenic aerosols decreased. The EC/OC ratio was constant at 0.2 during the whole event. This indicates that the anthropogenic aerosol continued to be transported even when Kosa dominated. The transport pattern of both anthropogenic aerosol and Kosa was as follows. The anthropogenic aerosol originated from Taiwan and southern China was transported associating with the cold front. Then, under the high pressure system moving toward east, both anthropogenic aerosol and Kosa were transported together from China. In this case, the anthropogenic aerosol arrived slightly earlier. Kosa was continuously transported under the second high pressure system. As a result, Kosa dominated the later period of this event.
The pyrolysis characteristics of particles composed of organic and inorganic carbons were studied. The smoke, polystyrene latex, polyoxyethylene (10) octylphenyl ether particles were selected as the organic carbon particles, while the carbon beads were employed as the inorganic carbon particles. The particles were heated by an infrared lamp equipped with the power controller, and the particles number concentration was measured by a light scattering particle counter. It was found that the temperature at which the particles of the polystyrene latex are completely volatilized ranges from 640 to 660 °C, whereas polyoxyethylene (10) octylphenyl ether particles disappears at about 300 °C, the smoke particles at 160 to 250 °C, and carbon beads at about 1,200 °C. These results suggest the possibility of qualitative real-time analysis of organic and the inorganic carbon particles in aerosol.
Concentrations of PM2.5 mass and inorganic ionic components measured in Komae, Tokyo, from FY1998 to FY2005 are analyzed for studying the interannual variability. It is found that the mass concentration tends to decrease annually, and this tendency is pronounced in winter. It is considered from the comparisons with other studies in South Kanto that the decreasing tendency is a relatively large-scale phenomenon, and the PM2.5 mass concentration is being uniformed in the area. Inorganic ionic species do not show such decreasing tendency in general. As a result, concentration ratios of secondary inorganic aerosol (SIA) to PM2.5 mass tend to increase. This could lead to the uniformity in PM2.5 mass concentration. The SIA concentration is high, very close to the US Air Quality Standard for PM2.5. Therefore, the SIA should be more concerned in order to reduce the PM2.5 mass concentration.