Liming Zhou

Advanced Factor Analysis on Pittsburgh Particle Size-Distribution Data Special Issue of Aerosol Science and Technology on Findings from the Fine Particulate Matter Supersites Program

Positive matrix factorization (PMF) method was applied to particle size-distribution data acquired during the Pittsburgh Air Quality Study (PAQS) from July 2001 to August 2001. After removing those days with nucleation events, a total of 1632 samples, each with 165 evenly-sized intervals from 0.003 to 2.5 μm, were obtained from scanning mobility particle spectrometer (SMPS) and aerodynamic particle sampler (APS). The temporal resolution was 15 min. The values for each set of five consecutive-size bins were averaged to produce 33 new size channels. The size distributions of particle number as well as volume were analyzed with a bilinear model. Three kinds of information were used to identify the sources: the number and volume size distributions associated with the factors, the time frequency properties of the contribution of each source (Fourier analysis of source contribution values) and the correlations of the contribution values with the gas-phase data and some composition data. Through these analyses, the sources were assigned as sparse nucleation, local traffic, stationary combustion, grown particles and remote traffic, and secondary aerosol in a sequence of decreasing number concentration contributions. Conditional probability function (CPF) analysis was performed for each source so as to ascertain the likely directions in which the sources were located.

On-Road Exposure to Highway Aerosols. 1. Aerosol and Gas Measurements

On-road experiments were conducted to determine the sensitivities of rats to real-world aerosol. This article summarizes the on-road aerosol and gas measurements and provides background information for the companion paper on the rat exposures. Measurements were carried out over 10 days, 6 h/day, driving a route from Rochester to Buffalo. Aerosol instrumentation used in this study included two scanning mobility particle sizers (SMPS) to determine the aerosol size distribution from 10 to 300 nm, 2 stand-alone condensation particle counters to determine the total aerosol number concentration, and an electrical aerosol detector to determine the aerosol length concentration. A thermal denuder (TD) was used with one of the SMPS instruments to determine the size distribution of the non-volatile fraction. Filter samples were collected and analyzed for elemental carbon, and gas analyzers measured ambient levels of CO, CO2, and NO. Average daily total aerosol number concentration ranged from 200,000 to 560,000 particles/cm3. Past studies on urban highways have measured total number concentrations ranging between 104 and 106 particles/cm3. The average daily NO concentration ranged from 0.10 to 0.24 ppm and the corresponding CO2 concentration ranged from 400 to 420 ppm. The average daily geometric number mean particle size determined by the SMPS ranged from 15 to 20 nm. The TD reduced the average SMPS number concentration between 87 and 95% and the SMPS volume between 54 and 83%, suggesting that most of the particles consisted of volatile material. The TD also increased the geometric number mean diameter from 15 to 20 nm to 30 to 40 nm.

Application of PSCF and CPF to PMF-Modeled Sources of PM2.5 in Pittsburgh

Ambient PM 2.5 composition data in Pittsburgh, PA have been used with Positive Matrix Factorization (PMF) to determine the major sources of PM 2.5 sampled. This paper describes the use of the potential source contribution function (PSCF) with the PMF-modeled source contributions to locate the sources in a grid of 0.1° × 0.1° cells. The domain extends from the Pittsburgh Supersite at 40.44°N, 79.94°W over the range 35°–50° north latitude and 75°–90° west longitude. Six-hour back trajectories have been obtained from HYSPLIT four times each day for the 13 months of the study for use with PSCF. Using the results, higher probability locations are compared with known locations of specific source types, based on information from the EPA Toxic Release Inventory (TRI) and the EPA AIRS Database. PSCF results for several sources are compared to the conditional probability function (CPF) analysis, which uses 15-minute wind direction data to determine the most probable direction of a source. Using PSCF and CPF together aids in interpretation of potential source regions. The selenium and sulfate factor source locations are regional, while the lead, cadmium, and specialty steel factor source locations are local. The gallium-rich and Fe, Mn, and Zn factor source locations are potentially both local and regional. The nitrate, vehicle emissions and road dust, wood combustion, vegetative detritus and cooking, and crustal material factor CPF and PSCF results were inconclusive as sources of these factors exist in all directions from the site and therefore one would not expect a clear probability field in any one direction.

Investigation of the relationship between chemical composition and size distribution of airborne particles by partial least squares and positive matrix factorization

[1] Two multivariate data analysis methods, partial least square (PLS) and positive matrix factorization (PMF), were used to analyze aerosol size distribution data and composition data. The relationships between the size distribution data and composition data were investigated by PLS. Three latent variables summarized chemical composition data and most variations in size distribution data especially for large particles and proved the existence of the linearity between the two data sets. The three latent variables were associated with traffic and local combustion sources, secondary aerosol, and coal-fired power plants. The size distribution, particle composition, and gas composition data were combined and analyzed by PMF. Source information was obtained for each source using size distribution and chemical composition simultaneously. Eleven sources were identified: secondary nitrate 1 and 2, remote traffic, secondary sulfate, lead, diesel traffic, coal-fired power plant, steel mill, nucleation, local traffic, and coke plant.

Major Source Categories for PM2.5 in Pittsburgh using PMF and UNMIX

An objective of the Pittsburgh Air Quality Study was to determine the major sources of PM2.5 in the Pittsburgh region. Daily 24-hour averaged filter-based data were collected for 13 months, starting in July 2001, including sulfate and nitrate data from IC analysis, trace element data from ICP-MS analysis, and organic and elemental carbon from the thermal optical transmittance (TOT) method and the NIOSH thermal evolution protocol. These data were used in two source-receptor models, Unmix and PMF. Unmix, which is limited to a maximum number of seven factors, resolved six source factors, including crustal material, a regional transport factor, secondary nitrate, an iron, zinc and manganese factor, specialty steel production and processing, and cadmium. PMF, which has no limit to the number of factors, apportioned the PM2.5 mass into ten factors, including crustal material, secondary sulfate, primary OC and EC, secondary nitrate, an iron, zinc and manganese factor, specialty steel production and processing, cadmium, selenium, lead, and a gallium-rich factor. The Unmix and PMF common factors agree reasonably well, both in composition and contributions to PM2.5. To further identify and apportion the sources of PM2.5, specific OC compounds that are known markers of some sources were added to the PMF analysis. The results were similar to the original solution, except that the primary OC and EC factor split into two factors. One factor was associated with vehicles as identified by the hopanes, PAHs, and other OC compounds. The other factor had strong correlations with the OC and EC ambient data as well as wood smoke markers such as levoglucosan, syringols, and resin acids.

An Intercomparison of Measurement Methods for Carbonaceous Aerosol in the Ambient Air in New York City

Measurement methods for fine carbonaceous aerosol were compared under field sampling conditions in Flushing, New York during the period of January and early February 2004. In-situ 5- to 60-minute average PM 2.5 organic carbon (OC), elemental carbon (EC), and black carbon (BC) concentrations were obtained by the following methods: Sunset Laboratory field OC/EC analyzer, Rupprecht and Patashnick (R&P) series 5400 ambient carbon particulate monitor, Aerodyne aerosol mass spectrometer (AMS) for total organic matter (OM), and a two-wavelength AE-20 Aethalometer. Twenty-four hour averaged PM 2.5 filter measurements for OC and EC were also made with a Speciation Trends Network (STN) sampler. The diurnal variations in OC/EC/BC concentrations peaked during the morning and afternoon rush hours indicating the dominant influence of vehicle emissions. BC/EC slopes are found to range between 0.86 and 1.23 with reasonably high correlations (r > 0.75). Low mixing heights and absence of significant transported carbonaceous aerosol are indicated by the measurements. Strong correlations are observed between BC and thermal EC as measured by the Sunset instrument and between Sunset BC and Aethalometer BC. Reasonable correlations are observed among collocated OC/EC measurements by the various instruments.

Source Apportionment of Airborne Particulate Matter for the Speciation Trends Network Site in Cleveland, OH

Abstract Aerosol composition data from the Speciation Trends Network (STN) site (East 14th Street) in Cleveland, OH, were analyzed by advanced receptor model methods for source apportionment as well as by the standard positive matrix factorization (PMF) using PMF2. These different models are used in combination to test model limitations. These data were 24-hr average mass concentrations and compositions obtained for samples taken every third day from 2001 to 2003. The Multilinear Engine (ME) was used to solve an expanded model to estimate the source profiles and source contributions and also to investigate the wind speed, wind direction, time-of-day, weekend/weekday, and seasonal effects. PMF2 was applied to the same data-set. Potential source contribution function (PSCF) and conditional probability function (CPF) analyses were used to locate the regional and local sources using the resolved source contributions and appropriate meteorological data. Very little difference was observed between the results of the expanded model and the PMF2 values for the profiles and source contribution time series. The identified sources were as ferrous smelter, secondary sulfate, secondary nitrate, soil/combustion mixture, steel mill, traffic, wood smoke, and coal burning. The CPF analysis was useful in helping to identify local sources, whereas the PSCF results were only useful for regional source areas. Both of these analyses were more useful than the wind directional factor derived from the expanded factor analysis. However, the expanded analysis provided direct information on seasonality and day-of-week behavior of the sources.