This paper presents an overview of a major, long-term program for tropospheric gas and aerosol research in the southeastern United States. Building on three existing ozone (O 3 )-focused research sites begun in mid-1992, the Southeastern Aerosol Research and Characterization Study (SEARCH) was initiated in mid-1998 as a 7-year observation and research program with a broader focus including aerosols and an expanded geographical coverage in the Southeast. The monitoring network comprises four urban-rural (or urban-suburban) site pairs at locations along the coast of the Gulf of Mexico and inland, including two moderately sized and two major urban areas (Pensacola, FL; Gulfport, MS; Atlanta, GA; and Birmingham, AL). The sites are equipped with an extensive suite of instruments for measuring particulate matter (PM), gases relevant to secondary O 3 and the production of secondary aerosol particles, and surface meteorology. The measurements taken to date have added substantially to the knowledge about the temporal behavior and geographic variability of tropospheric aerosols in the Southeast. Details are presented in four papers to follow.

The Southeastern Aerosol Research and Characterization Study (SEARCH) was implemented in 1998 -1999 to provide data and analyses for the investigation of the sources, chemical speciation, and long-term trends of fine particulate matter (PM 2.5 ) and coarse particulate matter (PM 10 -2.5) in the Southeastern United States. This work is an initial analysis of 5 years (1999 -2003) of filter-based PM 2.5 and PM 10 -2.5 data from SEARCH. We find that annual PM 2.5 design values were consistently above the National Ambient Air Quality Standards (NAAQS) 15 g/m 3 annual standard only at monitoring sites in the two largest urban areas (Atlanta, GA, and North Birmingham, AL). Other sites in the network had annual design values below the standard, and no site had daily design values above the NAAQS 65 g/m 3 daily standard. Using a particle composition monitor designed specifically for SEARCH, we found that volatilization losses of nitrate, ammonium, and organic carbon must be accounted for to accurately characterize atmospheric particulate matter. In particular, the federal reference method for PM 2.5 underestimates mass by 3-7% as a result of these volatilization losses. Organic matter (OM) and sulfate account for ϳ60% of PM 2.5 mass at SEARCH sites, whereas major metal oxides (MMO) and unidentified components ("other") account for Ն80% of PM 10 -2.5 mass. Limited data suggest that much of the unidentified mass in PM 10 -2.5 may be OM. For paired comparisons of urban-rural sites, differences in PM 2.5 mass are explained, in large part, by higher OM and black carbon at the urban site. For PM 10, higher urban concentrations are explained by higher MMO and "other." Annual means for PM 2.5 and PM 10 -2.5 mass and major components demonstrate substantial declines at all of the SEARCH sites over the 1999 -2003 period (10 -20% IMPLICATIONSParticulate matter (PM) continues to be implicated as a significant factor in studies of the effect of air pollution on human health and welfare. SEARCH was designed to better understand the sources, chemical speciation, and longterm trends of PM in the Southeastern United States. This work provides an initial analysis of 24-hr average, filterbased measurements covering the period 1999 -2003. These observations and findings will increase the scientific basis for the formulation of sound strategies and policies for addressing atmospheric PM and will also provide a rich data set for evaluating the performance of air quality models.

Deployment of continuous analyzers in the SoutheasternAerosol Research and Characterization Study (SEARCH) network began in 1998 and continues today as new technologies are developed. Measurement of fine particulate matter (PM 2.5 ) mass is performed using a dried, 30°C tapered element oscillating microbalance (TEOM). TEOM measurements are complemented by observations of light scattering by nephelometry. Measurements of major constituents include: (1) SO 4 2Ϫ via reduction to SO 2 ; (2) NH 4 ϩ and NO 3 Ϫ via respective catalytic oxidation and reduction to NO, (3) black carbon (BC) by optical absorption, (4) total carbon by combustion to CO 2 , and (5) organic carbon by difference between the latter two measurements. Several illustrative examples of continuous data from the SEARCH network are presented. A distinctive composite annual average diurnal pattern is observed for PM 2.5 mass, nitrate, and BC, likely indicating the influence of traffic-related emissions, growth, and break up of the boundary layer and formation of ammonium nitrate. Examination of PM 2.5 components indicates the need to better understand the continuous composition of the unmeasured "other" category, because it contributes a significant fraction to total mass during periods of high PM 2.5 loading. Selected episodes are presented to illustrate applications of SEARCH data. An SO 2 conversion rate of 0.2%/hr is derived from an observation of a plume from a coal-fired power plant during early spring, and the importance of local, rural sources of NH 3 to the formation of ammonium nitrate in particulate matter (PM) is demonstrated.

Abstract. An informal intercomparison of NOy measurement techniques was conducted from

Currently there is no standard method for measurement of atmospheric mercury dry deposition. While all operationally defined forms of atmospheric mercury (elemental, oxidized, and particulate) can be dry deposited, oxidized forms are of concern due to high deposition velocities, water solubility, and reactivity. This paper describes the development of a surrogate surface for characterizing potential dry deposition of gaseous oxidized mercury (GOM). Laboratory tests showed that the surface collected HgCl2, HgBr2, and HgO with equal efficiency, and deposition was not significantly influenced by temperature, humidity, or ozone concentrations. Deposition of mercury to surfaces in field deployments was correlated with GOM concentrations (r2 = 0.84, p < 0.01, n = 326. Weekly mean GOM deposition velocities from surface deployments (1.1 +/- 0.6 cm s(-1)) were higher than modeled values (0.4 +/- 0.2 cm s(-1)) at four field sites, but were within the range reported for direct measurements. Although the surfaces do not simulate the heterogeneity of natural surfaces and need to be validated by direct measurements, they do provide a physical means for estimating temporal trends and spatial variability of dry deposition of GOM.