Snow samples from southern New Mexico, west Texas, Antarctica, and Greenland were analyzed for aerosol and graphitic carbon. Graphitic carbon contents were found to be between 2.2 and 25 μg L−1 of snow meltwater; water‐insoluble aerosol content varied between 0.62 and 8.5 mg L−1. For comparison, two samples of Camp Century, Greenland, ice core, having approximate ages of 4,000 and 6,000 years, were also analyzed. Ice core graphitic carbon contents were found to be 2.5 and 1.1 μg L−1.

Near-surface and lower tropospheric aerosol characteristics have been determined at several remote sites near Orogrande, New Mexico, using a variety of methods and sensors including quartz fiber filter samplers, hi-vol samplers, ground-based and aircraft-mounted light-scattering aerosol counters, an aerosol counter equipped with a heated inlet, and an aethalometer (a device that measures aerosol absorption). The results of these measurements, which have been taken sporadically over the last 15 years but with a higher density of measurements during 1989-1991, suggest that regardless of season, aerosol consists of two modes: a submicron fraction composed primarily of ammonium/acid sulfates and elemental black carbon (BC) and a supermicron fraction composed mainly of quartz and clay minerals of soil origin. Aerosol mass concentrations determined from weekly or biweekly samples are dominated by wind-derived aerosol of soil origin and typically range from • 10/xg m -3 in the fall rainy season (August-October) to • 100/xg m -3 during the spring windy season (February-June). The mass concentration frequency of occurrence of both total aerosol and BC are well approximated by -3 lognormal probability distributions. Geometric mean masses of aerosol range from 16 to 32/xg m and are smaller for mountain sites than for basin sites; geometric mean BC mass concentrations at a single basin site are •0.15 /xg m -3. Aerosol loadings decrease slightly from 1975-1977 to 1990 measurement periods, whereas BC concentrations, within statistical error, remain unchanged. However, a statistically significant increase in BC did occur from 1990 to 1991. The fraction of (visible spectrum) aerosol absorption attributable to BC has a yearly mean value of •0.84, although during dusty conditions occurring in spring or early summer, non-BC aerosol absorption contributes up to •30% of the total. Sulfate and BC aerosol concentrations are sometimes highly correlated, suggesting they have a common source and may be internally mixed. Limited aircraft measurements in the lowest few kilometers of the troposphere reveal a well-mixed aerosol for a neutral atmospheric condition, and a significant decrease in aerosol concentration with altitude for a stable atmospheric condition. INTRODUCTION Previous measurements of the physiochemical properties of natural aerosol in the arid southwestern United States have been directed toward determination of aerosol composition [Hoidale and Blanco, 1969; Lindberg, 1975; Reagan et al., 1984; P. J. Sheridan et al., unpublished manuscript, 1992] and size distribution [Twomey, 1976; Patterson and Gillette, 1977a; Reagan et al., 1980], characteristic morphologies [Gillette and Walker, 1977], average refractive index [Grams et al., 1974; Herman et al., 1975; Lindberg and Gillespie, 1977], effect on visibility [Patterson et al., 1976; Patterson, 1977; Patterson and Gillette, 1977b], vertical structure [Spinhirne et al., 1980; Schwiesow et al., 1981; Reagan et al., 1984], backscatter to extinction ratios [Reagan et al., 1988], w...

No abstract

[1] Total aerosol mass loading, aerosol absorption, and black carbon (BC) content were determined from aerosol collected on 598 quartz fiber filters at a remote, semiarid site near Orogrande, New Mexico from December 1989 to October 1995. Aerosol mass was determined by weighing filters before and after exposure, and aerosol absorption was determined by measuring the visible light transmitted through loaded filter samples and converting these measurements to aerosol absorption. BC content was determined by measuring visible light transmitted through filter samples before and after firing and converting the absorption to BC mass, assuming a BC absorption cross section of 19 m 2 /g in the fiber filter medium. Two analyses were then performed on each of the logged variables: an autoregressive integrating moving average (ARIMA) analysis and a decomposition analysis using an autoregressive model to accommodate first-order autocorrelation. The two analyses reveal that BC mass has no statistically significant seasonal dependence at the 5% level of significance but only random fluctuations varying around an average annual value that has a long-term decreasing trend (from 0.16 to 0.11 mg/m 3 during 1990-1995). Aerosol absorption, which is dominated by BC, also displays random fluctuations about an average value, and decreases from 1.9 Mm À1 to 1.3 Mm À1 during the same period. Unlike BC, aerosol mass at the Orogrande site displays distinctly different character. The analyses reveal a pronounced seasonal dependence, but no long-term trend for aerosol mass. The seasonal indices resulting from the autoregression analysis have a minimum in January (À0.78) and maximum in June (+0.58). The geometric mean value over the 1990-1995 period for aerosol mass is 16.0 mg/m 3 . Since BC aerosol at the Orogrande site is a product of long-range atmospheric transport, a back trajectory analysis of air masses was conducted. Back trajectory analyses indicate that air masses traversing high population centres on the west coast of the USA can lead to significant BC loading, while air masses originating from the southeast of the site tend to have less BC. Additional aerosol measurements are reported for a remote maritime site located on the Atlantic side of the former Panama Canal Zone for the 1976-1979 period, although these data have been analyzed with less statistical rigor. The average geometric mean value of aerosol mass loading for this site is 9.7 mg/m 3 , with a significant decreasing trend of À19% per year. The mean value of aerosol absorption is 0.59 Mm À1 , with a decreasing trend of À15% per year.

In situ measurements of the sizes and concentrations of dust particles generated by the detonation of high explosives in clay soil near Leesville, La., sandy clay soil near Huntsville, Ala., and sandy soils near Orogrande, N.M. are reported. Measurements were generally made within 1 m of the surface (in one case 10 m) at distances ranging from 10 to approximately 50 m from the detonation point with a combination of Knollenberg lightscattering counters (for particles with equivalent radius in the submicron to 15-microm range) and a Knollenberg optical array probe (for particles of 10-150 microm). Measurements were made for periods of several tens of seconds following detonation. All dust size distributions, irrespective of soil or explosive type, exhibit a bimodal character with mass mean radii of approximately 7 and 70 microm. Peak aerosol mass loadings inferred from the distributions have values ranging from 0.05 to 10 g gm-3 with the larger mode of particles contributing most to the mass loading. Predictions of dust extinction coefficients at visible (0.55-microm) and IR (10.4-microm) wavelengths were made using the measured size distributions together with estimates of dust refractive indices. These predictions suggest that extinction should be nearly neutral (wavelength independent) in agreement with transmission measurements made during some of the tests. Predicted mass extinction coefficients, under the assumption of dust material density of 2.5 g cm-3, are of the order of 0.05 m2 g-1 at both visible and IR wavelengths. This value is also in good agreement with a test-averaged measured value of 0.03 m3 g-1 (at lambda = 10.6 microm) obtained using a short path transmissometer and hi-vol sampler.