© Author(s) 2008. This work is distributed under the Creative Commons Attribution 3.0 License. Atmospheric Chemistry and Physics

We report an extensive airborne characterization of aerosol downwind of a massive bovine source in the San Joaquin Valley (California) on two flights during July 2007. The Center for Interdisciplinary Remotely-Piloted Aircraft Studies (CIRPAS) Twin Otter probed chemical composition, particle size distribution, mixing state, sub- and supersaturated water uptake behavior, light scattering properties, and the interrelationship between these parameters and meteorology. Total PM_(1.0) levels and concentrations of organics. nitrate. and ammonium were enhanced in the plume from the source as compared to the background aerosol. Organics dominated the plume aerosol mass (~56-64%), followed either by sulfate or nitrate. and then ammonium. Particulate amines were detected in the plume aerosol by a particle-into-liquid sampler (PILS) and via mass spectral inarkers in the Aerodvne C-ToF-AMS. Amines were found to be a significant atmospheric base even in the presence of arnmonia; particulate amine concentrations are estimated as at least 14-23% of that of ammonium in the plume. Enhanced sub- and supersaturated water uptake and reduced refractive indices were coincident with lower organic mass fractions, higher nitrate mass fractions, and the detection of amines. The likelihood of suppressed droplet growth owing to kinetic limitations from hydrophobic organic material is explored. After removing effects associated with size distribution and mixing state, the normalized activated fraction of cloud condensation nuclei (CCN) increased as a function of the subsaturated hygroscopic growth factor, with the highest activated fractions being consistent with relatively lower organic mass fractions and higher nitrate mass fractions. Subsaturated hygroscopic growth factors for the organic fraction of the aerosol are estimated based on employing the Zdanovskii-Stokes Robinson (ZSR) mixing rule. Representative values for a parameterization treating particle water uptake in both the sub- and supersaturated regimes are reported for incorporation into atmospheric models.


Introduction
Bovine emissions are a major source of methane (CH 4 ), nitrous oxide (N 2 O), and ammonia (NH 3 ); they are also the dominant anthropogenic source for amines (Schade and Crutzen, 1995). Ammonia is the dominant base in the atmosphere, efficiently neutralizing acidic substances. The main global sources of ammonia are from livestock waste, 5 fertilizer applications, biomass burning, motor vehicle emissions, and coal combustion (Apsimon et al., 1987;Asman and Janssen, 1987;Kleeman et al., 1999;Anderson et al., 2003;Battye et al., 2003). Typical ammonia mixing ratios over continents range between 0.1 and 10 ppb (Edgerton et al., 2007, and references therein), while levels as high as a few ppm have been reported near areas of extensive livestock operations 10 (Rumburg et al., 2006).
The availability of a massive, concentrated source of ammonia and amines offers a unique opportunity to probe the response of the resulting aerosol. During July 2007, the Center for Interdisciplinary Remotely Piloted Aircraft Studies (CIRPAS) Twin Otter 20 probed the aerosol downwind of a major cattle feedlot in the San Joaquin Valley in California. The San Joaquin Valley, the major geographical feature in central California, is bordered on its west and east sides by mountain ranges and is characterized by relatively stagnant air circulation (Fig. 1a). Consequently, this region is one of the largest nonattainment areas for ozone and particulate matter in the United States (Chow et is inferred from established emissions inventories from animal husbandry sources and from measurements of the ammonium and amine content of the aerosol downwind of the plume source. The goal of the present study is to provide a comprehensive airborne characterization of the aerosol downwind of a major bovine source. First, the aircraft instrument 5 payload and flight path strategy are presented. Detailed measurements were obtained for: meteorology, aerosol size distributions and number concentrations, aerosol composition, mixing state, refractive index, hygroscopic growth factors at three different relative humidities, and cloud condensation nucleus (CCN) behavior. Special attention is given to the sub-and supersaturated water uptake properties of the aerosol, and how 10 these relate to chemical composition. Subsaturated hygroscopic growth factors for the organic fraction of the aerosol are reported based on a closure analysis employing the Zdanovskii-Stokes Robinson mixing rule. Subsaturated hygroscopic growth data are then compared to measured supersaturated CCN activity to evaluate the level of consistency between observed water uptake in the two regimes. Introduction Water-soluble aerosol chemical composition was measured by a particle-into-liquid sampler (PILS, Brechtel Mfg Inc.;Sorooshian et al., 2006a). In the PILS, submicrometer ambient particles are grown into droplets sufficiently large to be collected by inertial impaction for subsequent chemical analysis. At the entrance to the instrument a se-5 ries of three denuders (URG and Sunset Laboratories) remove inorganic (basic and acidic) and organic gases that would otherwise bias aerosol measurements. The denuders have been shown to successfully remove gaseous amine species (Murphy et al., 2007). The impacted droplets are delivered to a rotating carousel containing 72 vials, with each vial containing material representing a period of ∼5 min of flight, or 10 alternatively, a distance of 15 km in flight (aircraft speed ∼50 m/s). The contents of the vials are subsequently analyzed off-line using a dual ion chromatography (IC) system (ICS-2000, Dionex Inc.) for simultaneous anion and cation analysis. The PILS-IC instrument uncertainty has been established as±7%, and the detection limit (calculated as air-equivalent concentration of the lowest concentration standard 15 that is distinct from baseline noise in the IC plus three times the standard deviation of this measurement) is <0.1 µg/m 3 for the inorganic ions (Na + , NH boxylic acids with one to nine carbon atoms) (Sorooshian et al., 2007b). The PILS-IC technique has been demonstrated to speciate amines, including ethylamine, diethy-20 lamine, triethylamine, methylamine, dimethylamine, and trimethylamine (Murphy et al., 2007); however, only ethylamine and diethylamine were measured above detection limits (0.01 µg/m 3 ) in the present study. It should be noted that ammonium and ethylamine co-elute in the IC cation column; however, ethylamine was detected at sufficiently high concentrations for its peak to be distinguishable from that of ammonium. The reported acetate reported is likely an overestimate. As compared to acetate, using the calibration equation of glycolate would reduce the estimates by <10%.

Aerosol chemical composition (Aerodyne cToF-AMS)
Chemical composition measurements for non-refractory aerosol species (sulfate, nitrate, ammonium, and organics) were performed using an Aerodyne compact Time of 5 Flight Aerosol Mass Spectrometer (cToF-AMS; Drewnick et al., 2004aDrewnick et al., , 2004b. At the entrance to the instrument, an aerodynamic lens focuses particles with vacuum aerodynamic diameters between approximately 50 nm and 800 nm through a 3.5% chopper and onto a tungsten vaporizer (∼550 • C) (Murphy et al., 2007). The chopper can be operated in three modes to gather either background mass spectra, ensemble average 10 mass spectra over all particle sizes, or size-resolved mass spectra. Once vaporized, molecules undergo electron impact ionization and travel through a time of flight mass analyzer. The cToF-AMS detects the presence of amines in the form of characteristic amine peaks at m/z 30, 56, 58, 73, and 86 (McClafferty and Turecek, 1993;Angelino et al., 2001;Murphy et al., 2007). The detection limit, calculated as three times the 15 standard deviation of the noise for filtered air, is <0.05 µg/m 3 for all species measured. The cToF-AMS can be used to calculate a quantity that will be referred to subsequently as excess nitrate. Excess nitrate is defined as the nitrate mass, derived from cToF-AMS spectra, remaining after both sulfate and nitrate have been fully neutralized by ammonium. A zero or slightly negative value indicates that sufficient ammonia exists 20 to neutralize both sulfate and nitrate, while a positive value indicates that some nitrate is associated with other cations besides ammonium. cToF-AMS calibrations allow an assessment of the error associated with the excess nitrate calculation; introducing pure ammonium nitrate into the instrument should result in an excess nitrate value of zero. After flight A, calibrations were conducted with monodisperse ammonium nitrate parti-Introduction from NO + , can be overestimated. A relatively conservative approach is to calculate nitrate mass using the peak intensity at m/z 46 (NO + 2 ), which rarely corresponds to an organic fragment (McClafferty and Turecek, 1993). Calibration of the instrument with pure ammonium nitrate allows observation of the ratio of the peak intensities at m/z 30 and 46 when NO + and NO + 2 are present without organic interference. During such a 15 calibration conducted after flight A, the peak at m/z 30 was observed to be 2.2 times that at m/z 46. Thus, for this flight, nitrate mass at m/z 30 was calculated to be 2.2 times the mass at m/z 46. Based on calibrations on the day of flight B, the peak at m/z 30 was observed to be 2.8 times that at m/z 46. The two calibrations were seventeen days apart, and instrumental drift is responsible for the difference in the m/z 30:46 ratio 20 between the two days; typical values observed with this instrument range between two and three. The mass remaining at m/z 30 after the nitrate contribution is subtracted is assumed to be organic, including amine species, which often exhibit a major peak here.
2.3 Aerosol hygroscopicity and refractive index (DASH-SP) 25 A differential aerosol sizing and hygroscopicity spectrometer probe (DASH-SP; Brechtel Mfg Inc.; Sorooshian et al., 2008) was included in the instrument payload on the aircraft. The DASH-SP consists of a single classification differential mobility analyzer 10422 (DMA) followed by a set of parallel hygroscopic growth chambers operated at different relative humidities. A ∼0.5 LPM aerosol sample flow passes first through a Nafion drier, and then through a 210 Po neutralizer that brings the dried particles to a stable, steadystate charge distribution. A cylindrical DMA selects particles in a narrow interval of mobility-equivalent diameters in the 0.1 to 1.0 µm range. The classified aerosol leaving 5 the DMA is split into five separate flows. In one of the five streams, the total concentration of classified particles is determined using an integral TSI Model 3831 water-based condensation particle counter (CPC). The remaining four classified aerosol flows pass through a Nafion humidifier (Perma Pure, LLP, Model MD-070-24FS-4) to achieve thermodynamic equilibrium with water vapor at a constant, predetermined RH. The four 10 conditioned aerosol flows pass directly to dedicated, custom-built OPCs (λ =532 nm, World Star Technologies, Model TECGL-30) designed to size particles in the 100 nm to 3 µm diameter size range. An iterative data processing algorithm quantifies an 'effective' aerosol refractive index that is used to calculate hygroscopic growth factors (GF=D p,wet /D p,dry ) corrected for the refractive-index dependence of the OPC response 15 . During this study, the DASH-SP provided simultaneous measurements of GFs at different RHs for dry DMA-selected particle diameters between D p =150-200 nm. One humidifier was operated dry (RH <8%), and the other three were at RHs of 74%, 85%, and 92%. (No data from the RH=85% channel were available during flight A.) The uncertainty associated with growth factor measurements is Introduction ticles that activate and grow sufficiently large (D p >0.75 µm) for detection by an OPC were quantified. The activated fraction is determined as the ratio of the CCN number concentration to the total particle (CN) number concentration.

Size Distributions and particle number concentration
Aerosol size distribution data were obtained by a DMA (D p =10-800 nm) and an exter-5 nal passive cavity aerosol spectrometer probe (0.1-3 µm) (PCASP, PMS Inc., modified by DMT Inc.). Particle number concentrations were quantified with two condensation particle counters (CPC Model 3010, TSI Inc., D p >10 nm; UFCPC Model 3025, TSI Inc., D p >3 nm). When the two CPCs experienced electrical saturation, particle number concentrations from the DMA are reported. The DMA time resolution is 74 s as Sampling lasted for more than three hours starting just before noon during both flights. The feedlot operation is slightly larger than 3 km 2 in area (∼800 acres). The feedlot is bordered on the west side by the Interstate 5 roadway, which is a 25 major transportation route and source of vehicular emissions, connecting northern and 10424 southern California. Five-day back-trajectories, computed using the NOAA HYSPLIT model (Draxler and Rolph, 2003), show that the background air during flight A originated over the Pacific Ocean, whereas the background air during flight B was transported over land from the north (Fig. 5). This suggests that the background aerosol in flight A may have 20 carried the signature of cleaner marine air, while that measured during flight B was more exposed to urban and agricultural emissions.

Particle number concentrations and size distributions
Average submicrometer particle number concentrations in the plume were 30 528±8987 cm −3 (flight A) and 16 606±4286 cm −3 (flight B) ( ber concentrations in and out of the plume were similar, indicating the absence of significant emissions of particles from the source. (As noted in Sect. 3.3, nitrate enhancement is used to identify the location of the plume.) During the downwind plume transects, number concentrations decreased slightly with increasing altitude until a sharp decrease near the top of the boundary layer to below 300 cm −3 (flight A) and 5 800 cm −3 (flight B) at altitudes of 550 m and 400 m, respectively. The ratio of the number concentration of particles with D p >3 nm to the number concentration of particles with D p >10 nm was 1.1±0.1 and 1.2±0.1 for flights A and B, respectively. This ratio showed no difference in and out of the plume and decreased with altitude.
The vertical structure of the PCASP (D p =0.1-3 µm) and CPC number concentra-10 tions were similar. Average PCASP concentrations in the plume for flights A and B were 1065±330 and 675±220 cm −3 , respectively, indicating that most of the particles were smaller than 100 nm in diameter. The number of 0.1 to 3 µm diameter particles in the plume was enhanced by a factor of 2.5-3 times as much over that in the background aerosol. This is especially evident by the increase in number concentra-15 tion (D p >100 nm) observed when passing through the beginning of the plume over the perimeter of the source (upper panels of Fig. 2/ 3; see the flight segments labeled "2" when the aircraft was circling the perimeter of the plume source); this likely is a result of smaller particles growing into this size range. However, the increase in number concentration for particles with D p =0.1-3 µm was not sufficiently large to result in a 20 significant difference in the submicrometer number concentration (D p =0.01-1 µm) in and out of the plume.
Aerosol number and volume distributions were similar both between the two flights, and in and out of the plume. Multiple modes normally existed in the number and volume distributions. A number concentration mode was generally present between D p =20-25 60 nm, with a weaker mode between D p =60-100 nm (lower panels of Fig. 2/ 3). There was also a dominant number concentration mode, which will be referred to as the nucleation mode, at sizes smaller than the detection limit of the DMA (10 nm); this mode is evident from the difference in number concentration measured by the DMA and the UFCPC 3025. Volume concentration modes existed at D p =30-60 nm, D p ∼100 nm, and frequently at D p >100 nm. The number and volume distributions in the plume shifted slightly to larger sizes with downwind distance from the plume source.

Submicrometer aerosol chemical composition
Figures 6 (time series), 7/8 (spatial distribution), and 9 (vertical distribution) summa-5 rize the submicrometer aerosol composition data, in addition to Table 1, which reports the background and in-plume composition for both flights. The reported total organic mass is non-refractory organic mass that was measured by the cToF-AMS. The total aerosol mass is determined as the sum of inorganic mass, as determined by the PILS and cToF-AMS, and non-refractory organic mass from the cToF-AMS. Nitrate enhance-10 ment is used to define the location of the plume. Then, knowing where the plume is, local enhancements in other species concentrations and aerosol properties can be determined. Plume ages are noted on the spatial plots in Fig. 1B/C and were calculated using downwind distance and average wind speed in the vicinity of the source. The highest plume age encountered in flights A and B was 0.9 h and 1.2 h, respectively.

Total aerosol mass and major components
The average total aerosol mass in the boundary layer was 8.85±1.79 µg/m 3 and 3.40±0.98 µg/m 3 during flights A and B, respectively, with significant enhancements in the plume (Table 1) Overall, organic species dominated the total mass. Organics accounted for 61.9%±2.6% (flight A) and 55.5%±6.4% (flight B) of the plume aerosol mass, and 63.5%±3.3% (flight A) and 63.1%±11.9% (flight B) of the background aerosol mass. 25 The organic mass concentration in the plume was 6.48±0.98 µg/m 3 (flight A) and Introduction 2.46±0.29 µg/m 3 (flight B), and in the background aerosol was 5.10±1.07 µg/m 3 (flight A) and 1.73±0.70 µg/m 3 (flight B). The next largest contributor to the particulate mass was either sulfate or nitrate, depending on the day and aerosol type, followed by either ammonium or nitrate (see Table 1). The ratio of organic mass to inorganic mass in the plume was 1.64±0.19 (flight A) and 1.30±0.39 (flight B), while the ratio in the 5 background aerosol was 1.77±0.29 (flight A) and 1.92±0.68 (flight B). Previous measurements in the San Joaquin Valley have also shown that organic aerosol contributes significantly to the fine particle mass (Chow et al., 1996;Neuman et al., 2003).

Inorganic aerosol
Within the source plume, the levels of nitrate and ammonium increased significantly above their respective values in the background valley aerosol ( Fig. 6-8). The ammonium-to-sulfate molar ratio is an important indicator of the level of partitioning of ammonium nitrate between the gas and aerosol phases. Since this ratio usually exceeded two, ammonia was available to foster partitioning of nitrate to the aerosol phase in the plume. The vertical distribution of nitrate and ammonium exhibited similar trends 15 in each flight, unlike sulfate, which did not increase in concentration at plume altitudes (∼100-300 m) ( Fig. 9). Generally, other inorganic species, including chloride, sodium, potassium, calcium, magnesium, and nitrite, did not contribute significantly (>0.1 µg/m 3 ) to the aerosol mass. Many of these species are expected to be found primarily in the coarse parti- The concentration of total organics, as determined by the cToF-AMS, in the plume aerosol significantly exceeded those in the background valley aerosol ( Fig. 6-8). The vertical distribution of total organic concentrations was somewhat similar to those of nitrate and ammonium, with the exception that there was not as sharp an enhancement 5 in concentration at plume altitudes, as was especially evident in flight B (Fig. 9). Figure 10 shows the representative mass spectra of the organic fragments detected by the cToF-AMS in the background valley aerosol, the plume aerosol close to the source, and farther downwind. All non-organic contributions to the mass spectra have been removed using the methodology described in Allan et al. (2004); fragmentation 10 at m/z 30 was further modified as described in Sect. 2.2. Figure 11 indicates that the signal at m/z 30 represents a large fragment for organics, including amines. The chemical signatures of the organic aerosol in the three categories appear to be quite similar. One difference, which is highlighted also in Fig. 11, is that the m/z 30 (common amine marker) peak intensity is enhanced by ∼150% at the closest point to the plume 15 source as compared to background aerosol, and decreases by ∼25% at the farthest downwind distance. Figure 11 shows comparisons of the plume aerosol organic mass spectra to the background spectra for both flights. The overall organic aerosol appears to be similar in and out of the plume; however, peak intensities at m/z 30, 56, 74 and 86 are enhanced in plume. These are all peaks in the electron impact mass spectrum 20 of amines, including diethylamine and triethylamine (McClafferty and Turecek, 1993;Angelino et al., 2001;Murphy et al., 2007). A similar analysis for plume organics close to the feedlot and farther downwind reveals no significant difference in most peaks, although the intensity of the m/z 30 peak decreases with increasing plume age; as will be discussed subsequently, this is likely attributed to increased partitioning of particulate 25 amines to the gas phase to maintain thermodynamic equilibrium as the plume dilutes with background air. An analysis of the background valley aerosol spectra during the transit portions of the flights indicates that the organic aerosol composition was similar Introduction Interactive Discussion throughout the valley. Two amines were detected by the PILS, diethylamine and ethylamine; these were found only in the plume ( Fig. 6-8). Diethylamine reached higher concentrations (up to 0.18 µg/m 3 and 6.0% of the organic mass) and was more abundant farther downwind of the feedlot as compared to ethylamine; diethylamine was observed at plume ages 5 up to 0.9 h (flight A) and 0.7 h (flight B). Ethylamine was detected by the PILS only in three samples collected during the two flights. It was found immediately downwind of the feedlot up to plume ages of 0.7 h (flight A) and 0.3 h (flight B) at concentrations near 0.02 µg/m 3 , which corresponds to 0.8% of the total organic mass, as inferred from the cToF-AMS data. 10 The collective organic acid concentration, as determined by the PILS, reached levels of up to 0.23 µg/m 3 (flight A) and 0.41 µg/m 3 (flight B), accounting for 0.4% (±0.8%) and 0.4% (±0.6%) of the cToF-AMS total organic mass during flights A and B, respectively. Oxalate was the most abundant organic acid, followed by succinate, formate, and acetate. The concentration of the organic acids (C 1 -C 9 ) were not found to be 15 correlated with those of total organics, amines, or any inorganic species.
3.4 Aerosol mixing state Figure 12 shows speciated size distributions for the background and in-plume aerosol at various downwind distances for both flights. Organics, nitrate, and sulfate all appear to be externally mixed to some extent. This is especially clear when examining

Refractive index
The background aerosol exhibited a consistent average dry-particle refractive index of 1.54±0.07 and 1.54±0.04 for flights A and B, respectively (Table 2). Since these values are close to those of ammonium nitrate (1.55) and ammonium sulfate (1.52-1.53) (Weast, 1987;Tang, 1996), which are the dominant inorganic components of the 5 aerosol, assuming a volume-weighted overall refractive index, the organic component refractive index is calculated also to be 1.54. Notably, Zhang et al. (1994) reported a similar refractive index of 1.55 for particulate organic compounds in Grand Canyon aerosol. The overall aerosol refractive indices were slightly lower in the plume, with the lowest values observed closest to the feedlot during the aircraft circling maneuvers 10 (1.48±0.09 and 1.51±0.01 for flights A and B, respectively). Organic species, such as amines, may be responsible for this decrease as the organic mass fraction dominated the total mass; ethylamine and diethylamine have refractive indices of 1.37 and 1.39, respectively (Dean, 1999). Although only two particulate amines were speciated at low concentrations (<4% of total mass), other amine compounds may well have existed in 15 the total organic mass with comparable refractive indices.  Kreidenweis, 2000). Overall, the plume aerosol exhibited higher hygroscopic growth factors as compared to the background aerosol. Hygroscopic growth factors in the immediate vicinity of the source were usually between 1.75-1.90 at RH=92%. Hygroscopic growth factors are now related to the mass fractions of the aerosol components. Figure 15 shows the dependence of growth factors at an RH of 92% 5 on mass fractions of nitrate and organics measured for flight B. (Similar effects occur at the other RHs.) Increasing growth factors coincide with higher nitrate mass fractions. This effect is most evident during flight B, in terms of slope (0.31) and correlation (R 2 =0.43), partly because of the larger range of nitrate mass fractions observed. In addition, growth factors exhibited a negative correlation (R 2 =0.46, slope=-0.28) with 10 organic mass fractions. Less correlation exists between observed growth factors and mass fractions of ammonium and sulfate (R 2 <0.21). The data show that subsaturated hygroscopicity increases as a function of increasing fraction of ammonium nitrate, a highly hygroscopic salt, and decreasing fraction of organics, the growth factor of which will be explored subsequently.

CCN
The CCN data acquired are summarized in Table 3 and Fig. 16. Owing to the large number of particles with diameters below about 60 nm, in the background atmosphere as well as in the plume, the activated fractions were quite small. An enhancement in activated fraction was observed in the plume, which is consistent with the observed 20 behavior of subsaturated hygroscopic growth factors. An important issue is the extent to which aerosol composition influences CCN behavior. This can be manifested in two ways: (1) by affecting the critical supersaturation of the particles; and (2) by influencing the growth rate once the particle activates. Figure 17 shows the normalized activation fraction as a function of hygroscopic growth 25 factor for flight B. The normalization is done by computing the activation fraction assuming the CCN are composed of pure ammonium sulfate. The normalization removes any variations due to shifts in the shape of the size distribution. It is noted that, 10432 in general, higher supersaturations are required to activate particles composed of less hygroscopic material. Regarding the effect of particle composition on growth rate, consider two particles each having the same critical supersaturation. If the ambient supersaturation exceeds the critical supersaturation, then each particle will activate. The subsequent rate of 5 growth by water condensation depends on the uptake of water molecules. If the two particles have different composition, then the uptake coefficients for water vapor can be different; the particle with the smaller water uptake coefficient will exhibit a slower rate of growth after activation. In a CCN instrument, like the CCNc employed here, the more slowly growing particle may not reach its ultimate size before it exits the 10 growth chamber of the instrument and is detected by the OPC. The growth rate of pure ammonium sulfate particles of the same critical supersaturation as that of the particle in question can be taken as the standard against which particle growth rates can be compared. Since an entire distribution of particles enter the CCNc, with different critical supersaturations (as a result of size and composition), the standard used is ammonium 15 sulfate with a critical supersaturation equal to the supersaturation of the instrument. Based on this standard, if all particles grow as quickly as those composed entirely of ammonium sulfate, all particles will have droplet sizes equal to or larger than the standard. Hence, at a given supersaturation, the presence of droplets with a size less than that of the standard indicates retarded growth. We express this effect in terms of 20 the fraction of droplets less than the standard at the supersaturation of the instrument. Figure 18 shows the cToF-AMS-derived ratio of m/z 57:44 as a function of organic mass fraction. (A larger m/z 57:44 ratio is correlated with the organic material being less oxidized, and hence more hydrophobic.) The color coding of the data points corresponds to the fraction of droplets that have a size (D i ) less than the ammonium sulfate 25 standard (D AS ), as described above. The size of the symbols reflects the hygroscopic growth factor at 92% RH. The data indicate that at high organic mass fractions when the particles are composed of less oxidized material, there is a tendency, although weak, towards retardation of growth. The growth factor exhibits a clear anti-correlation with organic mass fraction.

Discussion
In this section we explore key findings in this study. The observations reveal significant differences in aerosol properties in and out of the plume, and as a function of plume age. Significant enhancements in nitrate, ammonium, and organic levels in the plume 5 were observed; this coincided with an increased potential for water uptake in both the sub-and supersaturated regimes. While trends in the data from the two flights were similar, particle number and mass concentrations were larger in flight A. Explanations for this discrepancy will be pursued.
Owing to the range of organic fractions observed, the present study provides an op-10 portunity to evaluate the sensitivity of mixed inorganic/organic particle hygroscopicity to the organic fraction. Subsaturated hygroscopic growth factors are calculated for the organic fraction based on a closure analysis using the Zdanovskii-Stokes Robinson (ZSR) mixing rule. Measurement of CCN activity in this study also presents an opportunity to assess the consistency of observed supersaturated water uptake with the 15 subsaturated water uptake measurements.
4.1 Enhancements in mass production and water uptake in the plume aerosol Significant production of ammonium nitrate and organic mass in the plume occurred during both flights. Ammonium nitrate production is expected due to the high ammonia levels and the presumed abundance of nitric acid from the daytime photochemistry. Or-20 ganic aerosol mass production results from both condensation of semi-volatile organic species and acid-base chemistry of amines, followed by condensation of low-volatility products onto pre-existing aerosols.
Enhanced sub-and supersaturated water uptake coincides with greater fractions of ammonium, nitrate, and amines. Dinar et al. (2008) have shown, for example, that the reactive uptake of ammonia by aerosols containing slightly soluble organics leads to substantial increases in hygroscopic growth and CCN activity. This observation appears to be consistent with the present measurements. Speciated size distributions show that the aerosol is, in part, externally mixed, with different species growing independently during plume aging (Fig. 12). The high organic fractions, particularly during 5 flight A, may have masked the expected and significant growth exhibited by pure ammonium nitrate and ammonium sulfate salts; however, amines, which represent one class of organic species in the plume, are thought to be highly hygroscopic. Aklilu et al. (2006) also suggested that the organic fraction of the aerosol can suppress the growth normally associated with nitrate based on ambient measurements at two rural,  (Blanchard et al., 2000;Pun and Seigneur, 1999). During an aircraft study in the San Joaquin Valley in May 2002, Neuman et al. (2003 reported simultaneous nitric acid depletion and aerosol mass enhancements when the aircraft either encountered large ammonia sources or reached lower temperatures at higher altitudes in the boundary layer. Ammonia is emitted from the ground, whereas nitric acid is efficiently produced 20 photochemically throughout the entire boundary layer, especially during the summer in the daytime San Joaquin Valley atmosphere. Since gas-phase ammonia and nitric acid were not measured in the present study, observed particulate levels of ammonium, nitrate, sulfate, and amines can help determine the limiting reactant in chemical processing inside the plume. Excess nitrate 25 is most abundant within the plume, reaching levels as high as 1. enough ammonia was present, on average, to neutralize both sulfate and nitrate. The data suggest either of two conclusions: (1) insufficient ammonia was present to neutralize both sulfate and nitrate within the plume, thereby distinguishing ammonia as the limiting reactant; or (2) sufficient ammonia was present, but a significant amount of nitric acid formed salts preferentially with amines rather than ammonia. The detection of amines by the PILS and the large amount of organic mass, with representative amine markers detected by the cToF-AMS, suggests that the second explanation may be more plausible. This is a significant finding in the atmosphere that is consistent with laboratory observations made in photooxidation experiments of aliphatic amines (Angelino et al., 2001;Murphy et al., 2007). The affinity of inorganic acids for amines 15 in the presence of ammonia has even greater implications during the winter and at night, when lower temperatures and higher RHs enable increasing partitioning of both ammonium nitrate and amine salts into the aerosol phase.

Sources and character of amines
The formation of particulate amine salts depends on temperature, the identity and con-20 centrations of the amine and acidic species present, and the concentration of ammonia that competes with amines for the acidic species. Once the particulate amine salts are formed, they may revolatilize, undergo subsequent particle-phase reactions including oxidation, or serve as a site for the condensation of other organic compounds. Chamber experiments performed by Murphy et al. (2007) showed that the dominant for- 25 mation mechanism for amines is that of acid-base reactions (amine+nitric acid) rather than from photooxidation to form non-salt condensable organics. These experiments showed that nitric acid preferentially reacts with amines, depending on the species, 10436 rather than ammonia. It is expected that particulate amines should be prevalent close to the source of amine emissions where gaseous amine concentrations are highest.
If the temperature dependence of amine salt equilibria resembles that of ammonium nitrate, then amines should partition more favorably to the aerosol phase at lower temperature (higher altitudes) within the plume. Of the six amines studied by Murphy et 5 al. (2007), diethylamine was shown to have the most favorable equilibrium constant for salt formation in the presence of ammonia, an observation that is consistent with the present field measurements since diethylamine was the most abundant amine detected in the aerosol. The amine salts produced in the laboratory chamber experiments eventually repartitioned back to the gas phase. In the present study, amine concentrations 10 decreased as a function of plume age, as evident in the ethylamine and diethylamine data and the m/z 30 peak intensity data from the cToF-AMS. The decreasing amine levels in the aerosol phase presumably occur because of two reasons: (1) amine concentrations, like those of ammonium and nitrate, decrease due to dilution as a function of plume age; and (2) amines partition back to the gas phase to maintain thermody- 15 namic equilibrium due to the decreasing gas-phase concentrations owing to dilution. Diethylamine, measured exclusively in the plume, exhibited a strong and positive correlation with nitrate, ammonium, sulfate, and total organics during flight A (n=8, R 2 : ni-trate=0.65, ammonium=0.73, sulfate=0.72, organics=0.68), but showed a weaker correlation with the same species during flight B (n=7, R 2 : nitrate=0.35, ammonium=0.04, 20 sulfate=0.13, total organics=0.40). The positive correlation between diethylamine and nitrate suggests that nitric acid exhibits an affinity for amines as an atmospheric base, even in the presence of ammonia. Diethylamine concentrations correlated more weakly with sulfate than to nitrate during flight B, possibly because nitric acid levels were higher than those of sulfuric acid causing the formation of particulate amines to proceed only 25 through the amine +HNO 3 acid-base reaction. When diethylamine was detected, its average mass ratio relative to nitrate was 0.31 (

Total amine mass calculations
One can estimate the total mass of amines present in the plume. This calculation assumes that excess nitrate is in a 1:1 molar ratio with amines. Amine mass is then determined by assuming a representative molecular weight for the amine population. We choose to use methylamine and triethylamine as lower and upper limits, respec-5 tively, since these species represent the smallest and largest amines that can be speciated using the PILS-IC technique (Murphy et al., 2007); the molecular weight of the two amines detected in this study, ethylamine and diethylamine, fall within the range of those of methylamine and triethylamine. On the basis of the molecular weight of methylamine (31.1 g/mol), average amine

Degree of oxidation and volatility in the aerosol
In the absence of strong signals in the data representing primary particulate emission 25 sources, the submicrometer aerosol in the sampling region is presumed to originate mainly from secondary production. No obvious signs of primary aerosol vehicular emissions existed based on organic markers in the cToF-AMS spectra. The ratio of m/z 57:44 peak intensities from the cToF-AMS can provide some insight into the relative ratio of hydrocarbon-like (HOA) and oxygenated organic (OOA) aerosols (Zhang et al., 2005). This ratio was ∼0.07±0.01 during both flights, with no major changes during the flights, nor between plume and background aerosol (Table 1). Based on this ratio, 5 it appears that the aerosol was highly oxygenated with relatively little hydrocarbon-like organic aerosol (HOA). The ratio of the peak intensity between m/z 44 and total organics was, on average, 0.10±0.01 in the plume aerosol and 0.11±0.02 in the background aerosol (Table 1). Peak intensities at m/z 44 (and 29 for flight A) are slightly greater in the background aerosol relative to plume aerosol, indicating a greater degree of oxidation out of the plume than within it (Fig. 11). This is presumably because the background aerosol had aged longer than the fresh emissions in the plume. Organic acids represent a pool of organic species that are water-soluble and highly oxidized. In previous aircraft measurements, organic acids (C 1 -C 9 ) contributed 3.4±3.7% to the total PILS mass in an urban atmosphere (Houston, Texas; 15 Sorooshian et al., 2007a) and 3.5±3.1% in a marine atmosphere (Eastern Pacific Ocean; Sorooshian et al., 2007b). In the present study, organic acids contributed 2.4±5.5% to the total PILS mass, indicating greater variability and a lower average mass fraction of organic acids than seen in the other field data. Since the lower organic acid contributions are likely not a result of less photochemical processing, gas-20 particle partitioning of these water-soluble organic species may have been affected by the high ambient temperatures in the present flights. In addition, the relatively low humidities and lack of clouds prevented organic acid production via aqueous-phase processing during the measurement period (Sorooshian et al., 2006b). Organic acid concentrations were not correlated with ammonium or nitrate, which represent semi- 25 volatile species, during the present flights. The ammonium nitrate levels varied significantly in the plume, but concentrations of the organic acids were relatively stable. Thus, the data do not allow one to conclude whether volatility or RH was the dominant factor controlling organic acid levels. Introduction such as ammonium nitrate and organics, to the aerosol phase. The back-trajectory analysis indicates that the sampled air mass during flight A originated three days previously over the Pacific Ocean, while the air sampled in flight B originated in a more polluted inland area. We conclude that ventilation of the valley during flight B was more effective than in flight A, reducing aerosol number and mass concentrations.
Over the flight durations, aerosol concentrations were influenced by competition between a growing boundary layer, decreasing RH, and increasing temperatures. Although the aerosol was well-mixed locally in the valley, the timescale for equilibration between the gas and particle phases is shorter than the boundary layer mixing time (Neuman et al., 2003); this may explain fluctuations in the concentrations of aerosol 20 species at various altitudes and distances downwind of the plume source. The topography of the sampling region downwind of the source and general buoyancy in the boundary layer facilitated vertical transport of emissions to lower temperature regions, reducing the dissociation constant of ammonium nitrate aerosol and, presumably, semivolatile organics. This might explain why the concentrations of organics (diethylamine 25 in particular), nitrate, and ammonium peaked at the highest altitude and farthest downwind distance from the plume source in flight A.

Estimated subsaturated hygroscopic growth factors for the organic fraction
Calculations were carried out to determine the effective growth factor for the organic fraction needed to achieve composition-hygroscopicity closure. Due to its simplicity and frequent application (Cruz and Pandis, 2000;Dick et al., 2000;Chan, 2002a, 2002b;Prenni et al., 2003;Wise et al., 2003;Clegg et al., 2003;Clegg and Seinfeld, 2004, 2006aKhlystov et al., 2005;Rissler et al., 2005;Aklilu et al., 2006;Svenningsson et al., 2006;Varutbangkul et al., 2006;Gysel et al., 2007;Sjogren et al., 2007;Dinar et al., 2008), the Zdanovskii-Stokes Robinson (ZSR) (Zdanovskii, 1948;Stokes and Robinson, 1966) mixing rule is employed to predict hygroscopic growth factors. This procedure of estimating hygroscopic growth based on specified composi-10 tion is based on the assumption that water uptake by each individual component of a particle is independent and additive. We use the following form of the ZSR mixing rule (Aklilu et al., 2006;Gysel et al., 2007): where GF mixed is the hygroscopic growth factor of the mixed particle, GF i is the hy-15 groscopic growth factor of pure compound i, a w is the activity coefficient of water, and ǫ i is the volume fraction of pure compound i in the dry particle. a w =RH in Eq. (1) (Seinfeld and Pandis, 2006). Growth factors for the pure inorganic components were obtained from the Aerosol Inorganics Model (AIM; http://mae.ucdavis.edu/ sclegg/aim.html; Clegg et al., 1998). A growth factor of unity is also assumed for EC, as 20 suggested by Aklilu et al. (2006).
Calculating the individual volume fractions requires an estimate of the organic density. If it is assumed that the aerosol is composed of ammonium sulfate (AS), ammonium nitrate (AN), organic carbon (OC), and elemental carbon (EC), then total aerosol density can be expressed as: where x i are mass fractions, m i are mass concentrations, and ρ is density of the multicomponent particle, as determined by the ratio of aerosol volume (via the DMA) to the aerosol mass (via the PILS and cToF-AMS); in Eq. (2), ρ OC is the unknown quantity that we desire to determine. Elemental carbon was not quantified in the present study; however, based on extensive chemical characterization of San Joaquin Valley PM 2.5 5 by Chow et al. (2006), the mass fraction of EC tends to be ∼5-10% near the present sampling site. In the absence of a quantitative measure of EC, it is assumed that 5% of the total submicrometer mass is composed of EC. It is assumed for the purpose of this calculation that all of the sulfate is neutralized by ammonium, and the remaining ammonium occurs as ammonium nitrate. Densities of 1.725, 1.769, and 1.9 g/cm 3 10 are used for AN, AS, and EC respectively. A wide range of densities are reported for EC (0.625-2.25 g/cm 3 ) (Fuller et al., 1999); here we assume a value of 1.9 g/cm 3 , similar to that employed by Dillner et al. (2001). From the mass concentrations and respective densities of AN, AS, OC, and OC, the volume fraction of each component can be calculated. 15 On average, the in-plume organic growth factors needed to match the data (flight A/B) are 1.07/1.02 (74% RH), NA/1.28 (85% RH), and 1.49/1.53 (92% RH) (Fig. 19). The background aerosol organic growth factors (flight A/B) are 1.08/1.03 (74% RH), NA/1.21 (85% RH), and 1.29/1.24 (92% RH). Flight B is characterized by a wider range in the mass fractions of organics, thus this flight presents a better indication of trends in 20 organic growth factor with changing mass fractions. During this flight, inferred organic growth factors increase in the plume as a function of decreasing organic fraction. The lowest organic fractions in flight B coincide with the detection of amines, which likely enhance the hygroscopicity of the organic fraction. There is a significant amount of variation in predicted growth factors at constant organic mass fractions and the abso- 25 lute values of some of the predicted organic growth factors (<1 and >2) are unrealistic. Possible explanations for the unrealistic organic hygroscopic growth factors include: (1) complex particle morphology; (2) complex interactions between the components in the particles leading to non-additive water uptake among the individual components; and (3) errors associated with the calculation of the volume fractions and uncertainties in the measurements.

Relationship between sub-and supersaturated water uptake
One anticipates a direct correspondence between subsaturated hygroscopic behavior and supersaturated CCN activity. For example, Mochida et al. (2006) explored the 5 relationship between hygroscopicity and CCN activity for urban aerosols using a hygroscopic tandem DMA (HTDMA) coupled in series to a CCNc; enhanced CCN activity coincided with higher subsaturated growth factors. A similar analysis for the present data (Fig. 17) shows that the normalized CCN activation ratio is generally consistent with water uptake in the subsaturated regime. Also, higher activated fractions are 10 consistent with smaller markers, which represent lower organic mass fractions and higher nitrate mass fractions. Enhancements in water uptake for aerosols with lower organic content are likely due to increasing dissolution of water-soluble species, including ammonium nitrate and amine salts, and the possible reduction in surface tension by surface-active species. Organics have previously been shown to influence CCN activity by adding solute and suppressing surface tension (Shulman et al., 1996;Facchini et al., 1999;Feingold and Chuang, 2002;Nenes et al., 2002).
Recent work has shown that kinetic limitations, including surface films and slow dissolution of particulate substances, can suppress droplet growth (Asa-Awuku and Nenes, 2007;Ruehl et al., 2008). With few exceptions including solute being physically 20 "trapped" within some type of waxy material, dissolution kinetics is governed by diffusion of solute from the solid "core" at the center of a droplet to the growing droplet. It has been argued that the latter process is slow enough for compounds with high molecular weights to influence droplet growth kinetics and the Köhler curve (Asa-Awuku and Nenes, 2007;Taraniuk et al., 2007;Moore et al., 2008). According to Fig. 18, droplet 25 growth was at times less than that expected for pure ammonium sulfate. In addition, it is shown to some extent that the droplet growth was suppressed for CCN with relatively greater amounts of hydrophobic organic material. This suggests kinetic limitations may 10443 Introduction have played a role in suppressing water uptake. This issue will be revisited in subsequent work that will address size-resolved CCN data from this experiment.

Parameterization for sub-and supersaturated water uptake
To effectively represent the process of water uptake by multicomponent particles in atmospheric models, parameterizations are used. A number of investigators have at-5 tempted to introduce parameters to describe water uptake in both the sub-and supersaturated regimes. Expanding upon the earlier work of Fitzgerald et al. (1982), Svenningsson et al. (1992) used a parameter termed ǫ to link subsaturated water uptake to cloud and fog activation. Subsequent work introduced closely related parameters for sub-and supersaturated regimes (Kreidenweis et al., 2005;Rissler et al., 2006;Petters 10 and Kreidenweis, 2007). A recently introduced parameter, κ (Petters and Kreidenweis, 2007), can be calculated without knowledge of the particle properties such as density, molecular weight, and surface tension. Kappa can be determined from either CCN activity data or subsaturated hygroscopic growth data. Since the subsaturated DASH-SP growth factors are measured for size-15 resolved particles, we use the subsaturated hygroscopicity data to predict the value of with the following equation (Petters and Kreidenweis, 2007): where A=(4σ s/a M w )/(RT ρ w ), D d is the dry particle diameter, GF is the growth factor at the corresponding RH, M w is the molecular weight of water, ρ w is the density of water, 20 R is the universal gas constant, T is temperature, and σ s/a is the surface tension at the air/water interface. The water surface tension of 0.072 J/m 2 is assumed, as in the analysis of Petters and Kreidenweis (2007). Briefly, κ values of 0.5 to 1.4 represent highly hygroscopic salts such as sodium chloride, values of 0.01 to 0.5 represent slightly to very hygroscopic organics, and a value of 0 represents a non-hygroscopic component

Printer-friendly Version
Interactive Discussion (Petters and Kreidenweis, 2007; see Table 1). κ values representative of urban, maritime, continental, and remote areas, as derived by Petters and Kreidenweis (2007) using data from previous ambient studies (Fitzgerald and Hoppel, 1982;Hudson and Da, 1996;Dusek et al., 2006), have been reported to range from 0.1 to 0.94. Table 4 summarizes the values of κ derived in the present study. A noticeable en-5 hancement in κ occurs within the plume as compared to the background aerosol, which is consistent with the enhancement in subsaturated growth factors. κ is enhanced by between 21% and 67% in the plume, with typical values being between 0.36-0.44. The range of κ values, based on DASH-SP data at RHs of 85% and 92%, is 0.11-0.87. The correlation between κ values and the mass fraction of organics is more pronounced for 10 flight B; κ increases as the mass fraction of organics decreases and that of nitrate increases. Representative κ values are assigned to two categories: aerosol from the strong bovine source (κ=0.40) and aerosol in an agricultural area (κ=0.30). Values of κ determined here fall within the range of those derived from previous ambient studies.

15
An extensive set of airborne aerosol and meteorological measurements were performed downwind of a massive bovine source in the San Joaquin Valley of California during two flights in July 2007; these include meteorology, particle size distributions, aerosol composition and mixing state, sub-and supersaturated water uptake behavior, aerosol refractive index, and interrelationships between these properties. 20 Concentrations of total mass, organics, nitrate, and ammonium were elevated within the plume as compared to the background aerosol during both flights. Evidence exists of some degree of external mixing of particles in the plume. Organics constituted the dominant fraction of the total mass in the plume and background aerosol (∼56-64%), followed either by sulfate or nitrate, and then ammonium. Particulate amines were 25 detected in the plume and are shown to be a significant atmospheric base even in the presence of ammonia; the total amine concentration accounted for at least 23% (flight A) and 14% (flight B) of that of ammonium. The refractive index of the background aerosol in the valley was on average 1.54, but reductions were observed in the plume, especially in the immediate vicinity of the plume source (flight A ∼1.48; flight B ∼1.51). Measurements indicate that increasing uptake of ammonia by aerosols, in the form of ammonium nitrate and ammonium sulfate, relative to the organic fraction, results in an enhancement in particle water uptake and a reduction in refractive index. Amine salts are also hypothesized to have contributed to significant hygroscopic growth in the plume. Hygroscopic growth factors in the immediate vicinity of the source were generally between 1.75-1.90 at RH=92%. Estimated hygroscopic growth factors (RH=92%) for the organic fraction on average were 10 1.49-1.53 in the plume and 1.24-1.29 in the background aerosol. It is shown that kinetic limitations associated with hydrophobic organic species likely suppressed droplet growth. After removing effects associated with size distribution and mixing state, enhanced CCN activated fractions were generally observed as a function of increasing subsaturated growth factors, with the highest activated fractions being consistent with 15 the lowest organic mass fractions. Representative values (Petters and Kreidenweis, 2007) are assigned to two categories: aerosol from the bovine source (κ=0.40) and aerosol in an agricultural area (κ=0.30). Since organics dominated the particle mass, these values of κ are indicative of fairly hygroscopic organics. Technol., 41, 5439-5446, 2007. Dick, W. D., Saxena, P., and McMurry, P. H.: Estimation of water uptake by organic compounds in submicron aerosols measured during the Southeastern Aerosol and Visibility Study, J. Geophys. Res., 105, 1471-1479: Measuring the mass extinction 5 efficiency of elemental carbon in rural aerosol, Aerosol Sci. Tech., 35, 1009-1021. Dinar, E., Anttila, T., and Rudich, Y.: CCN activity and hygroscopic growth of organic aerosols following reactive uptake of ammonia, Environ. Sci. Technol., 42, 793-799, 2008. Doyle, G. J., Tuazon, E. C., Graham, R. A., Mischke, T. M., Winer, A. M., and Pitts, J. N.: Simultaneous concentrations of ammonia and nitric-acid in a polluted atmosphere and their Feingold, G. and Chuang, P. Y.: Analysis of the influence of film-forming compounds on droplet growth: Implications for cloud microphysical processes and climate, J. Atmos. Sci., 59, 2006-2018, 2002  Milne, P. J. and Zika, R. G.: Amino-acid nitrogen in atmospheric aerosols-occurrence, sources and photochemical modification, J. Atmos. Chem., 16, 361-398, 1993. Mochida, M., Kuwata, M., Miyakawa, T., Takegawa, N., Kawamura, K., and Kondo, Y.: Relationship  S., Fuzzi, S., Zhou, J., Monster, J., and Rosenorn, T.: Hygroscopic growth and critical supersaturations for mixed aerosol particles of inorganic and organic compounds of atmospheric relevance, Atmos. Chem. Phys., 6, 1937Phys., 6, -1952Phys., 6, , 2006 Table 2. DASH-SP hygroscopicity and dry particle refractive index data categorized by the time of flight as shown in Fig. 1    and B (lower panel). The number labels for each shaded box correspond to the portion of each flight represented in Fig. 1. The shaded areas representing the plume are characterized by significant increases in organics, ammonium, nitrate, and amines. The ammonium-to-sulfate molar ratio exceeds 2.0 in the plume, allowing nitrate to partition into the aerosol phase. The multiple cToF-AMS spikes in the species concentrations cannot be resolved by the PILS since the 5-min time intervals for sample collection average out the quick plume passes with the longer legs outside of the plume. Agreement between the PILS and cToF-AMS is most evident for sulfate, since this species was relatively level in concentration during the flights.  The cToF-AMS marker sizes for "Total mass" are proportional to the organic:inorganic ratio. Marker sizes for the individual cToF-AMS species are proportional to the respective mass fraction of that species. Marker sizes for PILS "Organic acids" are proportional to the relative contribution by oxalate. Marker sizes for PILS ammonium are proportional to the ammonium-to-sulfate molar ratio. Total mass, nitrate, ammonium, and organics increase in concentration with increasing altitude up to ∼250-300 m, before decreasing in both flights. cToF-AMS concentrations exceed those of the PILS for commonly detected species in the plume, especially nitrate and ammonium, since the PILS averages 5-min worth of aerosol composition whereas the cToF-AMS has a time resolution of ∼20-30 s. Total PILS mass includes inorganics and organic acids, whereas total cToF-AMS mass includes inorganics and nonrefractory organic mass. Introduction  Fig. 10. Aerosol mass spectra from the cToF-AMS in the background aerosol and in the plume at various downwind distances from the feedlot for flight B. There is no significant difference in the chemical signature of the aerosol in the three categories shown, with the notable exception of an enhancement in the m/z 30 peak, a common amine marker, at the closest point to the source. The flight A spectra are similar to those presented here. Fig. 11. A comparison of the plume organic mass spectra versus the background aerosol organic mass spectra for flights A (upper panel) and b (lower panel). All non-organic contributions to the mass spectra have been removed by fragmentation calculations described in Sect. 3.3.3. This means that the signal at m/z 30 represents a fragment of organics, including amines. The data on the y-axis were generated by taking the difference in the organic spectra in the plume and out of the plume. The organic aerosol appears to be very similar except this plot shows that there is an increase in peak intensities at m/z 30, 56, 74 and 86, common amine peaks, in the plume. The peak intensity at m/z 44 (and 29 in flight A) are greater in the background aerosol spectra indicating increased oxidation out of the plume as compared to inside the plume. Introduction  . The x-axis is the vacuum aerodynamic diameter. There is evidence of externally mixed aerosols, evident by the independent shifts of various species as a function of increasing plume age. This is clearest in flight B where nitrate is shown to grow in size with plume age with less growth for organics and no growth for sulfate. Introduction