As indicated in Table 2, during the first several cruises (PSI-91, MAGE92, RITS93, RITS94), total particle number concentrations for Dp greater than 3 and 12 nm were measured with TSI 3025 and 3760 CPCs, respectively. The ultrafine particle number concentration was then defined as the difference between the number concentration measured with the 3025 and 3760 CPCs. For these initial cruises and all subsequent ones, diffusion driers (Permapure Inc.) were placed upstream of the CPCs to minimize particle diameter changes due to hygroscopic growth to less than 5 % (Swietlicki et al., 2008). The use of a diffusion drier also helped prevent the uptake of water in the CPC condenser that results when sampling in the humid marine atmosphere. Beginning with ACE-2 in 1997 and continuing through ATOMIC in 2020, particle number concentrations for Dp>12 nm were measured with a TSI 3010 CPC. Starting in 2008 for VOCALS, a water-based TSI 3785 CPC was added to also measure the concentrations of particles with diameters greater than 3 nm.

For MAGE92, RITS93, and RITS94, particle number size distributions from 0.02 to 0.6 µm were measured with a TSI 3071 Differential Mobility Analyzer (DMA) (Quinn et al., 1998) with the number concentration in each bin measured with a TSI 3760 CPC. The resulting number mobility distributions were inverted to a number size distribution by using the manufacturer-provided algorithm (Keady et al., 1983) and assuming that a Fuchs-Boltzman equilibrium charge distribution resulted from a 85Kr charge neutralizer (TSI model 3077) on the inlet of the DMA. The number concentration was corrected for the counting efficiency of the CPC (Zhang and Liu, 1991) and diffusion losses in the DMA (Reineking and Porstendorfer, 1986). The sample air passed through a diffusion drier to reduce the RH to less than 25 %.

A Vienna short column Ultrafine DMPS (UDMPS) coupled to a TSI 3025 CPC was added for ACE-1 and all subsequent cruises to extend the size distribution measurements to the 0.005 to 0.02 µm size range. For both ACE-1 and CSP, the UDMPS and the TSI 3071 DMA were located outside of the temperature-controlled box. Sheath air for both resulted in a measurement RH of less than 25 % RH. As for the earlier cruises, the mobility distributions were inverted to number size distributions by assuming that a Fuchs-Boltzman charge distribution resulted from the 85Kr charge neutralizers on the inlet of the DMAs. The data were corrected for diffusional losses (Covert et al., 1997) and size dependent counting efficiencies (Wiedensohler et al., 1997) based on pre-ACE-1 intercalibration exercises. In addition, an APS (TSI 3300) was added for ACE-1 and all subsequent cruises to measure the number size distribution between 0.6 and 9.6 µm. APS diameters measured at ∼40 % RH were converted to geometric diameters by dividing by the square root of the particle density for sea salt (1.9 gcm-3) and dried to 10 % RH assuming a sea salt growth factor of 1.5 between 10 % and 40 % RH (Berg et al., 1998).

For ACE-2 and all subsequent cruises, the UDMPS and TSI DMA were put into the temperature-controlled box to maintain an instrumental RH of greater than 45 % (see Table 2). The APS was also transferred to the temperature-controlled box where it measured at an RH of approximately 40 %.

For AEROSOLS99 and all subsequent cruises, a Vienna medium column DMPS was used to measure particles in the 0.02–0.9 µm size range instead of the TSI 3071 DMA. In addition, the APS model 3300 was replaced with an APS model 3320. Duplicate Vienna medium column DMPSs were deployed for AEROSOLS99 and INDOEX. One measured at 10 % RH outside of the temperature-controlled box and the other measured at 55 % RH inside the box.

Although the APS was located in the temperature-controlled box and its inlet was maintained at 55 % RH, internal heating of the sample flow by its sheath flow and waste heat likely reduced the measurement RH (Bates et al., 2004). For ACE-Asia and all subsequent cruises, the APS sheath flow was routed outside of the instrument to equilibrate with the air temperature in the temperature-controlled box and then reintroduced to the sheath and acceleration nozzle to lower the temperature and increase the measurement RH. Also starting with ACE-Asia, densities and the associated water masses at the instrumental RH were calculated with a thermodynamic equilibrium model (AeRho) using the measured inorganic ion composition (Quinn et al., 1998). These calculated densities were used to convert the APS data from aerodynamic to geometric diameters for merging with the DMPS data. Due to the atmospheric dust that was sampled during ACE Asia, the APS data were corrected for ultra-Stokesian conditions in the instrument jet and nonspherical shape (Wang and John, 1987).

A CCN counter (DMT CCN-100) was added for TexAQS and several later cruises (CalNex, NAAMES-1, NAAMES-2, NAAMES-3, NAAMES-4, and ATOMIC) (see Table 2). A CCN counter was onboard during WACS and WACS2 but it sampled nascent SSA with Sea Sweep for the majority of the time. CCN concentrations were measured at supersaturations between 0.2 % and 1.0 %. Details of the CCN counter can be found in Roberts and Nenes (2005). A multijet cascade impactor (Berner et al., 1979) with a 50 % aerodynamic cut-off diameter of 1.1 µm was upstream of the CCN counter. More details about the CCN measurements can be found in Quinn et al. (2008).

A seven stage multi-jet cascade impactor was used on all cruises to collect size segregated samples for ion chromatography analysis. The ions quantified were Na+, NH4+, K+, Ca2+, Mg2+, Cl-, NO3-, SO4=, and methane sulfonate or MSA⁻. A Millipore Fluoropore filter (1.0 µm pore size) was used for the final, smallest size range stage. The Millipore filter has a collection efficiency of 99 % or greater for particles with diameters larger than 0.035 µm (Liu and Lee, 1976). Tedlar films were used for the six largest stages. The films were cleaned in an ultrasonic bath in 10 % H2O2 for 30 min, rinsed 6 times in distilled, deionized water, and then dried in an NH3- and SO2-free glove box. Material collected on the filters and films was extracted by wetting with 1 mL of methanol and then adding 5 mL of distilled deionized water and sonicating for 30 min. Samples were handled in a glove box that was purged with air that had passed through a scrubber containing potassium carbonate, citric acid, and activated charcoal to remove SO2, NH3, and volatile organics, respectively. Sampling times varied between ∼12 and 36 h and were based on the amount of aerosol present.

For the first few cruises (MAGE92, RITS93, RITS94), higher time resolution (<12 h) submicron aerosol samples were collected using a filter holder downstream of a cyclone with a Daero,50 of 1 µm. Starting with ACE-1, a 2-stage impactor was used for higher time resolution sampling of sub- and supermicron aerosol for ion chromatography analysis. Substrates, sample handling, and blank determinations were the same as discussed above.

A Particle-Into-Liquid-Sampler (PILS) coupled to an ion chromatograph was used to sample submicron inorganic ions during NEAQS 2002, NEAQS 2004, and TexAQS at a higher time resolution (15 min) than any of the impactors (Bates et al., 2008). The common aerosol inlet was used to deliver aerosol to a PILS at 55 % RH. An impactor with a Daero,50 of 1.1 um was upstream of the PILS. Flow through the impactor was 30 slpm with 15 slpm through the PILS and 15 slpm through a bypass line. Two annular, glass denuders (URG) were in series downstream of the impactor and upstream of the PILS. One was coated with sodium carbonate for the removal of gas phase acids and the other was coated with citric acid to remove gas phase bases. Two Kloehn syringe pumps were used to deliver a solution of LiF to the top of the PILS impactor to correct for dilution of the sample within the PILS. Two additional pumps delivered sample from the PILS simultaneously to a cation and an anion IC. More information about the PILS can be found in (Weber et al., 2001). Between every 45 min to 2 h, sample air was passed through a HEPA filter for 15 min to remove particles and determine the measurement blank. This blank was subtracted from the sample concentrations.

For both the impactor and the PILS data, non-sea salt (nss) SO4= concentrations were calculated from Na+ concentrations and the ratio of sulfate to sodium in seawater. Sea salt concentrations were calculated from 1Sea salt(µgm-3)=Cl-(µgm-3)+Na+(µgm-3)×1.47 where 1.47 is the seawater ratio of (Na++K++Mg2++Ca2++SO4=+HCO3-)/Na+ (Holland, 1978). This approach prevents the inclusion of non-sea salt K+, Mg2+, Ca2+, SO4=, and HCO3- in the sea salt mass and allows for the loss of Cl- mass through Cl- depletion processes. It also assumes that all measured Na+ and Cl- is derived from seawater. Results of Savoie and Prospero (1980) indicate that soil dust has a minimal contribution to measured soluble sodium concentrations.

Starting with ACE-Asia in 2001, sub-1 and sub-10 µm samples were collected for OC/EC analysis using 2 and 1 stage impactors, respectively (Bates et al., 2004). Each impactor had 2 quartz backup filters. OC concentrations from both impactors were corrected for blanks and artifacts using the last quartz filter in line. Aluminum foil was used as a substrate on the 1.1 µm jet plate. All substrates, aluminum foil and quartz, were prebaked at 500° prior to sampling. One sub-10 µm and one sub-1 µm impactor were operated without a denuder upstream to avoid losses of large particles in the denuder. OC from the sub-1 µm impactor was subtracted from the OC from the sub-10 µm impactor to determine supermicron OC concentrations. A second sub-1 µm impactor was operated with a denuder upstream that contained strips of carbon-impregnated glass fiber filters to remove gas phase organics. Sub-1 µm OC and EC were quantified on the impactor samples downstream of the denuder.

During ACE-Asia and NEAQS 2002, a 7-stage impactor was used for the sampling of OC/EC providing greater size resolution. Aluminum foil substrates were used on all stages of the impactor along with 2 quartz backup filters. A 3-stage impactor (D50,aero of 0.18, 1.1, and 10 µm) was used during WACS, WACS-2, and NAAMES-1 through NAAMES-4. These cruises focused on the composition and properties of sea spray aerosol. The 3 size cuts allowed for the separation of the sub-0.18 µm aerosol from larger size ranges in recognition that smaller particle sizes are known to be enriched in organics through the sea spray aerosol production process (Keene et al., 2007).

OC and EC concentrations on the impactor substrates were measured with a Sunset Labs thermal/optical analyzer (Birch and Cary, 1996). Four temperature steps were used to achieve a final temperature of 870 °C in He to drive off OC. After cooling the sample down to 550 °C, a He/O2 mixture was introduced and the sample was heated in four temperature steps to 910 °C to drive off EC. The transmission of light through the filter was measured to separate EC from any OC that charred during the initial stages of heating. No correction was made for carbonate carbon so OC included both organic and inorganic carbon. The mass of particulate organic matter (POM) was determined by multiplying the measured organic carbon concentration in µgm-3 by a factor of 2.1 in marine regions and 1.6 elsewhere (Turpin and Lim, 2001).

A semi-continuous real-time Sunset Labs thermal/optical analyzer was used during NEAQS 2004 for higher time resolution measurements of OC concentrations. The OC/EC analyzer was downstream of a sub-1 µm impactor and a denuder. The analyzer collected air on a filter for 45 or 105 min depending on OC concentrations. At the end of the sampling time the instrument analyzed the filter using the same temperature program described above. The sampling times were not long enough to measure EC above the detection limit of 0.35 µgm-3.

A PILS coupled to a Total Organic Carbon (TOC) analyzer (Sievers Model 800 Turbo) was used during TexAQS to measure water soluble organic carbon (WSOC) (Bates et al., 2008). As for the PILS-IC, the PILS-WSOC was connected to the common aerosol inlet but with a stainless steel line. An impactor with a D50,aero of 1.1 µm was upstream of the PILS to sample submicron aerosols. A denuder identical to the one used in the thermal/optical analysis was downstream of the impactor and upstream of the PILS to remove gas phase organics. Two syringe pumps (Kloehn) delivered low-TOC water to the top of the PILS impactor. Two additional pumps were used to pull sample out of the PILS and into the TOC analyzer. The sample was passed through a 0.5 µm in-line filter before entering the TOC analyzer to measure WSOC. Between every 45 min to 2 h, sample air was passed through a HEPA filter for 15 min to remove particles and determine the measurement background. The measurement background was subtracted from the sample air to obtain ambient WSOC ambient atmospheric concentrations.

Starting with AEROSOLS99, sub-1 and sub-10 µm samples were collected for trace element analysis using impactors with a D50,aero of 1.1 and 10 µm, respectively (Quinn et al., 2001). Energy Dispersive X-RAY Fluorescence (ED-XRF) was used for quantification (Buck et al., 2021). Both impactors collected aerosol on 2.0 µm pore size PALL Teflo Membrane Disc Filters. Supermicron concentrations were determined by subtracting the sub-1.1 µm values from the sub-10 µm values. No corrections were made for particle size or loading. Samples for XRF analysis were collected during all subsequent cruises except for VOCALS, WACS, WACS-2, and NAAMES-1 to 4.

Concentrations of Al, Si, Ca, Fe, and Ti are included in the NCEI archive as these were used to calculate dust. Other trace elements were measured but were often below detection limit so they were not reported. Dust concentrations were calculated based on measured values of Al, Si, Ca, Fe, and Ti assuming that each element was present in the aerosol in its most common oxide form (Al2O3, SiO2, CaO, K2O, FeO, Fe2O3, TiO2) (Seinfeld, 1986). The measured elemental mass concentration was multiplied by the appropriate molar correction factor as shown below (Malm et al., 1994) 2Dust=2.2(Al)+2.49(Si)+1.63(Ca)+2.42(Fe)+1.94(Ti). This equation includes a 16 % correction factor to account for the presence of oxides of other elements such as K, Na, Mn, Mg, and V that are not included in the linear combination. In addition, the equation omits K from biomass burning by using Fe as a surrogate for soil K and an average K/Fe ratio of 0.6 in soil (Braaten and Cahill, 1986). Non-crustal K was calculated using the K/Al ratio (0.31) of Asian loess (Jahn et al., 2001) which is similar to the ratio in Saharan dust (0.24) and average crustal rock (0.32) (Formenti et al., 2003). Sea salt Ca was accounted for based on the ratio of Ca to Na in seawater.

Concentrations of submicron non-refractory (NR) NH4+, SO4=, NO3-, and particulate organic matter (POM) were measured on NEAQS 2004, TexAQS, ICEALOT, VOCALS, CalNex, DYNAMO, WACS and WACS-2 with a Quadrupole Aerosol Mass Spectrometer (Q-AMS, Aerodyne Research Inc., Billerica, MA, USA) (Jayne et al., 2000). The NR species measured by the AMS are defined here as all the chemical components that vaporize at 550 °C. The AMS was downstream of an impactor with a D50,aero of 1.1 µm. The ionization efficiency of the AMS was calibrated every few days with dry monodisperse ammonium nitrate particles. Particle losses due to transmission through the aerodynamic lens were corrected by using the DMPS and APS-measured size distributions. Particle losses due to bounce off of the impactor-vaporizer were corrected using simultaneously sampled NH4+ and non-sea salt SO4= concentrations from either the PILS-IC or the impactors (Quinn et al., 2008, 2006).

A filter holder with a Millipore Fluoropore filter (1.0 µm pore size) collected aerosol downstream of a cyclone with a D50,aero of 1 µm during RITS93 and RITS94. Filters were taken back to PMEL for gravimetric analysis to determine total submicron aerosol mass. The filters were weighed before and after sample collection with a Mettler UMT2 microbalance. The microbalance was housed in a glove box maintained at a constant RH to allow each sampled filter to come into equilibrium with the same vapor pressure of water, thus reducing experimental uncertainty due to a variable lab RH. For RITS93 and RITS94 the RH was maintained at less than 30 % by circulating air through a flat baffle box containing a saturated solution of MgCl₂ ⋅ 6 H₂O and then through the glove box (Young, 1967). The circulated air was cleaned by passing it through a scrubber containing activated charcoal, potassium carbonate, and citric acid. Filters were equilibrated overnight in the glove box prior to weighing. Static charging, which can result in balance instabilities, was minimized by coating the walls of the glove box with a static dissipative polymer (Tech Spray, Inc.), placing an antistatic mat on the glove box floor, and exposing the filters to a 210Po source to dissipate any built-up charge.

For ACE-1 and the other cruises listed in Table 3, a 2-stage impactor was used to collect submicron and supermicron aerosol for gravimetric analysis. Millipore Fluoropore filter (1.0 µm pore size) and Tedlar films were used for the collection of submicron and supermicron aerosol, respectively. The Tedlar films were cleaned as described in Sect. 3.3.1 prior to sample collection. Both the Millipore filters and the Tedlar films were weighed before and after sampling. Millipore filters were weighed on the Mettler UMT2 microbalance and Tedlar films were weighed on a Cahn Model 29 microbalance. Both balances were housed in the RH-controlled glove box described above. For cruises with higher sampling RHs of 55 % to 60 % (see Sect. 3.1), a saturated solution of KBr was used in the baffle box.

In addition to the 2-stage impactor used during ACE-Asia, a 7-stage impactor was used for higher size resolution total aerosol mass concentrations.

All reported mass concentrations include the water mass that is associated with the aerosol on the filter at the glove box RH.

During RITS93 and RITS94, sub-10 micron aerosol light scattering was measured with a newly developed, highly sensitive multiwavelength integrating nephelometer (Bodhaine et al., 1991). This nephelometer, with its high sensitivity, was combined with the closed geometry of the Ahlquist and Charlson (1967) nephelometer to develop the TSI, Inc. model 3563 (Anderson et al., 1996) that was used during the rest of the cruises reported on here. The enclosed geometry allows for the calibration of the nephelometer with gases with known scattering coefficients.

The TSI Inc. model 3563 integrating nephelometer was used for all remaining cruises to measure sub-1 and sub-10 micron scattering at three wavelengths (450, 550, and 700 nm). Sub-1 and sub-10 micron backscattering were measured on cruises between ACE-1 and WACS. Two single-stage impactors, one having a D50,aero of 1.1 µm and the other of 10 µm were placed upstream of the nephelometer. A valve automatically switched between the two impactors every 15 min so that sampling alternated between sub-1 micron and sub-10 micron aerosol. Scattering and backscattering by the supermicron aerosol was determined by difference.

During all cruises, the nephelometer was calibrated with CO2 and zeroed with particle-free air every 3 to 4 d (Quinn et al., 1996). The resulting zero offset and span factors were applied to the data. In addition, data were corrected for angular nonidealities of the nephelometer, including truncation errors and non-Lambertian illumination using the method of Anderson and Ogren (1998) or one similar to it.

The RH of the air sampled by the nephelometer was nominally that of the common inlet as described in Sect. 3.1. Heating within the nephelometer likely led to slightly lower RH's than for the sizing instruments detailed in Table 2. For ACE-Asia and all subsequent cruises the nephelometer flow path was modified so that the sheath flow was conditioned outside of the instrument case to equilibrate with the temperature-controlled box. It was then reintroduced into the sheath and acceleration nozzle to minimize heating of the sample air and lowering of the measurement RH. In addition, an RH sensor was placed in the nephelometer sensing volume.

As indicated in Table 4, the relative humidity dependence of scattering, f(RH), was measured on TexAQS, ICEALOT, VOCALS, CalNex, DYNAMO, WACS, WACS-2, NAAMES-1, NAAMES-2, and ATOMIC. A humidity-controlled system measured light scattering at two different relative humidities, ∼20 % and ∼85 %, with two nephelometers operated in series downstream of an impactor (D50,aero=1.1 µm). The first nephelometer in line measured sample air dried with a PermaPure, multiple-tube nafion dryer (model PR-94). Downstream of this nephelometer a humidifier was used to add water vapor to the sample flow using 6 microporous Teflon tubes surrounded by a heated water-jacket. Humidity was measured using a chilled mirror dew point hygrometer downstream of the second nephelometer in line. The same calibration procedure described in Sect. 3.4.2 was used (Quinn et al., 2022).

Between ACE-1 and INDOEX, the aerosol light absorption coefficient of sub-10 µm was measured with a Particle Soot Absorption Photometer (PSAP, Radiance Research) at a wavelength of 550 nm and ∼55 % RH. Measured values were corrected for a scattering artifact, the deposit spot size, flow rate, and the manufacturer's calibration (Bond et al., 1999). Beginning with NAURU99, the absorption coefficient was measured for sub-1 and sub-10 µm aerosol with the PSAP located downstream of the same impactors as the nephelometer. For NEAQS 2002 and all subsequent cruises, a modified PSAP was used to measure light absorption at three wavelengths (467, 530, and 700 nm) close to that of the TSI nephelometer for calculation of single scattering albedo (Virkkula et al., 2005). Beginning with TexAQS and all following cruises, a PermaPure nafion dryer was placed upstream of the PSAP so that the sample air was at ∼25 % RH. Measurement of dry air was found to reduce instrument noise.

Handheld sunphotometers were used to measure AOD for ACE-1 and all subsequent cruises. A single wavelength (550 nm) sunphotometer was used for ACE-1. A microtops unit (Solar Light Co.) was used for all other cruises measuring at 380, 440, 500, 675, and 870 nm. Units were calibrated before each cruise by either Solar Light Co. or NASA Goddard Space Flight Center (GSFC) using a Langley plot approach (Shaw, 1983). Initially, a NASA Sensor Intercomparison and Merger for Biological and Interdisciplinary Oceanic Studies (SIMBIOS) MATLAB routine was used to convert raw signal voltages to AOD. Included in the conversion is a correction for Rayleigh scattering (Penndorf, 1957) and air mass to account for the curvature of the Earth (Kasten and Young, 1989). Beginning with ICEALOT in 2008, data were reduced as part of NASA's Maritime Aerosol Network (Smirnov et al., 2009).

O3 was measured on all cruises with the exception of NAURU99 and WACS. Three different ozone UV analyzers were used over the years including a Dasibi 1008-AH UV photometer, a TECO Model 49 O3 Analyzer, and a TECO Model 49C O3 Analyzer (Table 5). For all cruises, a 0.635 cm ID Teflon sample line was used to draw air from the top of the aerosol common sampling mast to the O3 instrument located in the lab container at the base of the mast. The loss of O3 in a Teflon sampling line is approximately 5 % per 30 m indicating that losses were negligible (<3 %). At intervals of 1 to 4 d, a charcoal filter was placed in the sampling line for 1 h to determine a zero which was subtracted from the O3 signal. More details can be found in Johnson et al. (1990).

SO2 was measured during ACE-Asia, NEAQS 2002, NEAQS 2004, TexAQS, ICEALOT, VOCALS, and CalNex with a Thermo Environmental Instruments Model 43C trace level pulsed fluorescence analyzer. Air was drawn through the 18 m aerosol common sampling mast at 1 m3min-1. At the base of the mast, a 5.0 Lmin-1 flow was pulled in series through a 1 m long Teflon tube, a Millipore Fluoropore Teflon filter (1.0 µm pore size), a Perma Pure Inc. Nafion dryer (MD-070), a 2 m long Teflon tube, and then into the SO2 analyzer. The initial 1 m of tubing, filter, and drier were located in the humidity-controlled box at the base of the mast. Dry air was pulled through a charcoal trap and then through the outside of the Nafion dryer at 2 Lmin-1. The analyzer was run with two channels (0–20 ppb full scale and 0–100 ppb full scale) and a 20 s averaging time.

Zero air (scrubbed with a charcoal trap) was introduced into the sample line upstream of the Fluoropore filter for 10 min every hour to establish a zero baseline. An SO2 standard was generated with a permeation tube held at 50 °C. The flow over the permeation tube, diluted to 17.7 ppb, was introduced into the sample line upstream of the Fluoropore filter for 10 min every 6 h (Bates et al., 2004).

The rate of emission of radon from the ocean is ∼100 times less than over land. As a result, 222Rn is a qualitative tracer of an air mass that has been recently influenced by continental emissions (Carlson and Prospero, 1972).

Radon was measured on all cruises starting with ACE-1 except AEROSOL99, INDOEX, NAURU99, and DYNAMO. Radon (222Rn – half-life of 3.82 d) was measured using the two-filter detector method of Whittlestone and Zahorowski (1998). Air is drawn through a HEPA filter which removes all radon and thoron decay products (i.e., daughters), then through a delay chamber in which some daughters are produced. Finally, the air passes through a second filter which retains the daughters. These daughters have been produced in controlled conditions so their number is proportional to the radon concentration. A photomultiplier then counts the radon daughters produced in a 750 L decay tank for a 30 min period. The detector was standardized using radon emitted from a dry radon source (RN-25, PylonElectronics Corp). Background counts were measured under conditions of zero air flow (Quinn et al., 2022).

Air and seawater samples for DMS were analyzed using an automated purge and trap system. Air samples were collected through a Teflon line which ran approximately 60 m from the top of the aerosol sampling mast to the instrument. One hundred mLmin-1 of the 4 Lmin-1 flow were pulled through a KI solution at the instrument to eliminate oxidant interferences (Cooper and Saltzman, 1993). The air sample volume ranged from 0.5 to 1.5 L depending on the DMS concentration. Water vapor was removed by passing the flow through a -25 °C Teflon tube filled with silanized glass wool. DMS was then trapped in another -25 °C Teflon tube filled with Tenax. During the sample trapping period, methylethyl sulfide (MES) was added to the sample stream as an internal standard. At the end of the sampling/purge period, the coolant was pushed away from the trap and the trap was electrically heated. DMS was desorbed onto a DB-1 mega-bore fused silica column where the sulfur compounds were separated isothermally at 50 °C quantified with either a Flame Photo Detector (FPD) or a Sulfur Chemiluminescence Detector (SCD). The instrument was calibrated gravimetrically with calibrated permeation tubes. More details of the analysis can be found in Bates et al. (1998b).

Seawater samples for DMS analysis were collected from the ship's seawater pumping system at a depth of approximately 5 m below the ship's waterline. Periodically, a 5 mL water sample was valved from the ship's water line into a Teflon gas stripper. The sample was purged with hydrogen for 5 min. DMS and other sulfur gases in the hydrogen purge gas were collected on the Tenax trap held at -25 °C as for the air samples. Seawater and air sample analysis was identical.

Seawater samples for the analysis of NH4+ and NO3- were taken from a depth of ∼5 m using the ship's seawater pumping system. Samples were analyzed for NH4+ using the phenolhypochlorite colorimetric method of Solarzano (1969) and for NO3- using the method of Parsons et al. (1984). Both of the analyses were undertaken with a Technicon Autoanalyzer II (Technicon Corp., Tarrytown, New York).

Discrete seawater samples for chlorophyll a analysis were taken from the ship's seawater pumping system 2 to 6 times per day. Samples were immediately filtered, put into 10 mL of 90 % acetone, and frozen. Samples were analyzed with a fluorometer within 3 to 4 d onboard the ship. Depending on the cruise, the fluorometer was calibrated several times during, before, or after the experiment usually with algal chlorophyll a (Sigma Chemical Corp.). The discrete samples were used to calibrate continuous fluorescence measurements of seawater also from the ship's underway seawater pumping system.