SIMBIOS METHODS Flow-Through System Our flow-through system is equipped with several sensors, with data integration and logging performed by National Instruments LabVIeW software run on two pentium processor computers. The system receives time and the ship's geographic position continually from a Garmin 220 global positioning system, the antenna of which is mounted outside of the ship, off the stern. Seawater first enters a vortex debubbler (SUNY #VDB-1), then a 4 foot tall debubbler equipped with a 1mm screen to keep the largest zooplankton and salps out of the optical instrumentation. The flow then enters an InterOceans temperature and conductivity sensor. It measures salinity with an accuracy of +0.05 Practical Salinity Units and temperature to an accuracy of +0.1oC. Following the temperature/salinity measurements, the flow bifurcates into a Turner Designs 10-AU-005-CE fluorometer for monitoring underway fluorescence and a tertiary debubbler prior to measurements of light scattering. The fluorometer is equipped with a daylight white lamp, 340-500nm excitation filter, and >665nm emission filter. Due to fast growing bio-fouling organisms in the Gulf of Maine, the fluorometer is cleaned regularly. As for the light scattering measurements, the water passes from the tertiary debubbler, via diaphragm metering pump, into a Wyatt Technologies Model Dawn laser light scattering photometer at a flow-rate of ~15ml min-1. The photometer operates with an 10 mW Argon ion laser (514 nm) which is directed into the center of a flow-through cuvette, whereupon, seawater is viewed by 18 photodiodes arranged between 21.54o and 158.14o. Included in the 18 detectors are two photodiodes for laser power monitoring (one prior to passage through the viewing cuvette, and one post). The laser beam has a 1/e2 gaussian beam profile radius of 0.39 mm which makes the effective viewing volume of the light scattering photometer 0.25 ml. All detectors are scanned at rates up to 400 HZ. In fact, for most flow-through applications, we slow the scanning rate to 200 HZ in order to not sample the same seawater volume twice. The LabVIeW software, which controls the Wyatt light scattering photometer, can be programmed to calculate averages and standard deviations of seawater volume scattering data to any desired time period. For field applications, we typically average the data for about 50 seconds (which then represents an effective volume viewed of 13ml). The statistics are highly informative for understanding the variance of the optics due to particle types. Because of our interest in calcite, we also measure seawater pH immediately after the light scattering instrument to verify that the pH is sufficiently high such that calcite cannot dissolve. Following the first 50s of measurements done on each raw seawater sample, another peristaltic pump is activated by the LabVIeW control system, which injects 0.5% glacial acetic acid into the flow stream, and mixes it by running it through a Teflon mixing coil. This drops the pH to about 5.8 to dissolve any calcium carbonate. Once the pH stabilizes at the more acidic value, volume scattering is re-sampled, and average backscattering re-calculated. The difference between the raw and acidified backscattering values represents the "acid-labile" backscattering. Using field measurements, we have calibrated this acid-labile backscattering to atomic absorption estimates of suspended calcite concentration (r2=0.83). The time for a complete acidification cycle can be adjusted, but we have preferred to collect average backscattering values such that one complete raw/acidification cycle takes 4 minutes. This means that during any passage, we would be logging a data point about once every 2000 meters. For sea-truth measurements, this is adequate since typically a 3 pixel by 3 pixel area from SeaWiFS is viewed (10.9 km2). Water next flows into an AC-9 (Wet Labs, Oregon). This instrument simultaneously measures spectral beam attenuation and spectral absorption at nine wavelengths using a dual path optical scheme. Fundamentally, this consists of two pressure housings, with the absorption and attenuation beam paths in between. The absorption light path passes through a reflective tube while the attenuation light path passes through a non-reflective tube. A rotating filter wheel provides the 9 different wavelengths, between 412-715 nm. The accuracy of the attenuation and absorption measurements are +0.005m-1 with linearity error of +0.1%. With access to attenuation (c) and absorption (a) information, we calculate scattering (b) = c - a. Alternating every 128 seconds the water is filtered through a 1 micron then a .2 micron prior to passing through the ac9. This allows discrimination between the total inherent optical properties (IOP's; dissolved and particulate) versus dissolved IOP's. Samples taken during the 16 second sample surrounding the transitions between filtered an unfiltered water are discarded. The last in-flow measurement is with the HobiLabs Hydroscat-2. This instrument is set to view an enclosed, 20l, sandblasted, stainless steel container (painted flat black within). This vessel has a "sweepable" brush inside to remove bubbles from the viewing window. The instrument measures volume scattering at 142o, and extrapolates to backscattering using an assumed volume scattering function. It makes measurements at 470 and 676nm plus chlorophyll fluorescence. Water-leaving radiance and downwelling irradiance (for calculating remote sensing reflectance) is measured from the Scotia Prince ferry using a Satlantic SeaWiFS Aircraft Simulator (SAS). This consists of a total radiance sensor and a sky radiance sensor mounted on the bow, and an irradiance sensor mounted on the compass deck, aft of the bridge, as far from any potentially shading structures as possible. The total radiance detector views the water forward of the ship, at 40o from nadir, the sky radiance sensor views the sky at 40o from zenith at the same azimuth angle as the total radiance sensor. The distance of the total radiance sensor to the water is ~15m. The direction of the radiance sensors is changed periodically, as the sun's azimuth changes, so, when possible, the sensor is viewing the water between 80o and 120o from the sun's azimuth, minimizing sun glint. Protocols for operation were made according to SeaWiFS technical memorandum #25 and the NASA SeaWiFS SeaBOARR Experiments (Postlaunch TM's 3 and 8). Data are logged at 6Hz and all but the lowest 5% of the data in any 16 second time interval discarded (with 488 nm as the reference wavelength). Discrete Samples: Every hour a water sample was taken for suspended CaCO3, particulate organic carbon, chlorophyll, and microscope counts. The technique of Fernandez et al. (1993) was used to measure CaCO3 concentrations. Briefly, 500 ml samples were filtered onto 1um pore-size polycarbonate filters, and rinsing first with filtered sea water, then borate buffer (pH=8) to remove seawater calcium chloride. Filters were placed in trace metal free centrifuge tubes with 5 ml 1.0% Optima grade Nitric acid (this will also drive off any 14C activity of the coccoliths). Next, the Ca concentration was measured using a Mass Spec. OES (Optical Emission Spectrometer). Chlorophyll and particulate organic carbon were measured according to the JGOFS protocols (JGOFS, 1996) using Millipore .45um membrane and GF/F filters respectively. Microscope enumeration of coccolithophores and coccoliths is done by filtering a 50ml water sample through a Millipore HA filter, rinsed with borate buffer, and frozen in a petri dish until counted (Haidar and Thierstein, unpublished-a; Haidar and Thierstein, unpublished-b). The filter is placed on a glass microscope slide, and 60oC Canada Balsam is placed on top of the filter, followed by a cover slip. The clarified filter is examined with an Olympus BH2 microscope equipped with polarization optics. Birefringent coccoliths and plated coccolithophores can then be counted. For statistical reasons, 200 coccoliths or cells are counted from each sample, when available. Data manipulation The data produced by this system are: time latitude longitude total radiance (uW cm-2 nm-1 sr-1; at 412, 443, 490, 510, 555, 670 & 685 nm) sky radiance (uW cm-2 nm-1 sr-1; at 412, 443, 490, 510, 555, 670 & 685 nm) downwelling irradiance (uW cm-2 nm-1; at 412, 443, 490, 510, 555, 670 & 685 nm) fluorescence (volts) calibrated to hourly discrete chlorophyll samples backscattering (m-1; at 470, 514 and 676 nm) absorption (m-1; at 412, 440, 488, 510, 555, 630, 650, 676 and 715 nm) attenuation (m-1; at 412, 440, 488, 510, 555, 630, 650, 676 and 715 nm) In addition, we also take Microtops sun photometer measurements when the sun was visible, and the data are offloaded for distribution to SEABASS. Discrete samples for suspended calcite, POC, PON, and coccolith counts are also taken hourly along with an XBT temperature profile, and are processed over subsequent months. Hourly XBT profiles are used to construct one temperature section per trip from which we can calculate isotherm slope (strongly related to gradients in integrated primary production and chlorophyll).