USGS Home
Contact USGS
Search USGS

Methods

Sampling of water

During the 5-week (August 25 to September 27, 2012) UNCLOS (United Nations Convention Law of the Sea) Extended Continental Shelf Project cruise on the U.S. Coast Guard Cutter Healy (Co-Chief Scientists: Larry Mayer and Andrew Armstrong), discrete and continuous underway water samples were collected and analyzed to document the carbonate chemistry of the Arctic waters. These data serve as a baseline and are being used to test the saturation state of the seawater with respect to calcium carbonate. These data are critical to refine existing models.

Underway Continuous Measurements

Approximately 1,800 continuous measurements of pH, pCO₂, TCO₂, salinity, and temperature were collected from August 31, 2012, to September 2, 2012, using a flow-through Multiparameter Inorganic Carbon Analyzer (MICA), and the methods of Wang and others (2007) and a Sea-Bird SBE49 CTD attached to the flow-through system of the USCGC Healy. Geographic, salinity, temperature, and fluorometric data were also collected using a shipboard Ashtech ADU5 GPS system, a Sea-Bird SBE45 Thermosalinograph, and a Seapoint Chlorophyll Fluorometer (SCF). A complete description of these can be found in Chayes and others (2010). The intake of the shipboard flow-through system was located approximately 8 meters (m) below the sea surface on the port side of the vessel. Water entered the sampling baffles at depth, was pumped into an ice chest for separation of ice, and was then pumped into a multi-port sampling manifold located in the ship's laboratory. Filtered seawater was then fed to a custom made PVC de-bubbler containing a Sea-Bird SBE49, prior to being transported to the intake port of the MICA. Measurements were taken and logged approximately every 2 min except during a MICA flushing cycle which occurred for approximately 10 min each hour (h). The MICA was calibrated using Certified Reference Material from Professor Andrew Dickson of the University of California at San Diego. Precision and accuracy for each channel was 0.002 for pH, 2 parts per million (ppm) pCO₂, and 2 micromoles per kilogram (μmol/kg) for TCO₂. Quality assurance and quality control of data were performed by identifying data that deviated significantly from previous values and by identifying spectrophotometric intensity values that fell outside of acceptable ranges.

Discrete Surface Samples

Discrete water samples were collected while underway following protocols outlined in Dickson and others (2007) and analyzed shipboard. Surface water samples were collected for measurement of pH, total alkalinity/total carbon, nutrients (NH₄, silica, PO₄, and NO₂+N), stable carbon and oxygen isotopic composition, elemental analysis, dissolved organic carbon (DOC) and particulate organic carbon (POC). Water samples were removed from the sampling port of the vessel's flow-through seawater system in the main laboratory. Approximately 600 pH and 600 TA discrete samples were measured underway. The pH and TA samples were collected hourly and analyzed immediately. The remainder of the samples were taken every 2 h. (POC) samples were collected daily. An existing onboard underway air-seawater equilibrator was used to measure pCO₂ values.

Shipboard pH Analysis

Approximately 30 milliliters (mL) of seawater was collected directly into cylindrical optical glass cells for pH measurements on the total hydrogen ion scale (pH_T) following the procedure of SOP6b (Dickson and others, 2007). Cuvettes for pH were placed into an aluminum cell warmer attached to a water bath at 25°C for approximately 30 min. Shipboard pH measurements were performed using an Agilent 8453 spectrophotometer, purified metacresol purple indicator dye, and equations modified by Liu and others (2011).

Shipboard Total Alkalinity Analysis

Approximately 300 milliliters (mL) of seawater was collected into serum bottles for TA measurements following the methods of Dickson and others, 2007. Bottles were warmed to approximately 25°C before analysis.Shipboard TA measurements were performed using the automated procedure of Liu and others (2014) with an Ocean Optic spectrophotometer, a dosimat titrator, and bromocresol purple indicator dye.

Preserved Total Alkalinity and Total Carbon Analyses

Seawater samples were collected from the sampling manifold of the shipboard flow-through seawater system in 300-mL borosilicate glass biochemical oxygen demand (BOD) bottles. Samples were preserved by adding 100 microliters (μL) of a saturated solution of mercuric chloride (HgCl₂) and were sealed with a ground glass stopper lightly coated with Apiezon grease. Samples were transported to the USGS Carbon Chemistry Lab in St. Petersburg, Fla. Total alkalinity samples were analyzed using an Ocean Optics USB 2000 spectrophotometer, bromocresol purple indicator dye, and the methods of Yao and Byrne (1998). Total carbon was analyzed using coulometric methods of Dickson and others (2007). Precision and accuracy for these methods was 1 μmol/kg for TA and TCO₂.

Nutrient Analyses

A syringe that was pre-rinsed with sample water was used to collect approximately 60 mL of seawater. A Sterivex filter cartridge (pore size 0.22 micrometer, μm) was attached to the Luer-Lock fitting of the syringe and was rinsed with approximately 40 mL of sample. The remaining 20 mL of sample was then collected in a 20-mL glass scintillation vile and frozen. Samples were analyzed for NH₄, silica, PO₄, and NO₂+N at the Woods Hole Oceanographic Institution (WHOI) Nutrient Analytical Facility.

Trace Metal Analyses

A syringe that was pre-rinsed with sample water was used to collect approximately 60 mL of seawater. A 60 mL nalgene bottle was rinsed 3 times with seawater and then filled. The sample was acidified with 6 drops of trace metal grade HNO^3-. Samples were analyzed for lithium, sodium, magnesium, aluminum, potassium, calcium, vanadium, manganese, rubidium, strontium, molybdenum, cesium, barium, and uranium at the University of South Florida (USF) College of Marine Science.

Isotope Analyses

Seawater samples from HLY1202 were collected separately for oxygen and carbon isotopic analyses.  Those for intended for oxygen isotopic composition analyses were collected in 20 mL borosilicate glass vials with thread seal caps. Samples were refrigerated during storage prior to analysis. Those for carbon isotopic analyses were filtered through 45μm glass fiber filters, and collected in 60 mL syringes without contact with atmosphere in the headspace.  The water in the syringe was injected into amber Environmental Protection Agency-Volatile Organic Amber glass vials sealed with custom made butyl rubber septa.  The water was collected such that contact with atmospheric CO2 was minimized, and samples were fixed immediately with ~5mg of CuSO4.  Vials were overfilled, such that no headspace remained, and chilled continuously prior to analysis.  Analyses were completed at the USF Department of Geology Stable Isotope Laboratory using a Thermo-Finnigan Delta V 3-kiloelectron volt (keV) Isotope Ratio Mass Spectrometer coupled to a Finnigan GasBench II preparation device. Separate analyses of δD and δ¹⁸O of H₂O were completed by equilibrating 200 μL of sample with approximately 12 mL headspace of H₂ (approximately 1% in balance He for δD) and CO₂ (approximately 0.3% in balance He for δ¹⁸O) in septum-capped vials. After equilibration, the isotopic composition of the headspace gas was measured (methods following Epstein and Mayeda, 1953; Prosser and Scrimgeour, 1995).  Analyses of δ¹³C of dissolved inorganic carbon (DIC) was completed by injecting 1 mL> of sample into an approximately 12-mL> vial that was pre-flushed with He and pre-filled with 1 mL of 85 percent H₃PO₄ (methods following Assayag and others, 2006).  The CO₂ produced by this acid-stripping of the DIC was then measured after 24 h> of equilibration.  All stable isotope data are expressed in the conventional delta (δ) notation:

δ = [(R_sample-R_standard)/R_standard] × 1000‰

where R_sample and R_standard are the ¹⁸O/¹⁶O, D/H ¹³C/¹²C ratios of the sample and standard, respectively for δ¹⁸O, δD and δ¹³C.  The standards used as a reference for the δ-scale are VSMOW for H₂O and VPDB for DIC .  Internal standards were used in the calibration to the VSMOW and VPDB scales (VEEN and HTAMP waters with δ¹⁸O = -13.17 per mil (‰) and +15.05‰; δD = -96.8 and +40.5‰; Carrara Marble and NBS-18 calcites with δ¹³C = +2.01‰ and -5.04‰). Analytical precision (2σ) on these standards was better than 0.15, 1.0 and 0.2‰ for δ¹⁸O, δD, and δ¹³C, respectively. Analytical precision (2σ) on the samples was better than 0.2‰ for δ¹⁸O and 1‰ for δD and δ¹³C.

Discrete Vertical Profile Samples

Discrete samples from vertical profile casts were collected at 8 locations (see Maps section). A 24-bottle Niskin rosette (12-L bottle volume) with an electronic trigger was fitted with a Sea-Bird SBE 911plus CTD and altimeter (Table 1). The CTD provided salinity, temperature, depth, fluorescence, and dissolved oxygen data. The rosette was lowered to just above the sea floor, and bottles were filled at select depths as the rosette was brought to the surface. Water samples were collected from the Niskin bottles for the full suite of discrete analyses shipboard and ashore as described above for surface samples.

Geographic and Other Data

Geographic data were collected using the vessel's Ashtech ADU5 global positioning system. Latitude, longitude, date, and time were recorded at 1-min intervals and reconciled with underway and discrete data after the research expedition. Wind speed was acquired from Healy Yardarm Ultra (or JackStaff Ultra).

Data Analyses

Other carbonate system parameters including seawater calcite and aragonite saturation states were calculated using the carbonate speciation program called CO2Calc (Robbins and others, 2010). Equilibrium constants used for these calculations included Dickson (1990), Lueker and others (2000), and Ho and others (2006).

Table 1. CTD sensors in 24-bottle Rosette configuration, field activity L-01-12-AR. [TSG, thermosalinograph]