Dataset: Gulf of Alaska DOM Optics

Location: North Pacific Gulf of Alaska: 48N to 59N and 129W to 153W

Methodology:

Methods are described in detail in the peer reviewed paper Cao et al. (2020).

Sampling and analytical procedures: From Cao et al. (2020):

Sampling: As part of the Deep Ocean Refractory Carbon (DORC) field campaign aboard the R/V Melville in the Gulf of Alaska (August 4–21, 2013) we performed high-resolution surveys of DOM optical properties, collecting samples from the sea surface down to ~5000 m using a 24 Niskin bottle rosette with a Sea-Bird CTD (conductivity-temperature-depth) sensor. Corresponding hydrographic (depth, T, S) and chemical (nitrate, nitrite, ammonia, silicate, phosphate, oxygen) data were collected for each cast and submitted to BCO-DMO (http://data.bco-dmo.org/jg/serv/BCO/NorthPacifi c_RDOC/CTD_Profiles.brev0) by co-chief scientist (D. Hansell, University of Miami). The DOC data corresponding with the optical measurements were also generated in the Hansell lab. For optical measurements, all labware was acid soaked in 2N HCl overnight with glass bottles subsequently precombusted (450C, 3 h minimum), and rinsed thoroughly with Milli-Q water (MQ; >18.2 MΩ⋅cm; Millipore) that was freshly produced on the ship (within 1 day) prior to sample collection. Seawater was gravity filtered directly from the Niskin bottle via silicon tubing through a pre-cleaned, 0.2 μm Whatman Polycap AS 75 nylon cartridge filter into the glass bottles. Filtered samples were analyzed for optical properties within 4 h after sampling to avoid potential long-term storage artifacts.

Fluorescent measurement of DOM and PARAFAC modeling:

A fluorescent excitation-emission matrix (EEM) spectrum was obtained for each sample with an Aqualog spectrofluorometer (HORIBA Jobin Yvon Inc., NJ, USA) using a 1-cm quartz cell with MQ as the blank. EEM fluorescence intensities were measured by scanning across an excitation range of 240–450 nm (5 nm intervals) and capturing emission spectra over a wavelengths range of 280–500 nm at 3.2 nm intervals. We optimized data quality by determining the integration time for each sample that maximized emission intensity detection at short excitation wavelengths without saturating the CCD at long excitation wavelengths (Gentry-Shields et al., 2013).

CDOM absorption measurement and spectral slope calculations:

CDOM absorption spectra were obtained with filtered samples at room temperature using a single-beam liquid waveguide capillary flow cell (World Precision Instruments Inc., measured at 92.2 cm pathlength), coupled with quartz fiber optic cables to a deuterium-tungsten light source (DT-mini GS) and a MAYA2000 PRO spectrophotometer (Ocean Optics, Inc.). To eliminate contamination from pump tubing, cleaning solutions and samples were pulled slowly through the system using Teflon lines and a peristaltic pump (~0.5 ml/min) located downstream from the waveguide.
The flow cell was cleaned at regular intervals (~every 10 samples) using a rotating valve that sequentially rinsed with a 20:80 vol/vol acetone:water mixture, 2N HCl, and fresh MQ water. In addition, to alleviate the common problem of microbubbles that accumulate on the interior surfaces of the flow cell when analyzing seawater samples, we dissolved bubbles off the cell walls back into degassed MQ supplied through a Liqui-CelTM MM-0.5x1 Series Membrane Contactor (3MTM) attached to an oilless vacuum pump. This was routine after the cleaning sequence and also used when blanks remained high after cleaning or between seawater samples. Removal of microbubbles was confirmed when no measurable change in absorbance was observed due to bubble compression after pressure was applied to the flow cell with MQ using a separate syringe. While this procedure proved excellent for removing attached microbubbles, the internal epoxy used to attach the membranes to the polycarbonate body of the Liqui-CelTM contributed an observable “CDOM-like” absorbance that could not be eliminated through cleaning the cartridge. Consequently, we flushed the system with MQ water after cleaning, degassing, and between samples until no change was observed in the UV portion of the blank spectrum with repeated “re-blanking” of the MAYA2000 signal. This re-blanked MQ reference accounted for instrument drift.

Instruments:

FDOM: Aqualog spectrofluorometer (HORIBA Jobin Yvon Inc., NJ, USA)
CDOM: Single-beam liquid waveguide capillary flow cell (World Precision Instruments Inc., measured at 92.2 cm pathlength), coupled with quartz fiber optic cables to a deuterium-tungsten light source (DT-mini GS) and a MAYA2000 PRO spectrophotometer (both from Ocean Optics, Inc.).