Dataset: GP16 Dissolved Aluminum
Deployment: TN303

Sample collection and processing:
Water column samples for trace metal analyses were obtained from two sampling devices, a CTD rosette package and a surface "fish" sampler. The Rosette package contained 24 modified 12 L GO-FLO bottles (General Oceanics) mounted on a trace-metal clean conductivity-temperature-depth (CTD) carousel (SeaBird) that was deployed on a Kevlar conducting cable (Cutter and Bruland, 2012). Upon recovery, the GO-FLO samplers were brought into a shipboard Class-100 clean laboratory container for sub-sampling. For filtered samples, the samplers were pressurized to 10 psi using filtered, compressed air, and the seawater samples were filtered through pre-cleaned 0.2 µm Acropak Supor capsule filters (Pall) (Cutter and Bruland, 2012). The fish sampler collected water from ~ 4m depth and pumped it into a clean van were it was filtered through pre-cleaned 0.2 µm Acropak Supor capsule filters. For dissolved aluminum (DAl), 918 samples were collected from all stations and depths. The 0.2 µm filtered subsamples were stored in 100 mL low density polyethylene (LDPE) bottles (Bel-Art) with LDPE caps and were acidified to pH ~1.7 with 12 N ultrapure hydrochloric acid (Fisher Optima).

Analytical methods:
DAl was determined at sea by flow injection analysis with in-line pre-concentration and fluorimetric detection (Resing and Measures, 1994). Method modifications included replacing resin-immobilized 8-hydroxyquinoline with Toyopearl AF-Chelate 650M, and using acidified de-ionized water as the carrier instead of acidified seawater. Daily precision for repeat analysis of internal and primary standards was ± 0.1 nM or 4.2%, whichever is larger. Two internal reference (IR1 and IR2) standards were run during the cruise, with DAl concentrations for IR1= 1.78 ± 0.27 nM (± 15%; n = 118) and and IR2 = 1.98 ± 0.07 nM (± 3.4%; n = 75), respectively. IR1 was run first and IR2 second with no overlap. The SAFe reference samples were analyzed simultaneously during sample analysis; SAFe S was 2.54 ± 0.16 nM (n = 34), SAFe D2 was 1.80 ± 0.18 nM (n = 33) and SAFe D1 was 1.40 ± 0.20 nM (n = 39). The internal standards, Geotraces reference materials, and other samples run throughout the cruise were used to correct for drift in the data. These corrections resulted in small overall changes in IR1 (1.67 ± 0.18 nM; ± 11%; n = 118), SAFe S (2.45 ± 0.14 nM; n = 34; consensus value = 1.74 ± 0.09 nM); for SAFe D2 (1.70 ± 0.14 nM; n = 33; consensus value 1.04 ± 0.1 nM); and for SAFe D1 (1.31 ± 0.13 nM; n = 39; consensus value = 0.64 nM). IR2 was run over a period of time with no observable drift and thus was not corrected. The least squares best fit between the average value for each standard versus the consensus values yielded DAlsample = 1.04 DAlGT + 0.63 nM (r² = 0.99). The overall regression for all 116 data points yielded DAlsample = 1.06 DAlGT + 0.62 nM (r² = 0.92). Prior to this cruise our laboratory produced DAl determinations that were statistically indistinguishable from the SAFe consensus concentrations. However since this cruise our values continue to be elevated compared to the consensus value. A similar offset has been communicated to me by another Geotraces scientist. It should be noted that these references samples are stored in LDPE bottles with HDPE lids, thus allowing for the possibility of contamination. Samples collected throughout the cruise (n=125) where run on different days and/or from different sample bottles to check for overall accuracy and precision. Some of these samples required drift correction some did not. The average absolute difference between these replicate analyses was 0.16nM. We also note that for seven samples collected in LDPE bottles with LDPE lids yielded Al concentrations statistically indistinguishable from those in LDPE bottles with Polypropylene lids with an average difference of 0.07nM. This is not surprising given the short contact time between the sample and the upright bottle.

These data will be submitted to the Intermediate Data Product (IDP) when the crossover data is available from the US Geotraces PMT cruise. The PI currently has the preliminary data is waiting for the analyses of those data to be completed.

Problem report:
The data reported here have been corrected, based on reevaluating peak processing, drift corrections, and through the removal of errant data points. Additional missing data points have been also been added. Those data points are mostly from the Fish surface-ocean sampler. The Al from Station 1 was revised, however the use of internal standards and replicate analyses was limited for these data.