13 / 48
Ref. No. [UMCES] CBL 2015-012
ACT VS15-05
measurements following standard oceanic protocols (Dickson et al., 2007). In brief, a 500 mL
glass sample bottle (supplied by Andrew Dickson, SIO) was filled from the Van Dorn sampling
bottle using PTFE tubing and preserved with 110 µL of saturated HgCl
2
solution. Reference
samples shipped to Oregon State University (Burke Hales, OSU) were analyzed for pCO
2
and
TCO
2
. Reference samples shipped to Scripps Institution of Oceanography (Andrew Dickson)
were analyzed for TA and pH. Reference samples shipped to the University of South Florida
(Robert Byrne) were analyzed for pH. These external sample analyses provided further
estimation of the any possible uncertainty of ACT’s reference sample measurements through
both direct pH measurement and calculated pH from carbonate parameters using CO
2
-Sys
(Pierrot et al., 2006). However, because these samples were not handled in a manner identical to
the unpreserved reference samples that were analyzed within a few hours of collection, they will
not be used in any published direct assessment against the test instruments without agreement of
the TAC.
All sampling procedures were practiced by ACT staff prior to the beginning of the
evaluation, and the Chief Scientist verified that all staff were trained in both instrument and
sample collection protocols. A preservation test was done at each site to ensure samples that
were shipped for external analysis were stable.
Sample Handling and Custody Requirements -
All collected reference water samples for all four
field tests were handled in the same manner. All reference samples were processed according to
methods defined within these protocols and cited Standard Operating Procedure documents
(Dickson et.al. 2007). Each reference sample was dated and coded according to site, test
condition and sample sequence. The actual sample container was labeled with a number for
identification. The reference sample number was used in all laboratory records and Chain-of-
Custody (COC) forms to identify the sample. The COC specified time, date, sample location,
unique sample number, requested analyses, sampler name, required completion time, date and
time of sample transaction, and name of receiving party for the samples. Proper labeling of
sample bottles was critical. The COC was a mechanism by which samples could be tracked
through the various phases of the process: collection, shipping, receiving, logging, sample
prep/extraction, analysis and final data QA/QC review. Transfer of reference samples from field
personnel to laboratory personnel was also recorded on the COC and records were maintained in
the laboratory with the names and signatures of persons leaving and receiving the custody. All
logs were duplicated weekly. The original log was retained at the ACT Partner site and a copy
was sent to the ACT Chief Scientist. Accumulated samples to be analyzed by outside
laboratories were shipped for analysis at the end of the month long deployment tests, and
monthly for the extended HI deployment. Samples stored on site were routinely inspected by
ACT personnel to assure proper preservation and label integrity. All reference samples not
immediately analyzed on site by ACT staff were accompanied by the sample collection sheet and
COC forms.
Analytical Methods for Reference Samples -
Three spectrometer cells (10 cm cylindrical cell)
were filled at the deployment site for each field reference sample, and transported directly to the
lab. All analysis was done at a fixed temperature for a given test site. The measurement
temperature was 25°C for HIMB, 20°C for MLML, 15°C for CBL and 25°C for Lake Michigan.
The temperatures for CBL and Lake Michigan were estimated to be near the mid-point of the
expected 30 day ambient range. Filled cells were incubated in a cell warmer (manufactured at
USF according to Bob Byrne’s specifications) to reach the specified analytical temperature