Dataset: LB_2012_Mooring_Optics
Deployment: LB_2012_LB3

Mooring data records for optical sensors:   

• LB_2012_LB1 mooring (30 m depth):  Three ECO units were deployed, one on the bottom mooring frame (S/N 738) about 0.5 m above bottom, and two on a taut-line mooring attached to the frame, nominally at 15 m depth (S/N. 736) and 5 m depth (S/N 735).  Sampling for the three LB1 ECO units was three consecutive 1-second readings at 6-minute (360 second) intervals.
• LB_2012_LB2 mooring (76 m depth):  One ECO unit was deployed on the ADCP mooring frame about 0.5 m above bottom (S/N 739).  This unit had an apparent electronic failure during the deployment which resulted in a distinct positive baseline shift for turbidity and negative baseline shift for chlorophyll.  The data record was considered suspect at this point and only data from the initial deployment period (before the baseline shift) is reported.  Sampling for the LB2 unit was three consecutive 1 second readings at 6-minute (360 second) intervals. 
• LB_2012_LB3 mooring (171 m depth):  The ECO unit (S/N 740) was mounted about 0.5 m above the bottom on the lower portion of the Floatation Technologies AL-200 Trawl Resistant Bottom Mount frame which held a 150 kHz ADCP unit.  Due to the longer deployment period for the LB3 instrument and internal memory limitations, sampling was set for a single 1-second sample at 6-minute (360 second) intervals (instead of three 1-sec samples for the ECO units at LB1 and LB2). 

Instrument calibrations and inter-comparisons:  Manufacturer supplied calibration factors for chlorophyll concentration and turbidity were employed following pre- and post-deployment checks for consistency.  The five ECO FLNTUSB units were serviced at WET Labs and calibrated as a set in May, 2011. Instrument inter-calibrations and the factory chlorophyll calibration factors were checked by obtaining a concurrent set of measurements in the water column before deployment.  All ECO units were mounted on a piece of angle iron that was hung on a weighted line at the same depth as the ship’s CTD/carousel system (20 m below the surface at the shelf break station LB2 to avoid surface irradiance effects on the chlorophyll fluorescence signal).  The units were held at depth for 5 minutes (acquiring 4-5 readings).  Niskin bottle water samples were collected about mid-point of the sampling period for chlorophyll analyses (triplicate filtered samples frozen and stored in liquid nitrogen on the ship, analyzed in the shore lab by the fluorometric method).  After recovery of the moorings, the units were photographed to record bio-fouling.  The four functioning ECO units (except for S/N 739) were then mounted on the ship’s CTD/carousel frame for a post-deployment set of inter-comparison measurements at two depths in the water column (35 m and 10 m), again with Niskin samples collected for chlorophyll analyses.  For chlorophyll estimated from fluorescence, the pre- and post-deployment readings showed agreement to within 0.1-0.15 mg/m3 for readings at a measured concentration of 0.4 mg/mg3 (both pre- and post-deployment) and 0.2 mg/m3 at a measured concentration of 1.25 mg/m3.  For turbidity the pre-deployment check showed close agreement between all units (average 0.066 +/- 0.002 NTU at 20 m depth).  The post-deployment check appeared to be compromised by drying of a film on the optical windows after recovery prior to the post-deployment checks.  For the turbidity data records, baseline readings over the deployment periods (minimum values) only showed marked increase for S/N 738 deployed on the mooring frame at LB1 (increasing from about 0.4 to 1.2 NTU during the latter part of the deployment).  For the data reported here, there has been no correction for the apparent increase in baseline readings for this instrument. 

Wetlabs calibration sheets: See Supplemental Files

Internal Clock Drift:  Time was recorded as UTC with internal clocks set to a common GPS-based reference during the pre-deployment set-up.  Internal clock drift during deployments was checked against the reference after recovery.  Temporal offsets at recovery ranged from -9 seconds to +528 seconds with an average offset of +359 seconds (or about the sampling interval of 360 seconds).  Given that most events in the optical records occurred on time scales of hours to days, no clock drift correction was applied.  Time stamps in raw files in the format MM/DD/YY HH:MM:SS were converted to the Microsoft Excel epoch convention (01/01/1900 = 1) and to the Matlab serial date number format (01/01/0000 = 1) for Excel and Matlab files respectively.