Beginning with the selection of instruments and sensors to be used for data acquisition, ODF data technicians evaluate each detail that will result in the best possible hydrographic data. Data are reviewed and processed shipboard whenever possible, and corrected as necessary to meet or exceed established data standards.
The following conditions/problems can affect data that we collect/process:
- Manufacturer / model no. / serial no. of each sensor and CTD deployed
- Sensor calibration dates, drift between calibrations, and damage/repair history
- Equipment problems, ship-board repairs, change-outs (winch/wire, rosette/carousel, deck unit, CTD, sensors)
- Storage of rosette between casts ideally out of wind/sunlight, similar temperature to outside air
- Software / hardware problems and bugs during acquisition and processing
- Potential freezing issues in sub-zero air temperatures
- Organic matter contamination causing sensor noise or offsets (jellyfish, etc.)
-
Sea state - rough seas can cause:
- Collision of rosette / instruments with ship at deployment or recovery
- Lack of proper equilibration time for CTD/sensors at cast start
- Misleading density inversions caused by rosette agitating / mixing water
- Stress / kinks in wire (signal problems, wire breakage / loss of package)
-
Electrical problems - shorts caused by water leakage can cause:
- Signal spiking from intermittent bad connections
- Loss of power or signal (sensors, CTD, cabling between sensors and CTD, carousel, wire, slip-rings, deck unit, and/or computer)
- Bottle data: mis-trips or collection errors, analysis or calibration problems
Once issues have been documented, problems are fixed where possible and data are corrected.
- Despike/offset data by linear interpolation through nearby values
- Use upcast instead of downcast data if problems are not correctable
-
Apply polynomial corrections using fits of differences between:
- bottle-CTD values
- primary-secondary sensors
as a function of:
- time/station
- pressure
- temperature/conductivity
With corrections applied, data quality is evaluated by comparing:
-
Pre- and post- cruise calibrations - considering:
- Equipment mis-handling or damage during cruise or shipment
- Observed sensor shifts during cruise
- Historical drift tendencies in sensor calibrations
- down and up profiles of same cast
- data from nearby casts
- cruise data with historical data
Quality Control Links