126
of the primary sensor. With a typical response time of 0.065 s such a sensor, routinely deployed to operate autonomously for up to 1 year, can collect over two million data points. In practice, data averaging and storage and transmission limitations
constrain operational data yield of some sensor systems. Hourly binning of buoy
surface data, for example, amounts to a modest 8760 data points per year but shorter
sampling periods as required allow increased data rates. Many platforms perform
vertical profiling as well thus increasing data rates by the number of vertical bins.
Remote sensing systems providing synoptic coverage of 2-D fields yield orders of
magnitudes greater data density. Aircraft or satellite-borne optical raster scanners
and HF radar exemplify such remote sensing instruments. Additional large data sets
are presently produced as numerical model output, an essential product of ocean
observing. These large volumes of instrumental and virtual data are transmitted to
data assembly stations, archived, rendered into value-added data products such as
imagery and graphical output, and made available to end users.
7.2 Quality Assurance and Quality Control (QA/QC)
for In Situ Ocean Observing Data
Extreme measures are taken to ensure the quality of instrumental data. Standardized
best practices for instruments currently in use are codified in QA/QC manuals. The
US IOOS has to date developed a series of twelve QA/QC manuals for a variety of
sensors (https://ioos.noaa.gov/project/qartod/. Accessed 5/1/2018). It is the intent of
IOOS to develop authoritative QA/QC guidelines for all core variables addressing
each variable as funding permits.
Quality assurance refers to a number of active steps taken prior to, during, and
after instrument deployment. These include, among others, appropriate instrument
choice with the degree of resolution appropriate to the task at hand, rigorous instrument calibration, choice of instrument position aboard the platform free of interferences and abrasion, proper provision for error-free data transmission, steps for
minimization of corrosion and biofouling, and post-deployment recalibration.
Numerical values of calibration coefficients obtained in the laboratory prior to
sensor deployment allow conversion, for example, of the frequency output of a
Wein-bridge oscillator coupled to a thermistor or conductivity bridge, to temperature in degrees Kelvin and practical salinity, respectively. Calibration must be traceable to an appropriate standard such as a certified platinum resistance thermometer
for temperature or certified IAPSO standard seawater for salinity. It is common
practice to return instruments to the manufacturer at prescribed periods for recalibration at the plant.
Instrument stability, and thus the need for recalibration, has been discussed previously. A few instruments such as ADCPs exhibit intrinsic stability allowing
extended deployment (years), but others, notably chemical and bio-optical instrumentation subject to biofouling, are prone to rapid signal degradation requiring frequent cleaning and recalibration. IOOS manuals provide instructions for appropriate
measures to reduce corrosion and biofouling allowing extended instrument operation.
7 Coastal Ocean Observing Data Quality Assurance and Quality Control, Data…
of the primary sensor. With a typical response time of 0.065 s such a sensor, routinely deployed to operate autonomously for up to 1 year, can collect over two million data points. In practice, data averaging and storage and transmission limitations
constrain operational data yield of some sensor systems. Hourly binning of buoy
surface data, for example, amounts to a modest 8760 data points per year but shorter
sampling periods as required allow increased data rates. Many platforms perform
vertical profiling as well thus increasing data rates by the number of vertical bins.
Remote sensing systems providing synoptic coverage of 2-D fields yield orders of
magnitudes greater data density. Aircraft or satellite-borne optical raster scanners
and HF radar exemplify such remote sensing instruments. Additional large data sets
are presently produced as numerical model output, an essential product of ocean
observing. These large volumes of instrumental and virtual data are transmitted to
data assembly stations, archived, rendered into value-added data products such as
imagery and graphical output, and made available to end users.
7.2 Quality Assurance and Quality Control (QA/QC)
for In Situ Ocean Observing Data
Extreme measures are taken to ensure the quality of instrumental data. Standardized
best practices for instruments currently in use are codified in QA/QC manuals. The
US IOOS has to date developed a series of twelve QA/QC manuals for a variety of
sensors (https://ioos.noaa.gov/project/qartod/. Accessed 5/1/2018). It is the intent of
IOOS to develop authoritative QA/QC guidelines for all core variables addressing
each variable as funding permits.
Quality assurance refers to a number of active steps taken prior to, during, and
after instrument deployment. These include, among others, appropriate instrument
choice with the degree of resolution appropriate to the task at hand, rigorous instrument calibration, choice of instrument position aboard the platform free of interferences and abrasion, proper provision for error-free data transmission, steps for
minimization of corrosion and biofouling, and post-deployment recalibration.
Numerical values of calibration coefficients obtained in the laboratory prior to
sensor deployment allow conversion, for example, of the frequency output of a
Wein-bridge oscillator coupled to a thermistor or conductivity bridge, to temperature in degrees Kelvin and practical salinity, respectively. Calibration must be traceable to an appropriate standard such as a certified platinum resistance thermometer
for temperature or certified IAPSO standard seawater for salinity. It is common
practice to return instruments to the manufacturer at prescribed periods for recalibration at the plant.
Instrument stability, and thus the need for recalibration, has been discussed previously. A few instruments such as ADCPs exhibit intrinsic stability allowing
extended deployment (years), but others, notably chemical and bio-optical instrumentation subject to biofouling, are prone to rapid signal degradation requiring frequent cleaning and recalibration. IOOS manuals provide instructions for appropriate
measures to reduce corrosion and biofouling allowing extended instrument operation.
7 Coastal Ocean Observing Data Quality Assurance and Quality Control, Data…