The QC version of the datasets has been elaborated verifying sensor efficiency through post-calibration and standardized validation procedures. These procedures have been applied to all datasets following the QUARTOD (Quality Assurance/Quality Control for Real-Time Oceanographic Data) guidelines (https://ioos.noaa.gov/project/qartod/) and other recommendations for in situ data quality control adapting the protocols for deep-sea observations. Considering quality control has to account for the specificity of each location and sensor, the tests have been automated as much as possible aiming for a standardized procedure. To avoid confusion, the flagging scheme is univocal for all sensors and data types. Since in the data files, there are no interpolated nor calculated values, the representative flag values are: 1 = good data
3 = suspicious data correctable or high-interest data
9 = missing value or removed data. Thresholds and ranges for the quality check tests were adjusted for each sensor characteristic, location, and acquisition period (see the information available in the metadata). All datasets have undergone the following tests:
*Time test --> Checking the sensibility of observations date and time for the whole campaign duration according to sensor frequency
*Out of range test --> Removing values outside of sensor output range
*Gross range test --> Removing values outside expected extremes (reasonably adjusted for each case)
*Percent good (ADCP only)  --> This test uses the beams' percent good, which indicates what fraction of the pings were accepted for a given ensemble.
*Error velocity test (ADCP only)  --> Removing values outside error velocity thresholds 
*Regional/seasonal range test --> Flagging values outside selected thresholds (from seasonal averages, climatological ranges, expected variability from the area, etc.)
*Spike test --> Removing single value spikes, relative to adjacent data points
*Rate of change test --> Finding outliers (i.e., data over three standard deviations from the mean) and then removing or flagging (for Dissolved Oxygen only)
*Flatline test --> Flagging data series that is nearly a flatline
*Sensor tilt test (current meter and ADCP only) --> Removing data collected outside the correct range of tilt values for the current sensor used
*Echo intensity test (ADCP only)  --> Flagging echo intensities that may indicate interaction with the surface, bottom, or in-water structures
*Current gradient test (ADCP only)  --> Flagging excessive current speed or direction changes in the vertical profile
