Raw fDOM fluorescence output (counts) have been manipulated three ways:
Counts were converted to fDOM ppb quinine sulfate equivalents (QSEs) using the manufacturer calibration data provided for each unit.
Data were corrected for quenching of fluorescence at higher temperatures. Lab experiments showed that fluorescence of a given source water declined linearly by ~20% between 20-40C. A pooled regression correction was applied to all units to compensate for the lowered fluorescence response at higher temperatures. All fluorescence data are normalized to fluorescence at 20C.
Fluorescence offsets (as QSEs) between units during successive deployments were corrected using a multi-step procedure
- Obvious bad data were excised from the records
- Each individual deployment output was smoothed using a weighted 8 point local regression routine (Matlab ‘rloess’) to suppress spiky outliers.
- Each fluorometer output was then adjusted up or down based on their mean intercalibration and deployment overlap offsets to bring them all in rough alignment.
- The mean of the residual overlap gaps for each successive deployment was determined, and a Matlab pchip spline was applied to connect those points.
- Each deployment record was made to conform to the spline curve fit so that the overlapping endpoints had a mean difference -> 0 and possessed a smooth transition between successive deployments.
- The concatenated data file was linearly interpolated onto a 10 minute time vector to facilitate intercomparison of records.
- No attempt to date has been made to correct the fluorescence data for quenching due to turbidity in the water column.
BCO-DMO Processing notes:
- added conventional header with dataset name, PI name, version date
- modified parameter names to conform with BCO-DMO naming conventions
- added station, lat, lon, date, time, ISO_DateTIme columns
- ISO Date format generated from date and time values
- reduced decimal places of fDOM and temperature to 3 places