Blog Home » Q&A: Negative Oxygen Values

Blog Home » Q&A: Negative Oxygen Values

Q&A: Negative Oxygen Values

CTD profile showing unusual oxygen data

Oxygen at this location is historically around 90% saturation at the surface. However, after uploading data from this CTD, the data from the SBE 43 oxygen sensor appeared to be averaging around -1 mL/L – obviously incorrect. After running through common SBE 43 troubleshooting tests, this sensor and cable appeared to be working properly. Can you determine the error that created the unusual oxygen profile above?

Find the Answer

Answer: Incorrect Configuration File

The hardware used in this cast was working as expected – this problem was software-related. The .xmlcon configuration file was not arranged properly, resulting in dissolved oxygen data being processed as turbidity data (and vice versa) in post-processing . The error and necessary correction are shown below:

 corrections to sensor order in an .xmlcon configuration file

In the original configuration file, the order of sensors was incorrect – the position of the dissolved oxygen sensor was switched with the position of the turbidity sensor. This was incongruent with the order that sensors were physically connected to the CTD’s auxiliary sensor channels, where the SBE 43 dissolved oxygen sensor was physically connected to the CTD’s auxiliary channel 0 and the ECO-NTU turbidity sensor was physically connected to the CTD’s voltage channel 4. When the user loaded the incorrect .xmlcon file into SBE Data Processing, the software interpreted the analog output from the turbidity sensor as data from an oxygen sensor, resulting in the unusual oxygen profile shown above. Correcting the .xmlcon file and reprocessing data fixed the error.

Analog-output sensors like the SBE 43  and ECO-NTU can be easily integrated with many Sea-Bird Scientific CTDs, adding a vast array of parameters and customization options. But beware – changing the physical sensor array necessitates changes to the .xmlcon configuration file. Failure to do so will cause the software to misinterpret the data. Read our analog sensor integration guide to learn more.

February 24, 2021

Related Posts

Featured Posts

Video: Simulating an Inductive Mooring

Conceptually, inductive moorings seem complex - communicating with 100 instruments through a single cable doesn’t exactly sound intuitive. But, in practice, the setup is surprisingly simple. Watch our latest video to see how easy it is to test and simulate an...

5 Inductive Modem Configurations

A buoy, a jacketed wire rope, and some exposed metal. In its simplest form, that’s all you need to create an inductive modem (IM) mooring. The buoy supports the surface modem, power supply, and data logging/transmission. Some buoys may also contain weather sensors, or...

Photo Contest Winners

From intrepid underwater shots to time-worn photos of vintage sensors, the day-to-day of marine science produces some arresting photography. People from around the globe submitted their best shots for their chance to win in 5 categories: Best on Deck (or Dock)...