Improve sensors

1. CTD measurements

Accuracy :

At the first Euro-Argo users group (June 2008) there was interest in developing floats that would profile to greater depths (3000m was mentioned). The major limitations on such a development are a)-pressure case strength, b)-energy considerations,c)-sensor stability and accuracy. To achieve depth capability greater than 2000 dbar, SBE has proposed a titanium pump impeller housing, rather than plastic. This part is the same as used on SBE 49 FastCAT (adds about 255 grams to the weight). At the end of 2008, the new SBE 41CP will be available and will consume less than 180 milliwatts (300 mW today).

Other CTD sensors have been tested in the past (FSI) but its sensitivity to environment and its drifting caused by biofouling made us renounce to this technology. An alternative way could be the development of an optical density sensor which has started in 2010.

Acquisition :
  • On ARVOR float, if enabled by user, "economical" sampling method can be chosen. Instead of maintaining CTD pump always on, acquisition may be done by "spot sampling": the pump is only "on" around programmed pressure +/- 1 meter, in the middle of each slice. The "high speed" measurements (one every 2 second) are averaged on this 2 meters area, then the pump is put off until next slice. This can be applied to bottom area. On the other hand, pumping on in surface area is maintained to minimize thermal mass errors in the conductivity cell, which can be large in the thermocline. Like that, power is hugely conserved: 1 KJ instead of 6.5 KJ for default scheme. Approx. 25% of the number of cycles can be saved.
  • Arvor floats and Apex floats can be used with Iridium/GPS communications allowing high-resolution (2db) sampling throughout the profiling range. Around 150 such Apex floats are active in February 2010. The majority of these are operated by the University of Washington, USA with a smaller number by the Australian Argo project. In 2010, 2 Arvor floats are operated by Ifremer and 2 by OGS in Italy for Mediterranean EuroArgo contribution.
  • A recent development initiated by the UK allows Apex floats to make detailed near-surface profiles of temperature that are of value in allowing Argo data to be used in the study of surface mixing. The floats do this by continuing to sample temperature after the CTD pump has been switched off. This float’s firmware contains the Near Surface Monitoring feature. At depth greater than 20 dbar, PTS samples are taken as in the normal depth table. At 20dbar, 15 dbar, 10 dbar and 5 dbar cross-calibration samples are collected which consist of a non-pumped PT immediately followed by a pumped PTS. At depths less than 5 dbar, 12 non-pumped PT samples are taken at 6 seconds intervals. This time duration allows for sampling up to and including the surface. The near surface sampling option requires only one extra Argos message and has no significant power implications.
  • In a similar way, in 2010, Ifremer (France) has developped the Near Surface Measurement feature to the Arvor by adding one extra surface layer. Now, data (continuous sampled during profile) are averaged into three kinds of slices instead of two earlier. The surface layer (typically 10m to 0m) can be programmed into 10 slices of 1m thickness and the cut off pressure of the CTD pump can be programmed. This cut off pressure is taken into account in order to flag TS samples which are not pumped samples. This feature has been tested in 2010 on Arvor operated by Ifremer in Mediterranean Sea.

2. Evaluation of O2 Optodes for Argo application

Hampered by the fact that no funds were allocated to investigating O2 sensors on Argo floats, the main work was on accuracy issues of the sensors themselves and prepared and deployed floats equipped with O2 – Optodes funded by other resources. Three of these (2 APEX and 1 PROVOR) were deployed in the Oxygen Minimum Zone (OMZ) off West-Africa, and 10 additional APEX floats with Optodes are built for deployment off Peru in early 2009. We further rely on the activities performed in Carbo Ocean, where a full report of the oxygen float pilot study should be generated.Ten APEX floats with Optodes were deployed off Peru in early 2009. A second bunch of 10 floats had been deployed in 2011.

  • Long term stability is an important aspect for the quality of the measurements, and was investigated by performing calibrations (with sensors fixed to CTD’s and used on stations) with all aspects on optical measurements ( temperature, salinity and pressure effects) taken into account. This is not easily achieved with sensors on floats, and therefore we used sensors mounted on moored profiling CTD’s and in moored stations with pre and post deployment calibrations for evaluation. This work was performed during research cruises in the tropical Atlantic Ocean. Special emphasis was on investigating the pressure dependence of a bunch (8) of Optodes; this work was accompanied by Laboratory investigations in a pressure controlled environment and is summarized in a Batchelor Thesis.
  • Sensor type issues: two different sensor types are used today on Argo floats, Seabird electrochemical sensor and Aanderaa Optode 3830. Each of these has pros and cons; while for response times and initial accuracy the Seabird sensors are thought to be superior, it is the long term stability, the ability to measure in low O2 concentrations, and the robustness against biofouling that speaks for the Optode.
  • Accuracy considerations : Probably the most demanding O2-accuracy requirements are for climate related issues like the evolution of the oceanic oxygen minimum zones (OMZ). It will be one of the challenges to quantitatively resolve the changes in the OMZ’s and both existing O2-sensors have to be improved for long term stability, time constant, calibration etc., and some efforts are undertaken in that respect. For example a new version of the Optode with more exposed temperature sensor (Mk II) will likely reduce the mismatch of the optical and temperature measurements in the Optode (Beta tests are running).
  • Calibration issues : The Aanderaa Optode 3830 is used on various platforms, including moored fixed level instruments, moored profilers, autonomous gliders, and profiling floats. The instrument specifications claim long-term stability of measurements (more than one year) without recalibration. However, our comparisons with CTD measurements (RV l’Atalante cruise in the tropical Atlantic, February 2008) show that the factory settings require an instrument-specific calibration to satisfy the accuracy needs to measure oceanographically relevant signals. A sufficient pre-deployment calibration is therefore of great importance. Two possible ways are under consideration: First, factory calibrations with improved accuracy are underway in collaboration between manufacturer and the Bjerknes Centre (Bergen),and these will be evaluated soon. Second, an in-situ calibration of the Optode is presently under investigation – with convincing results. Comment: individual near-by CTDO2 profiles are not sufficient for e.g. OMZ studies.

As the recovery and the post-deployment in-situ calibrations were performed during spring 2010 on a research cruise into the Oxygen Minimum Zone of the eastern tropical Atlantic (RV Meteor M80/1) there is no final clue on the long term stability of the Optodes. However, the analysis is underway and with some of the sensors being for about 20 month in the water, there is a good chance to evaluate the quality of the sensors for a time period comparable to that of a profiling Argo float (for half of their typical life time).

  • O2-Float lifetime : 2 Generally the lifetime of a profiling float is reduced by a significant fraction when O2 sensors are implemented. It is too early to judge whether O2 sensors are reliable for durations extending beyond the approximately 3-year float-lifetime of any of the O2 sensor model; however, T,S sensors can be used beyond that duration. Therefore, it must be investigated whether total float life time should be extended by the use of Lithium Batteries to assure that the Argo requirements could be fulfilled. A Number of Optodes have been deployed during spring and summer 2008 in long term moorings (1-2 years deployment duration) – as these are to be recovered they offer a good basis for long term tests.
  • Sensor position on floats (on top of float?) : O2 sensors are mounted at different locations on floats; e.g. on top of APEX float, but at lower end of PROVOR float. However, it has been shown, that Optodes can measure in moist air, and surface measurements may be used for calibration purposes. Thus a careful evaluation of sensor position for both, electrochemical and optical sensors is necessary, and also necessary is the further evaluation of the potential for barometric pressure measurements from floats for oxygen sensor calibration purposes. This work has been done by using the Floats deployed during spring 2010 by RV Merian in the tropical / subtropical Atlantic.
  • Long-Term stability and Quality Control (delayed mode) : Compared to the effort with the salinity quality control (delayed mode QC) the O2-QC is in its infancy; one of the problems is the availability of reliable historical oxygen data. Here, Euro Argo should encourage the research community to timely make their measurements available. In parallel efforts should and will be undertaken to investigate the long term stability of available sensors on other platforms; e.g. moored stations, gliders, moored CTDO2-profilers – this will allow detailed long term (2 years at minimum) investigations of sensor behavior and post deployment calibrations.

There are aspects of existing technology which needs to be addressed before it can be mass deployed in open ocean studies.

* On Provor, concerning the 3830 Aanderaa optode, the objectives of the manufacturer are to reach the following specifications :

  • Absolute sensor accuracy of ±1 % or ±2 µM (whichever is greater )
  • 90 % response time of less than 15 s.
  • faster T sensor for temperature compensation will be placed in the proximity of the foil
  • the response time of the foil should be improved.
  • Improve its calibration procedure.

The position of the optode seems to be better if it is close to the CTD. In this case, the raw data of the oxygen could be combined to CTD temperature measurement, which is faster. The interest of this aspect should be confirmed. This will have consequences on transmission: more data to transmit.

* on Apex floats : The performance of oxygen sensors on Apex floats has been described by Kortzinger et al 2008. Much of the development work has been carried out by Steve Riser, University of Washington who has deployed floats fitted with both Aanderaa Optode and SBE sensors. This results have not yet been reported in the peer-reviewed literature.

First results of the O2 floats in the Pacific were presented at the Euro-Argo User Workshop in 2010.

3. Improvements concerning biogeochemical sensor :

The demand for nutrient measurements is growing. Nitrate measurement should be implemented in 2010 on Provor, using new MBARI/S-Atlantic SUNA sensor. Trade-off between power availability and amount of data collected shoud be analysed.