Long-Term Oxygen Measurements

Posted with permission of International Ocean Systems,
As seen in International Ocean Systems, Volume 12, Number 2, March/April 2008
Authors
Carol Janzen, Ph.D., Nordeen Larson, Ph.D., and David Murphy, Sea-Bird Electronics, Inc.

Numerous implementations of oxygen sensor technologies including galvanic, Clark electrode, optical, and others are being used for long-term measurements. Obtaining accurate long-term data from any of them requires a sensor with high initial (calibration) accuracy, inherent sensor stability, and an effective defense against fouling. The user also needs a practicable means of determining how well the initial accuracy was preserved at convenient intervals during deployment or post-recovery.

Sea-Bird’s SBE 43 dissolved oxygen (DO) sensor (Figure 1) was engineered to be integrated with a pumped CTD and provide rapid-response dynamically accurate DO measurements during profiling. The Clark electrode technology was chosen over optical methods for its high accuracy potential and fast response, both critical to profiling applications. However, the SBE 43, being a complete redesign of the Clark electrode sensor, incorporates features that eliminate previous causes of instability, and so is also able to deliver stable moored measurements for months in high fouling environments.


Figure 1. The SBE 43 DO sensor (without protective plenum housing)

The SBE 43 accuracy and stability is derived from its re-engineering, careful calibration, and effective bio-fouling controls. Modern electronics eliminate electronic acquisition error. Every SBE 43 is calibrated at three different oxygen concentrations at each of six temperature points (18 points in total) using Winkler titration, and temperature and salinity standards. The resulting sensor accuracy is unsurpassed by other technologies. Electrochemical drift, a limitation in previous Clark designs, exists somewhere below the calibration uncertainty of 1 µM/kg and has not been observed in several years of factory calibration data. In fact, upper estimates based on ocean deployments are less than 0.5 % over 1000 hours of sensor flushing time (Janzen and Larson, 2008). Sea-Bird’s unique implementation of flow controls significantly reduces bio-fouling impact on the sensor. This allows for much longer deployments (months versus weeks) than typically achieved by continuously exposed DO sensors.

The SBE 43 Defense Against Bio-fouling

Moored deployment data from many customers demonstrate that the SBE 43 DO sensor provides high resolution and low-drift measurements resulting in retained accuracy within a few percent for periods of three to five months with no servicing. This is observed in a variety of aquatic environments, including productive coastal waters with significant biological fouling pressure (Figures 2 and 3).


Figure 2. SBE 43 percent oxygen saturation values at a monitoring buoy in Cockburn Sound (Western Australia), a biologically productive coastal lagoon, June 26 – December 22, 2006. The SBE 43 sensor was recovered on November 5, 2006, and replaced with a new calibrated sensor on November 9, 2006. The mean measured difference at the time of the sensor swap indicates less than a 5% change in the initial sensor calibration over 4 months. The decline observed in percent oxygen saturation in December, similar to that observed in late September, is a natural trend of oxygen draw-down and rapid storm replacement observed at multiple mooring sites within Cockburn Sound. (data courtesy of Water Corporation)


Figure 3. Image of diver’s hand and an SBE 16plus with an integrated SBE 43 DO sensor 2 months after deployment in Cockburn Sound, Western Australia, August 2006. This image illustrates the fouling pressure experienced during winter. (photo courtesy of Greenspan Technology)

 

The SBE 43 DO sensor is flushed with a flow-controlled pump to deliver a fresh sample of ambient water to the sensor. For moored applications, the unique plumbing arrangement serves as a first line of defense against fouling, protecting the sensor from continuous exposure to external biological contamination (Figure 4). Pumping prior to each measurement flushes stagnant water out of the plumbing. Between measurements, anti-foulant placed at each end of the conductivity cell’s plumbed path diffuses into the stagnant water to neutralize any biota that enters the system during the previous flush. The flushing itself agitates and removes the neutralized biota from the sensor.


Figure 4. (left) External plenum SBE 43 DO sensor after a 4 month deployment at Shilshole Marina, March – July 2007, and (right) the same sensor with the plenum removed.

 

SBE 43 Calibration Drift and Its Correction

The SBE 43 sensor output is linear with respect to oxygen concentration and maintains a stable output at zero oxygen (Equation 1).

Oxygen (ml/L) = Soc * [V - Voffset] * [Tcor * Pcor * OXSOL]               (1)

  • SOC is the linear slope scaling coefficient.
  • V is the sensor output voltage; Voffset is a fixed sensor voltage at zero oxygen.
  • OXSOL is the oxygen solubility function and converts oxygen partial pressure (sensor measurement) to oxygen concentration (Garcia and Gordon, 1992).

The Tcor and Pcor functions correct for the effects of temperature and pressure. These are lower order terms and remain essentially constant with fouling and sensor age. With negligible electrochemical drift and stable electronics, any loss of sensitivity can be attributed to bio-fouling of the sensor itself. When the sensor does foul, the character of the change in sensor output is a simple loss of sensitivity, as evidenced in Figure 5 and Table 1, and the ratio of measured to true concentration remains constant over the whole range of the sensor. Therefore, adjusting the slope in the calibration equation (SOC) is the appropriate means of maintaining accuracy if fouling occurs.

The sensor’s characteristic drift pattern allows for easy data correction using a single in situ reference value to determine the slope (SOC) correction (Figure 5; Table 1). DO data can be corrected either in real-time or during post processing, depending on the availability and quality of an in situ reference value or a post-recovery calibration. This offers a powerful and scientifically defensible way to make residual corrections to data from unattended long-term deployments.


Figure 5. Original calibration of SBE 43 sensor SN 1114 used in Cockburn Sound (green dots); post-recovery calibration prior to sensor cleaning (blue triangles); re-calibration after sensor cleaning (red squares). Notice the loss of sensitivity in the post-recovery calibration (blue triangles) is strictly linear, corrected by a simple  multiplier to the slope coefficient (SOC) as demonstrated in Table 1.

 

Winkler DO of Bath
(ml/L)
SBE 43 Output
(ml/L)
Residual
(SBE 43 - Winkler)
Correction Factor
(Winkler / SBE 43)
6.80 6.75 -0.05 6.80/6.75=1.007
4.20 4.17 -0.03 4.20/4.17=1.007
1.20 1.19 -0.01 1.20/1.19=1.007

Table 1. Reference Winkler water samples and sensor readings at three dissolved oxygen concentrations during the post-recovery calibration of SBE 43 sensor SN 1114 (Figure 5). The ratio between the Winkler values and corresponding SBE 43 output can be used to calculate the SOC correction factor. Note the SOC correction factor remains constant at each validation point over the range of oxygen values shown, illustrating that any single validation point alone could be used to correct the slope.

 

Data Correction in High-Fouling Conditions

A SBE 43 dissolved oxygen sensor was deployed at Shilshole Bay Marina north of Seattle, Washington, USA for four months in 2007 during the biologically active spring and summer seasons. The integrated SBE 43-CTD (conductivity, temperature, and depth) was moored at 2 meters water depth and sampled every 10 minutes following a 30-second flush cycle. Replicate Winkler samples were collected bi-weekly from a 1.2-liter Niskin bottle adjacent to the moored SBE 43 sensor at the time of a sample. The purpose of the test was to monitor how long the sensor could maintain sample accuracy within five percent of Winkler references.

The SBE 43 measured dissolved oxygen within five percent of Winkler reference values for over 107 days (~three months) during high biological fouling conditions (Figure 6). Results from other long-term deployments (e.g., Western Australia data shown in Figure 2) corroborate this performance. This exceeds expectations of other DO sensors, which historically experience impaired coastal DO measurements within four to six weeks (ACT 2004a; ACT 2004b).


Figure 6. SBE 43 dissolved oxygen time series plotted in dark blue, Shilshole Bay Marina, Seattle, WA (USA), March 23 – July 31, 2007. Data with a slope adjustment made after May 13 are co-plotted in cyan. Average Winkler values are shown as open pink circles, and the percent difference between the SBE 43 and Winkler averages are co-plotted along the right y-axis as black solid squares (before the May 13 validation and slope adjustment), and as red open squares (from May 13 forward following the adjustment). The dashed curved lines are drawn to illustrate how correcting the in situ data can prolong deployment while maintaining accuracy in real-time (or post-processed) data. Mean standard deviation of the Winkler replicates is 0.03 ml/L.

 

To demonstrate how the correction is made, assume the single validation point made on May 13 (see arrow in Figure 6) is used to correct data after May 13. The average of the replicate Winkler values on that date was 9.737 ml/L, and the SBE 43 reported 9.308 ml/L. To adjust the calibration, a new SOC value is obtained by multiplying the pre-May 13 value of SOC by the ratio [(Winkler value)/(SBE 43 value)] (9.737/9.308 = 1.046) (Equation 2):

NewSOC = previousSOC * ([Winkler]/[SBE 43])               (2)

1.3866e-04 = 1.3256e-04 *1.046

The SBE 43 DO data calculated after May 13 using the NewSOC demonstrate how sensor accuracy is maintained near initial calibration accuracy by using a single quality reference sample to correct the calibration slope. Utilizing this approach can prolong an instrument deployment between mooring service intervals, and reduce service gaps in moored data streams.

To correct data in post-processing, the simple course is to assume a linear fouling adjustment per day (or week, or month) for the entire period or between any field validation data. This can be programmed into a simple script to calculate calibrated DO data with time.

Summary

The SBE 43 DO sensors have high initial calibration accuracy, low (non-detectable) electrochemical drift, an effective anti-fouling approach, and a predictable linear response to bio-fouling. Effective anti-fouling defenses maintain a nearly drift-free signal for months, and the predictable character of drift when fouling occurs allows a reliable means of data correction. These characteristics extend the accuracy and useful data life of the sensor, extending the interval between service visits and dramatically lowering the maintenance and data costs of the SBE 43 oxygen sensor. The proven strategy is being used successfully in critical monitoring applications, for example in Cockburn Sound, Western Australia.

Application notes documenting the SBE 43 performance, calibration, and methods for optimizing deployments are available at the Sea-Bird Electronics’ website.

Acknowledgment

The authors express appreciation to Wayne Farrell of Greenspan Technology, Australia, and the staff and engineers at the Perth Seawater Desalination Plant (Water Corporation, Western Australia) for sharing moored water quality data from the Cockburn Sound real-time monitoring project.

References

ACT (Alliance for Coastal Technologies), Performance Verification Statements for the Aanderaa Instruments Inc. Dissolved Oxygen Optode (ACT VS04-01),  Greenspan Technology Dissolved Oxygen Sensor (ACT VS04-02), In-Situ Inc. Dissolved Oxygen RDO Sensor (ACT VS04-03), YSI Inc. Rapid Pulse Dissolved Oxygen Sensor ACT VS04-04 (www.act-us.info/Download/Evaluations/DO/Aanderaa/), 2004a.

ACT (Alliance for Coastal Technologies), State of Technology in the Development and Application of DO Sensors, Workshop Proceedings, Savannah, GA, January 12-14, 2004, UMCES Technical Report Series: TS-444-04-CBL/Ref. No. [UMCES]CBL 04-089, (www.act-us.info/Download/Workshops/2004/SkIO_DO/), 2004b.

C Janzen, D Murphy, and N Larson. Getting more mileage out of dissolved oxygen sensors in long-term moored applications. In: Proceedings of the Oceans 2007 MTS/IEEE VANCOUVER Conference and Exhibition, Vancouver, B.C. Canada, September 29-October 4, 2007. 0-933587-35-1. (pdf), 2007.

C Janzen, D Murphy, and N Larson. Assessing the calibration stability of oxygen sensor data on ARGO profiling floats using routine WOCE monitoring data from HOT. Session: 182 - Variability and Trends in Oceanic Oxygen: From a Tracer of Biological Production to a Bellwether of Climate Change, Poster Presentation, 2008 Ocean Sciences Meeting, Orlando, FL, USA, March 3-7, 2008

HE Garcia, LI Gordon, “Oxygen solubility in seawater: better fitting equations,” Limnology and Oceanography, vol. 37, no. 6, pp. 1307-1312, 1992.