Calibration of 43I sensors is performed in dedicated bath systems, separately from the CTD. A submersible calibration module, incorporating a pump and dual sensor acquisition channels, is a temporary host for the 43I modules during calibration. The automated bath systems produce a high-precision 18-point calibration (3 oxygen levels at 6 temperatures) with a unique set of coefficients for each sensor.
|
SBE 43I mounted on |
Closeup of SBE 43I
Sensor
Closeup of SBE 43
Sensor |
SBE 43I Calibration |
Specifications for Argo Float Applications:
| Measurement range 1 | 0 -120% of saturation, minimum |
| Output signal frequency | 3000 - 20,000 Hz |
| Initial accuracy 2 | 1% of saturation |
| Observed stability 3 | Drift less than 1 micro-molar/year |
| Resolution 4 | 1 Hz (approximately 0.01% saturation, < 0.05 micro-molar) |
| Depth rating | 2000 dbars |
| Power | 50 milliwatts |
Footnotes:
1 The sensor output frequency signal is converted to oxygen concentration via the calibration equation. Inspection of the equation reveals that the frequency is proportional to percent saturation, and a span of about 10,000 Hz is 100% saturation. The minimum guaranteed measurement range and associated signal resolution depends strongly on temperature conditions:
| T,S | Saturation | 120% Saturation |
| -1.8C, 35 psu | 365 micro-molar | 440 micro-molar |
| 30C, fresh | 237 micro-molar | 285 micro-molar |
2 Initial absolute accuracy is influenced by the oxygen standard (Winkler titration standard), the calibration bath environment, and the accuracy with which the calibration equation captures the characteristics and sensitivities of the sensor. We believe the sum of errors is less than 1% saturation. The internal repeatability of calibrations, and agreement between different sensors, is better.
3 The calibration stability of the Clark sensor technology for measuring oxygen is affected by and related to the amount of oxygen gas converted to hydroxyl ion within the sensor electrolyte. This inherent sensor stability is a function of how the Clark sensor design is specifically implemented. We have assessed the inherent sensor drift rate from drift performance observed in deployed sensors on Argo floats. This drift rate is less than 0.5% saturation in 33 months (Janzen, C., Larson, N., Murphy, D., Examining the Calibration Stability of Sea-Bird’s Oxygen Sensor Technology: Drift < 1 mmol/kg per year, poster presentation, AGU 2006 Fall Meeting). When these SBE 43I sensors are paired with the SBE 41 and SBE 41CP CTD systems designed for Argo floats, the CTDs provide an unusually clean and well-antifouled environment that reduce the oxygen measurement drift rate to the inherent sensor stability. We believe SBE 43I sensors in this specific float environment will drift less than 1 micro-molar/year.
4 See footnote 1. The data resolution are reported rounded to 1 Hz. This is less than 0.05 micro-molar, depending on temperature and salinity. The sensor accurately resolves static equilibrium oxygen changes in the calibration bath environment at better than 0.1% of saturation.
Specifications are subject to change without notice.
