Development of CTD-systems

       The main tool for a physical oceanographer is an instrument for measuring electrical conductivity, temperature, and pressure/depth (CTD), whose output parameters can be used to calculate the practical salinity and density of seawater using an accepted “equation of state” based on these three physical parameters.

     

Bruce Hamon (1917-2014)

Restrictions on collecting seawater samples and analyzing them with salinometers led to the development in 1948 of the first in situ system developed by A. W. Jacobsen, which worked for the Bristol Corporation in Waterbury, Connecticut. Its use was limited to 400 meters. This device used a carrier cable and another multi-core cable to get data. Although this system was very simple, it opened the way for new measurement methods in Oceanography.

In 1958, in Australia, at CSIRO Division of Fisheries and Oceanography Bruce Hamon (1917-2014) and Neil Brown (1927-2005), the ”fathers” of the modern CTD-systems, described a “Temperature-Chlorinity-Depth”  recorder, designed for use at depth to 1000 meters. The accuracy for temperature was within ± 0.15°C,  for chlorinity ± 0.03ppt and depth ± 20m. Chlorinity was measured using a conductivity cell with two electrodes and a circuit including thermistors and resistors to compensate for the effect of temperature on conductivity. A simple phase shift oscillator converts the information coming from each sensor into a frequency-modulated audio signal. This signal was transmitted to a ship, where it was converted to direct DC voltage and recorded on a strip-chart recorder.

A.Bradshaw(L) and N.Brown(R) are setting up the WHOI STD-system, c.1960

    In 1959, Neil Brown left Australia to settle in Woods Hole (MA, USA) with WHOI and work with Bradshaw and Schleicher to understand what could be done with a more accurate in situ measuring instrument. It used an inductive conductivity sensor and a sealed compensation cell filled with standard seawater. This cell was intended to accurately compensate for the simultaneous effects of temperature and pressure.

    It was a brilliant solution in the absence of relations between conductivity, temperature and pressure of seawater, implemented in the form of measurements of relative conductivity (Rt). Such a measuring system has a significant time constant due to slow temperature stabilization in the sealed reference cell, and its application requires special methodological solutions when conducting measurements in profiling mode. I know it from my own experience of developing a Microsalinometer MS-310 with a similar ratiometric principle of measurement. Unfortunately, WHOI did not recognize the uniqueness of this measurement system and work to improve it was abandoned. N.Brown returned to Australia in 1961.

STD system rocketing launch.

In 1962 Neil Brown returned to the United States to join a company called Hytech in San Diego, California, to continue the development work started at WHOI. Hytech was formed in 1960 by Don Cretzler and initially, it made bathythermographs, current meters, and a wave and tide monitor. These were amongst the first electronic ocean instruments.  Neil Brown joined the company to develop a laboratory salinometer (see Development of Salinometers).

Hytech STD

Soon after, he designed the electronic components that were
the key to making in-situ measurements of salinity enabling the first STD to be built in 1964. The following year the Navy contracted with Hytech and Santa Monica, CA based Bissett-Berman, who was one of the few companies at the time with computer technology. Together they made the first automated ocean data collection system that was used to collect large
volumes of data. As a result of this collaboration, Bissett-Berman acquired Hytech. In 1964, the Navy awarded them a contract to develop the first oceanographic sensors capable of continuous operation for up to one year in-situ. This was followed by other Navy contracts for specialized sensor systems. In 1969, the San Diego division moved into a 33,000 square foot purpose-built building in Kearny Mesa, CA. In 1970, Bissett-Berman was purchased by the British company Plessey Ltd. and the San Diego Division was renamed Plessey Environmental Systems. 

This instrument measured Conductivity, Temperature, Pressure (CTP), but because the instrument implemented the “salinity and depth equations” in complex adjustments of internal analog electronics, and these values of Salinity, Temperature, and Depth were transmitted to the surface, the instrument became known as ” STD”, or “S/T/D”.

Commercial of Model 9040 S/T/D System, ca.1976

The Plessey S/T/D systems were on the market from 1964 to 1980,  with popular Telemetering S/T/D 9040 and Self-recording S/T/D 9060.

Commercial of Model 9060, ca.1976

      At that time, computers and their peripherals were expensive, not very reliable, and difficult to use at sea. Since it was essential that salinity measurements were available immediately, the complex relationships between salinity, temperature, pressure, and conductivity had to be simulated using a “salinity bridge”. This bridge consisted of two platinum temperature sensors and three thermistors, two pressure sensors and six transformers, as well as an inductive conductivity cell and important electronic circuits. Its calibration required numerous and complex settings and adjustments.    

Schematic of STD showing available sensors

 The STD’s accuracy was not adequate for depths much greater than 1,000 meters, and there was a severe “spiking” in the salinity data, due to the slow response of the temperature sensor relative to the conductivity sensors. Also, its accuracy was limited by systematic errors in the salinity bridge emulation of the salinity, pressure, temperature, and conductivity relationships.

     Two aspects of the inductive conductivity sensor have limited its commercial applications to the measurement of Oceanographic salinity and density. First, the large thermal mass of the sensor relative to the volume of water that passed through it from the inside, which led to a thermal error in the measured conductivity. The electrical conductivity of a liquid depends strongly on the temperature at which it is measured. If, due to the presence of the thermal mass of the sensor, the conductivity sensor heats or cools the volume of water it is trying to measure, this will lead to inaccuracies in the calculated salinity/density.

Inductive conductivity cell and temperature sensors of STD-system

       Since Oceanographic sensors are often used at significant depths where the surrounding water pressure can reach 700 bar, the magnetic cores of the sensors were installed in a pressure-proof metal housing. Then the pressure protection housing had to be covered with a dielectric material to prevent the metal housing from short-circuiting around the sensor.

      Second, the large non-contact external field can be affected by other structures associated with the CTD device, such as auxiliary sensors, protective frames, etc. These external structures change the free current flowing through seawater. Consequently, their effect cannot be distinguished from changes in the conductivity of seawater, and they introduce a mixing variable.

NBIS CTD – resolving space and time.

The rapid evolution of the digital computer prompted N.Brown to rejoin WHOI in 1969 to commence the development of a digital instrument to address the limitations of the STD.

NBIS MarkIIIB CTD

In 1970, Brown again developed an improved device that used a 4-electrode MK IIIB CTD conductivity sensor 4-electrode conducting cell in which the two electrodes were deployed internally, and the two electrodes were deployed externally of a small (4 mm square alumina tube×3 cm long) head. In this” contact-type “conductometric cell, the electrodes and tube are configured in the traditional” four-pole resistance measurement ” method. 

Four-pole resistance measurement is advantageous because the resistance of the measuring lead wires does not affect the measurement result. Thus, the sensors are connected based on the four measuring terminals can be excluded from the calculation of the conductivity of the fluid mixing variables arising both from the lead-in wires and more difficult to stabilize the impedances of the electrodes.

NBIS Mark III conductivity cell schematic

It measured temperature using a combination of a very stable platinum thermometer with a typical response time of 250 milliseconds and a thermistor with a response time of approximately of 50 milliseconds. The outputs of these two sensors were combined in an analog circuit that had the stability of the platinum thermometer and the speed of the thermistor with no sensitivity to the steady-state calibration errors in the thermistor. The combined output was digitized along with the pressure and conductivity sensors. Experience showed that due to the variability and complexity of the time response of these sensors, it was better to digitize their outputs separately and combine the outputs numerically in the computer. One key development in the CTD was the high resolution (16-bit) AC digitizer, which had a noise level of 0.1 microvolts at a rate of 100 samples per second.

Following the publication of his work in 1974, N.Brown left WHOI to form Neil Brown Instrument Systems, Incorporated (NBIS Inc.) to manufacture the Mark III CTD.

Neil Brown “adjusting” his CTD c.1970

The MK IIIB CTD and sensor were a significant commercial success and were the tool that established CTD as the primary tool for collecting high-resolution oceanographic salinity data. Since the original design goal of the MK IIIB was to provide a very high vertical resolution of salinity profiles, the small size of the MK IIIB conductivity sensor, combined with its partial external field, were acceptable technical compromises compared to the desire to achieve absolute measurement stability. It has become a common scientific practice to adjust  CTD data based on reference salinometry made on discrete water samples that were collected at the same time. The small length of the MK IIIB sensor provided the sensor with a short flushing interval, so that it responded quickly to changes in conductivity. It was known as the “microstructure ” sensor. The high response rate was problematic for users, as it required complex time/data matching with slower temperature sensors. This mismatch between the time constant of conductivity and temperature led to calculated” spikes “in salinity / density or “anomalies” in the resulting oceanographic profiles.

(To be continued)