CTD stands for conductivity, temperature and pressure (depth) of the seawater. A CTD chain consists of many CTD probes aligned on a cable, which by inductive coupling provides the power for the sensor electronics and the data transfer between underwater sensors and deck unit.

Each sensor fin is an autonomous CTD device with a platinum thermometer, a seven-pole conductivity cell and a piezoresistive pressure transducer. Voltages are multiplexed and digitized by a 16-bit analog to digital converter. The conversion rate is 20 cycles per second. Data can be internally low pass filtered for adjustment to the rate of external polls, typically every two seconds.

The vertical resolution of the acquired two-dimensional property fields is defined by the distance between sensor fins. A CTD chain can be configured with arbitrary sensor positions. The most appropriate spacing depends on the oceanographic situation. Typical sensor distances are between 1 and 10 metres. Several tens of sensor fins are required for a good representation of two-dimensional property fields. The theoretical upper limit of 254 underwater units is caused by their unique 8 bit addresses.

The horizontal resolution of a towed CTD chain is the product of the towing speed and the duration of a polling cycle, which is typically two seconds. The maximum towing speed must be adjusted to the breaking strength of the towing cable. The drag of a towed CTD chain system increases with the speed of the towing vessel, with the length of the cable, and with the efficiency of the depressor at its tail. Speeds in excess of 8 knots (4 m/s) do not challenge the sensors, they are however only recommended for very short chains (<25 m) or for long chains with weak and quite inefficient depressors. Realistic tow speeds for chains up to 200 m long are between 4 and 6 knots (2 to 3 m/s). Hence the typical horizontal resolution becomes 5 metres.

Temperature section acquired by a CTD chain system in Massachusetts Bay in June.

The data example above was taken from an acquisition in Massachusetts Bay in June. 200 seconds (100 records) from a long section are shown. Due to limited water depth only the lower 25 sensor fins of the CTD chain were deployed into the water, while the remaining 15 fins were still on deck and measured in air (data discarded). Each tile in the image represents one data point. The records appear slanted in accordance with the spatial lag of lower sensors in relation to the stern of the towing ship. Internal waves appearing in the records were produced further out at the shelf break by interaction of tidal currents with topography. While temperature is displayed above, salinity and density sections (not shown) look almost the same in this case because of a fixed T-S relation where the warmer surface water has lower salinity.

Another shallow water example was obtained from a CTD chain configured with 81 sensor fins. It took 90 minutes to run the distance of 13 km. A pixel width in the image on the left represents 15 seconds, while original data were acquired every 2 seconds. In contrast to the example from Mass Bay the T-S relation in the records from the Baltic Sea in February is not a monotonic function. See for instance the temperature structure in the halocline. A comprehensive description of the oceanographic situation is found in the article of J. Sellschopp et al., Direct observations of a medium-intensity inflow into the Baltic Sea, Continental Shelf Research 26, pp. 2393-2414 (2006).

13-Dec-2013 js