Metadata Report for BODC Series Reference Number 696130

• Metadata Summary

• Problem Reports

• Data Access Policy

• Narrative Documents

• Project Information

• Data Activity or Cruise Information

• Fixed Station Information

• BODC Quality Flags

• SeaDataNet Quality Flags

Metadata Summary

Problem Reports

No Problem Report Found in the Database

Data Access Policy

Open Data supplied by Natural Environment Research Council (NERC)

You must always use the following attribution statement to acknowledge the source of the information: "Contains data supplied by Natural Environment Research Council."

Narrative Documents

James Clark Ross 097 Idronaut OceanSeven 320 CTD Narrative

The following is adapted from the cruise report.

Description of the instrument

During JR097 a CTD probe from Idronaut, the OceanSeven 320, was mounted on the rosette to compare the performance of this instrument with a Sea-Bird Electronics (SBE) 911+ CTD system. The Idronaut CTD belongs to BAS, and has been used in the past for taking CTD casts through holes drilled in the Ronne-Filchner Ice Shelf. It has a 7000 m pressure casing, a 4000 dbar pressure sensor, and dual temperature/conductivity sensors. In addition, a mechanical bottom sensor switch and FSK telemetry circuitry are installed, although they were not used on this cruise. There is also a magnetic on/off switch on the end of the instrument. Data were stored in 1 MB of internal memory, as well as on a 128 MB memory expansion card. The probe was running the newest available firmware, version 8.017, and had been returned to the factory in November 2004 for recalibration following an upgrade of the conductivity preamplifier circuitry, the design of which has been recently improved.

Data were downloaded in three ways: using Idronaut's REDAS software, version 3.25, which is primarily intended for real-time data acquisition, as well as with Idronaut's terminal emulator Iterm, which also has functions for downloading and saving data from the probe. In addition, several Matlab scripts were written in order to download data directly from the probe using REDAS 5 binary mode. It was verified that data downloaded in this way matched the data from Iterm, albeit with more precision, while there were very small differences (rounding errors?) between these data and the data from REDAS.

Logging modes

Several different logging modes were utilized on the cruise. When the instrument is connected to a computer on an RS232 cable or through an FSK deck unit, it can transmit data at a rate near 20 Hz; the time stamps, which are added by the instrument to live data, have a mean time difference of 51.5 ms. However, when the instrument logs internally, the options are somewhat more limited and confusing.

Linear mode will supposedly log at full speed, taking a maximum of a certain number of samples within a preset pressure interval. Only the downcast is recorded. The settings recommended by Idronaut are 20 samples and a bin size of 0.2 dbar. We used this on some profiles, and on some profiles we used 100 samples per bin. In practice, the timestamps in this mode are spaced very irregularly, mostly oscillating between being 10 and 90 ms apart, but occasionally with large jumps. For instance, in cast 3, a sequence of time stamps (in hundredths of seconds) is:

5170 5171 5180 5181 5190 5191 5290 5291 5292 5293 5294 5295 5296 5297 5298 5299 5300 5301 5302 5303 5304 5305 5306 5307 5308 5310 5320 5321 5330 5331 5340

These large jumps are few and far between, and the time stamps do appear to try to catch up;, but it seems unlikely that they correspond to the actual time of measurement. Instead, they may be the time when data are written to memory. In linear mode, the mean time difference (during an uninterrupted downcast) was found to be 57.5 ms.

In timed mode, a constantly spaced time series can be recorded. Here, the time stamps are quite regularly spaced, but the highest sampling frequency available is 10 Hz. The time stamps are usually 100 ms apart, but with bursts of four samples spaced 200, 10, and finally 90 ms apart; occurring exactly every 11 seconds. This will average out to 100 ms. More worrying are longer gaps, which occasionally occur, and where data simply are missing. These occur in several stations, and appear to happen consistently within 1 s of timestamps 178890, 358890, and 539090 (i.e. half an hour apart), when there is a dropout of more than one second. An example (from station 24) is:

178870 178880 178890 179010 179011 179020 179030 179040 179050 179060

Timed mode and linear mode can both be started using the magnetic switch, making them much easier to use if the instrument is mounted on the rosette (without having to plug in a computer).

Finally, there is a fast acquisition mode, which cannot be started using the switch, but which is supposed to log to memory at full speed. In this mode, the time stamp differences also appear to oscillate between 90 and 10 ms, as in linear mode, and also exhibit the half-hourly data dropouts seen in timed mode. However, the effective sampling rate still appears to be best in this mode, with samples taken 50.5 ms apart on average.

Experiments

A summary of the casts taken is given in the table below. Parameters are only indicated when they were set, i.e. the magnetic switch was used between casts when nothing is indicated.

The instrument was in rosette position 5 except during station 80, when it was suspended by a kevlar rope approximately 10 m below the rosette. When the probe was mounted in the rosette, its pressure port was 48.5 cm above that of the SBE CTD. The SBE CTD was mounted horizontally near the bottom of the rosette; the outflow from its pumps pointed slightly upwards, possibly leading to a slight difference in flow rate between upcasts and downcasts.

The probe is powered by 2 packs containing 10 AA batteries each. The batteries only ran out once, after Station 37. However, there are some gaps in the sampling. During stations 1-14 the probe was in linear acquisition mode, and the probe was not connected to a computer between stations. When the probe was connected to a computer after station 14, it did not enter logging mode when the switch was turned, and only three casts had been stored. Apparently the probe had somehow left linear logging mode, which is very disconcerting if the probe is to be used as a closed box system by an untrained operator. The same thing may have happened during station 80. However, here the acquisition stopped in the middle of the profile, restarting briefly later in the cast. It is possible that the connector between the endcap and the circuit board came loose before or during this cast; the probe was not opened after Station 79, where it functioned correctly.

An initial look at the data shows that the sensors do give slightly different readings. There is an offset between the two sets of temperature/conductivity sensors, and this offset drifted between casts. It is surprising that the recently calibrated sensors have not maintained their calibrations better. While the conductivities should be readily correctable using bottles, it would be much harder to fix the temperatures; assuming that one sensor is correct, the other would have drifted over a range of 0.007 °C; a preliminary comparison with the SBE CTD in sections of near-constant temperature showed that both sensors appeared to change with respect to the Seabird sensors. However, at least once during the cruise, the SBE CTD did also suddenly shift its calibration. The temperature calibration of both instruments could be checked against a SBE 35 deep ocean standards thermometer, which was used on some casts.

Generally the OceanSeven 320 temperatures and conductivities appear to exhibit more variation (noise/fine-scale structure) than those of the SeaBird on the downcast. However, on the upcast the opposite appears to be the case. There is a slight lag between the temperature and conductivity sensors on the OS320 that does not appear to be present in the Seabird (presumably as a result of its pumped system). This lag does vary with descent rate; the lag is particularly high on station 80, where the descent rate was 0.5 m/s, and the mode of the lag is around 0.7 s. Generally the lag is much smaller. Although no clear correlation was found with rate of change of temperature, it is possible that the thermal mass of the conductivity cell is causing some of the lag.

Although the OS320 temperatures appear smoother than those of the Seabird, they are clearly not matched up properly against the conductivities, introducing salinity variations of up to 0.04. In contrast, the Seabird, whose conductivity appears to vary approximately the same amount as the Idronaut, but which has more variation in the temperature, produces a salinity signal that does not vary more than 0.004 from a straight line over the plotted range.

Conclusions

• The OS320 cannot provide a regularly spaced time series at 20 Hz; for unattended profiling this is a clear disadvantage compared with the constant 24 Hz logging rate of the Seabird. However, it does have more complex logging options, which would be preferable for other types of deployments. Perhaps a constant logging rate can be provided in future firmware updates.
• Data dropouts of 1 s length occur every half hour.
• The recorded time stamps do not correspond to the time of measurement.
• Unattended use of the OS320 did not prove reliable during the cruise; on two occasions the instrument stopped logging, when it was supposed to be in an unattended mode.
• The calibration of the two temperature sensors constantly changed over the cruise; this difference is much greater than the stated accuracy. The Seabird sensors did also shift, although this occurred only once (as a step), and was a much smaller shift, possibly due to an electrical fault in the CTD.
• The conductivities also shifted during the cruise; again, this difference is much larger than the stated accuracies. This could be checked against water bottles.
• The sensors on the OS320 exhibit slightly higher variability than on the Seabird on the downcast; on the upcast the opposite is true.
• There is a lag between the conductivity and temperature measurements, which is not present on the Seabird (after conductivities have been advanced in the deck unit). The lag is variable, and appears to be worse when profiling at 0.5 m/s compared with 1 m/s. In addition, it is different between the upcasts and downcasts.
• The lag introduces large errors in derived variables, which a simple smoothing would not remove. Instead, it may be better to perform some more complex filtering to remove the mismatch; more research into the time response of the sensors while profiling (not just in a calibration tank) would probably be warranted. Perhaps the thermal mass of the conductivity cell is responsible for some of the errors.

Idronaut Ocean Seven 304 CTD

The Ocean Seven 304 CTD is a high performance CTD probe with very small diameter and extremely low power consumption. It can be easily integrated/adapted to third-party systems like floating profilers and/or buoy-moored systems. It does not require pumps or any other external device to flush the sensors, which minimizes its power consumption. The 304 CTD offers a combination of 16-bit high resolution data accuracy, with long-term sensor stability. The user can select the proper conductivity range i.e. for salt or fresh water.

Sampling modes

User selectable sampling/operating modes include: continuous, pressure, timed, conditioned and burst.

Real time communications

The Ocean Seven 304 CTD communicates with a computer via a standard RS232C interface. Real-time data can be acquired by means of the REDAS Windows software. An optional RS422 interface overcomes the limitation of the RS232C cable maximum length (200 m) and allows the probe to transmit data through distances up to 1000 m. The communication speed is user selectable among: 9600, 19200, 38400 and 57600 bps. A Bluetooth wireless interface can be optionally added to the wired interfaces.

Software

Idronaut programmes operating under Windows 98se/ME/2000/XP allow the operator to configure the Ocean Seven 304 CTD data acquisition and logger functions and upload data from the 128-MByte internal memory. They are:
ITERM: terminal emulation programme to easily communicate with the Ocean Seven 304 CTD using the probe integrated operator interface.
REDAS: data processing and retrieval programme which allows the display and plotting of conductivity, temperature, pressure and derived variables such as salinity, sound speed, density, according to UNESCO formulas and recommendations.

Data storage and battery endurance

The Ocean Seven 304 CTD allows the storing of 4,100,000 data sets each one being composed of the reading of CTD sensors plus the acquisition date and time. The batteries are sufficient to keep the CTD continuously on for 36 hours in continuous sampling mode and at the maximum sampling rate. Further battery endurance can be obtained by using lithium batteries. Whenever the CTD operates in "Timed, Burst and Conditioned" modes, the battery endurance is considerably extended because the CTD waits for the interval between acquisitions in "Sleep mode".

CTD chains

Chains of Ocean Seven 304 CTDs are deployed by easily clamping the CTDs with a screwdriver to a rope. Chains of OS304 CTDs can be used to profile or perform long-term monitoring by properly configuring the CTD data acquisition method. Furthermore, the Bluetooth wireless connectivity option allows the instant recovery of data stored in the CTDs internal memory once they are back to the surface. OS304 CTD Bluetooth unique addressing identification code allows the operator to select one OS304 among the others present in the chain.

Specifications

Further information on the instrument can be found on the Idronaut website.

James Clark Ross 097 Idronaut OceanSeven 320 CTD BODC Processing

The Idronaut CTD casts were received by BODC in an ASCII format. They were copied to BODC's QXF format by transfer process 392. The table below shows the mapping of parameters from the raw data file to BODC's parameter dictionary

Following the transfer, the data failed BODC's standard checks as the time steps were not monotonic (i.e. the length of the time steps varied). In order that the data could pass these checks, the channels AADYAA01 and AAFDZZ01 were dropped from the QXF files. The data originator reports that there were issues with the time stamps of the data, so this deletion of the channels should not be a serious concern.

It should be noted that the upcast portion of cast 037 (series 696154) is incomplete, and the series ends with a minimum pressure of 34.081 decibars.

Following the trimming of the series so that only one upcast and one downcast remained in the record, the first and last cycles of the downcast were flagged 'B' and 'E' respectively.

Please note that the following pairs of casts were made at the same locations, so the latitude and longitude in their metadata will be the same : 80 and 80B (series 696178 and 696191) and series casts 999 and 999A (696209 and 696210).

The data were then screened using BODC's in-house visualisation software, EDSERPLO. The following problems were noted with the data:

Project Information

AutoSub Under Ice (AUI) Programme

AutoSub was an interdisciplinary Natural Environment Research Council (NERC) thematic programme conceived to investigate the marine environment of floating ice shelves with a view to advancing the understanding of their role in the climate system.

The AUI programme had the following aims:

• To attain the programme's scientific objectives through an integrated programme based on interdisciplinary collaborations and an international perspective
• To develop a data management system for the archiving and collation of data collected by the programme, and to facilitate the eventual exploitation of this record by the community
• To provide high-quality training to develop national expertise in the use of autonomous vehicles in the collection of data from remote environments and the integration of such tools in wider programmes of research
• To stimulate and facilitate the parameterising of sub-ice shelf processes in climate models, and to further demonstrate the value of autonomous vehicles as platforms for data collection among the wider oceanographic and polar community

Following the invitation of outline bids and peer review of fully developed proposals, eight research threads were funded as part of AUI:

Physical Oceanography

• ISOTOPE: Ice Shelf Oceanography: Transports, Oxygen-18 and Physical Exchanges.
• Evolution and impact of Circumpolar Deep Water on the Antarctic continental shelf.
• Oceanographic conditions and processes beneath Ronne Ice Shelf (OPRIS).

Glaciology and Sea Ice

• Autosub investigation of ice sheet boundary conditions beneath Pine Island Glacier.
• Observations and modelling of coastal polynya and sea ice processes in the Arctic and Antarctic.
• Sea ice thickness distribution in the Bellingshausen Sea.

Geology and Geophysics

• Marine geological processes and sediments beneath floating ice shelves in Greenland and Antarctica: investigations using the Autosub AUV.

Biology

• Controls on marine benthic biodiversity and standing stock in ice-covered environments.

The National Oceanography Centre Southampton (NOCS) hosted the AUI programme with ten further institutions collaborating in the project. The project ran from April 2000 until the end of March 2005, with some extensions to projects beyond this date because of research cruise delays. The following cruises were the fieldwork component of the AUI project:

Table 1: Details of the RRS James Clark Ross AUI cruises.

All the cruises utilised the AutoSub autonomous, unmanned and untethered underwater vehicle to collect observations beneath sea-ice and floating ice shelves. AutoSub can be fitted with a range of oceanographic sensors such as:

• Conductivity Temperature Depth (CTD) instruments
• Acoustic Doppler Current Profillers (ADCP)
• A water sampler
• Swath bathymetry systems
• Cameras

In addition to use of AutoSub during each cruise measurements were taken from ship. These varied by cruise but included:

• Ship underway measurements and sampling for parameters such as:
  • Salinity
  • Temperature
  • Fluorescence
  • Oxygen 18 isotope enrichment in water
  • Bathymetry using a swath bathymetry system
• Full-depth CTD casts for with observations of samples taken for parameters such as:
  • Salinity
  • Temperature
  • Fluorescence
  • Optical transmissivity
  • Dissolved oxygen
  • Oxygen 18 isotope enrichment in water
  • Water CFC content
• Sea floor photography and video using the WASP system
• Sea floor sampling with trawls/rock dredges
• Sea ice observations (ASPeCt), drifters and sampling

The AutoSub project also included numerical modelling work undertaken at University College London, UK.

The project included several firsts including the first along-track observations beneath an ice shelf using an autonomous underwater vehicle. The AutoSub vehicle was developed and enhanced throughout this programme and has now become part of the NERC equipment pool for general use by the scientific community. Further information for each cruise can be found in the respective cruise reports (links in Table 1).

Data Activity or Cruise Information

Cruise

Complete Cruise Metadata Report is available here

Fixed Station Information

No Fixed Station Information held for the Series

BODC Quality Control Flags

The following single character qualifying flags may be associated with one or more individual parameters with a data cycle:

SeaDataNet Quality Control Flags

The following single character qualifying flags may be associated with one or more individual parameters with a data cycle: