CoMSBlack 92 international cruise Dataset

CoMSBlack 92 international cruise in the Black Sea was held in July 1992. Intercalibration Workshop for physical and chemical data was held at the Institute of Marine Sciences of the Middle East Technical University, Erdemli, Turkey, from 15-29 January 1993.

CoMSBlack 92a acquired complete hydrographic, biological, and chemical data set for the entire Black Sea with the participation of all Black Sea riparian countries (except Georgia) as well as the US. Nearly 400 hydrographic stations were occupied to the nominal water depth of 500 db. This quasi-synoptic survey was accomplished using five ships during an interval of less than four weeks.
The data set on this server contains only intercalibrated and pooled physical data. Requests for other types of data have to be sent to the data originators.

Ship and CTD inventory is shown in Table 1. The cruise coverage was essentially the same as for the companion cruise, HydroBlack 91 of September 1991, with approximately 20 nm between the stations.

In general, results revealed good agreement between CTD's from the different regions (an improvement over HydroBlack 91), and the intercalibrated results show a consistent and high resolution detail of the dynamic topography and other physical characteristics of the entire Black Sea basin.

Table 1

Specifications for the CTD probes used during the experiment are shown in Table 2.

Table 2.

Physical data intercalibration was accomplished using the data from the intercalibration stations occupied by two, three or four ships within a reasonable interval of time. The comparison was done only for data recorded below 1000 db, since the spatial and temporal variability at these depths is very small so that the dynamic range is very narrow. These data, as in HydroBlack 91, proved to be suitable for drift and noise analysis and removal.

For Bilim: The CTD data recorded with the SBE-9 instrument are of good quality. The data are practically noise-free and there is no evidence of sensor drift for temperature or conductivity, and, as a result, the salinity profiles are also smooth.

For Professor Kolesnikov : The ISTOK VII CTD instrument recorded data which are of good quality, without noise or time drift, both in temperature and conductivity/salinity.

For Piri Reis: The ship used the same instrument as Bilim (SBE-9) and occupied the same stations as Bilim in the western part of the network. The data are of good quality, with no differences in the bottom layer.

For Akademik: The data collected with an ISTOK V CTD are of good quality, but they show the same problems as the HydroBlack 91 data recorded by Professor Kolesnikov with the same instrument. Both temperature and conductivity data were noisy due to coarse discretization, but this fact has little influence on the data quality, the noise level being below the instrument accuracy. The response of the temperature sensor experienced a drift during the cruise, but it was not linear in time, showing instead erratic evolution from station to station. However, in each intercalibration station, the difference due to the drift was not depth dependent. Also, the sensor "stuck" at a temperature of about 9.025 for a depth range of more than 25 m at every station. There was practically no drift in the conductivity sensor, the shift in the recorded values being well below the desired accuracy (see Table 2).

For Professor Vodianitsky: The data set was processed during fish egg & larvae intercalibration meeting held in Erdemli in December 1992. The ship has covered the same area as Professor Kolesnikov but on a coarser grid and its data are not included here.

Water samples were obtained at intercalibration stations for intercalibration of the salinity values obtained from the CTD casts. The titration of the bottle samples was carried out on board of the R/V Professor Kolesnikov during the cruise. The titration results made on board were found to be consistent with the CTD values.

The results of data intercomparison for Bilim, Professor Kolesnikov , and Piri Reis show that in the deep layers the profiles are remarkably smooth and their concordance meets the experiment requirements. For the Akademik data, the above mentioned noise, drift (with a range exceeding slightly 0.01°) and the 'step' in the temperature were obvious. Conductivity profiles were smooth, without time or depth dependent drift.

At the common intercalibration station (N30N45), the vertical profiles of the temperature and salinity from all the four ships showed a good agreement for the first parameter (except for the Akademik data, where temperatures were higher), and a fairly good one for the second (again, the data from Akademik being higher and noisier).

Temperatures and conductivities recorded by Akademik at the intercalibration stations were first smoothed using a centered three point running average filter, applied three times. It was not considered appropriate to use more sophisticated filters, because the final goal was to reduce the noise in the differences with respect to other data sets. Contrary to the HydroBlack 91 data set, the present observations were made within the upper 500 db layer, except few deep intercalibration stations. That is why the 'step' in the temperature response of the ISTOK V sensor (Akademik), which was removed in the HydroBlack 91 data and replaced by linearly interpolated values, was left unchanged in the CoMSBlackB '92 data set.

The second test was to compute the statistical parameters (mean and standard deviation) of the differences between Bilim data and other data, using the temperature and salinity/conductivity at all the levels below 1000 db. The results show that Akademik temperature data are approximately 0.013°C higher than that of Bilim, the mean differences ranging from -0.007 to -0.020°C. For Professor Kolesnikov and Piri Reis, the temperatures exceed those of Bilim by, roughly, 0.003°C, that means that they are practically the same at the given precision. For the Bilim - Akademik conductivity differences, the mean values at every station ranged from -0.0017 to -0.0033 S/m, with an average of -0.0025 S/m. The differences between Bilim and Prof. Kolesnikov or Piri Reis data did not exceed an absolute value of 0.00036 S/m.

After the comparison made using the data from the intercalibration stations, the agreed correction procedures were applied to all the data sets:

For TEMPERATURE: Owing to the fact that Professor Kolesnikov and Piri Reis temperatures were practically identical, they were taken as a reference and left unchanged. The Bilim temperatures were reduced by 0.003°C (the mean difference relative to the previous two) for all the stations. An average value of 0.013°C was subtracted from all Akademik temperature values.

For CONDUCTIVITY: Due to good quality of data, only Akademik conductivities have been changed, by subtracting the value of the mean difference found in the intercalibration stations, i.e. 0.0025 S/m.

At some stations, 'positive' as well as 'negative' salinity spikes have been recorded leading sometimes to unrealistic density 'instability'. After discussions it was decided to leave the data unchanged.

When refering to the data, you must acknowledge the source of data which is the Cooperative Marine Science Program for the Black Sea.

Download CoMSBlack 92 Dataset (1324K zip file)

The archive will be loaded from the main BSEIN server located at bsein.mhi.iuf.net (Sevastopol, Ukraine). The internet connection can be slow.