HYDROBLACK 91 international cruise Dataset

HYDROBLACK 91 international cruise in the Black Sea (Aubrey et al., 1992, IOC Workshop Report No.91) accomplished 307 hydrographic stations with the ships from three Black Sea riparian countries. Two Ukrainian vessels (Professor Kolesnikov and Parshin), two Turkish ships (Bilim and Piri Reis), and one Bulgarian vessel (Akademik) took part, occupying stations quasi-synoptically over the entire Black Sea within a period of four weeks. Stations spacing was approximately 20 nm. Ships and CTD inventory is shown in Table 1.

Since ships used different types of CTDs (basic specifications are shown in table 2) data comparison was considered as essential part of the intercalibration exercise. The deep data (below 1000 m) from intercalibration stations were used for intercomparison and final intercalibration of the pooled data set. Stations at which at least two ships acquired measurements to full water depth or to 2000 db as synoptically as possible were planned to be intercalibration stations. The deep data from other stations were also used for examination of time drift and noise problems with the instruments. The results were compared with data from R/V Knorr deep stations. Examination of the data from intercalibration stations revealed the following major problems of different instruments:

• spikes in both temperature and salinity profiles for SBE-9 used on Akademik which occurred due to cable problems;
• space/time variability with a span of about 0.02°C for Professor Kolesnikov Istok V temperature sensor with a non linear time drift;
• poor accuracy of Hydrozond compared to Sea Birds or Istok.

Table 1.

Table 2.

Analysis of the data indicated that some filtering for spike removal was required for Akademik and Professor Kolesnikov data. Filtering was limited to levels greater than 500 db. A tenth order Butterworth filter with a cutoff at 0.05 per meter was applied. Only conductivity and temperature were filtered. The ‘staircases’ which appeared in Professor Kolesnikov temperature profiles were judged to be a faulty thermistor. To correct the data, noticeable ramps were filtered by applying a linear correction over the length of the ramp. To smooth out discretization ramps, both temperature and conductivity below 500 m were filtered with the tenth order low-pass Butterworth filter. No filtering was applied for Parshin Hydrozond data because of their coarse resolution.

The filtered and de-spiked data were then compared to fix intercalibration offsets for all instruments. For temperatures, the Bilim, Piri Reis and Akademik data were left unchanged, and Professor Kolesnikov temperatures were reduced by 0.01°C. At the same time the scatter about that value is about 0.005°C. For conductivity, the sensors of Professor Kolesnikov and Piri Reis were intercalibrated to Bilim sensor which showed nearly same values as for Akademik. The Professor Kolesnikov conductivity was reduced by 0.0023 S/m, Piri Reis conductivity was reduced by 0.001 S/m, and Akademik conductivity was reduced by 0.0006 S/m which is close to the accuracy of the data. The Parshin data for deep stations were discarded. The final agreement between data from different vessels for intercalibration stations is within 0.005°C for temperature and 0.003 for salinity, though the uncertainties for the entire data set should be estimated as 0.01°C and 0.005 respectively. Comparison of HYDROBLACK 91 data with other data (Knorr 88 cruise, CoMSBlack 92, and CoMSBlack 93 cruises) reveal agreements within the same limits.

When refering to the data, you must acknowledge the source of data which is the 'HydroBlack'91' research program.

Download HYDROBLACK 91 Dataset (1152K 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.