38 
Sy, A. et al. (2002): Upper Ocean Climate Shlp-of-Opportunity Programme of BSH - A Status Report 
5. Data processing of delayed mode XBT data, quality control and hardware testing 
XBT data processing is carried out within the framework of our standard CTD data processing, it 
is part of the BSH quality system which has been certified according to the international 
standard ISO 9001. An overview of the main processing steps is given in Fig. 29. In this context, 
the filter coefficients for the moving average and median applications are only meant as 
guideline values. The processing of individual profiles of particularly poor quality may differ 
under certain circumstances. For instance, in the case of profiles disturbed by heavy noise, 
stronger median filtering (Sy, 1985) was carried out using q > 15 filter weights. Such doubtful 
profiles are indicated in the station lists. The final depth-averaging at 1 m intervals does not take 
into account the reduced statistical independence caused by previous filtering. Compaction on 
1 m steps should guarantee a uniform standard for archived data. Whatever the requirements of 
scientific analysis may be, the relatively high nominal vertical resolution of the final data allows 
investigators to adjust data according to their own needs. The final data are archived in the 
Deutsches Ozeanographisches Datenzentrum (DOD) which is operated by BSH. 
Quality assessment of the XBTs and XCTDs currently in use and of any new expendable probe 
that may enter the market is a very important part of assuring a high-quality data base. Provided 
that the system is working correctly, there are three main error sources which may seriously 
affect the data quality. They are 
digitizer errors which are responsible for inaccuracies in the conversion of the 
electronic signal into temperature and/or conductivity units. 
Sensor errors due to various causes, e.g. wrong thermistor response, air bubbles in 
the conductivity cell (bubble adhesion) or calibration failure. 
Depth fall rate errors, e.g. due to an inaccurate depth time equation or production 
changes. 
A well designed quality test strategy should be aimed at keeping all three main error sources 
under control. 
Systematic field tests of expendable probes, which had been initiated at the 3 rd SOOP meeting 
held in Hamburg in 1989 (IGOSS, 1989; Sy and Ulrich, 1990) and were aimed at checking the 
manufacturer’s specification independently, i.e. from the customer’s point of view (TT/QCAS, 
1992), have been carried out in the past by members of the Task Team of Quality Control of 
Automated Systems (TT/QCAS). They resulted in significant equipment improvements by the 
manufacturer and also helped to improve data quality through internationally co-ordinated and 
controlled XBT depth fall rate experiments (Hanawa et al., 1994; 1995; Rual et al., 1996). As 
Sippican Inc., Marion, USA, monopolizes the market of XBT probes, the importance of this type 
of co-ordinated quality tests cannot be overemphasized. 
Since the early days of WOCE, there have been demands for an improvement of the quality 
of industrially produced expendable CTDs (XCTD) in order to ensure the accuracy and 
precision needed for large-scale measurements of heat and salt storage of the upper ocean 
in the WOCE voluntary observing ship programme. Comparative data from research vessels 
were to be used to check the accuracy of XCTDs (WOCE, 1988). Therefore, during WOCE 
cruises, independent scientists took the opportunity to test this new expendable device 
against a controlled and accurate CTD reference (Sy, 1992; 1993b; 1995; 1996; 1998a; 
1998b; Johnson, 1995; TT/QCAS, 1997; Watanabe et al., 1998; Gilson et al., 2000), which 
helped to improve the system’s performance so that it eventually met the claimed 
specifications and became a valuable tool for upper ocean thermal and salinity investigation 
meeting the requirements of the oceanographic community.