Calculation of a new depth equations for expendable bathythermographs using a temperatureerrorfree method (Application to Sippican/T SK T7, T6 and T4 XBTs)
Hanawa, K.; Rual, P.; Bailey, R.; Sy, A.; Szabados, M. (1994). Calculation of a new depth equations for expendable bathythermographs using a temperatureerrorfree method (Application to Sippican/T SK T7, T6 and T4 XBTs). UNESCO Technical Papers in Marine Science = Documents techniques de l'Unesco sur les sciences de la mer, 67. UNESCO: Paris. x, 30 + appendices pp.
Part of: UNESCO Technical Papers in Marine Science = Documents techniques de l'Unesco sur les sciences de la mer. UNESCO: Paris. ISSN 05034299; eISSN 02515024, more
 
Authors   Top 
 Hanawa, K.
 Rual, P.
 Bailey, R.



Abstract 
A new depthtime equation for Sippican and TsurumiSeiki (TSK) T7, T6, and T4 type expendable bathythermographs (XBTS ) is presented based on the results of an internationally coordinated set of controlled XBTCTD comparison experiments geographically distributed over as many different oceanic water masses as possible. A newly developed temperatureerrorfree method is applied to the data set to obtain the timedepth pairs necessary for the calculation of the individual depthtime equations by the method of least squares. The accuracies in depth were found in general to be outside of the manufacturers’ specified accuracies. The mean deptherror for the T7 type of probe was found to be about +25m at 750m, whereas the manufacturers’ depthaccuracy specification at 750m is only ±15m.Since the T4/T6 and the T7 data sets were found not to be statistically different at the 95% confidence level, a unique new SippicanTSK T4/T6/T7 depthtime equation, whose coefficients are the means of the individual depthtime equation coefficients, is given (at a 95% confidence level) by: Z= (6.691 ± 0.021) t  (2.25+ 0.30) 10^{3} t^{2},where Z is the new corrected XBT depth in metres at the elapsed time t in seconds. It replaces the manufacturers’ equation ( z_{m}= 6.472 t 22.16 10^{3} t^{2}). The new equation reduces the mean deptherrors to less than ±lm in the range 0800m. However the individual scatter of the probes is much larger than the manufacturers’ specifications. The observed scatter at 3 standard deviations (99.7%) is approximately ±10m at 100m and ± 25m at 800m. The relationship between the scatter of the individual quadratic depthtime equation coefficients and the deptherror is discussed. It is shown that depthdifferences between respective depthtime equations can be small, even if the two coefficients of the equations have apparently very different values. The effect of the region, of the onboard equipment or of the manufacturer, is also discussed and generally found not to be statistically significant, i.e. smaller than the above probe to probe scatter. An approximate linear correction formula is determined for correcting (with a maximum error of ±0.lm) the depths recorded using the manufacturers’ original depthtime equation (z_{m}). It is given by : Z_{l} = 1.0336 Z_{m}.Until an international mechanism is established to implement the general use of the new equation, it is of the utmost importance not to use the new T4/T6/T7 equation when archiving or exchanging XBT data. The mixing of data in the data archives must be absolutely avoided. 
