The procedures used in the Synoptic Ocean Prediction experiment for ca
libration of inverted echo sounder (IES) travel time tau data have led
to substantially improved accuracy. In previous work, tau was convert
ed to main thermocline depth Z(tau) using point measurements of the 12
degrees C depth from coincident expendable bathythermograph (XBT) cas
ts. The new method presented in this study vertically integrates the X
BT temperature to produce the quantity Q(T), which is very tightly rel
ated to tau. The advantage of this method is that the calibration cons
tants have roughly half the uncertainty obtained with the traditional
method. Seasonal changes in tau determined from hydrographic data in t
he Gulf Stream region were found to vary with geographic region; near
68 degrees W tau changed by 1.8 ms from March to September, whereas ne
ar Cape Hatteras the annual change was only 1.0 ms. The fully calibrat
ed IES Z(T) has an estimated error of 19 m (one standard deviation).
An iterative optimal interpolation scheme is described, by which the G
ulf Stream thermocline depth was mapped on a two-dimensional grid. Ini
tially, daily maps were generated using established mapping techniques
. Subsequently, these maps were averaged for 31-day periods and in tur
n used as the new ''mean fields'' for the final maps. Verification of
the mapped fields against independent XBT data gives rms differences o
f 31-46 m, which is only 3%-5% of the 900-m range of the observed Z(T)
.