The WSTP tool was deployed to collect in-situ temperature data at 55.2 mbsf in Hole 897A and at 117.9, 166.1, and 214.4 mbsf in Hole 897C (WSTP depth is based on the depth of penetration of the bottom of the previous core barrel). We did not use the measurement at 117.9 mbsf because the probe was apparently inserted into fill at the bottom of the hole. The tool deployed at 166.1 mbsf experienced an electronic failure. We judged that only the data from 55.2 and 214.4 mbsf warranted further analysis.
Interpretations of data from the tool deployed at 55.2 mbsf yielded a bottom-water temperature of 3.6°±0.1°C and an in-situ temperature of 7.3°±0.8°C (error temperatures are informed guesses; Fig. 67). The bottom-water temperature was obtained by averaging temperature readings between 2900 and 3100 s, when the tool was stopped at the seafloor. The in-situ temperature was extrapolated using 4753 s as the insertion time and modeling the data over the interval from 4770 to 4982 s. The recorded temperatures indicate that the tool was inserted at least twice, originally at about 4400 s and again at about 5900 s. The tool also may have been moving during the measurements, which may have introduced errors because of frictional heating of the probe. We chose to model the data over the interval prior to the second penetration, when motion apparently had stopped. Because of these problems, we have assumed that a larger than normal error range is appropriate for this in-situ temperature measurement.
Analyses of the measurements at 214.4 mbsf yielded a bottom-water temperature of 3.6°±0.1C° and an in-situ temperature of 13.1°±0.1C° (Fig. 68). The bottom-water temperature was obtained by averaging temperature readings between 2200 and 2400 s, when the tool had stopped at the seafloor. The in-situ temperature was extrapolated using 3168 s as the insertion time and modeling the data over the interval from 3233 to 4102 s. The shape of the temperature curve suggests that the tool was inserted only once and that the tool was stationary during the measurement, which suggests that this measurement is more reliable than the one at 55.2 mbsf.
The slope of a linear least-squares fit of the temperature to depth (Table 19) yields an estimate of 43 ± 10 mK/m (95% confidence level) for the temperature gradient in the upper 215 mbsf at Site 897. The slope of a linear least-squares fit of the temperature to vertically integrated thermal resistivity (Table 19) yields an estimate of 54 ± 13 mW/m2 (95% confidence level) for the heat flow (see "Explanatory Notes" chapter, this volume).