Calculated [CH4]0 values (Table T1; Fig. F1) are given as the mean calculated from equations 7 and 8, [CH4]0mean, because total pressure and Ar were not measured (see previous sections). The error bars are the limits given by equations 7 and 8 and thus are conservative estimates of the uncertainty. Reproducibility was examined by collecting duplicate samples from two core sections. The average deviation from the mean was 10% for [CH4]0 based on either the low or high limit. Of the 23 samples collected, 4 were contaminated with air to an extent that resulted in an uncertainty of >30% using an assumed 5% analytical precision for each gas.
The [CH4]0 values we report, [CH4]0mean, are the mean values based on equations 7 and 8. [CH4]0mean values vary from 36 to 75 mM/kg between 36 and 66 mbsf.
At greater depths, concentrations are more scattered. Much of the scatter at these greater depths is presumably due to the nonuniform distribution of methane hydrate. Chloride concentrations at this site become scattered to lower values at depths >20 mbsf, indicating the presence of methane hydrate that decomposed during core recovery. The additional CH4 added to the fluid due to hydrate decomposition can be estimated based on the chloride anomaly at a given depth and assuming a H2O to CH4 ratio of 5.9 (Davidson, 1983). Between 35 and 66 mbsf, the additional CH4 due to hydrate decomposition is only ~1–3 mmoles/kg, <13% of the total [CH4]0. This can be taken as a maximum, as not all of the hydrate decomposed prior to the collection of the vapor phase in the core liner. Below this depth, the greater scatter in the chloride data indicates locally higher abundances of hydrate, indicating potentially large contributions to the measured CH4.
These results can be compared to the CH4 concentrations determined using the PCS and to the methane hydrate saturation boundary expected for in situ pressure and temperature. For PCS samples collected in the same depth interval at Site 1230, there is a general agreement with the PCS data (Fig. F1) although a detailed comparison is not possible because the samples were not collected from the exact same depth intervals and the nonuniform distribution of hydrate results in scatter (Dickens et al., 2003).
Davie et al. (2004) reported a method for estimating CH4 solubility with respect to hydrate for seafloor conditions. However, Davie et al. (2004) presented data for pressures up to only 50 MPa, whereas pressures at Site 1230 exceed this number. We have extrapolated their data for use in this study. Based on this extrapolation, and the geotherm at Site 1230, we have plotted the methane hydrate solubility in Figure F1. Between 35 and 66 mbsf, the estimated saturation CH4 concentration varies from ~45 to 49 mM/kg. Our calculated values span this concentration range over this depth interval. Concentrations in excess of those based on solubility can be used to infer the presence of hydrate, and, if the hydrate completely decomposes prior to sampling, its abundance can be quantified.
In summary, in situ dissolved CH4 concentrations can be determined, with good reproducibility, based on the analysis of exsolved vapor phase CH4 and N2 in core liners. Atmospheric contamination can be quantified, and therefore raw data can be quality controlled. Preliminary testing of this method results in values that are consistent with PCS-based measurements and expectations based on methane phase relationships. The method may be further refined and tested, if Ar and/or total pressure are also determined.