INTRODUCTION

Stable isotope analyses of pore waters are important for gas hydrate studies because both the formation and dissociation of hydrates are associated with isotope fractionation effects. Chemical-potential gradients generated by hydrate formation and melting will also give rise to diffusion, which is another process that causes isotope fractionation. Shore-based isotope studies of interstitial waters from hydrate-bearing sediments can thus provide significant additional information to the major-element chemistry largely determined on board the drill ship.

Chlorine stable isotopes are useful as tracers of solute transport by advection and/or diffusion. Because chloride rarely seems to participate in chemical reactions during diagenesis, physical transport processes may become traceable isotopically, if pore water with a distinct isotopic source signal is involved. During hydrate formation, the heavy isotopes of oxygen and hydrogen are enriched in the solid phase; the pore waters are depleted correspondingly. Conversely, dissociation of hydrate at the base of the hydrate stability zone causes the pore-water values for ¹⁸O and D to increase due to release of water enriched in the heavy isotopes. Hydrate melting as an artifact of the sampling process has the same effect. Heavy oxygen and hydrogen isotope enrichment of the pore water is typically associated with lowered chlorinity, and vice versa. Pore-water salting results from salt exclusion during crystallization of hydrates, whereas freshening results from freshwater release during melting, and these effects parallel the oxygen and hydrogen isotopic changes during hydrate formation and dissociation. The isotopic effects can thus aid to identify the presence of hydrates and provide insight into the processes of hydrate formation and dissociation, although the hydrate-related signatures can be overprinted by other fractionation effects related to diagenesis, such as alteration of volcanic glass to zeolites or clay minerals and other hydration reactions (Perry et al., 1976), and are therefore not unambiguous. Connate-water variations in isotopic composition can also be inherited from seawater isotopic fluctuations during glacial/interglacial cycles (McDuff, 1985) and can overprint hydrate-related signatures. A coupled anomaly of downward increasing oxygen isotope (and hydrogen isotope) ratios and decreasing chlorinity in drill hole samples invariably attests to the presence of hydrate. No other geochemical processes are known to produce the same coupled signals (Hesse, 1990). The reverse, obviously, is not always true due to the possible overprints mentioned. Consequently, not all hydrate zones show a positive O-isotope (or D) anomaly coupled with a chloride decrease in their pore waters.

The ultimate aim then is to separate isotopic and chemical effects related to hydrate formation and hydrate decomposition from diagenetic effects and from those of diffusion and advection and use the hydrate-related pore-water signatures to estimate hydrate concentration in the pore space. This has not been possible previously due to the complex interaction of the various processes referred to above, thus one of the aims of Ocean Drilling Program (ODP) Leg 164 was to tackle this problem. Egeberg and Dickens (1999) have developed a mathematical model that successfully simulates the measured chlorinity for Site 997 using advection as transport mechanism counteracted by diffusion. In the present paper the model is tested with chlorine isotope data.