Figure F5. A. Amplitude of Horizon A. Gas saturation of 50%–90% of pore space is inferred from low in situ density values measured with LWD at Sites 1245, 1247, 1248, and 1250. White lines indicate the depth of Horizon A below sea level. The onset of large-amplitude anomalies along Horizon A is interpreted to indicate the base of a connected free gas column. Although Horizon A continues to the west of where it intersects the bottom-simulating reflector (BSR), its amplitude decreases abruptly and it cannot be unambiguously identified. B. Lithostatic stress (sigmaL) compared to gas pressure (Pg) in Horizon A. Assuming that gas pressure is in equilibrium with hydrostatic pressure (Pw) at the base of the connected gas column, gas pressure in Horizon A exceeds the lithostatic stress at Site 1250. C. Effective gas stress on Horizon A. Topographic contours at a 20-m interval are also shown (black lines). Effective gas stress is negative beneath most of the summit. (A–C adapted from Tréhu et al., 2004a.) D. In situ bulk density (from LWD) and grain size distribution at Site 1245, which illustrate the anomalous nature of Horizon A. E. C1/C2 ratio and isotopic composition of methane at Sites 1245 and 1247, which shows anomalous geochemistry (low C1/C2; high delta13C) associated with Horizon A. Gas chemistry within the GHSZ is characteristic of biogenic gas. F. C1/C2 ratio and isotopic composition of methane at the summit. Gas chemistry indicates migrated gas both in Horizon A and in the upper 20–30 mbsf. The BSR is located at approximately the same depth at all three sites shown here. (D adapted from Tréhu et al., 2004a; E and F adapted from Claypool et al., this volume.) G. P- and S-wave velocities from sonic logs and a vertical seismic profile (VSP). Horizon A appears as a pronounced low-velocity zone in VS and VP (VSP only). The absence of a VP anomaly in the sonic log data may result from intrusion of drilling fluid near the borehole. The VS anomaly is unexpected, since VS is not sensitive to whether the pore space is occupied by water or gas, and suggests that overpressures are large enough to affect the sediment shear strength. H. Increase in effective stress (sigmae) with depth, as calculated from sediment bulk density at Site 1249 compared to the internal pressures of gas hydrate crystallite (Ph) and gas bubbles (Pg) calculated following Clennell et al. (1999) assuming an average pore radius of 0.5 mm, compared to Cl– enhancement and gas hydrate saturation predicted by a one-dimensional transport reaction model with enhanced gas dissolution and hydrate precipitation rates in near-surface sediments. Model parameters are described in Torres et al. (2004b). Cl– concentrations measured at Site 1249 (squares) and IR anomalies at Sites 1249 (green bars) and 1250 (blue bars) are also shown. I. Seismic cross section at the summit showing the steady-state gas migration path suggested by Liu and Flemings (2006) and Milkov and Xu (2005) as a green line, compared to two possible paths (from among a wide range of possible paths) in the distributed, temporally variable scenario hypothesized by Tréhu et al. (2004b), Torres et al. (2004b), and Weinberger and Brown (2006). (G adapted from Tréhu et al., this volume; H adapted from Torres et al., 2004b; I adapted from Tréhu et al., 2004a).

Figure F5. A. Amplitude of Horizon A. Gas saturation of 50%–90% of pore space is inferred from low in situ density values measured with LWD at Sites 1245, 1247, 1248, and 1250. White lines indicate the depth of Horizon A below sea level. The onset of large-amplitude anomalies along Horizon A is interpreted to indicate the base of a connected free gas column. Although Horizon A continues to the west of where it intersects the bottom-simulating reflector (BSR), its amplitude decreases abruptly and it cannot be unambiguously identified. B. Lithostatic stress (

_L) compared to gas pressure (P_g) in Horizon A. Assuming that gas pressure is in equilibrium with hydrostatic pressure (P_w) at the base of the connected gas column, gas pressure in Horizon A exceeds the lithostatic stress at Site 1250. C. Effective gas stress on Horizon A. Topographic contours at a 20-m interval are also shown (black lines). Effective gas stress is negative beneath most of the summit. (A–C adapted from Tréhu et al., 2004a.) D. In situ bulk density (from LWD) and grain size distribution at Site 1245, which illustrate the anomalous nature of Horizon A. E. C₁/C₂ ratio and isotopic composition of methane at Sites 1245 and 1247, which shows anomalous geochemistry (low C₁/C₂; high

¹³C) associated with Horizon A. Gas chemistry within the GHSZ is characteristic of biogenic gas. F. C₁/C₂ ratio and isotopic composition of methane at the summit. Gas chemistry indicates migrated gas both in Horizon A and in the upper 20–30 mbsf. The BSR is located at approximately the same depth at all three sites shown here. (D adapted from Tréhu et al., 2004a; E and F adapted from Claypool et al., this volume.) G. P- and S-wave velocities from sonic logs and a vertical seismic profile (VSP). Horizon A appears as a pronounced low-velocity zone in V_S and V_P (VSP only). The absence of a V_P anomaly in the sonic log data may result from intrusion of drilling fluid near the borehole. The V_S anomaly is unexpected, since V_S is not sensitive to whether the pore space is occupied by water or gas, and suggests that overpressures are large enough to affect the sediment shear strength. H. Increase in effective stress (

_e) with depth, as calculated from sediment bulk density at Site 1249 compared to the internal pressures of gas hydrate crystallite (P_h) and gas bubbles (P_g) calculated following Clennell et al. (1999) assuming an average pore radius of 0.5 mm, compared to Cl^– enhancement and gas hydrate saturation predicted by a one-dimensional transport reaction model with enhanced gas dissolution and hydrate precipitation rates in near-surface sediments. Model parameters are described in Torres et al. (2004b). Cl^– concentrations measured at Site 1249 (squares) and IR anomalies at Sites 1249 (green bars) and 1250 (blue bars) are also shown. I. Seismic cross section at the summit showing the steady-state gas migration path suggested by Liu and Flemings (2006) and Milkov and Xu (2005) as a green line, compared to two possible paths (from among a wide range of possible paths) in the distributed, temporally variable scenario hypothesized by Tréhu et al. (2004b), Torres et al. (2004b), and Weinberger and Brown (2006). (G adapted from Tréhu et al., this volume; H adapted from Torres et al., 2004b; I adapted from Tréhu et al., 2004a).