Implications of Measured Thermal Gradients for Water Movement Through the Northeast Japan Accretionary Prism

Abstract

Thermal gradients of 24°–36°C/km have been found in the slope cover sediments of the accretionary prism off northeast Japan (Deep Sea Drilling Project (DSDP) sites 438A, 439, and 440). These gradients lead to the prediction of temperatures in excess of 400°C at depths of 15 km using an increase in thermal conductivity with depth due to the decrease in porosity. Such temperatures conflict with the predictions of temperatures of 150°C at even greater depths from conductive thermal models and petrologic studies of low‐temperature blueschists at similar ancient margins. One‐dimensional, steady state conduction and advection via Darcian water flow, of heat, up through the prism was simulated numerically. The measured thermal gradients can be reconciled with temperatures less than 125°C at the prism base by water flowing upward at rates of the order of 10−10 m/s. Such velocities imply that the total volume of water available from compaction of sediments subducted beneath the northeast Japan prism would flow up through a <10 km width of the prism (perpendicular to the trench) and would require the DSDP sites to have been located fortuitously over the localized zone(s) of water flow. Smaller velocities will reproduce the measured gradients, and thus greater widths through which water flow occurs are predicted, if the temperature at the prism base is higher and/or if the prism is thinner and a subduction channel several kilometers high separates the accreted sediments from the descending plate. Where water flow is significant, as likely wherever abundant, water‐rich sediments are subducted, thermal gradients below a few kilometers in accretionary prisms may be much lower than near‐surface values. This has profound implications for the thermal structure of the prism off northeast Japan and elsewhere.

Thermal maturity and constraints on deformation from illite crystallinity and vitrinite reflectance in the shallow levels of an accretionary prism: Eocene-Oligocene Shimanto Belt, southwest Japan

“…Such a gradient seems rather high when compared to some models of the thermal structure of accretionary prisms that indi-cate much lower geothermal gradients of 10 to 15°C/km (e.g., Oxburgh and Turcotte, 1970;Wang and Shi, 1984;van den Beukel and Wortel, 1988). However, recent studies have shown that geothermal gradients in active prisms, at least in the toe region, may be much higher than previously thought due to the important role of advective heat transfer by fluid flow (e.g., Davis and Langseth, 1986;Reck, 1987;Moore and others, 1988). Studies of heat flow at active margins using the gas hydrate method as well as direct temperature measurements in boreholes show that many accretionary prisms are characterized by geothermal gradients of from 30 to 60°C/km or more (e.g., Yamano and others, 1982; and see Table 1 in Underwood, Hibbard, and DiTullio, this volume).…”

Sediment Geochemistry, Clay Mineralogy, and Diagenesis: A Synthesis of Data from Leg 131, Nankai Trough

“…Many subduction zones are characterized by relatively low geothermal gradients due to the subduction of old oceanic lithosphere (e.g., Yamano et al, 1982;Reck, 1987). Under these typical circum-stances, the documentation of important diagenetic reactions is impeded because the critical temperature conditions occur beyond the depth limits of conventional Deep Sea Drilling Project (DSDP)/ODP boreholes.…”

The thermal imprint of spreading ridge subduction on the upper structural levels of an accretionary prism, southwest Japan

“…One problem with a straightforward determination of the geothermal gradient based on limited data (i.e., temperature/depth and no control on thermal conductivity) is the necessary assumption that the gradient is constant with depth. Also, recent modeling suggests that geothermal gradients are linear for middle to lower depths of accretionary prisms, but this appears untenable for upper structural levels subject to fluid advection (Wang and Shi, 1984;Reck, 1987). With these precaution noted, we present paleogeothermal gradient calculations for the Nabae subbelt only as first approximations of the actual thermal system.…”

“…However, these models can only be considered as rough approximations because the thermal structure of accretionary prisms appears to result from the com-plex interplay of many factors, including radiogenic heating, accretion and erosion in the prism, internal strain heating, frictional heating on the décollement, and advection by fluid flow, in addition to heat conduction from the mantle and the velocity of subduction (Wang and Shi, 1984;Reck, 1987;Dahlen and Barr, 1989). While it is intuitive that the first-order effect of spreading ridge subduction is to increase conductive heat flow in the overriding prism, the effects of such elevated heat flow on other processes, such as frictional heating on the décollement and fluid flow are unknown.…”