The impact of splay faults on fluid flow, solute transport, and pore pressure distribution in subduction zones: A case study offshore the Nicoya Peninsula, Costa Rica

Abstract

Observations of seafloor seeps on the continental slope of many subduction zones illustrate that splay faults represent a primary hydraulic connection to the plate boundary at depth, carry deeply sourced fluids to the seafloor, and are in some cases associated with mud volcanoes. However, the role of these structures in forearc hydrogeology remains poorly quantified. We use a 2-D numerical model that simulates coupled fluid flow and solute transport driven by fluid sources from tectonically driven compaction and smectite transformation to investigate the effects of permeable splay faults on solute transport and pore pressure distribution. We focus on the Nicoya margin of Costa Rica as a case study, where previous modeling and field studies constrain flow rates, thermal structure, and margin geology. In our simulations, splay faults accommodate up to 33% of the total dewatering flux, primarily along faults that outcrop within 25 km of the trench. The distribution and fate of dehydration-derived fluids is strongly dependent on thermal structure, which determines the locus of smectite transformation. In simulations of a cold end-member margin, smectite transformation initiates 30 km from the trench, and 64% of the dehydration-derived fluids are intercepted by splay faults and carried to the middle and upper slope, rather than exiting at the trench. For a warm end-member, smectite transformation initiates 7 km from the trench, and the associated fluids are primarily transmitted to the trench via the decollement (50%), and faults intercept only 21% of these fluids. For a wide range of splay fault permeabilities, simulated fluid pressures are near lithostatic where the faults intersect overlying slope sediments, providing a viable mechanism for the formation of mud volcanoes.

Coupled Evolution of Deformation, Pore Fluid Pressure, and Fluid Flow in Shallow Subduction Forearcs

“…In these models, upper plate fault‐zone k ranges from 10 −13 –10 −14 m 2 at the seafloor to 10 −16 –10 −17 m 2 where they intersect the décollement at depth. This range of fault permeability is consistent with previous estimates based on in situ single‐ or across‐borehole measurements (Screaton et al, 2000; Fisher & Zwart, 1996, 1997; Kinoshita & Saffer, 2018) and modeling studies of thermal and chemical budgets (Lauer & Saffer, 2012, 2015; Saffer, 2015; Spinelli et al, 2006 and references therein). Modeled flow rates across the seafloor vary over three orders of magnitude, with the peak surface seepage rates (up to 10 cm yr −1 ) corresponding to fault locations.…”

“…Deformation and faulting, in turn, influence fault and sediment permeability—and therefore affect drainage and pore pressure—as well as the evolution of stress and mechanical loading. For example, interconnected fractures near fault zones form effective pathways that accommodate volatile fluxes and focus fluid seepage (drainage), enhancing the rates of consolidation and dewatering at depth (e.g., Lauer & Saffer, 2012, 2015) and potentially affecting megathrust slip behavior within the seismogenic and tremor zones (Halpaap et al, 2019; Ranero et al, 2008; Wells et al, 2017). In this study, we systematically investigate these processes and their fundamental feedbacks, using a numerical approach that fully couples mechanical loading, deformation, and fluid drainage.…”

“…Some studies suggest the likely widespread presence of fluid overpressure at many subduction margins (e.g., Gamage & Screaton, 2006; Saffer & Bekins, 1998; Screaton et al, 1990), as an outcome of rapid mechanical loading that outpaces drainage, together with fluid release from low‐temperature diagenetic and metamorphic reactions (Saffer & Tobin, 2011). Other studies have highlighted the role of interplate and intraplate faults as permeable conduits in governing distributions of fluid pressure and surface seepage (e.g., Ellis et al, 2015; Lauer & Saffer, 2012, 2015). These existing studies typically did not fully couple mechanical and hydrological processes; instead, they either prescribed a steady‐state porosity field and fixed sediment flux trajectories (e.g., Bekins & Dreiss, 1992; Saffer & Bekins, 1998; Screaton et al, 1990; Wang, 1994) or used a simplified representation of in situ stress state to approximately account for mechanical loading (e.g., Screaton & Saffer, 2005; Shi & Wang, 1988; Skarbek & Saffer, 2009; Stauffer & Bekins, 2001) (see Text S1 in the supporting information for a summary of previous modeling studies).…”

Thermal environment of the Southern Washington region of the Cascadia subduction zone

“…With high temperatures occurring at shallow depths within Cascadia, temperature‐dependent sediment dehydration reactions will initiate further updip than on other subduction zone margins [ Lauer and Saffer , ]. The relatively shallow depth of these strongly temperature‐dependent reactions will result in the generation of unbound fluid within the accretionary sediments of the offshore section of the accretionary wedge.…”

Anomalous Concentration of Methane Emissions at the Continental Shelf Edge of the Northern Cascadia Margin

“…Listric and normal faults likely play a similar role in transporting fluid from deep within the accretionary wedge on the upper Cascadia margin as they do in other subduction zones (e.g., Hensen et al, ; Melnick et al ; Moore & Saffer, ). These quasi‐vertical faults have been shown to be conduits for gas, fluid, and solute, transporting a high percentage of the total dewatering flux from deep in the accretionary wedge to the overlying seafloor, and in doing so, weaken and segment the upper plate (Lauer & Saffer, ). Although the MCS reflections we are interpreting as listric and normal faults are not visible in all of the archive along‐strike MCS images, most profiles demonstrate that the uppermost margin and western continental shelf edge is a zone that has experienced extension, including placement of subsurface diapirs and tectonic disturbances that are co‐located with active bubble plume emissions (Figures a, a, b, and S2–S12 in the supporting information).…”