Citation
The Effect of Incoming Sedimentation Changes on Pore Pressure Development and Dehydration Reaction Progress in the Aleutian Margin

Material Information

Title:
The Effect of Incoming Sedimentation Changes on Pore Pressure Development and Dehydration Reaction Progress in the Aleutian Margin
Creator:
Meridth, Lanie N
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (118 p.)

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Geology
Geological Sciences
Committee Chair:
SCREATON,ELIZABETH JANE
Committee Co-Chair:
JAEGER,JOHN M
Committee Members:
MARTIN,ELLEN ECKELS
Graduation Date:
4/30/2016

Subjects

Subjects / Keywords:
Dehydration ( jstor )
Drilling ( jstor )
Geologic deformation ( jstor )
Modeling ( jstor )
Opal ( jstor )
Porosity ( jstor )
Quartz ( jstor )
Sediments ( jstor )
Smectite ( jstor )
Subduction ( jstor )
Geological Sciences -- Dissertations, Academic -- UF
dehydration -- hydrology -- opal -- smectite -- subduction
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Geology thesis, M.S.

Notes

Abstract:
Sediment inputs to subduction zones impart a significant control on hydrologic and tectonic behavior during subduction. Rapid sediment accumulation (>1 km/my) in the Aleutian Trench due to intensified glaciation increases overburden and should accelerate dehydration of hydrous minerals by elevating temperatures in the incoming sediment column. These processes have the potential to generate fluid overpressures in the low permeability incoming sediments to the Aleutian Trench. Mineralogical analyses on incoming sediments show that both smectite and opal-A are present as hydrous mineral phases. A 1-D sedimentation model was developed to track dehydration reaction progress and increment excess pore pressures in the incoming sediments to the Aleutian Trench. Temperatures increase due to the insulating effect of trench sediments. As a result, opal-A dehydrates to quartz at the deformation front while smectite remains mostly unreacted. Excess pore pressures increase at the deformation front due to loading of trench sediments. The 1-D modeling results at the deformation front were incorporated into a subduction model that tracks the incoming sediments into the subduction zone. Excess pore pressures approach ~70% of lithostatic pressure by 60 km landward of the deformation front. Modeling results suggest that dehydration reactions are not a significant component of excess pore pressure generation, likely because the reactions occur in the same region that is most affected by loading. In other margins, such as Barbados and the Japan Trench, dehydration is shifted significantly landward relative to Alaska, and could have a greater effect on excess pore pressure generation during subduction. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (M.S.)--University of Florida, 2016.
Local:
Adviser: SCREATON,ELIZABETH JANE.
Local:
Co-adviser: JAEGER,JOHN M.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2017-05-31
Statement of Responsibility:
by Lanie N Meridth.

Record Information

Source Institution:
UFRGP
Rights Management:
Copyright Lanie N Meridth. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
5/31/2017
Classification:
LD1780 2016 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

THE EFFECT OF INCOMING SEDIMENTATION CHANGES ON PORE PRESSURE DEVELOPMENT AND DEHYDRATION REACTION PROGRESS IN THE ALEUTIAN MARGIN By LANIE MERIDTH A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2016

PAGE 2

2016 Lanie Meridth

PAGE 3

To my family and friends for their never ending support. And for my grandfather, the biggest Gator fan there ever was

PAGE 4

4 ACKNOWLEDGMENTS I thank my fellow Department of Geological Science s graduate students, the faculty and my advisor, Dr Elizabeth Screaton, for her encouragement and continual support throughout my studies. I would also like to thank my parents and friends for supporting me and keeping me sane through graduate school. I would like to thank my sister, Jennifer, and my room mate, Andrea, for always being there and for their great taste in music. I would like to thank my loving boyfriend, David, for his encouragement, love and sense of humor.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 7 LIST OF FIGURE S ................................ ................................ ................................ ......................... 8 ABSTRACT ................................ ................................ ................................ ................................ ... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 12 2 BACKGROUND ................................ ................................ ................................ .................... 20 Study Area ................................ ................................ ................................ .............................. 20 Incoming Sediments ................................ ................................ ................................ ............... 23 Trench Sedimentation ................................ ................................ ................................ ............. 25 Accretionary Prism ................................ ................................ ................................ ................. 27 3 METHODS ................................ ................................ ................................ ............................. 32 Model Overview ................................ ................................ ................................ ..................... 32 1 D Modeling Approach ................................ ................................ ................................ ......... 33 Fluid Flow and Pore Pressure Simulation ................................ ................................ ....... 33 Rate of Fluid Release from Dehydration ................................ ................................ ......... 35 Sedimentation and Porosity Evolution ................................ ................................ ............ 36 Permeability Evolution ................................ ................................ ................................ .... 39 Decollement Zone Permeability ................................ ................................ ...................... 40 Temperature Evolution ................................ ................................ ................................ .... 40 Dehydr ation Reaction Progress ................................ ................................ ....................... 43 Benchmarking the 1 D MATLAB Code ................................ ................................ ................ 45 Sample Analysis ................................ ................................ ................................ ..................... 47 Biogenic Silica ................................ ................................ ................................ ................. 48 Mixed Layer Illite Smectite Clay ................................ ................................ .................... 48 Application of the Code to the Aleutian Trench ................................ ................................ ..... 49 Model Parameters ................................ ................................ ................................ ................... 50 4 1 D MODELING RESULTS ................................ ................................ ................................ 59 Seaward of the Deformation Front ................................ ................................ ......................... 59 Influx of Water at the Site U1417 Transect ................................ ................................ ..... 61 Sensitivity Analyses ................................ ................................ ................................ ........ 62 Trench sedimentation rate ................................ ................................ ........................ 62 Initial smectite mole fraction ................................ ................................ .................... 63

PAGE 6

6 Incoming permeability porosity relationship ................................ ........................... 64 Deformation Front to 60 Kilometers Landward ................................ ................................ ..... 65 Sensitivity Analyses ................................ ................................ ................................ ................ 68 Temperature ................................ ................................ ................................ ..................... 68 Permeability Porosity Relationship ................................ ................................ ................. 69 5 DISCUSSION ................................ ................................ ................................ ......................... 91 Comparison to Other Subduction Margins ................................ ................................ ............. 95 Nankai Margin, Japan ................................ ................................ ................................ ...... 95 Costa Rica Margin (Nicoya) ................................ ................................ ............................ 97 Ba rbados Accretionary Prism ................................ ................................ .......................... 99 Japan Trench, Tohoku ................................ ................................ ................................ ..... 99 Implications ................................ ................................ ................................ .......................... 101 6 CONCLUSIONS ................................ ................................ ................................ .................. 105 LIST OF REFERENCES ................................ ................................ ................................ ............. 109 BIOGRAPH ICAL SKETCH ................................ ................................ ................................ ....... 118

PAGE 7

7 LIST OF TABLES Table page 3 1 Site U1417 incoming sediment packet characterization. ................................ ................... 55 3 2 Trench packet characterization. ................................ ................................ ......................... 55 4 1 Fluid influx rate from pore water, smectite and opal dehydration at Site U1417. ............. 71 4 2 Fluid influx rate from pore water, smectite and opal dehydration at the deformation front ................................ ................................ ................................ ................................ .... 71

PAGE 8

8 LIST OF FIGURES Figure page 1 1 Geographic and tectonic setting of the Aleutian margin ................................ ................... 19 2 1 Incoming section at Site U1417. ................................ ................................ ........................ 29 2 2 Shipboard data from IODP Site U1417 and DSDP Site 178 ................................ ............. 29 2 3 Conceptual model of the Aleutian margin ................................ ................................ ......... 30 2 4 Grain size analysis on sediments recovered from Site 180 ................................ ................ 30 2 5 Seismic profile view of Aleutian margin ................................ ................................ ........... 31 3 1 1 D modeling steps. ................................ ................................ ................................ ........... 56 3 2 Conceptual diagram of 1 D sedimentation and subduction model ................................ .... 56 3 3 Benchmark modeling results of the Nankai Trough ................................ .......................... 57 3 4 Sequence III thickness from Site U1417 to the deformation front ................................ .... 57 3 5 Trench sedimentation rates based on Sequence III thickness data ................................ .... 58 3 6 Site U1417 sample analysis ................................ ................................ ............................... 58 4 1 Modeling results at Site U1417 plotted with IODP Exp edition 341 shipboard and drilling data ................................ ................................ ................................ ........................ 72 4 2 Dehydration reaction progress of smectite and opal at Site U1417 ................................ ... 72 4 3 Modeling results with the deposition of trench sediments at the deformation front ......... 73 4 4 Dehydration reaction progress with the deposition of trench sediments ........................... 73 4 5 Excess pore pressure distribution with the deposition of trench sediments ....................... 74 4 6 Sensit ivity to trench sedimentation rate ................................ ................................ ............. 74 4 7 Sensitivity of excess pore pressure to trench sedimentation rate ................................ ....... 75 4 8 Sensitivity to smectite content for the trench wedge packets ................................ ............ 75 4 9 Sensitivity to incoming k n relationship ................................ ................................ ............ 76 4 10 Sensitivity to trench k n relationship. ................................ ................................ ................ 76

PAGE 9

9 4 11 Depth of potential decollement locations in the incoming section at the deformation front and 60 km landward. ................................ ................................ ................................ 77 4 12 Three decollement zone depth scenarios used in the subduction model base run ............. 77 4 13 Porosity with subduction distance for each decollement location. ................................ .... 78 4 14 Temperature with subduction distance for each decollement zone. ................................ .. 78 4 15 Smectite dehydration reaction progress and fluid production during subduction ............. 79 4 16 Opal dehydration reaction progress and fluid production during subduction .................... 80 4 17 Excess pore pressure distribution during subduction for each decollement location. ....... 81 4 18 Effective stress distribution during subduction for each decollement location. ................ 82 4 19 Excess pore pressure dist ribution without dehydration included as a fluid source ........... 83 4 20 Effective stress distribution without dehydration included as a fluid source. ................... 84 4 21 Sensitivity to temperature gradient with subduction distance (dT/dx). ............................. 84 4 22 Sensitivity of smectite reaction progress and fluid production to temperature ................. 85 4 23 Sensitivity of opal dehydration reaction progress and fluid production to temperature .... 86 4 24 Sensitivity of excess pore pressure to temperature ................................ ............................ 87 4 25 Sensitivity of effective stress to temperature. ................................ ................................ .... 88 4 26 Excess pore pressure distribution assuming k n relationship determined from Site U1417 sample analysis ................................ ................................ ................................ ...... 89 4 27 Effective stress distribution assuming k n determined from Site U1417 sample analysis. ................................ ................................ ................................ .............................. 90 5 1 Simulated porosity compared to the empirical porosity velocity relationship determined by von Huene et al. (1998) ................................ ................................ ............ 103 5 2 Locations of subduction margins used for dehydration reaction comparison ................. 103 5 3 Distance landward of deformation front along the decollement of projected peak fluid release from smectite dehydration based on previous modeling studies ................. 104

PAGE 10

10 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degre e of Master of Science THE EFFECT OF INCOMING SEDIMENTATION CHANGES ON PORE PRESSURE DEVELOPMENT AND DEHYDRATION REACTION PROGRESS IN THE ALEUTIAN MARGIN By Lanie Meridth May 2016 Chair: Elizabeth Screaton Major: Geology Sediment inputs to subduction zones impart a significant control on hydrologic and tectonic behavior during subduction. Rapid sediment accumulation (>1 km/my) in the Aleutian Trench due to intensified glaciation increases overburden and should accelerate d ehydration of hydrous minerals by elevating temperatures in the incoming sediment column. These processes have the potential to generate fluid overpressures in the low permeability incoming sediments to the Aleutian Trench. Mineralogical analyses on incomi ng sediments show that both smectite and opal A are present as hydrous mineral phases. A 1 D sedimentation model was developed to track dehydration reaction progress and increment excess pore pressures in the incoming sediments to the Aleutian Trench. Temp eratures increase due to the insulating effect of trench sediments. As a result, opal A dehydrates to quartz at the deformation front while smectite remains mostly unreacted. Excess pore pressures increase at the deformation front due to loading of trench sediments. The 1 D modeling results at the deformation front were incorporated into a subduction model that tracks the incoming sediments into the subduction zone. Excess pore pressures approach ~70% of lithostatic pressure by 60 km landward of the deforma tion front. Modeling results suggest that dehydration reactions are not a significant component of excess

PAGE 11

11 pore pressure generation, likely because the reactions occur in the same region that is most affected by loading. In other margins, such as Barbados a nd the Japan Trench, dehydration is shifted significantly landward relative to Alaska, and could have a greater effect on excess pore pressure generation during subduction.

PAGE 12

12 CHAPTER 1 INTRODUCTION Excess pore pressures develop in subduction zone sediments when fluid escape is outpaced by pore pressure increases from mineral dehydration and tectonic loading (Saffer and Tobin, 2011). Sediment inputs impart a significant control on dehydration reaction progress and fluid production during subduction via their control on the amount of hydrous sediments and permeability (Underwood, 2007). In subduction zone sediments, matrix permeability generally decreases with burial depth due to consolidation and cementation. Fluids sourced from mineral dehydration are often released into sediment characterized by low permeability, which may enhance the development of excess pore pressure (Saffer and Tobin, 2011). Furthermore, during subduction, sediment compaction within the underthrust sediments decreases with progressive po rosity loss (Saffer and Tobin, 2011). Thus, compaction sourced fluids are typically most significant within 20 30 km landward of the forearc (Saffer and Tobin, 2011). Fluids liberated from dehydration of hydrous mineral components can provide a secondary f luid source after compaction fluids have dissipated (Saffer and Tobin, 2011). In the shallow subduction zone, the largest dehydration driven fluid source is generally from the reaction of smectite group clays to form the clay mineral illite (Bekins et al. 1994; Saffer et al., 2008; Saffer and Tobin, 2011). Smectite is common within hemipelagic marine sediments, comprising up to 45 50 wt% of the bulk sediment (Underwood, 2007) which translates to a water volume of 15 20%. This can be even higher in pelagic clay, where smectite has been shown to comprise ~8 0 wt% of a pelagic clay layer entering the Japan Trench (Ujiie et al., 2013). Biogenic opal also is also an important constituent of deep sea sediments and is the dominant sediment type where surface pl ankton productivity is high and the seafloor lies below the CCD (Spinelli et al., 2004). The amount of amorphous opal available for dehydration is

PAGE 13

13 affected by preservation, which is controlled by the balance between dissolution in silica under saturated wa ters and burial (Kastner et al., 2014). Both the transformation of smectite and biogenic opal are strongly controlled by reaction kinetics, mainly exposure time and temperature (Mizutani, 1970; Pytte and Reynolds, 1989; Bekins et al., 1994; Saffer et al., 2008). Fluids liberated from hydrous minerals during dehydration are freshened with respect to seawater and have distinctive geochemical signatures (Bekins et al., 1994; Saffer and Tobin, 2011; Kastner et al., 2014). Smectite is altered to illite in a ser ies of reactions over a temperature window of 60 160C (Bekins et al., 1994). Although typically secondary to smectite group clays, s iliceous microfossils may be an important source of fluids within sediments through the transformation of opal A to opal CT to quartz (Kastner, 1981; Moore and Vrolijk, 1992; DeMaster, 2003). The transformation of amorphous opal A (as precipitated by diatoms, radiolaria, sponges and silico flagellates) to opal CT generally begins at 10 20C and the reaction to quartz typical ly reaches completion prior to 100C (Moore and Vrolijk, 1992; DeMaster, 2003; Kastner et al., 2014). Like smectite dehydration, opal quartz reactions are highly time and temperature dependent with low activation energies (Ernst and Calvert, 1969; Kastner, 1981; Moore and Vrolijk, 1992). Researchers disagree concerning the importance of dehydration reactions in generating excess pore pressure, and studies suggest that the effect of dehydration fluids may be highly variable across subduction margins (Park et al., 2002; Spinelli et al., 2006; Ranero et al., 2008; Kameda et al., 2015). Previous work has utilized indicators such as the presence of negative polarity seismic reflectors, borehole geochemical data, extensional veining in exhumed rocks, and mud vol canism driven by the expulsion of fluids through the prism and seafloor to speculate on a relationship between dehydration fluids and zones of excess pore pressure development

PAGE 14

14 along the plate interface (Ranero et al., 2008; Saffer and McKiernan, 2009; Saff er and Tobin, 2011; Kimura et al., 2012,). Modeling of the Costa Rica margin indicates that a significant volume of fluids is produced from dehydration, which would increase pore pressure in low permeability sediment and decrease effective stress (Moore an d Saffer, 2001; Spinelli and Saffer, 2004). Ranero et al. (2008) suggest that the observed zones of high amplitude reflectivity at seafloor seeps in the Costa Rica subduction zone correlate with zones of highly pressurized fluids due to dehydration reactio ns. They further argue that the dissipation of these fluids controls where the fault zone becomes seismogenic. Kameda et al. (2015) found that peak dehydration of smectite and opal in the Costa Rica margin would occur landward of the observed updip limit o f seismicity, suggesting these reactions may play a small role in controlling mechanical behavior. Kimura et al. (2012) used dehydration and fluid flow modeling to show that the location of peak mineral dehydration coincides with the location of a high neg ative polarity reflector along the plate boundary megathrust landward of the Japan Trench. The relationship between dehydration reactions and excess pore pressures is ambiguous in the Nankai Trough, where a high amplitude negative polarity reflector exist s landward and at a deeper depth than the simulated zone of clay dehydration (Park et al., 2002). This relationship is complicated by modeling results of Saffer and Bekins (1998) who modeled smectite dehydration and peak pore water freshening within 40 50 km arcward of the deformation front in the Nankai margin, which coincides with a regional negative polarity reflector inferred to represent overpressured fault zones (Moore et al., 1990; Park et al., 2002). Excess pore pressures affect fault strength by r educing effective stress and shear strength, which increases the tendency to slip along a fault plane (Saffer and Tobin, 2011). As a result, excess pore pressures have been suggested to affect strain localization (Saffer et al., 2012),

PAGE 15

15 rupture propagation (Kimura et al., 2012), and the taper angle of accretionary wedges (Saffer and Bekins, 2006; Rowe et al., 2012). In addition to its effects on pore pressures, smectite dehydration has also been speculated to directly affect sediment frictional properties ( Kimura et al., 2012; Ujiie et al., 2013; Wang et al., 2013). Previous studies suggest that the transition to seismic behavior from aseismic stable along the interface is controlled by the replacement of mechanically weak smectite by the stronger clay miner al illite (Underwood, 2007). In this hypothesis, smectite dehydration causes a change from rate strengthening (stable slip) to rate weakening behavior (unstable slip) along the plate boundary interface (Underwood, 2007; Saffer et al., 2012). In examining t he transition from aseismic to seismic behavior along the Muroto transect in SW Japan, Moore and Saffer (2001) argue that diagenetic processes act in combination with decreasing fluid pressure ratios and temperature distribution of the incoming sediments t o control this updip seismic limit on seismicity. Laboratory experiments suggest that subduction sediments with a smectite content greater than 30 50% are needed to have a significant effect on fault zone frictional properties during subduction (Moore and Saffer, 2001). Localization of the plate boundary fault within a smectite rich pelagic clay was speculated to have allowed the large scale displacement and catastrophic tsunami of 2011 Tohoku earthquake rupture offshore Honshu, Japan. Due to the mechanical weakness of smectite group clays, subduction zone plate boundary faults may preferentially localize in smectite rich horizons (Wang et al., 2013). Ujiie et al. (2013) hypothesized that an abundance of unreacted smectite in pelagic clay where the plate bou ndary localized reduced the sliding stability of the fault, causing runaway slip up to the Japan Trench and tsunamigenesis.

PAGE 16

16 Kimura et al. (2012) report that the incoming section to the Japan Trench consists of ~20% volume fraction of diatom frustules (opa l A) which dehydrates to form chert. Friction experiments on chert using velocity measurement show that it exhibits a rate weakening behavior and promotes unstable slip consistent with the onset of the seismogenic zone in the Japan Trench. Kimura et al. (2 012) further suggest that anomalously high fluid pressures from mainly opal dehydration and a secondary component from smectite dehydration allowed runaway slip to breakthrough to the trench during the 2011 Tohoku earthquake. The majority of subduction zo ne modeling investigations of dehydration reaction progress and excess pore pressures have focused on the Costa Rica margin (Spinelli and Saffer, 2004; Screaton and Saffer, 2005), Nankai Trough (Screaton et al., 2002; Gamage and Screaton, 2006; Saffer et a l., 2008; Skarbek and Saffer, 2009; Saffer and McKiernan, 2009) and Barbados margins (Bekins et al., 1994; Screaton and Ge, 2000). These margins have been extensively drilled during Deep Sea Drilling Project (DSDP) (Legs 13, 67, 78, 87) and Ocean Drilling Program (ODP) (Legs 110, 156, 170, 171, 190, 196, 205) investigations and have been well characterized in terms of thermal structure through direct heat flow measurements and thermal modeling (e.g. Ferguson et al., 1993; Saffer and Bekins, 1998; Oleskevich 1999; Spinelli et al., 2006; Harris et al., 2010). Furthermore, previous studies have constrained excess pore pressures in the Costa Rican, Nankai and Barbados subduction zones using laboratory consolidation tests (Saffer et al., 2000; Saffer, 2003) and by inversion of porosity data from drilling and seismic reflection data (Screaton et al., 2002; Tobin and Saffer, 2009). In contrast to these regions, the Aleutian margin has relatively little data available for model implementation. At the Aleutian margi n, offshore Alaska, recent drilling during Integrated Ocean Drilling Program (IODP) Expedition 341 supplements data recovered during drilling of

PAGE 17

17 DSDP Leg 18, which provided a limited age, depth and lithology control due to poor recovery rates (Figure 1 1; Shipboard Scientific Party, 1973). These data enable characterization of the incoming section to the Aleutian Trench. However, with only two ocean drilling efforts, there are limited data available in the region. Unlike the Nankai margin, where more than 1 00 surface heat flux measurement exist perpendicular to the trench, there are limited heat flux observations on the entire ~3200 km length of the Aleutian subduction zone. Previous thermal models of the Aleutian margin (e.g. Ponko and Peacock, 1995; Oleske vich et al., 1999; Gutscher and Peacock, 2003) are also limited by the lack of heat flow data to validate thermal model results. Pore pressures have not been constrained in the Aleutian Margin and to date there has been no dehydration reaction progress mod eling of incoming and subducting sediments. von Huene et al. (1998) used estimates of seismic velocity to develop an empirical velocity porosity relationship for the accreted and underthrust sediments along the EDGE Line 302 transect (Figure 1 1; Moore et al., 1991). From this relationship, von Huene et al., (1998) developed a model of porosity reduction and inferred that pore pressures were likely elevated seaward of the deformation front offshore Alaska, however, no quantification of pore pressures were performed, nor was dehydration considered as an additional fluid source. Thus, this modeling effort will aim to provide some of the first estimates of fault zone pore pressure and fluids generated from loading and dehydration through numerical modeling. I n doing so, this modeling investigation will determine whether dehydration reactions are likely to be a significant source of excess pore pressure, relative to compaction fluid sources. Limited data provides a stimulus to use numerical modeling to better understand the complex tectonic and hydrologic processes in the Aleutian margin. The availability of drilling data enables the exploration of the interactions between tectonics and climate in this region. This

PAGE 18

18 is advanced by the proximity of the St. Elias coastal mountain range to the Pacific Ocean (Figure 1 1), which records these tectonic and climatic signals through time in the Gulf of Alaska. Furthermore, the margin is seismically active, and modeling results of pore pressure may have implications for m echanical behavior during subduction. Ongoing uplift of the St. Elias mountain range (Figure 1 1) and intensified glaciation following the mid Pleistocene transition (MPT), ~1.2 0.7 Ma, have resulted in increased sediment flux and newly carved pathways fo r sediment delivery to the Gulf of Alaska and rapid sediment accumulation (>1km/my) in the Aleutian trench (Reece et al., 2011; Jaeger et al., 2014; Gulick et al., 2015). Thus, an additional goal of this investigation was to assess the effect of rapid tren ch sedimentation (>1 km/ my) offshore southern Alaska on dehydration reaction progress and excess pore pressures. In this investigation, I modeled dehydration reaction progress and fluid in the incoming sediment section to the Aleutian Trench to determine if hydrous phases remain unreacted at the deformation front, which has implications for the volume of subducted hy drous fluids, the potential for excess pore pressure development and mechanical behavior during subduction. In particular, I investigated whether, a thick incoming sediment section drives mineral dewatering and pore pressure generation by elevating tempera tures. A second phase of modeling tracked subducting sediments to assess the importance of fluids liberated from dehydration reactions on excess pore pressure development, relative to fluids sourced from compaction of the underthrust sediments by loading b y the accretionary prism.

PAGE 19

19 Figure 1 1. Geographic and tectonic setting of the Aleutian margin. DSDP Leg 18 and IODP Expedition 341 drill sites shown. The modeling transect is boxed in red. Position of the EDGE seismic line based on Moore et al. (1991). Fan and channel locations are estimated from Reece et al. (2011). PZ= Pamplona Zone, YSV= Yakutat Sea Valley, ASV= Alsek Sea Valley.

PAGE 20

20 CHAPTER 2 BACKGROUND Study Area The Aleutian Trench forms where the Pacific Plate subducts beneath North America (Fig ure 1 1). The tectonic evolution of the present day Aleutian margin was strongly influenced by the migration and collision of the Yakutat Terrane, a 15 25 km thick mafic terrane that formed as an oceanic plateau, with North America as early as 10 Ma (Reece et al., 2011). In the Gulf of Alaska, the Pacific plate, Yakutat terrane and North American plate interact in a deformed zone near Kayak Island (Gulick et al., 2015). A transition fault marks the boundary between the Pacific Plate and Yakutat Terrane, whi ch subducts beneath the North American Plate along the Pamplona Zone (Figure 1 1; Worthington, 2008; Gulick et al., 2015). The Pacific Plate and Yakutat Terrane are both moving ~5 cm/yr relative to the North American Plate and this rate is assumed to have remained constant in the last ~6 Ma based on kinematic models of Pacific plate, Yakutat Terrane and North American Plate evolution since the Late Cretaceous (Doubrovine and Tarduno, 2008). A tectonic model of the region proposed by Gulick et al. (2015) su ggests that at this time, the Pacific Plate, North American Plate and Yakutat Terrane triple junction was stable and located near the Kayak Island Fault Zone. Approximately 600 km of Yakutat Terrane has been subducted in the last 10 Ma, and ongoing flat sl ab subduction forms the ongoing active uplift of the Chugach St. Elias orogen (Plafker and Berg, 1994; Rea and Snoeckx, 1995; Worthington et al., 2010; Reece et al., 2011). Subduction of the trailing edge of the Yakutat terrane beneath the central Gulf of Alaska ~3.5 Ma has been suggested to have created conditions favorable for accretion at the margin ~3 Ma (von Huene et al., 1998).

PAGE 21

21 The Aleutian margin is seismically active (von Huene et al., 2012). Earthquakes have sourced tsunamis in this region include the 1964 Kodiak Island, 1946 Unimak and the 1938 Semidi Islands earthquake segment ruptures (von Huene et al., 2012). To explain the spatial distribution of earthquake ruptures and aftershock distribution through time across the Aleutian margin, von Huene et al. (2012) invoke Quaternary Gulf of Alaska plate history and along strike variability in subducting plate material and relief. These factors are suggested to exert a geologic control on where and when large rupture occurs and help to identify potentia l earthquake and tsunami sources (von Huene et al., 2012). In addition to active tectonic deformation, climatic changes significantly altered the nature and rate of material flux from the margin to the ocean basin (Reece et al., 2011; Gulick et al., 2015; Jaeger et al., 2014). Alpine glaciation in the margin may have initiated as early as 7 Ma, but was enhanced ~5.5 Ma due to ongoing deformation of the Chugach St. Elias mountain range due to the feedback between elevation and trapped precipitation (Lagoe e t al., 1993; Worthington et al., 2010; Reece et al., 2011; Jaeger et al., 2014). The first appearance of ice rafted debris is concurrent with this alpine glaciation event, termed Glacial Interval A, and is exposed in the exposed Yakataga formation (Lagoe e t al., 1993; Reece et al., 2011). Glacial Interval B describes the onset of northern hemisphere glaciation between 2.9 and 2.4 Ma (Raymo, 1994) and the appearance of ice rafted debris in deep sea drill sites offshore Alaska (Rea and Snoeckx, 1995). Glacial Interval C describes further glacial intensification ~1 Ma, referred to as the mid Pleistocene transition, which could have been the result of a transition from 40 ka to 100 ka orbital glacial cycles (Berger et al., 2008). Since this interval, erosion and transport of sediment out of the St. Elias orogen has outpaced sediment influx through subduction/accretion by 50 80% (Gulick et al., 2015). Ongoing tectonic uplift of the Chugach St.

PAGE 22

22 Elias Mountains and intensification of glacial erosion are linked to ob served increases in terrigeneous sedimentation rates in the deep sea drilling sites offshore Alaska, providing evidence for a significant shift in incoming sedimentation patterns to the Aleutian Margin during the Pleistocene (Figure 1 1; Rea and Snoeckx, 1 995; Worthington et al., 2010; Reece et al., 2011; von Huene et al., 2012; Jaeger et al., 2014; Gulick et al., 2015). Varying degrees of glacial erosion and exhumation on land in the St. Elias mountain range have played an important role in the distributi on and transport of sediments into the Gulf of Alaska since the Late Miocene (Lagoe et al., 1993; Rea and Snoeckx, 1995). The Alaskan abyssal plain is characterized by sedimentary basins of Neogene age (Suess et al., 1998). The relative contributions of te ctonic inputs to the margin and sediment mass flux by erosion and transport of material out of the orogen have controlled the periodic growth cycle of these basins through time (Reece et al., 2011; Jaeger et al., 2014; Gulick et al., 2015). Sediment accum ulation on the abyssal plain is distributed in three major deep sea fan systems. In the Gulf of Alaska, the Surveyor Fan system comprises the majority of the abyssal plain and is characterized by sediments as old as ~20 Ma (Figure 1 1; Berger et al., 2008; Reece et al., 2011). Sediment delivery to the Surveyor Fan is through the 900 km long Surveyor Channel system, the inception of which is believed to be concurrent with an acceleration of glacial erosion in the Pleistocene (~2.6 Ma). (Figure 1 1; Reece et al., 2011; Gulick et al., 2015). The two main tributaries of the Surveyor Channel are the Yakutat Leg associated with the Yakutat Sea Valley and Malaspina and Hubbard Glaciers and the Alsek Leg, associated with the Alsek sea valley and glacier system (Fig ure 1 1; Reece et al., 2011). A large volume of sediment is also sourced from the Bering Bagley drainage via the Surveyor Channel to the Surveyor Fan (Gulick et al., 2015).The Zodiac Fan (Figure 1 1) is the oldest deep sea fan system

PAGE 23

23 (45 24 Ma; Suess et al ., 1998). The youngest fan system in the study region is the Baranoff Fan, which has a non distinct boundary with the Surveyor Fan system in the southeastern Gulf of Alaska and is believed to have originated in the Late Miocene (Figure 1 1; Suess et al., 1 998). Fan boundaries are estimated based on the locations of high seamount chains, including the Kodiak Bowie and Patton Murray seamount chain, which act as important natural controls of sediment transport and subducting plate geology through time (Figure 1 1, von Huene et al., 2012). Incoming Sediments Ocean drilling sites seaward of the deformation front characterize the sediments being carried to the Aleutian Trench (Figure 1 1). Site U1417 was drilled in the distal portion of the Surveyor Fan during Int egrated Ocean Drilling Program (IODP) Expedition 341 and is approximately 75 km seaward of the Aleutian Trench, offshore Southeast Alaska (Figure 1 1; Jaeger et al., 2014). The drilling site is ~2 km from Site 178 of Leg 18 of the Deep Sea Drilling Project (DSDP), thus Site U1417 supplements drilling data recovered from Site 178 (Figure 1 1; Shipboard Scientific Party, 1973). Crustal age of the Pacific Plate at Site U1417 is estimated to be 40 Ma based on shipboard age models (Jaeger et al., 2014). Directly above basement, Site 178 drilling recovered a 29 m thick section of pelagic clay overlain by ~7 m of marine chalk (Figure 2 1; Shipboard Scientific Party, 1973). Overlying these sediments are ~742 m of muds, diatomaceous sediments, turbidites and localize d volcanic ash beds that comprise the Surveyor Fan section (Figure 2 1; Shipboard Scientific Party, 1973). von Huene and Kulm (1973) and Stevenson and Embley (1987) divided the Surveyor Fan sediments into two sequences based on the integration of drilling data and 2 D seismic reflection profiles. By correlating an updated, regionally extensive seismic reflection dataset (acquired by the R.V. Marcus Langseth in 2008) to stratigraphy and age at DSDP Site 178 (Figure 1 1), Reece et al. (2011) delineated three

PAGE 24

24 sequences (Sequence I, II and III). Seismic sequences I III are regionally extensive across Surveyor Fan, and Sequence I directly overlies basement while Sequence III terminates at the seafloor (Reece et al., 2011). These sequence boundaries are likely rel ated to changes in exhumation on land and global changes in climate (Reece et al., 2011; Jaeger et al., 2014). At Site U1417, the Sequence I/II boundary (depth of ~300 mbsf) was dated to ~2.8 Ma and corresponds to the first appearance of ice rafted debris in the core (Figure 2 1; Gulick et al., 2015). The Sequence II/III boundary (depth of 160 mbsf) was correlated to an age of 1.2 Ma at Site U1417 (Figure 2 1). Reece et al. (2011) and Gulick et al. (2015) relate this boundary to glacial intensification ass ociated with the mid Pleistocene climate transition, which supports the idea that this transition was an important threshold in which climate exceeded tectonic uplift as a main control on sediment erosion (Berger et al., 2008). Site U1417 porewater geoche mical results provide insight into geochemical transformations occurring in the incoming sediment column. Dissolved silica concentration profiles show abrupt transitions, with a steep increase in the uppermost 15 m to a gradual increase at ~200 mbsf (Figur e 2 downcore, except for three distinct intervals of lower silica concentration: ~200 300 mbsf (as igure 2 2). Processes that can remove biogenic silica include the precipitation of authigenic clay minerals or opal diagenesis (DeMaster et al., 2003). Pore water chlorinity values suggest that there is slight downhole pore water freshening at Site U1417 (Figure 2 2; Jaeger et al., 2014). Freshened pore water with respect to seawater values are an indicator of possible opal A or smectite dehydration. Downcore chloride concentrations range from 546 to 565 mM and maximum pore water freshening occurs at a dep th

PAGE 25

25 of 300 mbsf at U1417 (Chloride = 546 mM), which is ~2% fresher than seawater (558 mM) (Figure 2 2). Samples from Site 180, drilled in the eastern Aleutian Trench (Figure 1 1) show a similar range of chloride concentrations (546 to 563 mM) (Shipboard sci entific party, 1973). Chloride concentrations higher than modern seawater values are likely from the burial of higher salinity Last Glacial Maximum water and chemical diffusion in the water column (Jaeger et al., 2014). Downhole temperature measurements f rom the upper 120 m of Site U1417 yield a Ponko and Peacock, 1995; Oleskevich et al., 1999 and Gutscher and Peacock, 2003) provide constraints on the thermal str ucture of the Aleutian margin. These thermal models are limited by a lack of direct heat flow measurements with which to validate modeling results and provide a range of computed temperatures along the plate interface. A 1 D thermal conduction model develo ped by Oleskevich et al. (1999) for southern Alaska computed a thrust temperature at the trench of 50C and a temperature of 150C, 160 km landward of the trench. A 2 D kinematic thermal model of southern Alaska (Ponko and Peacock, 1995) estimated that the 150C isotherm is 150 km landward of the trench. A 2 D thermal model of a Kodiak Island transect computed by Gutscher and Peacock (2003) estimated that the 150C isotherm is 60 km landward of the trench. Trench Sedimentation Figure 2 3 shows an idealized facies pattern for the Aleutian Trench at the deformation front, grading seaward from sand turbidites to abyssal plain silt mud deposits (Piper et al., 1973). Drilling of Site 180 partially characterized sediments deposited in the trench (Figure 1 1; Ship board Scientific Party, 1973). Drilling at Site 180 recovered graded silt turbidites and ice rafted erratics (Shipboard Scientific Party, 1973). Graded sand turbidites were also recovered at Sites 181 and 182, drilled on the continental slope (Figure 1 1; Shipboard Scientific Party, 1973).

PAGE 26

26 Despite the fact that Site 180 was drilled only ~5 km seaward of the deformation front (von Huene et al., 1998), grain size analyses show that Site 180 is dominated by mud and silt (Figure 2 4; Bode, 1973). Grain size dis tribution for sediments recovered at Site 180 show an average of by mass (Figure 2 4; Bode, 1973). For comparison, grain size distribution for the entire section re of mud and silt at Site 180 to the fact that the site was drilled i n the non axial part of the trench and thus, recovered overbank deposits from the main sediment channel (Figure 2 4). von Huene et al. (1998) interpreted that ~ 1 km of trench turbidites are accreted while ~1/3 of the trench sediment section and Surveyor Fan section are underthrust at the deformation front. Seismic data along the modeling transect (Figure 1 1) suggest that Sequence III sediments (0 1.2 Ma) thicken to ~1 km at the deformation front from ~200 m at Site U1417. Based on the estimate that ~1 k m of trench turbidites are accreted by von Huene et al. (1998), the decollement horizon would exist at ~1000 mbsf at the deformation front. Based on the location of a potential decollement (plate boundary fault) horizon (Figure 2 5) from a 2011 seismic sur vey onboard the R/V Marcus G. Langseth, it is also possible that as much as half of the incoming Surveyor Fan sediments (~400 m at Site U1417) are subducted (Gulick, personal communication). Due to uncertainty in the decollement zone location at the deform ation front, modeling results from several potential decollement horizons will be reported. These decollement zones will include the estimate of ~1 km accreted by von Huene et al. (1998) and that as much as half of the incoming Surveyor Fan section will be accreted (Gulick, personal communication).

PAGE 27

27 Accretionary Prism In southeastern Alaska, the accretionary prism consists of graded mud sand turbidites that overthrust deposits of the Surveyor Fan section (von Huene et al., 1998). Interbedded muds and extens ive sands were recovered from Site 182, drilled in upper continental slope of the margin, supporting that the prism consists of sediments deposited within the 75 km between Site U1417 and the deformation front (Figure 1 1; Shipboard Scientific Party, 1973) von Huene et al. (1998) utilized pre stack depth migration of a seismic reflection transect acquired during the EDGE Alaska project (e.g. Moore et al., 1991; Figure 1 1) in addition to wide angle seismic and swath bathymetry data acquired during a 1994 R /V Sonne cruise to characterize the structure of the accretionary prism at the Alaskan margin. Accreted sediment was recovered at DSDP Sites 181 and 182 and Kodiak seismic stratigraphic drilling (KSSD) drill holes (von Huene et al., 1998). The EDGE line cr osses the shelf north of the Kodiak Islands (Moore et al., 1991) and passes through the modeling transect used in this study (Figure 1 1). Across the EDGE line, von Huene et al. (1998) identified three segments within the accretionary prism with differing geometry and structural style due to varying degrees of sediment consolidation, morphology, porosity reduction and dewatering. Segment I characterizes the proto thrust zone, seaward of the deformation front that exhibits early stages of tectonic thickenin g (von Huene et al., 1998). Segment II begins at the deformation front and is characterized by low ridges and thrust faulting. A surface slope of 1 2 and a basal decollement dip of 2 in Segment II were used to compute a taper angle of 4 for this section of the accretionary prism. Segment III characterizes the main ridge and trough morphology of the lower continental slope, where accreted sediments are organized in a classic imbricate accretionary prism structure. A taper angle of 8 9 was computed for th is segment (von Huene et al., 1998).

PAGE 28

28 The accretionary wedge is stacked against a backstop approximately 50 km landward of the deformation front (von Huene et al., 1998). von Huene et al. (1998) developed an empirical seismic velocity (v in m/s ) porosity (n) relationship of n = ( 12ln ( v 1500 ) + 104 )/100 for the accreted and underthrust sediments along the EDGE Line 302 transect (Figure 1 1; Moore et al. 1991 ). von Huene et al. (1998) estimated maximum porosity reduction and rapid dewatering within the upper tr ench sediments and upper Surveyor Fan sediments approaching the deformation front. At the deformation front, a seismic velocity of ~2 000 m/s for the underthrust sediments was estimated, which results in a porosity of 0.30. Expulsion of fluids near the d eformation front is evidenced by seep biota, methane plumes and precipitates observed by Suess et al. (1998) along 800 km of the Alaskan margin offshore Kodiak Island. These sites were confined to the deformation front and were characterized as structurall y controlled (Suess et al., 1998). This was supported by video surveys along segments of the deformation front that correlated active venting sites to fault scarps, over steepened folds and bedding planes of consolidated rock (Suess et al., 1998). Suess e t al. (1998) and von Huene et al. (1998) suggest that this tectonically driven fluid flow is sourced from extensive shortening and sediment consolidation that cause rapid dewatering of sediments near the deformation front. By 25 km landward of the deformat ion front, underthrust sediments are well consolidated and von Huene et al. (1998) report velocities between 3300 3900 km/s, which suggests a porosity of 0.10 0.15.

PAGE 29

29 Figure 2 1. Incoming section at Site U1417. Samples obtained from Site U1417 are shown as red squares and Site 178, green star. Sequence boundaries are estimated from Reece et al. (2011) and Jaege r et al. (2014). Figure 2 2. Shipboard data from IODP Site U1417 and DSDP Site 178. A) Downhole sedimentation rate at Site U1417, B) Titrated chloride concentration (in mM), C) Silica concentration (mM) and D) Percentage of downhole clay sized grains by mass for Site 178 Bode, 1973).

PAGE 30

30 Figure 2 3. Conceptual model of the Aleutian margin. Sediments on the incoming plate (orange) and trench fill (blue) deposited in between Site U1417 and the deformation front grade from mud to sand turbidites towards the Aleutian margin. Based on facies model of trench deposition from Piper et al (1973). Figure 2 4. Grain size analysis on sediments rec overed from Site 180, drilled in the overbank deposits of the trench axial channel in the eastern Aleutian Trench (Bode, 1973). A) Percentage of clay sized grains by mass. B) Percentage of silt sized grains by mass. C) Percentage of sand sized grains by ma ss.

PAGE 31

31 Figure 2 5. Seismic profile view of Aleutian margin acquired from a 2011 seismic survey onboard the R/V Marcus G. Langseth. Sequence III (pink), Sequence II (Green) and basement (blue) mark seismic sequence boundaries in Surveyor Fan. The location of a potential decollement based on a seismic reflector is shown in white, truncating into basement. From Gulick, personal communication.

PAGE 32

32 CHAPTER 3 METHODS Model Overview This investigation uses numerical modeling in MATLAB to determine the effect o f rapid sediment accumulation in the Aleutian Trench on dehydration reaction progress and excess pore pressure generation. A 1 D sedimentation model was developed to track the incoming sediment section from bare crust to the Aleutian Trench. A combination of previous ocean drilling results and sample analyses were used to characterize the incoming sediment section to the Aleutian Trench for input into the model. At the deformation front, the prism was incorporated into the 1 D model by applying a prism thic kening rate. The evolution of the underthrust sediments was tracked during subduction from the deformation front to 60 km landward. This chapter first presents a detailed overview of the modeling steps implemented using the MATLAB code. The general flow o f modeling steps (Figure 3 1) was as follows: hydraulic head was incremented at the beginning of each time step to represent loading and fluid flow was simulated using a 1 D transient finite difference approximation. Then, effective stress ( e ) was calcula ted. If the calculated effective stress exceeded the value from the previous time step, a new porosity was calculated and permeability was recomputed as a function of the new porosity. With the new compacted porosity, thickness was recalculated and was use d to compute the new depth. The new porosity distribution was used to solve for thermal conductivity and temperature using a 1 D transient thermal conduction finite difference approximation during sedimentation. Landward of the deformation front, a linear temperature gradient was applied based on previous thermal models to compute temperatures within the underthrust sediment column. Updated temperatures were used to compute reaction progress with each time step and compute the rate of fluids released from dehydration.

PAGE 33

33 The 1 D modeling approach is followed by an evaluation of the MATLAB code by means of benchmarking to previously developed data and models for the Nankai Trough (Muroto). Model parameters are then specified for the incoming sediment section to the Aleutian Trench, offshore southeastern Alaska. 1 D Modeling Approach Fluid Flow and Pore Pressure Simulation Past studies have suggested that dewatering of sediments on the incoming plate in subduction margins is primarily vertical (e.g. Gamage and Screaton, 2006; Skarbek and Saffer, 2009), particularly at large distances from the trench where seaward lateral flo w from the decollement is limited. Thus, a 1 D transient modeling approach was used to simulate fluid flow and excess pore pressures in the incoming sediment section to the deformation front. The model top boundary condition is the seafloor and is hydrosta tically pressured, or excess pressure head (h*) is equal to zero. (Figure 3 2). A no flow boundary characterizes the base of the model, representative of basaltic crust (Figure 3 2). Fluid sources in the model include those from compaction, which expels existing pore fluids, and those sourced from mineral dehydration. At the beginning of each time step, the MATLAB code increments pore pressures in the sediment column through time based on the added pressure due to increased overburden as sediment is added following the method of Screaton and Saffer (2005) and Gamage and Screaton (2006). The added pressure from increased sediment overburden is calculated by multiplying the additional load by the loading efficiency onal vertical stress that is added to the pore fluid versus added to the sediment framework (Neuzil, 1986) The loading efficiency for incompressible sediments is defined by (Neuzil, 1986; Neuzil, 1995) by the expression:

PAGE 34

34 (3 1) in whi ch represents the matrix compressibility, n is porosity and denotes fluid compressibility (assumed 4.4 x 10 10 Pa 1 ). The expression for matrix compressibility e e represents the change in effective stress, is combined with expre ssions for porosity and effective stress (see Sedimentation and Porosity Evolution) to assign compressibility values in the underthrust section. After van der Kamp and Gale (1983), the pore fluid pressure change ( P) due to a change in the vertical load ( z ) is related to the 1 D loading efficiency by: (3 2) Incremented excess pore pressure (P*) from the additional sediment load and dehydration was converted to excess pressure head (h*) by dividing by gravitational acceleration (9.81 m/s 2 ) and the pore fluid density (assumed 1030 kg/m 3 ). Fluid flow in the incoming column to the trench was simulated using a 1 D transient finite difference approximation. The governing equation for 1 D transient, heterogeneous fluid flow is (Neuzil, 1995 ): (3 3) in which Ss represents specific storage, h represents hydraulic head (m), t is time (s), K is hydraulic conductivity (m/s), and (m/s) represents a fluid source term from compaction ( compaction ) and min eral dehydration ( smectite + opal ). Matrix compressibility ( ) values were used to calculate specific storage (Ss) at the center of each cell, which is defined as the volume of water that an aquifer releases into or takes from storage per unit surface a rea of the aquifer thickness per unit change in head. (3 4)

PAGE 35

35 in which n represents fractional porosity, denotes fluid compressibility (assumed 4.4 x 10 10 Pa 1 ), g represents gravitational acceleration (9.81 m/s 2 ) and w re presents the density of seawater (assumed 1030 kg/m 3 ). Rate of Fluid Release from Dehydration Previous studies suggest that fluid production by mineral dehydration is a discontinuous process, with fluids released within restricted temperature intervals (e .g. Moore and Vrolijk, 1992). However, comparison of dehydration reaction modeling with observed changes in hydrous mineralogy and porewater chemistry suggest that the reaction takes places gradually, rather than stepwise, releasing fluids over a continuou s distance from where the reaction is initiated (Bekins et al., 1994; Saffer and McKiernan, 2009). To determine the fluid source from smectite and opal dehydration ( smectite and opal ), our model takes into the account the water expelled from smectite in terlayers into the pore space during the reaction (Bekins et al., 1995) and the expansion of interlayer water when expelled (Bethke, 1985). Our modeling approach follows that of Saffer et al. (2008) and assumes that the porosity formed by the loss of solid mineral volume during dehydration accounts for 5% volumetric expansion of interlayer water upon expulsion. As a result, this should be considered a maximum estimate of the sources due to dehydration. We assigned a volumetric water content (H) of 40% (H=0. 40) which corresponds to two interlayers of bound water, in accordance with previous dehydration modeling studies (e.g. Bekins et al., 1994; Spinelli et al., 2004; Spinelli et al., 2006; Saffer et al., 2008). Adjusted for 5% interlayer expansion of water, H=0.42. For our simulation, we assigned a variable initial content of detrital smectite (C smectite ) ranging from 10% to 90% based on XRD analysis on samples (Hayes, 1973) and a smectite mole fraction (S) of 0.8 in mixed layer clay. Previous modeling (Bekin s et al., 1994; Spinelli and Saffer, 2004; Saffer et

PAGE 36

36 al., 2008; Saffer and McKiernan, 2009) has consistently assigned a mole fraction of S=0.8 for smectite in sediments entering the Barbados, Costa Rica and Nankai trough, respectively. The volume of fluids released from smectite dehydration ( were calculated using the following equation after Saffer et al. (2008) where t represents time in s and n represents fractional porosity: (3 5) Following the approach of Kimura et al., 2012, we assumed an average water content (by weight) of 7.3% for opal A, 5.5% for opal CT and 0.4% for quartz. These values are converted to volume percentages by assuming an average density of ~2.4 g/cm 3 for opal (Spinelli and Underwood, 2004) and expressed as a decimal to yield 0.175 for opal A (f a ), 0.136 for opal CT (f CT ) and 0.01 for quartz (f Q ). A range of initial bulk opal A contents (C opal ) are assigned in the incoming sediment section based on sample analysis and a mole fraction of opal A (A) is assumed throughout. From these parameters, the volume of fluids produced from opal dehydration ( were calculated after Kimura et al. (2012) where t represents time in s: (3 6) Sedimentation and Porosity Evolution Sedimentation was modeled by building a one dimensional sediment column outboard of the trench through time since crust formation (Figure 3 1). During subduction, the ac cretionary prism was incorporated into the model by applying a prism thickening rate computed from wedge geometry and plate convergence rate. This 1 D approach is similar to that of previous subduction zone models that enable variation of incoming physica l and hydrologic properties for the incoming sediment section (e.g. Screaton and Saffer, 2005; Gamage and Screaton, 2006; Saffer et al., 2008; Skarbek and Saffer, 2009). This approach also enables the user to specify the

PAGE 37

37 landward model boundary distance ba sed on a known plate convergence rate. With each time step, sediment deposition was simulated by adding a new row with a thickness based on sedimentation rates estimated from drilling and seismic data. The underlying sediments and crust are shifted downwa rd (Figure 3 1). The user selects the vertical discretization (dz) used in the 1 D model, which was assigned to be 20 m. With thickness specified, time step size (dt) is calculated based on the sedimentation rate for a sediment packet. Time step length var ies inversely with sedimentation rate and becomes smaller as sedimentation rates increase. The number of time steps for each sediment packet is determined from the total time for the packet to be deposited and the time step length. Due to the consolidatio n driven by the weight of each new sediment layer, porosity and thickness of the underlying sediments decrease. Sediments were assumed to compact depending on effective stress ( e ), which is calculated as a function of total stress ( t ) and pore fluid pres sure (P f ). (3 7) Slowly deposited sediments on the incoming plate are assumed to have hydrostatic pore pressure (P h ) (e.g. Screaton et al., 1990; Saffer and Bekins, 1998) and under this assumption, porosity has been observed to decay exponentially with depth after Athy (1930): (3 8) where n o is initial fractional porosity, b is the compaction coefficient (m 1 ) and a constant and z is burial depth (m) (Athy, 1930). This approach is consistent with a number of porosity evolution models that assume porosity varies exponentially with depth for a normally consolidated sediment column in the absence of excess pore f luid pressures (e.g. Bekins and Dreiss, 1992; Saffer and Bekins, 1998; Screaton et al., 2002; Screaton and Saffer, 2005; Gamage and Screaton,

PAGE 38

38 2006; Spinelli et al., 2006). Previous models (e.g. Screaton et al., 2002; Saffer, 2003; Gamage and Screaton, 2006 ; Spinelli et al., 2006; Skarbek and Saffer, 2009) justify the assumption of hydrostatic pressure at a reference site seaward of the deformation front due to large distance from the trench and low sedimentation rates, indicating sediments are likely to be normally consolidated. Assuming that compaction is vertical, total stress ( t is assumed to be equal to the weight of overlying sediments or lithostatic pressure ( P L ) (Screaton et al., 2002). We subtract out the hydrostatic pressure ( P h ) to yield the foll owing relationship: (3 9) where is equal to (P L P h ) and is equal to (P f P h ). Thus, under hydrostatic conditions, e with depth is given by: (3 10) where s represents solid grain density (assumed 2700 kg/m 3 ). Combining the porosity depth relationship (equation 3 8) and the effective stress depth relationship (equation 3 10), the porosity effective stress relationship for sediments undergoing compacti on can be expressed as: (3 11) With each time step, porosity, sediment thickness and depth are solved iteratively as sediment is added and the sediment below compacts. As porosity decreases, the associated volume change (dV) in sediments was computed as a function of porosity change (dn) from the previous time step (e.g. Screaton et al., 2002; Gamage and Screaton, 2006). (3 12) In the 1 D model, this volume change represents the change i n vertical thickness. Depths were then re computed based on the compacted thicknesses of each cell in the column.

PAGE 39

39 As sediments undergo compaction and porosity decreases, bulk compressibility is reduced. The relationship between change in porosity and effe ctive stress (equation 3 11) and the relationship between volume change and porosity change (equation 3 12) is combined with the e ) to assign compressibility values in the underthrust column: (3 13) Permeability Evolution Laboratory and field test results suggest that for clay rich sediments, the logarithm of permeability varies linearly with porosity (Neuzil, 1995): (3 14) where k o represents t he permeability at a porosity of (n=0) and describes the permeability change with a given porosity change, or the slope of the curve. Permeability porosity relationships were assigned based on global subduction zone compilation studies (e.g. Daigle and S creaton, 2015) and laboratory permeability measurements on samples collected from Site U1417. Vertical permeability tests were performed using the constant flow permeability method (e.g. Gamage and Screaton, 2006) that induces a hydraulic head gradient acr oss the measured sample core. Measured pressure differences enable the calculation of intrinsic permeability (k), a property of the rock, which can be used to compute hydraulic conductivity (K), which incorporates properties of the fluid medium: (3 15) s) and is a function of temperature (e.g. Spinelli et al., 2006):

PAGE 40

40 (3 16) Decollement Zone Permeability A limitation to the 1 D modeling approach is that it does not incorporate lateral flow or a permeable decollement zone. Previous subduction models assigned a decollement permeability higher than the underlying sediments based on observations that the decollement zone acts as a fluid conduit (Screaton and Saffer, 2005; Spinelli et al., 2006). Moore and Vrolijk (1992) used borehole thermal and geochemical evidence to support the idea that accretionary fault zones serve as highly permeable pathways for fluid migration. Several investigations have examine d the hydrogeologic properties of decollement fault zones using borehole packer tests, submersible based slug tests and constant rate flow tests. Single well tests and CORK installation in the Barbados Ridge Complex (e.g. Screaton et al., 1997) in addition to shipboard packer testing (Fisher and Zwart, 1997) show a 4 5 order of magnitude increase in decollement zone permeability as fluid pressure varies from hydrostatic to lithostatic. Screaton et al. (2000) used CORK observation data from ODP Leg 156 and L eg 171 in the Barbados Accretionary Complex to infer a decollement zone permeability 2 4 orders of magnitude higher than the overlying and underlying sediments, sufficient to focus fluid flow across two sealed boreholes ~50 m apart. Previous studies in Cos ta Rica (e.g. Saffer et al., 2000) suggest that underthrust permeabilities are insufficient to account for fluid drainage inferred from porosity loss, providing support to the idea that the decollement may be a preferential flow pathway in subduction margi ns. Temperature Evolution Incoming sediments. In the 1 D modeling approach, temperatures were calculated with a 1 D conductive heat transport model that excludes the effects of advective heat transpor t (e.g.

PAGE 41

41 Ferguson et al., 1993). Thermal boundaries inclu (top model boundary) (Figure 3 1) and a basal heat flow, q (mW/m 2 ) that decays as a function of crustal age (t, in m.y.) after Parsons and Sclater (1977): (3 17) Thermal conductivity (k t ) was calculated as an average of the fluid thermal conductivity (k w ) and solid thermal conductivity (k s ), 0.7 W/mC and 3.0 W/mC, respectively, as a function of porosity: (3 18) The crust was assigned a uniform thermal conductivity of 2.9 W/mC. Heat capacity (c) was calculated as a weighted average of the fluid heat capacity (c w assumed 1000 J/kg C) and solid heat capacity (c s assumed 4180 J/kg C): (3 19) Thermal conductivity and heat capacity were used to calculate thermal diffusivity, which b ), calculated as a weighted average of the fluid and solid density. (3 20) n of thermal conductivity, heat capacity and bulk density: (3 21) Initial temperatures were calculated assuming 1 m.y. old crust using the Parsons and Sclater (1977) relationship (equation 3 17) based on calculated heat flow (q), thermal conductivity (k t ) and cell thickness:

PAGE 42

42 (3 22) For subsequent time steps, a 1 D finite difference approximation was used to simulate transient heat flow. Following the methods of Ferguson et al. (1 993), the heat flow equation is: (3 23) where and c represent density and specific heat capacity as a function of depth (z), T is temperature (C), Q is a general heat source term (W/m 3 ) as a function of depth, k t is thermal conductivity as a function of depth, dt is a time increment and dz is a depth increment. Subducted sediments. Temperatures in the 1 D model during subduction were defined as a function of linear distance (dT/dx) based on the results o f previous thermal models of the margin. Landward of the deformation front (dT in Figure 3 2), temperatures were based on previous modeling results and reported locations of the 150 C isotherm along the plate interface. To calculate temperature, the respec tive dT/dx value was multiplied by the time step length (dt) and the convergence rate and added to the initial temperature calculated by the 1 D thermal conduction model at the deformation front. This approach is a simplification because it assumes that te mperatures increase linearly with depth, following the same trend as at Site U1417. It does not consider the influence of changes in the thermal properties of the subducting material or frictional heating along the plate boundary fault. For each time step, the sediment column was buried to greater depths simulating its landward movement under the prism. The sediments lose porosity and thin as the overlying accretionary wedge sediments load the underlying column (Figure 3 2). Temperatures were computed and s hifted with each time step and reaction progress was calculated from tracking sediment from its initial position outboard of the deformation front through the simulated temperature field during subduction.

PAGE 43

43 Dehydration Reaction Progress Temperatures were u sed to compute reaction progress of smectite and opal in the incoming sediments based on kinetic expressions. In the shallow subduction zone, dehydration of smectite group clays and biogenic opal represent the primary hydrous mineral phases in terms of flu id volume. Smectite dehydration. Smectite group clays can be of the dioctahedral or trioctahedral form, though the most common smectite group mineral is dioctahedral montmorillonite (Brindley and Brown, 1980). Montmorillonite consists of a 2:1 layer with charge balancing and interlayered cations that is most noted for its capability of interlayer expansion, or swelling, when saturated with certain ions (Brindley and Brown, 1980). Smectite can be sourced from detrital weathering and in situ alteration of di sseminated ash (Underwood and Pickering, 1996) and illite can be precipitated from the dissolution and dehydration of smectite (Pytte and Reynolds, 1989; Moore and Reynolds, 1989). Mixed layering is common in clay minerals (Moore and Reynolds, 1989) and ra ndom mixed layer mica montmorillonite (I/S) clays are the most common in nature (Hayes, 1973; Brindley and Brown, 1980; Moore and Reynolds, 1989). Kastner et al. (2014) report that the majority of detrital smectite that enters subduction zones is 70 75% mi xed layer illite smectite clay (I/S) rather than pure smectite. As smectite is converted to illite through a series of reactions, it forms mixed layer illite smectite clay. The reaction is controlled mainly by temperature and time, though secondary factor s include pressure, water/rock ratio and pore water chemistry (Bekins et al., 1994). The smectite dehydration kinetic expressions developed by Pytte and Reynolds (1989) is used in this modeling investigation. Pytte and Reynolds (1989) used a 6 th order kine tic expression (equation 3 24), and was able to successfully reproduce observed percentages of illite in sediments exposed to temperatures in between 70 250C. Pytte and Reynolds (1989) modeled the reaction

PAGE 44

44 based on data from a natural system, which may ac count for rate controls not understood in the illitization process and spans an appropriate temperature range for this study. The Pytte and Reynolds (1989) model for smectite dehydration has been shown to reproduce XRD derived illite percentages in sedimen ts from the Muroto transect in the Nankai Margin (Saffer et al., 2008), the Barbados accretionary prism (Bekins et al., 1994) and the Costa Rica Margin (Spinelli and Saffer, 2004). In addition, the model was applied to the underthrust sediments in the Japa n Trench by Kimura et al. (2012). (3 24) where S is the initial mole fraction of smectite in mixed layer I/S clay, A is a constant of 5.2 x 10 7 1/s, R is the gas constant (8.31 J/mol K), T is temperature (K) and E is activation energy, equal to 1.38 x10 5 J/mol. The ratio of activities of [K + /Na + ] assumes that albite is in equilibrium with K feldspar and is a function of temperature (Pytte and Reynolds, 1989). Pore water chemistry, specifica lly the ratio of K + and Na + activities, plays a lesser role in controlling the reaction rate of smectite than exposure time and temperature (Pytte and Reynolds, 1989; Bekins et al., 1994). Perry and Hower (1970) and Hower (1976) compared reaction progress with observations and concluded that K + is not a significant control on reaction rate until mica and detrital feldspar are consumed. For this modeling investigation, the effects of pore water chemistry on reaction progress are not considered. Opal dehydra tion. We follow the method outlined by Kimura et al. (2012) who use the kinetic expression developed by Mizutani (1970) to compute reaction progress and fluid production from the dehydration of opal A to quartz through opal CT. Mineralogically, biogenic si lica is called opal A, a highly disordered, nearly amorphous mineral (DeMaster, 2003). Opal A is a hydrated silicon (SiO 2 nH 2 O) that releases water and becomes more structurally ordered in

PAGE 45

45 its dehydration to quartz through opal CT (Mizutani, 1970; DeMaster 2003). As biogenic silica is buried beneath sediment, exposure to higher temperatures and pressures transforms it from amorphous opal A to opal CT (cristobalite) and in some cases to chert (DeMaster, 2003). The dehydration of opal CT is highly temperatu re dependent and involves a solid state dehydration that converts opal CT to microcrystalline quartz (Calvert, 1983). Mizutani (1970) performed a series of hydrothermal experiments using chemical kinetics to model the proportion of amorphous opal A, opal C T and quartz in 0.077N potassium hydroxide solutions at temperatures between 86 280C and 10 MPa of pressure. Based on the observation of conversion to 50% quartz within 3000 hr., the following kinetic expression and rate constants were derived: (3 25) where OA is the initial mole fraction of opal A, OCT is the initial mole fraction of opal CT, A is a rate constant equal to 7.51522 x 10 4 s 1 (for opal A) and 2.25x10 3 s 1 (for opal CT), and E is the activation energy, 6.72 x 10 4 J/mol. In this derivation, the ordering of opal CT was not considered to affect reaction rate and the surface area of the phase present, which may have an important control on dissolution precipitat ion reactions like that of opal A to quartz. Benchmarking the 1 D MATLAB Code The 1 D modeling code was benchmarked to previous modeling of smectite dehydration reaction progress and pore pressure at Site 1174, drilled during Leg 190 of Ocean Drilling Pro gram offshore of the Muroto Peninsula of Japan (Shipboard Scientific Party, 2001). The Nankai margin was chosen as a benchmark site due to the availability of both reaction progress modeling using the same kinetics (e.g. Pytte and Reynolds, 1989) as our mo deling investigation and modeling of pore pressure. Furthermore, the incoming sediments are well characterized in this margin and were penetrated along two transects by the DSDP and ODP during Legs 31, 87,

PAGE 46

46 190 and 196 which provide extensive shipboard and XRD data to characterize the incoming section. Site 1174 was drilled in the proto thrust zone to a total depth of ~1150 mbsf and recovered sediments on the incoming plate and the overlying trench wedge sequence. The section recovered at Site 1174 includes an upper trench facies sequence, the upper Shikoku Basin facies (USB) consisting of hemipelagic mudstone and ash, the lower Shikoku Basin facies (LSB) consisting of hemipelagic mudstone and claystone, and a lower unit of volcaniclastics (Shipboard Scienti fic Party, 2001). The decollement at Site 1174 is localized in the LSB facies, within a deformed zone ranging from ~808 840 mbsf (Shipboard Scientific Party, 2001). An exponential porosity depth relationship of n= 0.77 0.0011z was assigned to the LSB and U SB facies at Site 1174 after Gamage and Screaton (2006). A porosity depth relationship of n= 0.65 0.0007z was assigned to the trench turbidite sediments (Gamage and Screaton, 2006). Simulated porosity fits well with downhole porosity measurements in the trench wedge and USB facies, but deviates from the exponential trend at the top of the LSB and below the decollement (~808 to 840 mbsf) (Figure 3 3). Below the decollement, porosities are ~5 7 % higher, suggesting that this deviation is due to under compac tion (Saffer and McKiernan, 2009). Temperatures and thermal conductivity values match reasonably well with shipboard data (Figure 3 Previous modeling of dehydration along the Muroto transect at Site 1174 (e.g. Saffer et al., 2008) suggests that the reaction begins at ~680 mbsf. Overall, simulated results show a good match to observed changes in percent illite in mixed layer clay at S ite 1174 (Figure 3 3). Simulations show that diagenetic reaction progress of smectite begins at ~650 mbsf (Figure 3 3), which is consistent with modeling results by Saffer et

PAGE 47

47 al., (2008). Results show that the simulated mole fraction of smectite decreases from 0.8 to 0.3 (illite ~ 70% of interlayered illite smectite (I/S) clay) at a depth of ~1100 mbsf, which matches well with simulations of Saffer et al (2008) and XRD clay analyses (Steurer and Underwood, 2003). The progression of opal A to quartz through opal CT was not included in the modeling study of Saffer et al. (2008), likely due to predominantly hemipelagic mudstone dominated by smectite, which increases down section to >35 bulk wt % (Underwood, 2007). Previous computations of the excess pore pressu f P h / P L P h ) beneath the decollement at Site 1174 are variable and range between 0.42 (Screaton et al., 2002) and 0.14 (Gamage and Screaton, 2006). Screaton et al. (2002) used porosity measurements of core samples from Site 1174 t o estimate pore pressures in the underthrust sediments by relating volume change to porosity change in sediments from a seaward reference site (Site 1173) to Site 1174. Gamage and Screaton (2006) used numerical modeling to increment pore pressures as a fun 1174 using our MATLAB script (Figure 3 This result is very similar to that predicted by previous modeling by Gamag e and Screaton Screaton et al. (2002) which was based on interpretation of porosity data. Sample Analysis Incoming hydrous mineralogy was characterized throu gh sample analysis (biogenic silica) and previous X ray diffraction clay analyses on samples retrieved from DSDP Leg 18 (Hayes, 1973).

PAGE 48

48 Biogenic Silica To determine the amount of biogenic silica in 5 samples retrieved from Site U1417, an alkaline leaching method was used, as outlined by DeMaster (1981) and Spinelli and Underwood (2004). Samples (25 mg) were digested in 40 mL of 0.0316M NaOH at 85C. Biogenic opal digests more rapidly in alkaline solution than clay minerals, resulting in a rapid increase in silica concentration in the alkaline solution (Spinelli and Underwood, 2004). However, this signal can be overprinted by dissolution of silicate clay minerals. Thus, continuous dissolution of the samples was monitored over a total of 5 hours. DeMaster (198 1) found that alumino silicates from clay dissolution release silicates Si linearly over time, and most of the Si is dissolved within 2 hours of digestion. The concentration of silica (mg/g) in the leachate after 3, 4 and 5 hours was determined by spectro photometry. Mixed Layer Illite Smectite Clay Hayes (1973) used X Ray diffraction to analyze bulk sediment samples from DSDP Site 178 (Figure 1 1), and the following provides a brief methodology overview. Preparation of the clay sized (<2 m) fraction followed the pre treatment methods of Duncan et al. (1970). Clay minerals were separated by centrifugation, Mg saturated and analyzed with XRD before and after solvation with ethylene glycol. Samples were K saturated and heated to 525C to test for the pre sence of mixed layer minerals. Hayes (1973) interprets a broad peak at 17 with ethylene glycol solvation as random mixed layer illite smectite (I/S) clay with small vermiculite like layers which act to prevent a fully expanded spacing of mixed layer cla y. Increasing the number of mica layers causes the intensity and valley peak ratio (Biscaye, 1965) of the 17 peak to decrease, however, the peak remains at 17 (Hayes, 1973). The composition of mixed layer clay was determined by spacing of the hybrid 00 2 mica/003 montmorillonite peak, which is used

PAGE 49

49 to interpret the percentage of illite in mixed layer clay (Reynolds and Hower, 1970). Hayes (1973) quantified the relative proportion of chlorite, vermiculite, illite and mixed layer clay using peak heights of 7, 10, 14 and 17 above background levels and adding ratios of 14/10, 17/(3 X 10) and10/10 (equal to 1) and normalizing to 100 percent. Hayes (1973) addressed the overlap between chlorite and vermiculite basal reflections by computing the 7/14 pe ak ratio, assumed to be 0.2 for pure vermiculite and 2.0 for pure chlorite, to generate a mixing line and the percent chlorite for a given ratio. Application of the Code to the Aleutian Trench A 1 D sediment column was built through time and included sedi ments deposited at Site U1417 and ~1 km of trench sediments deposited in between Site U1417 and the deformation front (Figure 1 1). The left boundary of the model and the seafloor were assumed to be hydrostatic (h*=0). The incoming sediments were divided i nto packets of similar physical and hydrological properties, which enabled direct testing of the effects of thick trench sedimentation on excess pore pressures and reaction progress at the deformation front. A seismic reflection profile (Gulick, personal communication) across the modeling transect shown suggests that the trench wedge extends ~45 km outboard of the deformation front, which is ~30 km landward of Site U1417 (Figure 2 5). Two way travel time data of Sequence III sediments (0 1.2 Ma) deposited in between Site U1417 and the deformation front, were converted to sediment thickness by assuming a uniform average seismic velocity of 1750 m/s. The results suggest that Sequence III sediments reach ~1 km to the deformation front (Figure 3 5). At Site U14 17, the Site U1417 boundary was placed at ~200 mbsf. Thus, the increase in sediment thickness from Site U1417 to the deformation front is attributed to the accumulation of trench sediments outboard of the deformation front since the mid Pleistocene

PAGE 50

50 transit ion ~1 Ma. The trench wedge is observed in seismic profiles as continuous, flat lying strata (Figure 2 5). At the deformation front, wedge geometry was defined by von Huene et al. (1998) (Table 3 2) that computed changes in geometry of the continental slop e along the EDGE Alaska line that crosses the shelf north of the Kodiak seamount chain (Figure 1 1). This geometry enabled calculation of prism thickening rates from the deformation front to ~60 km landward, representing a total time of ~1.2 m.y. based on a convergence rate of 5 cm/yr. Model Parameters Ocean drilling results from DSDP Site 178 and IODP Site U1417 were used to characterize the physical and hydrological properties of the incoming section to the Aleutian Trench. Incoming sediments were divided into individual sediment packets based on mappe d seismic sequences (I III) identified by Reece et al. (2011) (Figure 2 5, Table 3 1). In the 1 D model, Sequence I was further subdivided to individually simulate the pelagic clay unit above igneous basement, which has been suggested to have distinct fri ctional properties during subduction (Wang et al., 2013; Ujiie et al., 2013). X ray diffraction clay analyses (Hayes, 1973) and grain size analyses (Bode, 1973) indicate that illite smectite mixed layer clay content (~90 wt %) and clay sized grains by mass (~95 wt %) are much higher in this unit than the overlying Surveyor Fan sediments. The average initial sedimentation rate for each unit was based on shipboard reported sedimentation rates at Site U1417 (Figure 2 3; Jaeger et al., 2014). The shipboard sed imentation rates were corrected for porosity loss to determine initial sedimentation rates (Table 3 1). Sedimentation rates for sediments deposited in between Site U1417 and the deformation front were calculated using change in sediment thickness (based on migrated Sequence III seismic data) and a convergence rate of 5 cm/yr (Figure 3 5; Gulick et al., 2015). Sequence III thickness

PAGE 51

51 was calculated from an average seismic velocity of 1750 m/s and two way travel time (TWT) along a transect from Site U1417 to t he deformation front. To determine sensitivity to uncertainty in seismic velocity and TWT, the seismic velocity of 1750 m/s was varied by 10%, yielding velocities of 1925 m/s and 1575 m/s. Within this velocity range, the maximum velocity (1925 m/s) was u sed to compute the maximum sediment thickness and sedimentation rate for each unit. The minimum (1575 m/s) and average (1925 m/s) seismic velocities were then used to compute the minimum and average sedimentation rates for each sediment unit. Calculated sedimentation rates were plotted with distance from Site U1417 to determine trends through time (Figure 3 5). From this plot, six distinguishable trends in sedimentation rate were observed for each of the three scenarios (maximum, minimum or average seismi c velocity assumed in calculation) (Figure 3 5). An average sedimentation rate was computed for each observable trend line (as shown in Figure 3 5). These average sedimentation rates were input as individual sediment packets in the 1 D model, which enabled variation of physical and hydrologic parameters for each unit (Table 3 3). Trench sedimentation rates were adjusted for porosity loss using an assumed porosity depth relationship of n o = 0.6 and b = 0.0007 (Table 3 2) based on porosity data recovered from drilling at Site 180 (Figure 1 1). This porosity depth relationship is consistent with Sequence II and Sequence III sediments analyzed at Site U1417 (Table 3 1). The different porosity depth relationship from Sequence I is thought to reflect differences in incoming lithology and grain size (Figure 2 1, Figure 2 2). Permeability porosity relationships for the incoming section and the trench wedge were calculated using grain size results from Site 178 and Site 180 (Bode, 1973) based on the global subduction zone porosity permeability dataset compiled by Daigle and Screaton (2015). Daigle and Screaton (2015) assembled a permeability dataset from 317 subduction zone samples and

PAGE 52

52 concluded that porosity permeability behavior is a function of clay sized fraction. Permeability was also directly measured on eleven samples retrieved from IODP Expedition 341 at Sites U1417 and U1418. The locations of these samples are shown in Figure 2 1. For the incoming sediments, the average percentage of clay size grains by mass fr om Site 178 (Figure 2 2) and Site 180 (Figure 2 4) was used to classify each sediment packet as a Group 1 clay (>80% clay size grains by mass; log k=5.75n + 20.924), Group 2 clay (60 80% clay size grains by mass; log k=9.78n+ 21.971) or Group 3 clay (<60% clay size grains by mass; log k = 8.39n + 20.862) (Table 3 1; Table 3 2). Permeability measurements on samples retrieved from Site U1417 were compared to permeability porosity relationships established using grain size results after Daigle and Screaton (2015) as shown in Figure 3 6. Samples analyzed for permeability were plotted with measured porosity and fit to a curve that yielded a best fit k n relationship of log k= 20.92+8.1132n with samples from Site U1417 (Figure 3 6). Sensitivity analyses were us ed to assess how applying the k n relationship estimated from sample analyses will affect pore pressure distribution during sedimentation and subduction. Results from Site U1418, located on an elevated region of the Surveyor Fan in between the Aleutian Tre nch and the Bering Channel which once terminated into the active trench, (Figure 1 1) were also evaluated, but suggest a different relationship, likely due to differences in grain size. Thus, these data were not included in the relationship used for the m odeling. In the trench wedge, sediment deposited in between Site U1417 and the deformation front was assigned a k n relationship consistent with those developed by Daigle and Screaton (2015) and assigned at Site U1417 (Table 3 2). Packet 6, landward of Si te 180 (Figure 1 1), was assigned a log (k o ) value two orders of magnitude higher to account for a transition from mud

PAGE 53

53 and silt dominated sediment to sandy turbidites near the deformation front (Figure 2 4; Table 3 2). The accretionary prism was incorpora ted as two individual sediment packets in the 1 D model to account for an increase in taper angle from 4 at the deformation front to 9 ~10 km landward (von Huene et al., 1998). Based on the given taper angle, sediment thickness at a specified distance fr om the deformation front can be determined. From this thickness, a prism thickening rate can be computed with an assumed plate convergence rate. For the first accretionary sediment packet, from the deformation front to 10 km landward, a prism thickening ra te of 0.0035 m/yr was calculated. A second prism thickening rate of 0.0079 m/yr was calculated to represent sediment accumulation in the accretionary prism from 10 to 60 km landward. The fluid source from compaction or mineral dehydration of accreted sedim ents was not included in the model. The accretionary prism was assigned an n depth relationship consistent with the relationship assigned to the trench wedge, n o =0.6 and b=0.0007. A k o of 18.862 and a was applied, consistent with the trench wedg e sediments. Landward of the deformation front in the 1 D model, a linear temperature change with distance (dT/dx) was assigned (Figure 3 2). This temperature gradient was assigned based on previous thermal modeling (Figure 3 2; Table 3 3; e.g. Ponko and Peacock, 1995; Oleskevich et al., 1999 and Gutscher and Peacock, 2003). For the model base run scenario, an average dT/dx (Table 3 3). By assigning a range of dT/dx v alues for the underthrust column, the effect of variable thermal regimes on reaction progress and excess pore pressure development was explored.

PAGE 54

54 XRD clay studies on samples from DSDP Leg 18 (Hayes, 1973) were used to characterize the incoming clay mineral ogy (Figure 3 6; Table 3 1, 3 2). Samples retrieved from Site U1417 (see Figure 2 1) were characterized for the mass percentage of biogenic silica in the bulk sediment following the alkaline leaching method of DeMaster (1981) and Spinelli and Underwood (20 04) (Figure 3 6). For calculation of fluid production, the volume fraction of I/S clay and biogenic opal (opal A) in the bulk sediment was also incorporated. Mass percentages of mixed layer I/S clay (Hayes, 1973) and biogenic silica (Kenney, personal commu nication) were converted into volume percentages using an average density of opal=2.47 g/cm 3 smectite= 2.02 g/cm 3 (Spinelli and Saffer, 2004). XRD analyses by Hayes (1973) suggests that the smectite content is variable in the incoming sediment section, wi th the mole fraction of smectite ranges from 0.5 to 0.8 with increasing depth at Site 178 (Table 3 1). For the purpose of this study, a smectite mole fraction of 0.8 is assigned to be consistent with previous modeling of dehydration reaction progress in ot her margins (e.g. Spinelli and Saffer, 2004; Saffer et al., 2008; Spinelli et al., 2006) however, the sensitivity of model results to smectite content was tested by varying the mole fraction of smectite in the 1 D simulations. Following the methods of Kimu ra et al., 2012, a molar fraction of 1 was assumed for opal A (Table 3 1; Table 3 2).

PAGE 55

55 Table 3 1. Site U1417 incoming sediment packet characterization. Note: Permeability porosity (k n) relationships from Daigle and Screaton (2015), percentage of I/S clay from Hayes (1973). S= mole fraction of smectite in illite smectite clay (I/S). Table 3 2. Trench packet characterization. Note: Permeability porosity ( k n ) relationships from Daigle and Screaton (2015), percentage of I/S clay from Hayes (1973). S= mole fraction of smectite in illite smectite clay (I/S).

PAGE 56

56 Figure 3 1. 1 D modeling steps. Figure 3 2. Conceptual diagram of 1 D sedimentation and subduction model. As new sediment packets are deposited, deeper layers shift down. Heat flow is simulated as a function of age after Parsons and Sclater (1977). Incoming sediments are underthrust at the deformation front. The accretionary prism is incorpor ated with a prism thickening rate computed using wedge geometry (taper angle) and plate convergence rate. During subduction, temperatures are computed using a linear temperature gradient (dT/dx).

PAGE 57

57 Figure 3 3. Benchmark modeling results of the Nankai Trou gh. A) Downhole porosity, B) temperature, C) % illite and D) excess pore pressure ratios from Site 1174, drilled along the Muroto transect in the Nankai Trough. Modeling results are compared to shipboard and drilling data (Ocean drilling program, Leg 190). Figure 3 4. Sequence III thickness from Site U1417 to the deformation front, as defined by seismic two way travel time data converted to thickness assuming a seismic velocity of 1750 m/s.

PAGE 58

58 Figure 3 5. Trench sedimentation rates based on Sequence III thickness data (Jaeger et al, personal comm.). Sediment packets are distinguished by numbers 1 6, and thickening rates were computed assuming a convergence rate of 5 cm/yr. Figure 3 6. Site U1417 sample analysis A) Mass percent biogenic silica from Site U1417 samples and mixed layer smectite illite clay from Site 178 (Hayes, 1973). B) Sample permeability and porosity plotted against Daigle and Screaton (2015) k n relationships with Site U1417 samples. C) Site U1418 permeability and porosity included.

PAGE 59

59 CHAPTER 4 1 D MODELING RESULTS Seaward of the Deformation Front To examine the effects of trench deposition between Site U1417 and the deformation front offshore Alaska on reaction progress and pore pressures, 1 D modeling results seaward of the tren ch wedge (Site U1417) and at the deformation front are compared. Simulated porosity, downhole temperature and thermal conductivity at Site U1417 are plotted with observed shipboard data collected during IODP Expedition 341 (Figure 4 1). The simulated poros ity profiles decrease exponentially with depth to ~0.4 at the base of the simulated section. There is a slight underestimate of porosity at ~300 400 mbsf (Figure 4 1). Simulated temperature reaches a maximum of 46C (Figure 4 1). Dehydration reaction progr ess of smectite to illite and opal A to quartz (through opal CT) are shown in Figure 4 2. Two scenarios for smectite are plotted; one in which smectite mole fraction is set constant for all sediment packets (S=0.8) and one in which S=0.5 for the Surveyor F an section and 0.8 for the pelagic clay unit above basement. Smectite remains unreacted at the base of Site U1417. Opal A progresses to opal CT and quartz below 400 m, and quartz reaches a mole fraction of ~0.9 at the base of Site U1417 (Figure 4 2). Due to sedimentation of the trench wedge, the total sediment thickness increases from ~800 m at Site U1417 to ~1600 m at the deformation front (Figure 4 3). At the base of the sediment column, simulated porosity has been reduced from ~0.4 at the base of Site U 1417 to ~0.25 at the deformation front (Figure 4 3) and temperatures have increased from ~50C to ~90C. Simulation results suggest that smectite undergoes slight dehydration to illite prior to subduction, and the mole fraction of smectite (S) has decrease d from S= 0.8 at the base of Site U1417 to S=0.77 at the base of the incoming section at the deformation front (Figure 4 4). At the base of the sediment column at Site U1417 (~800 mbsf), the mole fraction of opal A is ~0.1 and

PAGE 60

60 the mole fraction of quartz i s ~0.9 (Figure 4 4). Opal CT reaches a maximum mole fraction of ~0.2 at a depth of ~600 mbsf (Figure 4 4). The opal A conversion to quartz through opal CT shifts deeper with the addition of the trench wedge (Figure 4 4). At the deformation front, opal A tr ansformation to opal CT steepens at ~800 mbsf and the mole fraction of both forms of opal approaches zero at the base of the sediment column (Figure 4 4). Opal CT reaches its maximum composition of 0.2 at a depth of ~1300 mbsf and quartz approaches a mole fraction of ~1 at the base of the section (~1600 mbsf) (Figure 4 4). However, at shallower depths (<600 mbsf) opal A remains mostly unreacted (Figure 4 4). Excess pore pressures increase significantly with the addition of the trench sediments at the defor mation front (Figure 4 5). In this study, excess pore pressures are reported in terms of defined by Shi and Wang (1988) as: (4 1) is 0, pore pressures are hydrostatic a is 1, pore pressures equal the lithostatic pressure. Excess pore pressure ratios at Site U1417 are near hydrostatic with values reaching a maximum of 0.03 at a depth of ~30 mbsf and generally decreasing with depth (Figure 4 5). von Huene et al. (1998) imply that as much as 1 km of sediments are subducted at the deformation front. It has also been estimated that as much as half of the incoming Surveyor Fan section is subducted (Gulick, personal communication). Three potential decollement locat ions are highlighted at the deformation front (Figure 4 3 Figure 4 5). These depths assume that ~1000 meters of sediment are subducted (shallowest location), ~600 meters are subducted (middle location) and ~300 meters are subducted (deepest location). Exce ss pore pressure ratios at the deformation front range from depth of ~600 mbsf) to at a depth of ~1300 mbsf Figure 4 5). Excess pore pressures decrease slightly with depth in the incoming column,

PAGE 61

61 but remain above hydrostatic pressure at t (Figure 4 5). Influx of Water at the Site U1417 Transect The fluid influx along the transect was estimated by using the porosity distribution, mineral assemblage, sediment thickness and plate convergence rate of the incoming sediment section at Site U1417 (Table 4 1). The incoming sediment section was divided up into packets with differing physic al and hydrogeological properties (Table 3 1). Sequence III is approximately 160 m thick at Site U1417 and has an average porosity of 0.55 (Figure 4 1). Assuming a plate convergence rate of 5 cm/yr (Reece et al., 2011; Gulick et al., 2015), Sequence III ca rries ~4.4 m 3 /yr of pore water per m along strike trench length. Sequence III was determined to consist of on average ~10 vol. % smectite illite mixed layer clay (Hayes, 1973) and ~10 vol. % biogenic silica, assuming an average density for opal of g/ cm 3 (Spinelli and Underwood, 2004). The water content in smectite and biogenic silica is ~42 vol. % (assuming two layer hydration and accounting for 5% interlayer expansion; e.g. Saffer et al. 2008) and ~17.5 vol. % (assuming an average water weight percen tage of 7.3 wt. %; e.g. Kimura et al., 2012), respectively. Accordingly, Sequence III hydrous sediments deliver ~0.5 m 3 /yr of mineral fluid. Sequence II is approximately 140 m thick and has an average porosity of ~0.50 at Site U1417. With ~10 vol. % mixed layer clay (Hayes, 1973) and 2 vol. % biogenic silica, Sequence II hydrous minerals deliver ~0.3 m 3 /yr of fluid while ~3.6 m 3 /yr is delivered as pore fluid. Sequence I is ~450 m thick and has an average porosity equal to ~0.50 ~20 vol. % smectite (Hayes, 1973) and ~10 vol. % biogenic silica. Sequence I delivers ~11 m 3 /yr of pore fluid and ~2.3 m 3 /yr of mineral fluids. Incoming pelagic clay is ~29 m thick (Shipboard scientific party,

PAGE 62

62 1973) and has an average porosity of ~0.40, with an average of ~90 vol. % mixed layer clay and ~2 vol. % biogenic silica. Accordingly, the lowermost pelagic clay unit delivers ~0.6 m 3 /yr of pore and mineral fluid. The total water influx of the incoming sedimentary package to the Aleutian ma rgin is ~22.9 m 3 /yr. At the deformation front, the incoming sediment section has compacted due to the deposition of ~1 km of trench sediments. The thickness of the incoming sedimentary package has decreased from ~780 m at Site U1417 to ~530 m at the defor mation front (Table 4 2). The total fluid influx at the deformation front is ~11 m 3 /yr sourced from pore water expulsion (~80% of total fluids) and mineral dehydration. Based on the simulations, approximately of the total fluids arriving at Site U1417 ar e expelled before reaching the deformation front. Sensitivity Analyses Trench sedimentation rate As described in the modeling methods section, minimum, average and maximum sedimentation rates were calculated for each trench packet by assuming a range of s eismic velocities used for the time to depth conversion (1750 m/s 10 %). Simulation results show that the distribution of temperature, porosity and reaction progress with depth are slightly sensitive to changes in trench sedimentation rate (Figure 4 6). Temperature advances to 90C assuming a maximum rate of trench sedimentation for all packets, and to 80C assuming a minimum rate of sedimentation (Figure 4 6). Smectite undergoes the maximum conversion to illite under conditions of maximum sedimentation ( Figure 4 6). Assuming a maximum trench sedimentation rate, the mole fraction of smectite (S) approaches 0.74 at the base of the sediment column at the deformation front (Figure 4 6). Assuming a minimum trench sedimentation rate, S=0.78 at the base of the s ediment column at the deformation front (Figure 4 6).

PAGE 63

63 Diagenetic reaction progress of opal is similar for all three sedimentation sequences from the seafloor to a depth of ~800 mbsf, below which compositional changes in opal A, opal CT and quartz deviate ( Figure 4 6). Opal A is nearly converted to quartz at the base of the section for all simulations, although lowering the sedimentation rate increases the time step length (dt) for the trench wedge sediments and produces a shallower depth of conversion (Figu re 4 6). Assuming a minimum rate of sedimentation, quartz approaches a mole fraction of 1 at a depth of ~1500 m (Figure 4 6). Assuming a maximum rate of sedimentation, quartz approaches a mole fraction of 1 at a deeper depth (~1700 m), corresponding to the base of the sediment column (Figure 4 6). Excess pore pressures are slightly sensitive to changes in trench sedimentation rate (Figure 4 7). Excess pore pressures increase at depths >500 mbsf as sedimentation rates increase (Figure 4 7). Increasing the s edimentation rate increases the at the base of the incoming section from *=0.27 (assuming a minimum rate of sedimentation) to *=0.33 (assuming a maximum rate of sedimentation) at the deformation front (Figure 4 7). As a result of slightly higher exces s pore pressures, which reduce compaction due to loading, simulated porosity slightly increases as trench sedimentation rate increases (Figure 4 6). Initial smectite mole fraction Excess pore pressure ratios are not very sensitive to decreasing the assume d initial smectite mole fraction of the Surveyor Fan sediments from 0.8 to 0.5 (Figure 4 8). Smectite is mostly unreacted but undergoes slight diagenesis prior to subduction, with S~0.75 at the base of the incoming section at the deformation front. As a re sult, varying the smectite content resulted in no significant change in excess pore pressure at the deformation front.

PAGE 64

64 Incoming permeability porosity relationship Permeability and porosity measurements on eleven samples retrieved from Site U1417 and Site U1418 yield a data best fit k n relationship of log k= 18.823+4.1527n (Figure 3 6). This relationship differs from the k n relationships compiled for global subduction zone sediments by Daigle and Screaton (2015) based on the percentage of clay size grains predicting a higher permeability for a given porosity for samples U1417E 15R1 and U1417E 25R1 at depths of 478.82 and 574.75 mbsf, respectively (Figure 3 6). Sample U1417D 43X3 yielded a lower permeability for the measured porosity of 0.5 than the corres ponding Daigle and Screaton k n relationship based on percentage of clay sized grains by mass. Samples U1417D 54X2 and U1417E 39R7 correspond well with the Daigle and Screaton k n relationships. All samples from Site U1418 yield a higher permeability for a given porosity relative to the Daigle and Screaton k n relationships (Figure 3 6). This may reflect larger grain size and the proximity of Site U1418 to the shelf, outside of the modeling transect used in this investigation (Figure 1 1). Thus, it is possi ble that the samples analyzed from Site U1418 may not reflect the incoming section to the Aleutian Trench along the modeling transact. For this reason, permeability measurements from Site U1418 were excluded for sensitivity testing. A best fit k n relation ship of log k= 20.92+8.1132n was determined by fitting just the Site U1417 permeability, porosity measurements to a curve (Figure 3 6). Sensitivity to this relationship is tested in this modeling investigation. Sensitivity to the Site U1417 relationship w as tested by applying a uniform k n relationship equal to the best fit of the data (log k= 20.92+8.1132n) to the incoming section as well as the trench wedge sediments. The best fit k n relationship based on sample measurements, results in overall lower ex cess pore pressures at the deformation front (Figure 4 9). At a potential

PAGE 65

65 n applied (Figure 4 9). At the base of the incom ing 9). In the base run simulation, sediments recovered at Site U1417 were assigned k n relationships consistent with Daigle and Screaton (2015) while the trench wedge permeability distribution was varied to approximate the upward coarsening of sediments near the margin (Piper et al., 1973). In the base run simulation, a permeability porosity relationship log k= 20.862++8.39n was establishe d for trench packet 5 (sediment accumulation 50 60 km landward of Site U1417) and log k= 18.862+8.39n for Packet 6 (60 75 km landward of Site U1417) (Table 3 2). To isolate the effect of trench wedge permeability on pore pressure distribution at the deform ation front, this sensitivity simulation assumes a uniform k n relationship (log k= 21.971+9.78n) in the trench wedge, consistent with sediments recovered from Sequence III at Site U1417 (Table 3 2). Simulated excess pore pressures are only slightly sensit ive to changes in trench wedge sediment permeability (Figure 4 10). At a potential decollement depth of 1300 porosity distributions (Figure 4 10). At the base of the incoming colum n and variable k n relationship (Figure 4 10). Deformation Front to 60 Kilometers Landward The results of the 1 D sedimentation model at the deformation front were used as input for the subduction model, which tracked the final sediment column at the deformation front beneath the accretionary prism (Figure 3 2). Due to uncertainty in decollement location at the deformation front, several decollement scenarios were examined (Figure 4 11). Results are shown for three possibl e decollement zone locations assuming: ~1000 m of sediments are subducted at the deformation front (von Huene et al., 1998); ~600 meters are subducted (Gulick,

PAGE 66

66 personal communication) and ~300 meters are subducted. These depths are plotted in the 1 D model ing results at the deformation front (Figure 4 3 Figure 4 5) and tracked during subduction. Subduction results will be reported for the shallowest decollement location (~1000 m subducted at the deformation front) and the deepest decollement location (~30 0 m subducted at the deformation front) (Figure 4 12). Porosity reduction is greatest along the upper decollement zone, in which ~1000 m is subducted at the deformation front (Figure 4 13). Along this horizon, porosity is reduced from 0.4 at the deformati on front to 0.06 by 60 km landward (Figure 4 13). Along the deepest decollement zone, porosity decreases from ~0.35 at the deformation front to ~0.2 by 60 km landward (Figure 4 13). Simulated temperatures reach a maximum of ~150 C by 60 km landward assumin g that half of the incoming section is subducted (Figure 4 14). Smectite dehydration reaction progress and peak fluid production shift seaward with greater depth of the decollement zone (Figure 4 15). Peak fluid production from smectite dehydration is com puted to be ~3.4 x 10 16 s 1 (S=0.6), and the location shifts from ~25 km landward of the deformation front along the deepest decollement depth to ~50 km landward for the shallowest decollement depth (Figure 4 15). Opal A dehydration to quartz through opal CT shows a similar seaward shift in reaction progress and fluid production as decollement depth increases (Figure 4 16). Opal A dehydration to quartz is complete within ~25 km to ~45 km landward of the deformation front depending on decollement location ( Figure 4 16). Peak fluid production from opal dehydration is computed to be ~5.3 x 10 17 s 1 at ~30 km landward of the deformation front along the upper decollement zone (Figure 4 16). Opal dehydration is shifted seaward if the decollement follows a deeper location,

PAGE 67

67 in which peak fluids approach ~8x10 17 s 1 at ~10 km landward of the deformation front (Figure 4 16). Excess pore pressure and the excess pore pressure ratio ( *) generally increase during subduction along all simulated decollement horizons (Figure 4 17). At the deformation front, ranges from *=0.08 (1000 m subducted) to *=0.36 (300 m subducted) (Figure 4 17). Excess pore pressure ratios generally increase d uring subduction and range between *=0.32 (1000 m subducted) to *=0.73 (300 m subducted) (Figure 4 17). Effective stress (Lithostatic overburden pressure excess pore pressure) increases during subduction. At the deformation front, effective stress is l owest (~4 x 10 6 Pa) along the upper decollement zone that assumes 1000 m of sediments are subducted (Figure 4 18). Effective stress is slightly higher along the deepest decollement horizon at the deformation front, reaching ~7 x 10 6 Pa (Figure 4 18). By 60 km landward, however, effective stress increases to 4 x 10 7 Pa along the shallowest decollement zone (Figure 4 18). At this subduction distance, the deeper decollement horizon has a lower effective stress (~2 x 10 7 Pa) (Figure 4 18). To determine the sign ificance of dehydration reactions on excess pore pressures during subduction, smectite and opal dehydration reactions were removed from the model as a fluid source and results were compared to the base run simulation. Modeling results suggest that dehydrat ion reactions have a minor effect on excess pore pressure during subduction (Figure 4 19). Along the deepest decollement zone, decreases from 0.73 to by 60 km landward when dehydration reactions are excluded and is unchanged at the deformation f ront ( (Figure 4 19). The removal of dehydration reactions slightly increases effective stress along the decollement at 60 km landward of the deformation front, from ~1.8 x 10 7 Pa in the base run (Figure 4 18) to ~2.0 x 10 7 Pa (Figure 4 20).

PAGE 68

68 Sensi tivity Analyses Sensitivity runs for sedimentation seaward of the deformation front suggested that modeling results were most sensitive to temperature and sediment permeability. Thus, sensitivity to these parameters was tested in the 1 D subduction model. For base run comparisons, modeling results from the decollement zone representing subduction of ~300 m of sediment at the deformation front are reported (Figure 4 12). Temperature Sensitivity tests were performed to determine how varying the downgoing t emperature gradient at the deformation front affected reaction progress and pore pressure modeling results. Following sedimentation, a linear temperature gradient (dT/dx) of 0.001375 C/m was applied to the incoming sediment column landward of the deformati on front, which was computed based on an average of thermal modeling results. Based on these thermal models a high dT/dx (0.0025 C/m) and low dT/dx range (0.000625 C) were applied to the subducted sediments. The maximum temperature generated is ~200 C and ~100 C with the application of the high and low dT/dx value, respectively (Figure 4 21). Smectite reaction progress shifts seaward and fluid production increases with an increase in dT/dx (Figure 4 22). The peak fluid production of smectite ranges from ~6 .5 x 10 16 s 1 (16 km landward, high dT/dx) to ~1.6 x 10 16 (~50 km landward, low dT/dx) compare to the base run peak rate of ~3.4 x 10 16 ~25 km landward of the deformation front (Figure 4 22). By 60 km of subduction distance, the mole fraction of smect ite is highest (S=0.4) with the lowest dT/dx value and smallest (S=0.03) with the highest dT/dx value applied relative to the base run (S=0.1). Similar to the smectite dehydration curve, opal A dehydration to quartz shifts seaward with warmer temperatures (high dT/dx), although the effects are more subtle. The conversion of opal A to quartz occurs more rapidly with a warmer temperature gradient ap plied to the

PAGE 69

69 underthrust sediments (Figure 4 23). Opal A is converted to quartz by ~20 km and ~35 km landward of the deformation front assuming a high dT/dx and low dT/dx, respectively, relative to the base run dT/dx which reaches completion by ~25 km (Fig ure 4 23). Peak opal dehydration shifts between 10 km landward of the deformation front (high dT/dx) to ~11 km landward (low dT/dx), which is only a slight deviation from the base run peak distance of ~10.5 km (Figure 4 23). Peak fluid production is highes t with a warmer temperature gradient applied (~1.4 x 10 16 s 1 ) relative to the base run value of ~8 x 10 17 s 1 (Figure 4 23). Fluid production from opal decreases with colder temperatures, peaking at ~5 x 10 17 s 1 (Figure 4 23). Temperatures have a sli ght effect on excess pore pressures during subduction. Warmer temperatures slightly reduce from 0.73 (base run) to 0.68 by 60 km of subduction (Figure 4 24). Lowering temperatures slightly increases to by 60 km landward (Figure 4 24). Warmer temperatures increase effective stress relative to the base run (Figure 4 25). The effective stress by 60 km landward has increased by roughly 3.5 times (~2 x10 7 Pa to ~7 x10 7 Pa) the value computed in the base run simulation with a warmer temperature grad ient applied (Figure 4 25). Lowering temperatures does not affect effective stress significantly relative to the base run (Figure 4 25). Permeability Porosity Relationship porosity relationships was test ed by applying the best fit k n relationship from Site U1417 sample analysis (Figure 3 6) to the incoming sediments and accretionary prism. Generally, permeability is higher with the assumption of the Site U1417 k n relative to the grain size relationships (e.g. Daigle and Screaton, 2015) applied in the base run (Figure 4 9).

PAGE 70

70 Excess pore pressures decrease slightly relative to the base run with the Site U1417 k n relationship applied uniformly to the incoming sediments and prism (Figure 4 26). The excess p ore pressure ratio decreases from to at the deformation front and from to by 60 km landward (Figure 4 26). Effective stress is higher relative to the base run simulation at the deformation front and by 60 km, in which effect ive stress has increased from ~1.8 x 10 7 Pa (base run) to ~2.2 x 10 7 Pa (Figure 4 27).

PAGE 71

71 Table 4 1. Fluid influx rate from pore water, smectite and opal dehydration at Site U1417. Table 4 2. Fluid influx rate from pore water, smectite and opal dehydration at the deformation front, after the deposition of ~1 km of trench sediments.

PAGE 72

72 Figure 4 1. Modeling results at Site U1417 plotted with IODP Expedition 341 shipboard and drilling data (circles) (Jaeger et al., 2014). A) Simulated downhole porosity, B) thermal conductivity and C) temperature. Figure 4 2. Dehydration reaction progress of smectite and opal at Site U1417. A) Simulat ed mole fraction of smectite (S) in mixed layer smectite illite clay (I/S). A mole fraction of S=0.5(dashed) and S= 0.8 (solid) are assumed for sensitivity testing. B) Mole fraction of opal A, opal CT and quartz.

PAGE 73

73 Figure 4 3. Modeling results with the de position of trench sediments at the deformation front. A) Porosity, B) thermal conductivity and C) temperature plotted with depth below seafloor. Potential decollement locations are shown with horizontal dashed lines, assuming ~1000 m of sediment are under thrust at the deformation front, ~600 m of underthrust and ~300 m are underthrust. Figure 4 4. Dehydration reaction progress with the deposition of trench sediments. A) Mole fraction of smectite and B) opal A to quartz through opal CT. Potential decoll ement locations are shown with horizontal dashed lines, assuming ~1000 m of sediment are underthrust at the deformation front, ~600 m of underthrust and ~300 m are underthrust.

PAGE 74

74 Figure 4 5. Excess pore pressure distribution with the deposition of trench sediments. A) Excess pore pressure at Site U1417 (red) and the deformation front (blue) relative to lithostatic pressure (black). B) Excess pore pressure ratio ( at Site U1417 and the deformation front. Potential decollement locations are shown with ho rizontal dashed lines, assuming ~1000 m of sediment are underthrust at the deformation front, ~600 m of underthrust and ~300 m are underthrust. Figure 4 6. Sensitivity to trench sedimentation rate. A) Temperature, B) porosity and reaction progress of C) smectite and D) opal A to quartz through opal CT with maximum (red), minimum (blue) and average (green) sedimentation rates applied. Potential decollement locations are shown with horizontal dashed lines, assuming ~1000 m of sediment are underthrust at th e deformation front, ~600 m of underthrust and ~300 m are underthrust.

PAGE 75

75 Figure 4 7. Sensitivity of excess pore pressure to trench sedimentation rate (average, maximum and minimum). A) Excess pore pressure (dashed) relative to lithostatic pressure (solid) and B) excess pore pressure ratio with depth. Potential decollement locations are shown with horizontal dashed lines, assuming ~1000 m of sediment are underthrust at the deformation front, ~600 m of underthrust and ~300 m are underthrust. Figure 4 8. Sensitivity to smectite content for the trench wedge packets. A) Assumes that S=0.8 for the entire column and that S=0.5 for the Surveyor Fan section (after Hayes, 1973) and S=0.8 for pelagic clay. B) Excess pore pressure and lithostatic pressure at the de formation front. Potential decollement locations are shown with horizontal dashed lines, assuming ~1000 m of sediment are underthrust at the deformation front, ~600 m of underthrust and ~300 m are underthrust.

PAGE 76

76 Figure 4 9. Sensitivity to incoming k n rel ationship. A) Permeability as a function of depth at the deformation front. B) Excess pore pressure relative to lithostatic pressure (black) at the deformation front and C) excess pore pressure ratio ( at the deformation front. Potential decollement loc ations are shown with horizontal dashed lines, assuming ~1000 m of sediment are underthrust at the deformation front, ~600 m of underthrust and ~300 m are underthrust. Figure 4 10. Sensitivity to trench k n relationship. A) Excess pore pressure relative to lithostatic pressure (black) at the deformation front. B) Excess pore pressure ratio ( *) assuming a uniform k n relationship throughout the trench wedge (k o = 9.78; gamma= 21.971) and a variable k n relationship that reflects changing sediment compos ition near the deformation front (see Table 3 2).

PAGE 77

77 Figure 4 11. Depth of potential decollement locations in the incoming section at the deformation front and 60 km landward. Figure 4 12. Three decollement zone depth scenarios used in the subduction model base run. The upper and lower black line represent the top and base of the incoming section, respectively.

PAGE 78

78 Figure 4 13. Porosity with subduction distance for each decollement location. Figure 4 14. Temperature with subduction distance for eac h decollement zone.

PAGE 79

79 Figure 4 15. Smectite dehydration reaction progress and fluid production during subduction. A) Mole fraction of smectite with subduction distance for each decollement zone simulated. B) Peak fluid production (s 1 ) from smectite dehyd ration for each decollement zone.

PAGE 80

80 Figure 4 16. Opal dehydration reaction progress and fluid production during subduction. A) Mole fraction of opal A, opal CT and quartz with subduction distance for each decollement zone simulated. B) Peak fluid product ion (s 1 ) from opal dehydration for each decollement zone.

PAGE 81

81 Figure 4 17. Excess pore pressure distribution during subduction for each decollement location. A) Lithostatic overburden pressure (solid) and excess pore pressure (dashed) for each horizon. B ) Excess pore pressure ratio ( for each decollement horizon.

PAGE 82

82 Figure 4 18. Effective stress distribution during subduction for each decollement location.

PAGE 83

83 Figure 4 19. Excess pore pressure distribution without dehydration included as a fluid source. A) Lithostatic overburden pressure (black) and excess pore pressure (EPP) distribution (blue) without dehydration compared to the base run simulation. B) Excess pore pressure ratio without dehydration compared to the base run.

PAGE 84

84 Figure 4 20. Eff ective stress distribution without dehydration included as a fluid source. Figure 4 21. Sensitivity to temperature gradient with subduction distance (dT/dx). High dT/dx= 0.0025C/m, Low dT/dx=0.000625C/m, base run=0.001375C/m.

PAGE 85

85 Figure 4 22. Sen sitivity of smectite reaction progress and fluid production to temperature. A) Mole fraction of smectite in mixed layer clay. B) Peak fluid production from smectite.

PAGE 86

86 Figure 4 23. Sensitivity of opal dehydration reaction progress and fluid production to temperature. A) Mole fraction of opal A, opal CT and quartz. B) Peak fluid production from opal.

PAGE 87

87 Figure 4 24. Sensitivity of excess pore pressure to temperature. A) Lithostatic overburden pressure and excess pore pressure (EPP) distribution relative to the base run. B) Excess pore pressure ratio.

PAGE 88

88 Figure 4 25. Sensitivity of effective stress to temperature.

PAGE 89

89 Figure 4 26. Excess pore pressure distribution assuming k n relationship determined from Site U1417 sample analysis. A) Lithostatic overbur den pressure (solid) and excess pore pressure (dashed) distribution along decollement compared to base run. B) Excess pore pressure ratio along decollement compared to base run.

PAGE 90

90 Figure 4 27. Effective stress distribution assuming k n determined from Site U1417 sample analysis.

PAGE 91

91 CHAPTER 5 DISCUSSION The effects of trench sedimentation on the incoming sediment section become apparent when comparing model results from Site U1417 to model results from the deformation front (Figure 4 3 Figure 4 5). Trench sediments deposited in between Site U1417 and the deformation front insulate the underthrust sediments prior to subduction, causing temperatures to elevate (Figure 4 3) and reactions to begin prior to subduction (Figure 4 4). Sensitivity tests sugge st that results are not sensitive to initial smectite content in the incoming sediment section, which may be due to the limited reaction progress of smectite that has occurred at the deformation front. Uncertainty in smectite content is more likely to affe ct fluid production and excess pore pressure generation following subduction, where temperatures are highly elevated and dehydration is advanced. Trench sedimentation can result in excess pore pressure generation when the loading rate exceeds the drainage rate of fluids in low permeability sediments (Saffer and Bekins, 1998). Low permeability sediments, like those observed at Site U1417 and within the trench wedge, delay the 1 D model (Figure 4 5). Simulated overpressures at the deformation front are mostly due to compaction fluid sources. A minor component is from smectite, which is just beginning at the deformation front, and opal A, which is nearly completely transform ed to opal CT at the deformation front (Figure 4 4). Despite the influence of the trench wedge on advancing reaction progress, a majority of incoming smectite remains unreacted at the deformation front (Figure 4 4). The fluid influx budget suggests that ~ 1/2 of the total pore and mineral fluids are expelled in between Site U1417 and the deformation front due to trench sedimentation (Table 4

PAGE 92

92 1, Table 4 2). This result is consistent with observations by Suess et al. (1998), who cited seep biota, methane plum es and precipitates as evidence for active venting and expulsion of fluids near the deformation front. The suggestion that fluid expulsion near the deformation front is also consistent with estimates of maximum dewatering near the deformation front by von Huene et al., (1998). According to the porosity velocity relationship defined by von Huene et al. (1998), a porosity of 0.30 is estimated for the underthrust sediments near the deformation front, which is based on a seismic velocity of ~2000 m/s. Modeling results are consistent with this estimate, and suggest a porosity of 0.30 along the top of the underthrust section with the assumption that ~ 1/2 of the incoming section recovered at Site U1417 is subducted at the deformation front (Figure 4 14). von Huene et al. (1998) estimate a seismic velocity of 3300 3900 m/s for the underthrust sediments by ~25 km landward, which would suggest a porosity of 0.10 0.15. Modeling results suggest a porosity closer to 0.20 by this distance (Figure 5 1). Simulated porosity i s likely higher than estimated by von Huene et al. (1998) due to the fact that a permeable decollement and lateral flow are not included in the 1 D subduction model, which would drain excess pore pressures, increase effective stress and drive greater sedim ent consolidation. Decollement depth affects the distribution of excess pore pressure along the top of the underthrust section during subduction. There is a significant difference in the maximum generated at the deformation front and further landward d epending on whether the decollement forms above ~1 km of subducted sediments (max ~0.3) or above ~300 meters of underthrust sediment at the deformation front (max ~0.7). This difference may be due to greater upward fluid drainage through the accretiona ry prism and seafloor. During subduction, the expulsion of pore fluids is driven by tectonic loading of the accretionary prism. The porosity reduction along the upper decollement location (~1 km subducted at the deformation front) from the deformation

PAGE 93

93 fron t (n~0.4) to 60 km landward (n~0.1) supports the importance of upward fluid drainage to the seafloor through the accretionary prism is important. The significance of dehydration reactions on excess pore pressures was determined by removing dehydration rea ctions as a fluid source in the 1 D model with subduction. Comparing this simulation with the base run suggests that dehydration reactions have a small, but noticeable, effect on the location and magnitude of maximum excess pore pressure (Figure 4 19). The maximum decreases from 0.73 (base run) to without dehydration by 60 km of subduction (Figure 4 19). The fact that excess pore pressures are still near lithostatic by 60 km landward suggests that dehydration reactions are likely not as important as fluids sourced from compaction on excess pore pressure generation. This result is consistent with fluid influx estimates at the deformation front, which suggest that ~80% of the total fluid influx is sourced from pore fluids (Table 4 2). The excess po re pressure ratio distribution with and without dehydration is nearly indistinguishable until ~10 20 km landward, which is within the range of peak smectite and opal dehydration (Figure 4 15, Figure 4 16). This suggests that fluid sourced from dehydration become more significant after compaction sources driven by tectonic loading of the accretionary prism have dissipated. The location of dehydration reactions can slightly impact excess pore pressure distribution. Sensitivity tests suggest that dehydration reaction progress and excess pore pressures are sensitive to the downgoing temperature field during subduction (Figure 4 22 Figure 4 25). Higher temperatures drive smectite and opal dehydration seaward of the base run and generate a slightly larger peak fl uid source (Figure 4 22, Figure 4 23). Due to the fact that dehydration is shifted closer to the deformation front, where porosity and permeability are higher (Figure 4 13), excess pore pressures are lowered with respect to the base run simulation. As a

PAGE 94

94 re (Figure 4 24). Higher temperatures also reduce viscosity, which would effectively increase transmissivity and lower excess pore pressures. Lowering temperatures shi fts reaction progress and peak dehydration landward of the base run simulation (Figure 4 22, Figure 4 23), releasing fluids into material with lower porosity and permeability. Decreasing temperatures relative to the base run slows the reaction rate of smec tite and opal, releasing a lower volume of fluids into a similar volume of sediment over a longer period of time. However, shifting the reaction downdip releases fluids in lower permeability sediments. Thus, excess pore pressures are slightly elevated rela tive to the base run, 24). Excess pore pressure distribution in the 1 D simulations are sensitive to the permeability porosity relationships of the incoming sediments to the Aleutian Trench. The k n relationship developed from Site U1417 sample analysis (excluding the Site U1418 samples) had a small, but noticeable, effect on excess pore pressures seaward of the deformation front (Figure 4 9) and during subduction (Figure 4 2 6). The relative decrease in excess pore pressures associated with the Site U1417 k n relationship can be explained by the permeability distribution at the deformation front (Figure 4 9) which shows that, generally, the Site U1417 k n relationship yields a slightly higher permeability for a given porosity than the base run simulation that defined permeability based on grain size (Daigle and Screaton, 2015). Elevated sediment permeability reduces excess pore pressure generation by enabling greater fluid dra inage to the seafloor. Overall, the location and magnitude of maximum excess pore pressures during subduction is not significantly altered by changing the k n relationship of the underthrust section. In both the base run and Site U1417 k n sensitivity test

PAGE 95

95 26). This small difference may be due to the similar permeability and porosity distribution between Site U1417 samples and the Daigle and Screaton relationships (Figure 3 6), which results in o nly a slight change in excess pore pressures. Comparison to Other Subduction Margins To put the Aleutian Margin in the context of other subduction margins in terms of dehydration reaction progress, the following section compares results to those of previo us reaction progress modeling studies in four locations (Figure 5 2): 1) the Nankai margin of Japan (Saffer et al., 2008); 2) the Nicoya margin of Costa Rica (Spinelli and Saffer, 2004; Spinelli et al., 2006); 3) the Barbados accretionary complex (Bekins e t al., 1994); and 4) the Japan Trench (Kimura et al., 2012). All of the studies examined use the empirically derived kinetic expression of Pytte and Reynolds (1989) to compute reaction progress of smectite. Modeling results from these margins are compared to projected reaction progress and fluid production modeling presented in this study and discussed in the context of where reactions are occurring along the decollement zone. Nankai Margin, Japan Saffer et al. (2008) projected downdip reaction progress and fluid release of the subducting section at ODP Site 1173, drilled in the Nankai Trough offshore the Eastern Muroto Peninsula (Figure 5 2). Reaction progress modeling at the drilling site, ~10 km seaward of the deformation front, suggests that smectite und ergoes diagenesis outboard of the trench. Along the decollement zone, projected reaction progress indicates that the mole fraction of smectite approaches ~0.1 within 40 km landward the deformation front in Nankai, compared to ~60 km landward in the Aleutia n Margin (Saffer et al., 2008). Saffer et al. (2008) simulate the dcollement intersecting the modeled 150C isotherm ~25 km landward of the trench and peak

PAGE 96

96 smectite fluid sources of ~6 x 10 15 s 1 generated 16 km landward of the trench (Figure 5 3). In t he Aleutian Margin, smectite remains mostly unreacted at the deformation front and projected temperatures approach 150C by ~60 km landward (Figure 4 15). Peak smectite fluid sources are generated ~10 km landward and of a lower value (~3 x 10 16 s 1 ) in the Aleutian Margin than the fluid source from smectite projected in Nankai (Figure 5 3; Saffer et al., 2008). The seaward shift in reaction progress in Muroto is a response to a higher heat flow in Muroto, which has been attributed to extensive hydr othermal circulation and redistribution of heat in the oceanic crust (Spinelli and Wang, 2008). This redistribution acts to increase temperatures outboard of the trench and drive pre subduction diagenesis relative to the Aleutian Margin, where much of the incoming smectite remains unreacted at the deformation front despite the insulating effect of trench sediments. Taper angle also contributes to the thermal structure of the forearc. A steeper taper angle at Muroto drives diagenesis closer to the trench wit h all other factors constant (Saffer et al., 2008). Both Muroto and the Aleutian Margin have a similar taper angle of ~4 5 at the deformation front, although von Huene et al. (1998) report a steepening to 9 approximately 10 km landward of the trench alon g the modeling transect in the Aleutian Margin. Thus, higher heat flow at Muroto compensates for the lower taper angle, and acts to drive diagenesis in Muroto closer to the deformation front relative to the Aleutian Margin. The Muroto transect is located on a basement high formed by the Kinan seamount chain and has a larger percentage of mud/clay within the incoming section (82 7 % after Saffer and Bekins, 1998 and Moore and Saffer, 2001) relative to the eastern Aleutian margin sediments (66 30% after v on Huene and Kulm, 1973; Davis and von Huene, 1987) which show larger variability. Furthermore, XRD analysis on samples from Site 1173 indicate that smectite comprises between 35 to 60 wt % of the clay sized fraction (Steurer and Underwood, 2003)

PAGE 97

97 which is less than that reported for most of the incoming section offshore Alaska (Hayes, 1973). Thus, the larger peak fluid production in Nankai is likely a response to a greater smectite abundance, which causes a larger volume of bound water. Costa Rica Margin ( Nicoya) Costa Rica, offshore the Nicoya Peninsula, lacks a well developed accretionary prism and is characterized as an erosional margin with a thin (~350 m) incoming sediment section that includes hemipelagic sediments overlying carbonates (Figure 5 2; Sh ipboard Scientific Party, 1997). Offshore Nicoya, a triple junction divides the Cocos plate into crust formed at the East Pacific Rise (EPR) and the Cocos Nazca Spreading Center (CNS). Seafloor heat flow is anomalously cool on the EPR crust, dated at ~24 Ma offshore Nicoya, which has been attributed to hydrothermal circulation of seawater through basaltic outcrops (Fisher et al., 2003; Spinelli and Saffer, 2004; Spinelli et al., 2006). Spinelli and Saffer (2004) modeled temperatures, reaction progress of s mectite and opal and fluid production in underthrust sediment seaward of ODP Site 1039 on the EPR and CNS portion of the crust. In their model, a number of hydrothermal cooling scenarios (1 2 km of upper crust cooled) were examined to account for lower hea t flow in the EPR transect. Modeling of incoming sediments along the cooler, EPR oceanic crust project a peak smectite fluid source of ~5 x 10 14 s 1 ~40 km landward of the deformation front, assuming hydrothermal cooling of 1 km of upper crustal material (Figure 5 3; Spinelli and Saffer, 2004). Spinelli and Saffer (2004) modeled the reaction progress of opal A to quartz using the kinetic expressions developed by Ernst and Calvert (1969). Ernst and Calvert (1969) studied the conversion of opal CT to quartz under hydrothermal conditions with experiments carried out in distilled water at 300, 400 and 500C. T he kinetic expression developed by Mizutani (1970) that was applied in this modeling investigation models the conversion of opal A to opal CT and opal

PAGE 98

98 CT to quartz. Spinelli and Saffer (2004) simulate a steep conversion of opal CT to quartz by 35 km landward of the deformation front, producing a peak fluid source of ~10 14 s 1 Fundamental differences in temperature of the subducted sediment, convergence rate and incoming sedimentation controls variability in smectite reaction progress. Offshore southern Alaska, the transition of opal CT to quartz is more gradual than projected in Costa Rica, which could reflect the differences in kinetic expressions used to model dehydration reaction progress. Because the Ernst and Calvert (1969) kinetic expression is a zero order reaction, it quickly runs to completion. A peak fluid source from opal dehydration of 8 x 10 17 s 1 10 km landward of the deformation front is projected in the Aleutian Margin, which is seaward (by ~25 km) and three orders of magnitude lower than projected for Costa Rica (Spinelli and Saffer, 2004). Offshore Alaska, thick trench deposition insulates the incoming sediment column, resulting in elev ated temperatures relative to subduction inputs on the EPR crust offshore Costa Rica, which undergoes extensive hydrothermal cooling (Spinelli et al., 2006). Rotman and Spinelli (2013) project that hydrothermal circulation has an insignificant effect on te mperatures in the Aleutian Margin. Differences in convergence rate may also affect reaction progress of smectite and opal in these margins. In Costa Rica, the Cocos Plate subducts beneath the Caribbean Plate at a rate of ~8.5 cm/yr (Spinelli and Saffer, 20 04) which is higher than the convergence rate offshore southern Alaska of ~5 cm/yr. A higher convergence rate leads to delayed reaction, due to a lower residence time, which factors into the empirically derived kinetic expression for smectite transformatio n derived by Pytte and Reynolds (1989) and opal (Mizutani, 1970). Differences in the magnitude of fluids produced from opal dehydration may in part be due to the kinetic expressions used to model reaction progress as well as the higher initial incoming opa l content offshore Costa Rica (10 wt%).

PAGE 99

99 Barbados Accretionary Prism The Barbados accretionary complex forms where the Atlantic Plate subducts beneath the Caribbean Plate (Figure 5 2; Bekins et al., 1994). The incoming sediment section is ~700 m thick and c onsists of mainly calcareous clays and mudstone with localized ash layers (Bekins et al., 1994). Smectite measures 78 100% of interlayered bulk I/S clay within samples from ODP Site 672, which penetrated the incoming sediment section offshore the accretion ary complex (Bekins et al., 1994). Bekins et al. (1994) report a prism 125 km wide with a taper angle of 3.2 along the modeling transect. Modeled reaction progress of smectite projected that the transformation of smectite to illite takes place over 60 80 km and the peak rate of fluid release from smectite in the underthrust sediments is 2 x 10 15 s 1 and are located ~75 km landward of the deformation front (Figure 5 3). Bekins et al. (1994) model a maximum temperature of 130C by ~120 km landward of the d eformation front, reflecting the much colder temperature regime in Barbados relative to the modeled temperatures in the Aleutian Margin (temperatures reach 150C by ~60 km landward) due to the effects of trench sedimentation. Colder temperatures also refle ct the older subducting crust in Barbados, of probable Cretaceous age (Shipboard scientific party, 2001). In addition, the shallower taper angle offshore Barbados relative to the Aleutian Margin (4 9) leads to delayed reaction progress due to the effect o f taper angle on burial rate of underthrust sediments. Japan Trench, Tohoku The Japan Trench forms where the Pacific Plate subducts beneath northeast Japan at a rate of 8 cm/yr (Figure 5 2; Kimura et al., 2012). Kimura et al. (2012) suggest that shallow slip to the trench during the 2011 Tohoku earthquake was driven by the localiz ation of highly pressurized fluids resulting from dehydration of hydrous minerals opal and smectite. Kimura et al. (2012) projected downdip reaction progress of smectite and opal landward of the Japan

PAGE 100

100 Trench using the same kinetic expression for the transf ormation of opal A to quartz through opal CT (Mizutani, 1970) as this modeling study. Tohoku presents an important comparison with the Aleutian margin due to the subduction of a thin layer of smectite rich pelagic clay in both margins, which has been hypot hesized to affect frictional properties at the Japan Trench (Wang et al., 2013; Ujiie et al., 2013). Offshore northeast Japan, drilling of the plate boundary megathrust revealed that the decollement is localized in pelagic clay. While the location of the d ecollement is not precisely known in the Aleutian Margin, pelagic clay was recovered above basement during drilling of DSDP Site 178. Kimura et al. (2012) project that most of the opal A transformation to quartz occurs within 80 km of subduction, while the dehydration of smectite occurs farther landward. Kimura et al. (2012) report peak dehydration rate in (g/m 2 /yr), however, assuming a vertical underthrust sediment thickness of 400 m, a peak smectite fluid source of ~3 x 10 16 s 1 and a peak opal fluid sou rce of ~7 x 10 16 s 1 were calculated (assuming a seawater density of 1030 kg/m 3 ). Peak fluids sourced from opal dehydration occur within 50 km of the deformation front while peak smectite dehydration occurs within ~100 km (Kimura et al., 2012). These dist ances are significantly landward of where these reactions occur in the Aleutian Margin, due to the insulating effect of trench sediments offshore Alaska. Comparison of reaction modeling results to other subduction margins highlights the complexity of fact ors that control the spatial and temporal distribution of fluids from mineral dehydration and compaction. Primarily, heat flow is critical, and depends on crustal age and hydrothermal circulation. Incoming sedimentation plays an important role in the distr ibution of temperature, permeability, porosity and initial hydrous mineralogy of the underthrust sediments.

PAGE 101

101 In addition, taper angle and convergence rate can play a role on controlling where fluids sourced from dehydration are released along the plate int erface. Implications At other subduction margins, dehydration reactions have been proposed to affect excess pore pressures and mechanical behavior during subduction (Ranero et al., 2008; Kimura et al., 2012). The simulated effects at the Aleutian margin a re small, perhaps due to low abundances of incoming smectite and the reactions taking place near the deformation front relative to previously studied margins excluding the Muroto Transect of Nankai (Figure 5 3). Offshore Alaska, the insulating effect of t he trench sediments deposited in between Site U1417 and the deformation front act to drive dehydration of smectite and opal and bound water release close to the deformation front. As a result, smectite and Opal A become exhausted as a significant fluid sou rce rapidly, and do not contribute significantly to excess pore pressure generation further landward. A comparison of reaction progress modeling results from several subduction margins (Figure 5 3) suggests that dehydration reactions may have a more signi ficant impact on excess pore pressure where they dehydrate farther downdip, where permeability and porosity are lower. In the Japan Trench and the Barbados accretionary prism, projected peak smectite dehydration is within 80 100 km landward of the deformat ion front (Figure 5 3). At these margins, dehydration reactions are likely to have a greater impact on excess pore pressure. In the Nankai Trough, reaction progress modeling places peak smectite dehydration seaward of the Aleutian margin, suggesting that t hese reactions are not likely to be significant in this margin, due to dewatering close to the trench (Figure 5 3). Excess pore pressures have been linked to a number of subduction zone mechanical behaviors, including the location of the transition to sei smic behavior along the plate interface

PAGE 102

102 (Saffer and Tobin, 2011). Elevated fluid pressures control fault strength by reducing the effective and shear stress acting on the fault plane, which promotes slip tendency (Saffer and Tobin, 2011). Modeling results suggest that excess pore pressures build to near lithostatic ( =0.73) by 60 km of subduction. However, this is likely an overestimate due to the fact that the model does not include the simulation of a permeable decollement or faulting near the deformatio n front and within the accretionary prism. Fault zones provide higher transmissivity pathways for fluids to migrate and multiple lines of evidence suggest that the decollement acts as a fluid pathway, channeling fluids laterally and seaward. These processe s would act to drain fluids and reduce excess pore pressures. Previous studies suggest that excess pore pressures and effective stress are an important control on decollement zone propagation and sliding behavior along the plate interface (Bangs et al. 2004; Saffer and Tobin, 2009; Saffer and Tobin, 2011; Rowe et al., 2012). An examination of potential decollement locations suggests that the lowest simulated effective stress occurs with the subduction of the thickest sediment package at the deformation front (~1 km). Previous studies suggest that the decollement propagates along zones of mechanical weakness, characterized by high pore pressure and low effective stress (Morgan and Karig, 1995). Following this decollement location landward, effective str ess increases relative to the other potential decollement locations (Figure 4 18), which could contribute to decollement downstepping during subduction.

PAGE 103

103 Figure 5 1. Simulated porosity compared to the empirical porosity velocity relationship determined by von Huene et al. (1998). Porosities in the lower figure from von Huene et al. (1998) are interpreted from seismic velocity. Figure 5 2. Locations of subduction margins used for dehydration reaction comparison. Ocean drill sites used to characterize the incoming section for modeling are shown in green. Based on modeling of the Barbados Accretionary Complex (Bekins et al., 1994); Costa Rica Margin (Spinelli et al., 2006); Nankai Trough (Saffer et al., 2008); Japan Trench (Kimura et al., 2012) and Aleu tian Margin (this study).

PAGE 104

104 Figure 5 3. Distance landward of deformation front along the decollement of projected peak fluid release from smectite dehydration based on previous modeling studies of the Barbados Accretionary Complex (Bekins et al., 1994); Costa Rica Margin (Spinelli et al., 2006); Nankai Trough (Saffer et al., 2008); Japan Trench (Kimura et al., 2012) and Aleutian Margin (this study).

PAGE 105

105 CHAPTER 6 CONCLUSIONS Numerical modeling was used to investigate the effect of trench sedimentation on dehydration reaction progress and excess pore pressure development of sediments on the incoming plate to the Aleutian margin. In this region, ongoing uplift of the St. Elias Mountain range, carved pathways for sediment transport and intensification of gla ciation following the mid Pleistocene transition (~1 Ma) resulted in an increase in sediment flux to the Gulf of Alaska. This increase is recorded in distal DSDP and IODP ocean drill sites by a near doubling of sedimentation rates and an increase in the ab undance of ice rafted debris. The incoming section to the Aleutian Trench is mud dominated, and, sample analysis indicates that unreacted smectite and biogenic silica are present as hydrous phases on the incoming plate. This modeling investigation provide s the first estimates of dehydration reaction progress and pore pressures along the modeling transect (Figure 1 1). To address the effects of trench sedimentation, a 1 D sedimentation was developed in MATLAB to track the incoming sediment section from bar e crust to the deformation front. In the model, porosity was extrapolated from IODP Expedition 341 Site U1417 and permeability porosity relationships were assigned based on sample analysis and a global subduction zone permeability dataset developed by Daig le and Screaton (2015). Seaward of the deformation front, temperatures were calculated from a 1 D transient thermal conduction model. Dehydration reaction progress of smectite and opal were computed using kinetic expressions from previous modeling investig ations. A 1 D transient finite difference approximation was used to solve for hydraulic head and increment excess pore pressure in the incoming and underthrust sediment section.

PAGE 106

106 Landward of the deformation front, a 1 D subduction model incorporated sedime ntation of the accretionary prism and tracked the underthrust sediment column beneath the prism. During subduction, temperatures were computed based on a temperature gradient determined from a compilation of previous thermal modeling results of the region. This approach does not include the effects of frictional heating on thermal properties. A limitation of this modeling approach is that it is does not simulate the effects of a permeable decollement and lateral flow, which affect pore fluid pressures by re stricting fluid drainage upward through the seafloor. Model results illustrate that trench sedimentation has a significant effect on dehydration reaction progress and excess pore pressure distribution. Trench sediments insulate the underthrust sediments p rior to subduction, causing temperatures to elevate (Figure 4 3) and reaction of smectite to begin prior to subduction (Figure 4 4). However, smectite remains mostly unreacted at the deformation front (Figure 4 4). The estimated total fluid influx from por e, smectite and opal fluids at Site U1417 is ~23 m 3 /yr (Table 4 1). Approximately half of the fluids in the incoming sediment section are expelled in between Site U1417 and the deformation front due to loading of the trench sediments and compaction of the underlying sediments. Along the proto decollement, excess pore pressures are above hydrostatic and ~0.4 (Figure 4 5). To track the downgoing sediment column during subduction, a number of decollement locations were examined (Figure 4 11). Assuming ~half of the incoming Surveyor Fan section is subducted at the deformation front, porosity is ~0.3 at the deformation front and approaches ~0.2 by 60 km landward, which matches reasonably well with the velocity porosity relation developed by von Huene et al. (19 98). Temperatures approach ~150C by 60 km landward (Figure 4 14). Peak smectite fluid production (~3 x 10 16 s 1 ) occurs ~25 km landward of the deformation front (Figure 4 15). Projected opal dehydration occurs seaward of smectite

PAGE 107

107 dehydration, and peak f luid production (~8 x 10 17 s 1 ) is at ~10 km landward of the deformation front (Figure 4 16). Excess pore pressures are near lithostatic (maximum =0.73) by 60 km landward of the deformation front (Figure 4 17), although this estimate does not include th e effect of a permeable decollement or faulting at the deformation front and within the accretionary prism, which would act to increase fluid drainage and reduce excess pore pressures. Model results suggest that dehydration reactions have a minor effect on excess pore pressures during subduction, likely due to thick trench sedimentation that drives mineral dehydration near the trench (Figure 4 19). Thus, fluids sourced from compaction appear to have the greatest control on excess pore pressure distribution during subduction in Alaska. Sensitivity analyses suggest that model results are most sensitive to temperature gradient and sediment permeability. Increasing the downgoing temperature gradient shifts smectite and opal dehydration seaward and produces a la rger peak fluid production. As a result, excess pore pressures are slightly decreased, due to fluid release in higher porosity and permeability sediment. Seaward of the deformation front, model results are most sensitive to the permeability porosity relati onship determined from permeability measurements on samples from Site U1417 and Site U1418 (Figure 3 6). During subduction, the permeability porosity relationship determined from Site U1417 sample analysis has a minor effect on excess pore pressures relati ve to the base run simulation, which applies permeability porosity relationship from a global subduction zone dataset developed by Daigle and Screaton (2015). At other subduction margins, dehydration reactions have been proposed to affect excess pore pres sures and mechanical behavior during subduction (Ranero et al, 2008; Kimura et al, 2012). The simulated effects at the Alaska margin are small, perhaps due to the reactions taking place near the deformation front. Comparison of reaction modeling results in Alaska to previous

PAGE 108

108 results from other subduction margins illustrates the variety of factors including incoming sedimentation, heat flow, hydrous mineralogy and taper angle that control where bound fluids are released along the plate interface, which has i mportant implications for excess pore pressure smectite within 25 km of the trench in Alaska, due to the trench sedimentation. In contrast, similar simulations of the Japan Trench and Barbados ridge complex indicate that smectite dehydration peaks farther landward (~80 100 km). The thick trench section offshore Alaska insulates and rapidly consolidates the underthrust sediments at the deformation front, which drive s dewatering close to the trench. Previous studies suggest that excess pore pressures are an important control on mechanical behavior along the plate interface. Modeling results suggest that the lowest simulated effective stress occurs with the subduction of ~1 km of sediment at the deformation front, which may promote decollement propagation at this horizon. Smectite alteration has also been proposed to affect frictional properties and slip behavior during subduction due to intrinsically weak nature of sm ectite rich clay and its tendency to weaken during slip (Wang et al., 2013); Ujiie et al., 2013). At the deformation front, smectite is mostly unreacted, but dehydration peaks with 25 km of the trench.

PAGE 109

109 LIST OF REFERENCES Athy, L.F., 1930, Density, porosity, and compaction of sedimentary rocks, Am. Assoc. Pet. Geol. Bull., 14, p. 1 23. Bangs, N. L., Shipley, T. H., Gulick, S. P. S., Moore, G. F., Kuromoto, S., and Nakamura, Y., 2004c, Evolution of the Nankai Trough decollement from the trench into the seismogenic zone: Inferences from three dimensional seismic reflection imaging: Geology, v. 32, no. 4, p. 273 276. Bekins, B. A., and Dreiss, S. J., 1992, A simplified analysis of parameters controlling dewatering in accretionary prisms: Earth and Pla netary Science Letters, v. 109, no. 3 4, p. 275 287. Bekins, B., McCaffrey, A. M., and Dreiss, S. J., 1994, Influence of kinetics on the smectite to illite transition in the Barbados accretionary prism: Journal of Geophysical Research Solid Earth, v. 99, no. B9, p. 18147 18158. Bekins, B. A., McCaffrey, A. M., and Dreiss, S. J., 1995, Episodic and constant flow models for the origin of low chloride waters in a modern accretionary complex: Water Resources Research, v. 31, no. 12, p. 3205 3215. Berger, A. L., Gulick, S. P. S., Spotila, J. A., Upton, P., Jaeger, J. M., Chapman, J. B., Worthington, L. A., Pavlis, T. L., Ridgway, K. D., Willems, B. A., and McAleer, R. J., 2008, Quaternary tectonic response to intensified glacial erosion in an orogenic wedge: N ature Geoscience, v. 1, no. 11, p. 793 799. Bethke, C. M., 1985, A numerical model of compaction driven groundwater flow and heat transfer and its application to the paleohydrology of intracratonic sedimentary basins: Journal of Geophysical Research Solid Earth and Planets, v. 90, no. NB8, p. 6817 6828. Biscaye, P. E., 1965, Mineralogy and sedimentation of recent deep sea clay in Atlantic Ocean and adjacent seas and oceans: Geological Society of America Bulletin, v. 76, no. 7, p. 803 &. Bode, G.W., 1973, Appendix V: Grain size analyses, Leg 18, Deep Sea Drilling Project, Initial Reports of the Deep Sea Drilling Project, Vol. XVIII: Washington, D.C., U.S. Government Printing Office. Brindley, G.W., Brown G., 1980, Crystal Structures of Clay Minerals and t heir X ray Identification. Mineralogical Society Monograph no. 5. Mineralogical Society, London, 495 pp. Calvert, S. E., 1983, Sedimentary geochemistry of silicon: Silicon geochemistry and biogeochemistry, Academic Press, London. p. 143 186.

PAGE 110

110 Clift, P., and Vannucchi, P., 2004a, Controls on tectonic accretion versus erosion in subduction zones: Implications for the origin and recycling of the continental crust: Reviews of Geophysics, v. 42, no. 2. Daigle, H., and Screaton, E. J., 2015a, Evolution of sediment permeability during burial and subduction: Geofluids, v. 15, no. 1 2, p. 84 105. Davis, D. M., and von Huene, R., 1987a, Inferences on sediment strength and fault friction from structures at th e Aleutian trench: Geology, v. 15, no. 6, p. 517 522. DeMaster, D.J., 1981, The supply and removal of silica from the marine environment: Geochimica et Cosmochimica Acta 45, p. 1715 1732. DeMaster, D. J., 2003, The diagenesis of biogenic silica: chemical transformations occurring in the water column, seabed, and crust: Treatise on Geochemistry, v. 7, p.97 98. Doubrovine, P. V., and Tarduno, J. A., 2008a, A revised kinematic model for the relative motion between Pacific oceanic plates and North America si nce the Late Cretaceous: Journal of Geophysical Research Solid Earth, v. 113, no. B12. Duncan, J. R., Kulm, L. D., and Griggs, G. B., 1970, Clay mineral composition of late Pleistocene and Holocene sediments of Cascadia basin, northeastern pacific ocean: Journal of Geology, v. 78, no. 2, p. 213 &. Ernst, W. G. and S. E. Calvert ., 1969 An experimental study of the recrystallization of porcelanite and its bearing on the origin of some bedded cherts : Am. J. Sci. 267A p. 114 133 Ferguson, I. J., Westbrook, G. K., Langseth, M. G., and Thomas, G. P., 1993, Heat flow and thermal models of the Barbados ridge accretionary complex: Journal of Geophysical Research Solid Earth, v. 98, no. B3, p. 4121 4142. Fisher, A.T., and Zwart, G., 1997, Packer experiments along the dcollement of the Barbados accretionary complex: measurements of in situ permeability: Proc. ODP, Sci. Results, 156, p. 199 218. Fisher, A. T., Stein, C. A., Harris, R. N., Wang, K., Silver, E. A., Pfender, M., Hutnak, M., Cherkaoui, A., Bodzin, R., and Villinger, H., 2003, Abrupt thermal transition reveals hydrothermal boundary and role of seamounts within the Cocos Plate: Geophysical Research Letters, v. 30, no. 11, p. 4.

PAGE 111

111 Gulick, S. P. S., Jaeger, J. M., Mix, A. C., Asahi, H., Bahlburg, H., Belanger, C. L., Berbel, G. B. B., Childress, L., Cowan, E., Drab, L., Forwick, M., Fukumura, A., Ge, S. L., Gupta, S., Kioka, A., Konno, S., LeVay, L. J., Marz, C., Matsuzaki, K. M., McClymont, E. L ., Moy, C., Muller, J., Nakamura, A., Ojima, T., Ribeiro, F. R., Ridgway, K. D., Romero, O. E., Slagle, A. L., Stoner, J. S., St Onge, G., Suto, I., Walczak, M. D., Worthington, L. L., Bailey, I., Enkelmann, E., Reece, R., and Swartz, J. M., 2015, Mid Plei stocene climate transition drives net mass loss from rapidly uplifting St. Elias Mountains, Alaska: Proceedings of the National Academy of Sciences of the United States of America, v. 112, no. 49, p. 15042 15047. Gutscher, M. A., and Peacock, S. M., 2003, Thermal models of flat subduction and the rupture zone of great subduction earthquakes: Journal of Geophysical Research Solid Earth, v. 108, no. B1. Gamage, K., and Screaton, E., 2006, Characterization of excess pore pressures at the toe of the Nankai accretionary complex, Ocean Drilling Program sites 1173, 1174, and 808: Results of one dimensional modeling: Journal of Geophysical Research Solid Earth, v. 111, no. B4, p. 13. Harris, R. N., Grevemeyer, I., Ranero, C. R., Villinger, H., Barckhausen, U., Henke, T., Mueller, C., and Neben, S., 2010a, Thermal regime of the Costa Rican convergent margin: 1. Along strike variations in heat flow from probe measurements and estimated from bottom simulating reflectors: Geochemistry Geophysics Geosystems, v. 11. Hayes, J.B., 1973, Clay petrology of mudstones, Leg 18, Deep Sea Drilling Project, Initial Reports of the Deep Sea Drilling Project, Vol. XVIII: Washington, D.C., U.S. Government Printing Office, p. 903 924. Hower, J., Eslinger, E. V., Hower, M. E., and Perry, E. A., 1976, Mechanism of burial metamorphism of argillaceous sediment .1. Mineralogical and chemical evidence: Geological Society of America Bulletin, v. 87, no. 5, p. 725 737. Jaeger, J.M., Gulick, S.P.S., LeVay, L.J., and the Expedition 341 Scie ntists, 2014, S outhern Alaska Margin: Proceedings of IODP Exp. 341 College Station, TX, Integrated Ocean Drilling Program. Kastner, M., 1981, Authigenic silicates in deep sea sediments, in The Sea, vol. 7, edited by C. Emiliani, Wiley Interscience, New York., p. 915 980. Kastner, M. Solomon E.A., Harris R.N., and Torres M.E., 2014 Fluid origins, thermal regimes, and flui d and solute fluxes in the forearc of subduction zones in Earth and Life Processes Discovered from Subseafloor Environments: A Decade of Science Achieved by the Integrated Ocean Drilling Program (IODP) in Mar. Geol. vol. 7 edited by R. Stein et al., Elsevier, Amsterdam, Netherlands p. 671 733

PAGE 112

112 Kimura, G., Hina, S., Hamada, Y., Kameda, J., Tsuji, T., Kinoshita, M., and Yamaguchi, A., 2012, Runaway slip to the trench due to rupture of highly pressurized megathrust beneath the middle trench slope: The tsunamigenesis of the 2011 Tohoku earthquake off the east coast of northern Japan: Earth and Planetary Science Letters, v. 339, p. 32 45. Huene, R., and Kulm, L. D., 1973, Tectonic summary of Leg 18, Initial reports of the Deep Sea Drilling Pr oject, v. XVIII: Washington, D.C., U.S. Government Printing Office p. 961 976. Lagoe, M. B., Eyles, C. H., Eyles, N., and Hale, C., 1993, Timing of late Cenozoic tidewater glaciation in the far North Pacific: Geological Society of America Bulletin, v. 105, no. 12, p. 1542 1560. Mizutani, S., 1970. Silica minerals in the early stage of diagenesis: Sedimentology, v. 15, p. 419 436. Moore, D.M. and Reynolds, R.C., 1989, X ray Diffraction and the Identification and Analysis of Clay Minerals, Oxford Univer sity Press, Oxford. Moore, G. F., Shipley, T. H., Stoffa, P. L., Karig, D. E., Taira, A., Kuramoto, S., Tokuyama, H., and Suyehiro, K., 1990, Structure of the Nankai Trough accretionary zone from multichannel seismic reflection data: Journal of Geophysical Research Solid Earth and Planets, v. 95, no. B6, p. 8753 8765. Moore, J. C., Diebold, J., Fisher, M. A., Sample, J., Brocher, T., Talwani, M., Ewing, J., Vonhuene, R., Rowe, C., Stone, D., Stevens, C., and Sawyer, D., 1991b, EDGE deep seismic reflection transect of the eastern Aleutian arc trench layered lower crust reveals underplating and continental growth: Geology, v. 19, no. 5, p. 420 424. Moore, J. C., 1992, Fluids in accretionary prisms (reviews of geophysics, vol 30, pg 113, 1992): Rev iews of Geophysics, v. 30, no. 4, p. 353 353. Moore, J. C., and Saffer, D., 2001, Updip limit of the seismogenic zone beneath the accretionary prism of southwest Japan: An effect of diagenetic to low grade metamorphic processes and increasing effective st ress: Geology, v. 29, no. 2, p. 183 186. Morgan, J. K., and Karig, D. E., 1995, Decollement processes at the Nankai accretionary margin, southeast Japan propagation, deformation, and dewatering: Journal of Geophysical Research Solid Earth, v. 100, no. B8, p. 15221 15231. Morgan, J. K., and Ask, M. V. S., 2004, Consolidation state and strength of underthrust sediments and evolution of the decollement at the Nankai accretionary margin: Results of uniaxial reconsolidation experiments: Journal of Geophysic al Research Solid Earth, v. 109, no. B3.

PAGE 113

113 Neuzil, C. E., 1986, Groundwater flow in low permeability environments: Water Resources Research, v. 22, no. 8, p. 1163 1195. Neuzil, C.E., 1994, How permeable are clays and shales: Water Resources Research, v. 30, no. 2, p. 145 150. Neuzil, C. E., 1995, Abnormal pressures as hydrodynamic phenomena: American Journal of Science, v. 295, no. 6, p. 742 786 Oleskevich, D. A., Hyndman, R. D., and Wang, K. 1999, The updip and downdip limits to great subduction earthquakes: Thermal and structural models of Cascadia, south Alaska, SW Japan, and Chile: Journal of Geophysical Research Solid Earth, v. 104, no. B7, p. 14965 14991. Park, J. O., Tsuru, T., Takaha shi, N., Hori, T., Kodaira, S., Nakanishi, A., Miura, S., and Kaneda, Y., 2002, A deep strong reflector in the Nankai accretionary wedge from multichannel seismic data: Implications for underplating and interseismic shear stress release: Journal of Geophys ical Research Solid Earth, v. 107, no. B4. Parsons, B., and Sclater, J. G., 1977, Analysis of variation of ocean floor bathymetry and heat flow with age: Journal of Geophysical Research, v. 82, no. 5, p. 803 827. Perry, E., and Hower, J., 1970, Burial di agenesis in gulf coast pelitic sediments: Clays and Clay Minerals, v. 18, no. 3, p. 165 &. Piper, D. J.W., Von Huene, R., and Duncan, J. R., 1973, Late Quaternary sedimentation in the active eastern Aleutian Trench: Geology, p. 19 22. Plafker, G., and Berg, H.C., 1994, Introduction, in Plafker, George, and Berg, H.C., eds., The Geology of Alaska: Geological Society of America, p. 1 16. Ponko, S. C., and Peacock, S. M., 1995, Thermal modeling of the southern Alaska subduction zone ins ight into the petrology of the subducting slab and overlying mantle wedge: Journal of Geophysical Research Solid Earth, v. 100, no. B11, p. 22117 22128. Pytte, A.M., and Reynolds, R.C., 1989, The thermal transformation of smectite to illite, in Thermal Hi stories of Sedimentary Basins, edited by N.D. Naeser and T.H. McCulloh, Springer Verlag, New York, p. 133 140. Ranero, C. R., Grevemeyer, I., Sahling, H., Barckhausen, U., Hensen, C., Wallmann, K., Weinrebe, W., Vannucchi, P., von Huene, R., and McIntosh, K., 2008, Hydrogeological system of erosional convergent margins and its influence on tectonics and interplate seismogenesis: Geochemistry Geophysics Geosystems, v. 9.

PAGE 114

114 Rea, D. K., and H. Snoeckx, 1995, Sediment fluxes in the Gulf of Alaska: Paleoceanogra phic record from Site 887 on the Patton Murray Seamount Platform: Proceedings of the Ocean Drilling Program, Scientific Results, vol. 145 edited by D. K. Rea et al., p. 247 256. Reynolds, R. C. and Hower, J., 1970, The nature of interlayering in mixed layer illite montmorillonites: Clays and Clay Minerals, p. 25 36. Raymo, M. E., 1994, The initiation of northern hemisphere glaciation: Annual Review of Earth and Planetary Sciences, v. 22, p. 353 383. Rotman, H. M. M., an d Spinelli, G. A., 2013, Global analysis of the effect of fluid flow on subduction zone temperatures: Geochemistry Geophysics Geosystems, v. 14, no. 8, p. 3268 3281. Rowe, K. T., Screaton, E. J., and Ge, S. M., 2012, Coupled fluid flow and deformation mod eling of the frontal thrust region of the Kumano Basin transect, Japan: Implications for fluid pressures and decollement downstepping: Geochemistry Geophysics Geosystems, v. 13, p. 18. Reece, R. S., Gulick, S. P. S., Horton, B. K., Christeson, G. L., and Worthington, L. L., 2011, Tectonic and climatic influence on the evolution of the Surveyor Fan and Channel system, Gulf of Alaska: Geosphere, v. 7, no. 4, p. 830 844. Saffer, D. M., and Bekins, B. A., 1998, Episodic fluid flow in the Nankai accretionary c omplex: Timescale, geochemistry, flow rates, and fluid budget: Journal of Geophysical Research Solid Earth, v. 103, no. B12, p. 30351 30370. Saffer, D. M., Silver, E. A., Fisher, A. T., Tobin, H., and Moran, K., 2000, Inferred pore pressures at the Costa Rica subduction zone: implications for dewatering processes: Earth and Planetary Science Letters, v. 177, no. 3 4, p. 193 207. Saffer, D. M., and Bekins, B.A., 2002, Hydrologic controls on the morphology and mechanics of accretionary wedges: Geology, 30, p. 271 274 Saffer, D. M., 2003, Pore pressure development and progressive dewatering in underthrust sediments at the Costa Rican subduction margin: Comparison with northern Barbados and Nankai: Journal of Geophysical Research Solid Earth, v. 108, no. B5 p. 16. Saffer, D. M., and B.A. Bekins, 2006, An evaluation of factors influencing pore pressure in accretionary complexes: Implications for taper angle and wedge mechanics: Journal of Geophysical Research Solid Earth, v. 111, no. B4. Saffer, D. M., Underwood, M. B., and McKiernan, A. W., 2008, Evaluation of factors controlling smectite transformation and fluid production in subduction zones: Application to the Nankai Trough: Island Arc, v. 17, no. 2, p. 208 230.

PAGE 115

115 Saffer, D. M., and McKiernan, A. W., 2 009, Evaluation of in situ smectite dehydration as a pore water freshening mechanism in the Nankai Trough, offshore southwest Japan: Geochemistry Geophysics Geosystems, v. 10, p. 24. Saffer, D. M., and Tobin, H. J., 2011, Hydrogeology and Mechanics of Subduction Zone Forearcs: Fluid Flow and Pore Pressure, in Jeanloz, R., and Freeman, K. H., eds., Annual Review of Earth and Planetary Sciences, Vol 39, Volume 39: Palo Alto, Annual Reviews, p. 157 186. Saffer, D. M., Lockner, D. A., and McKiernan, A., 20 12, Effects of smectite to illite transformation on the frictional strength and sliding stability of intact marine mudstones: Geophysical Research Letters, v. 39, p. 6. Screaton, E. J., Wuthrich, D. R., and Dreiss, S. J., 1990a, Permeabilities, fluid pres sures, and flow rates in the barbados ridge complex: Journal of Geophysical Research Solid Earth and Planets, v. 95, no. B6, p. 8997 9007. Screaton, E., and Ge, S. M., 1997, An assessment of along strike fluid and heat transport within the Barbados Ridge accretionary complex: Results of preliminary modeling: Geophysical Research Letters, v. 24, no. 23, p. 3085 3088. Screaton, E., and Ge, S. M., 2000, Anomalously high porosities in the proto decollement zone of the Barbados accretionary complex: Do they in dicate overpressures?: Geophysical Research Letters, v. 27, no. 13, p. 1993 1996. Screaton, E., Carson, B., Davis, E., and Becker, K., 2000, Permeability of a decollement zone: Results from a two well experiment in the Barbados accretionary complex: Journal of Geophysical Research Solid Earth, v. 105, no. B9, p. 21403 21410. Screaton, E., Saffer, D., Henry, P., Hunze, S., and Leg 190 Shipboard Sci, P., 2002a, Porosity loss within the underthrust sediments of the Nankai accretionary complex: Implicati ons for overpressures: Geology, v. 30, no. 1, p. 19 22. Screaton, E. J., and Saffer, D. M., 2005, Fluid expulsion and overpressure development during initial subduction at the Costa Rica convergent margin: Earth and Planetary Science Letters, v. 233, no. 3 4, p. 361 374. Shi, Y., and Wang, C. Y., 1988, Generation of high pore pressures in accretionary prisms: Inferences from the Barbados subduction complex: Journal of Geophysical Research, v. 93, p. 8893 8910. Shipboard Scientific Party, 1973, Initial Re port of the Deep Sea Drilling Project, Leg 18, Deep Sea Drilling Project, Vol. XVIII: Washington, D.C., U.S. Government Printing Office. Shipboard Scientific Party, 1997, i n Kimura, G., Silver, E., Blum, P., et al., Proc. ODP, Initial Reports, 170: College Station, TX (Ocean Drilling Program), p. 45 93.

PAGE 116

116 Shipboard Scientific Party, 2001. Leg 190 summary, i n Moore, G.F., Taira, A., Klaus, A., et al., Proc. ODP, Initial Reports, 190: College Station, TX (Ocean Drilling Program), p. 1 87. Skarbek, R. M., and Saffer, D. M., 2009a, Pore pressure development beneath the decollement at the Nankai subduction zone: Implications for plate boundary fault strength and sediment dewatering: Journal of Geophysical Research Solid Earth, v. 114, p. 20. Spinelli, G. A., and Saffer, D. M., 2004, Along strike variations in underthrust sediment dewatering on the Nicoya margin, Costa Rica related to the updip limit of seismicity: Geophysical Research Letters, v. 31, no. 4. Spinelli, G. A., and Underwood, M. B., 2004, Ch aracter of sediments entering the Costa Rica subduction zone: Implications for partitioning of water along the plate interface: Island Arc, v. 13, no. 3, p. 432 451. Spinelli, G. A., Giambalvo, E. R., Fisher, A.T., 2004, Sediment permeability, distribution, and influence on fluxes in oceanic basement: Hydrogeology of the Oceanic Lithosphere, p. 151 188. Spinelli, G. A., Saffer, D. M., and Underwood, M. B., 2006, Hydrogeologic responses to three dimensional temperature variability, Costa Rica su bduction margin: Journal of Geophysical Research Solid Earth, v. 111, no. B4. Spinelli, G. A., and Wang, K., 2008, Effects of fluid circulation in subducting crust on Nankai margin seismogenic zone temperatures: Geology, v. 36, no. 11, p. 887 890. Steure r, J.F., and Underwood, M.B., 2003, Clay mineralogy of mudstones from the Nankai Trough reference Sites 1173 and 1177 and frontal accretionary prism Site 1174, i n Mikada, H., Moore, G.F., Taira, A., Becker, K., Moore, J.C., and Klaus, A. (Eds.), Proc. ODP, Sci. Results, 190/196, p. 1 37. Suess, E., Bohrmann, G., von Huene, R., Linke, P., Wallmann, K., Lammers, S., Sahling, H., Winckler, G., Lutz, R. A., and Orange, D., 1998, Fluid venting in the eastern Aleutian subduction zone: Journal of Geophysical Re search Solid Earth, v. 103, no. B2, p. 2597 2614. Stevenson, AJ., and Embley, R., 1987, Deep sea fan bodies, terrigenous turbidite sedimentation, and petroleum geology, Gulf of Alaska, in Scholl, D.W., Grantz, A., and Vedder, J.G., eds., Geology and resou rce potential of the continental margin of western North America and adjacent ocean basins Beaufort Sea to Baja California: Circum Pacific Council for Energy and Mineral Resources Earth Science Series, v. 6, p. 503 522.

PAGE 117

117 Tobin, H. J., and Saffer, D. M., 20 09, Elevated fluid pressure and extreme mechanical weakness of a plate boundary thrust, Nankai Trough subduction zone: Geology, v. 37, no. 8, p. 679 682. Ujiie, K., Tanaka, H., Saito, T., Tsutsumi, A., Mori, J. J., Kameda, J., Brodsky, E. E., Chester, F. M., Eguchi, N., Toczko, S., Expedition, S., and Expedition, T. S., 2013, Low Coseismic Shear Stress on the Tohoku Oki Megathrust Determined from Laboratory Experiments: Science, v. 342, no. 6163, p. 1211 1214. Underwood, M. B., and Pickering, K. T., 1996, Clay mineral provenance, sediment dispersal patterns, and mudrock diagenesis in the Nankai accretionary prism, southwest Japan: Clays and Clay Minerals, v. 44, no. 3, p. 339 356. Underwood, M. B., 2007, Sediment inputs to subduction zones: Why lithostrat igraphy and clay mineralogy matter, in The Seismogenic Zone of Subduction Thrust Faults, edited by T. H. Dixon and J. C. Moore, Columbia Univ. Press, New York. von Huene, R., Klaeschen, D., Gutscher, M., and Fruehn, J., 1998, Mass and fluid flux during accretion at the Alaskan margin: Geological Society of America Bulletin, v. 110, no. 4, p. 468 482. von Huene, R., Miller, J. J., and Weinrebe, W., 2012, Subducting plate geology in three great earthquake ruptures of the western Alaska margin, Kodiak to U nimak: Geosphere, v. 8, no. 3, p. 628 644. Vanderkamp, G., and Gale, J. E., 1983, Theory of earth tide and barometric effects in porous formations with compressible grains: Water Resources Research, v. 19, no. 2, p. 538 544. Wang, K. L., and Kinoshita, M ., 2013a, Dangers of Being Thin and Weak: Science, v. 342, no. 6163, p. 1178 1180. Worthington, L. L., S. P. S. Gulick, and T. L. Pavlis, 2008, Identifying active structures in the Kayak Island and Pamplona zones: Implications for offshore tectonics of th e Yakutat Microplate, Gulf of Alaska: Active Tectonics and Seismic Potential of Alaska, Geophys. Monogr. Ser., vol. 179, edited by J. T. Freymueller et al., p. 257 268. Worthington, L. L., Gulick, S. P. S., and Pavlis, T. L., 2010, Coupled stratigraphic a nd structural evolution of a glaciated orogenic wedge, offshore St. Elias orogen, Alaska: Tectonics, v. 29, p. 27.

PAGE 118

118 BIOGRAPHICAL SKETCH Lanie Meridth was born and raised in Jacksonville, FL and grew up with an intense interest in nature and the environment. She decided to pursue this interest at the University of Florida, where she received her Bachelor of Science degree in geology with a minor in the pur suit of a Master of Science degree in geology. She plans to use her knowledge of hydrology to help preserve water resources.