California State University, Northridge Estimation of Permeability, Porosity, and Grain-Size Distributions Across the San Andreas Fault Zone in Northwest Coachella Valley, CA (Riverside County) A thesis submitted in partial fulfillment of the requirements For the degree of Master of Science in Geology By Simarjit K. Chehal August 2014 The thesis of Simarjit K. Chehal is approved: ______________________________________ Dr. Janice M. Gillespie, Ph.D. _______________ Date ______________________________________ Dr. M. Ali Tabidian, Ph.D. _______________ Date ______________________________________ Dr. Richard V. Heermance, Ph.D., Chair _______________ Date California State University, Northridge ii Dedication Dedicated to Marilyn Hanna and My Parents Thank you for your support iii Acknowledgements Funding for this project was made possible due to the generous donations of others. Thank you to all those who have donated to the Hanna Fellowship, Hanna Summer Research Award, Los Angeles Geological Society Student Education Scholarship, Peter W. Weigand Memorial Scholarship in Geochemistry, and the Graduate Thesis/Project/ Performance Support Program 2013-2014. There are a number of people that I would like to thank, including Casey Polon and Michael Vadman for their assistance in conducting fieldwork in the scorching desert heat. To all the graduate students who participated in the Saturday grad-office brunches, thank you for the company. A special thanks to Dave Liggett and Mike Tacsik for helping me find the equipment to conduct a thesis project. Mari Flores and Terry Dunn, thank you for all the encouragement. Most of all, I would like to thank all those who made this thesis a reality. I express my gratitude to all the professors who have helped me in the conception of this study, especially Dr. M. Ali Tabidian, Dr. Kathleen Marsaglia, Dr. Vicki Pedone and Dr. Janice M. Gillespie (CSUB). Thank you for your guidance and providing resources and answers even when I did not know what I was asking. Last, but not least, I would like to thank my advisor for believing in me, even when I did not believe in myself. Dr. Richard V. Heermance, thank you for teaching me self-reliance, independence, and selfmanagement- a skill set that goes beyond a master’s degree. iv Table of Contents Signature Page Dedication Acknowledgements List of Figures List of Tables List of Equations Abstract ii iii iv vii x xi xii Chapter 1- Introduction 1 Chapter 2- Background 2.1- Description of Area 2.2- Climate 2.3- Water Importation 2.4- Geologic Structure 2.5- Stratigraphy 2.5.1- Non-water Bearing Units 2.5.2- Semi-Water Bearing Units 2.5.3- Water-Bearing Units 2.6- Groundwater Basins 2.6.1- Desert Hot Springs Subbasin 2.6.2- Mission Creek Subbasin 2.6.3- Indio Subbasin 2.7- Fault Zone Architecture 9 9 9 12 13 23 30 39 Chapter 3- Methods 42 3.1- Field Methods 42 3.1.2- Fault Zone Architecture Characterized by Numerical Measurements 3.2- Laboratory Methods 44 3.2.1- Core Laboratories 3.2.2- Thin Section Analysis 3.2.3- Porosity- Density Analysis 3.2.4- Grain-size Analysis 3.2.5- Hydraulic Conductivity Analysis 3.2.6- Permeability Analysis Chapter 4- Results 4.1- Field Observations 4.2- Results from Core Labs 4.3- Thin Section Analysis 4.3.1- Thin Section Analysis using Petrographic Microscope 4.3.2- JMicroVision Software 4.4- Porosity- Density Data 4.5- Grain Size Analysis 4.6 Hydraulic Conductivity v 54 54 68 68 82 86 86 4.7 Intrinsic Permeability 91 Chapter 5- Discussion 5.1 Field Data 5.2 Core Labs Data 5.3 Thin Section Data 5.4 Porosity-Density Relationship 5.5 Grain Size Analysis 5.6 Hydraulic Conductivity and Permeability 5.7 Implications 5.8 Fractoconformity vs. Fault Zone Controlled Fluid Flow 92 92 93 94 98 99 101 109 111 Chapter 6- Conclusions 113 References 115 Appendix A Appendix B Appendix C Appendix D Appendix E Appendix F Appendix G Appendix H Appendix I 119 122 133 149 151 159 163 167 202 vi List of Figures Figure 1-1: Schematic Fault Zone Architecture 3 Figure 1-2: Schematic of Perched Water Tables 5 Figure 1-3: Groundwater Subbasins and Subarea Boundaries 6 Figure 1-4: Schematic Comparison of Unfaulted and Faulted Water Bearing Units 7 Figure 1-5: 1936 Groundwater Surface Contour 8 Figure 1-6: 2009 Groundwater Surface Contour 8 Figure 2-1: Study Region 10 Figure 2-2: Study Region with respect to Pacific-North American Plate Boundary 11 Figure 2-3: Geomorphic Regions in California 15 Figure 2-4: Fault Strands in Study Region 17 Figure 2-5: Vegetation in Study Region 18 Figure 2-6: Geomorphic Surface Expression in Study Region 20 Figure 2-7: Alluvial Deposits in Study Region 21 Figure 2-8: Seismic Survey Interpretation of Mission Creek Strand 22 Figure 2-9: Stratigraphic Column of Northwest Coachella Valley, CA 24 Figure 2-10: Watersheds and Flow Path in Study Region 32 Figure 2-11: Gypsum Deposits along the Mission Creek Strand 34 Figure 2-12: Relationship between Faults and Fluids 41 Figure 3-1: Schematic Thin Section 46 Figure 3-2: Ernst Leitz PRADO- 500 46 Figure 3-3: Schematic Volume Water Displacement Test 47 Figure 4-1: Sampling Locations 55 vii Figure 4-2: Location B: B4 LiDAR Analysis 56 Figure 4-3: Location RCB:B4 LiDAR Analysis 57 Figure 4-4: Location MF: B4 LiDAR Analysis 58 Figure 4-5: Location P: B4 LiDAR Analysis 59 Figure 4-6: Banning Strand at Location B 61 Figure 4-7: Location B- Fault Zone Architecture 62 Figure 4-8: Location RDB- Fault Zone Architecture 64 Figure 4-9: Location MF- Fault Zone Architecture 65 Figure 4-10: Location P- Fault Zone Architecture 67 Figure 4-11: Banning Strand Fault Zone Indices 70 Figure 4-12: Mission Creek Strand Fault Zone Indices 70 Figure 4-13: Petrographic Microscope Porosity Calculation 72 Figure 4-14: Sample B2- Thin Section 73 Figure 4-15: Sample B3- Thin Section 74 Figure 4-16: Sample MF4- Thin Section 75 Figure 4-17: Sample MF7- Thin Section 76 Figure 4-18: Sample P1- Thin Section 78 Figure 4-19: Sample P3- Thin Section 79 Figure 4-20: Sample P4- Thin Section 80 Figure 4-21: Sample P5- Thin Section 81 Figure 4-22: JMicroVision Porosity Calculation 83 Figure 4-23: Density Relationship Porosity Calculation 84 Figure 5-1: Lithofacies Identification at Location P 95 viii Figure 5-2: Geologic Map of Study Region 100 Figure 5-3: Location B- Hydraulic Conductivity and Intrinsic Permeability 103 Figure 5-4: Location RCB- Hydraulic Conductivity and Intrinsic Permeability 102 Figure 5-5: Location MF- Hydraulic Conductivity and Intrinsic Permeability 103 Figure 5-6: Location P- Hydraulic Conductivity and Intrinsic Permeability 106 Figure 5-7: Fault Core and Damage Zone Hydraulic Conductivity VS Protolith 108 Rock Hydraulic Conductivity Figure 5-8: Fault Core Hydraulic Conductivity VS Damage Zone Hydraulic Conductivity ix 110 List of Tables Table 2-1: Lithofacies Associated with SGIM Complex 26 Table 2-2: DWR Classification Number of the Groundwater Basin 31 Table 2-3: Stratigraphy of Mission Creek Subbasin 36 Table 3-1: Sieves Used for Dry Sieving 49 Table 4-1: Fault Zone Index Calculation 69 Table 4-2: Porosity Comparison 85 Table 4-3: Effective Grain Sizes 87 Table 4-4: Uniformity Coefficients 88 Table 4-5: Breyer Hydraulic Conductivity and Intrinsic Permeability 89 Table 4-6: Slichter Hydraulic Conductivity and Intrinsic Permeability 90 Table 5-1: Uniformity Coefficients by Fault Strands 102 Table 5-2: Breyer and Slichter Hydraulic Conductivity and Intrinsic Permeability 107 by Fault Zone Architecture x List of Equations Equation 3-1: Fault Zone Architectural Index 43 Equation 3-2: Average Fault Zone Architectural Index 43 Equation 3-3: Spatial Variability of Fault Zone Architectural Index 43 Equation 3-4: Water Displacement Test 47 Equation 3-5: Bulk Density 48 Equation 3-6: Total Porosity 48 Equation 3-7: Hydraulic Conductivity 51 Equation 3-8: Uniformity Coefficient 51 Equation 3-9: Breyer Equation 52 Equation 3-10: Slichter Equation 52 Equation 3-11: Intrinsic Permeability and Hydraulic Conductivity Relationship 53 xi Abstract Estimation of Permeability, Porosity, and Grain-size Distribution Across the San Andreas Fault Zone in northwest Coachella Valley, California (Riverside County) By: Simarjit Kaur Chehal Masters of Science in Geology The groundwater aquifer system in the northwestern region of Coachella Valley, California serves as a major natural resource for agricultural and municipal uses. In this region, the aquifer system is partitioned into four groundwater sub-basins due to the presence of the San Andreas fault zone. Previous investigation involving land surface deformation, seismic data, and groundwater data indicate there are at least three main strands of the San Andreas Fault- Mission Creek Strand, Banning Strand, and Garnet Hill Strand. For years, these faults have been characterized as simple barriers to fluid flow due to measureable offsets in the water table across the fault strands and hydrochemistry variations. The ability for a fault to act as a barrier to flow in an aquifer system is the result of significant development of gouge causing lateral variations in fault zone permeability but the mechanisms for gouge development in the Coachella Valley in unconsolidated-to-weakly consolidated sediments is unclear. Another explanation for variations in water chemistry, temperature, and groundwater levels is due to displacement of impermeable bedrock or relative offset of water bearing units. This study proposes to document variations of permeability and porosity across the San Andreas Fault zone. Field mapping, sampling and descriptions of fault zone, damage zone, and gouge width were recorded at four fault outcrop locations in the region. Analysis of porosity, hydraulic conductivity, intrinsic permeability, and grain size distribution across the fault zone associated with each fault strand indicate that different regions of the fault zone can xii act to impede and/or enhance fluid flow across the faults. This data analysis of permeability and porosity variations across fault zones will help develop a better understanding of fault and aquifer interactions for future groundwater models, recharge activities, fault displacement, and development of gouge in unconsolidated sediments. xiii Chapter 1- Introduction Subsurface fluid flow systems are often modeled by geologists through a series of differential equations and/or simplified abstractions. This approach is ubiquitous for groundwater flow models in the southwestern United States. However, the numerical schemes employed to create subsurface models present a sense ambiguity to the representative model. The ambiguity associated with subsurface modeling is often a result of sparse sampling locations and lack of data related to actual porosity and permeability of the subsurface sediments. This presents a fundamental challenge to modeling, where modelers must assume the physical properties associated with porous media. Physical properties, in sedimentary rocks, such as porosity and permeability are readily influenced by lithological variations and structural features (Freeze and Cherry, 1979, Fetter, 2001, Boggs, 2006). Petrophysical properties associated with subsurface units can be affected by faults. Faults can disrupt the petrophysical properties of porous media by juxtaposing strata of varying permeability, porosity, and/or lithology. Additionally, faults are capable of producing regions of varying petrophysical properties, near the fault plane. An increase in stress and temperature along a fault plane during a rupture event provides the opportunity for the formation of fractures, deformation of grains, grain realignment, and grain-size reduction (cataclasism), hence, directly influencing fluid flow across the fault plane and causing heterogeneity in the porous media. Fluid flow can be enhanced due to the formation of fractures and/or it can be impeded due to the formation of fault gouge. Faults play an important role in influencing shallow subsurface fluid migration (i.e. Faulkner et al., 2010). Faults are composed of three distinctive zones: fault core, 1 damage zone, and protolith rock, as displayed in Figure 1-1 (Caine et al., 1996). The fault zone can either impede or enhance fluid flow (Caine et al., 1996). Fluid flow across fault zones is a function of porosity and permeability which in turn are dependent upon the lithology and fault movement in a given fault zone (Chester and Logan, 1986, Caine et al., 1996). Previous studies conducted regarding fault zones and fluid flow by Caine et al., 1996, Goodwin et al., 1999, Faulkner et al., 2010, have indicated that a contrast exists between the permeability field of a fault zone and the protolith rock (Caine, et al., 1996, Goodwin et al., 1999, Faulkner et al., 2010). The contrast exists when the protolith rock is relatively more permeable than the fault zone; thus, resulting in a barrier-forming fault with respect to fluid flow. The contrast in the permeability field is also present when the fault zone is relatively more permeable than the surrounding protolith rock; thus, resulting in a conduit-forming fault with respect to fluid flow. However, no correlations have been made indicating what causes the permeability variations that can cause some faults to behave as barriers or conduit to fluid flow. Purpose of Study Groundwater resources are critical to civilization, especially in desert regions, like the northwestern portion of Coachella Valley, California. Nearly 200,000 people, in the cities of Desert Hot Springs, Palm Springs, Cathedral City, Rancho Mirage, Thousand Palms, and Palm Desert, rely on groundwater resources and artificial groundwater recharge activities implemented in the region (MWH, 2013). Therefore, proper management of groundwater resources in this region is essential. The purpose of this study is to estimate porosity and permeability variations across fault zone architecture present along fault outcrops associated with the Mission 2 Creek, Banning, and Garnet Hill Strands using grain size distributions. Characterizing the porosity and permeability changes within the different zones of the fault architecture is important in order to evaluate a fault ability to behave as a barrier or conduit to fluid flow. Figure 1-1: The figure above displays a schematic diagram of the fault zone architecture. The fault zone architecture is characterized as a centralized fault core which is surrounded with a damage zone. The protolith zone is regarded as the region that is not affected by faulting. (Revised figure from Caine et al., 1996) Location of Study Area The study region is located in the northwestern corner of Coachella Valley, California (Riverside County). In this region, the San Andreas Fault zone splays into three primary fault strands, (as displayed in Figure 1-2a): Mission Creek Strand, Banning Strand, and 3 Garnet Hill Strand. The fault strands partition the groundwater aquifer into three distinctive basins: Desert Hot Springs subbasin, Mission Creek subbasin, and Indio subbasin (which contains the Garnet Hill subarea and Whitewater subarea), as shown in Figure 1-3. Some regions along the fault strands display vegetation alignment which indicates shallow groundwater resources. However, the surface feature is not traceable along the entire length of the fault strand, thus, not providing any information regarding the fault’s ability to act as a barrier or conduit along strike. Vegetation alignments in the region can result from: (1) Vertical fluid migration towards the surface due to a fault strand’s ability to create a barrier to fluid flow, as displayed in Figure 1-4. or (2) The presence of a perched aquifer resulting from a fractoconformity, as displayed in Figure 1-2b. Variations across the water table are also noted in groundwater well measurements. Historical records of groundwater table measurements are inputted into subsurface modeling programs (i.e. MODFlow) and overseeing water agencies in the region continue to record water-depth measurements from monitoring wells in the region. Reverse modeling techniques indicate offsets in the groundwater table across the fault strands existed in 1936 (Figure 1-5, MWH, 2013). After nearly 100 years of groundwater extraction and implantation of artificial recharge activities, offsets in the groundwater table were still present in 2009 (Figure 1-6, MWH, 2013). 4 Figure 1-2 (A and B): A- Study region outlined in black with the three main strands in the San Andreas Fault zone. B- Interpretative schematic diagram with a series of perched aquifers which are not hydraulically connected. The perched aquifers are a result of a fractoconformity. A fractoconformity is defined as the relation between conformable strata where faulting of older beds proceeds contemporaneously with deposition of newer strata, also referred to as syn-tectonic deposition. 5 N Figure 1-3: Groundwater subbasins and subareas within the study region. Subbain and subarea boundaries are attributed to the presence of faults offsetting the groundwater table. Artificial recharge ponds are located in the Indio and Mission Creek subbasins. (Modified from MWH, 2013) 6 Figure 1-4: A schematic comparison of unfaulted and faulted water bearing units. Diagram A displaying the unfaulted units and diagram B displaying a schematic where a barrier-forming fault is causing fluid migration towards the surface, resulting in the formation of an oasis at the surface. (Modified figure from Freeze and Cherry, 1979). 7 Figure 1-5: Groundwater contour surface in feet from 1936. Contour modeling is constructed from historical records of water well measurements. (From MWH, 2013) Figure 1-6: Groundwater contour surface in feet from 2009. Contour modeling is constructed from measurements of water depth taken at monitoring wells. (From MWH, 2013) 8 2. Coachella Valley 2.1 Description of Area The study area is located in the northwestern corner of the Coachella Valley, CA (Figure 2-1). The Coachella Valley is a part of Riverside County in Southern California, extending for nearly ~70 km (~43 miles) from the San Gorgonio Pass to the northern shore of the Salton Sea. It is centered in the northern portion of the Salton Trough bounded by the San Jacinto and the Santa Rosa Mountains along the southwest and the Little San Bernardino Mountains along the northeast. The Coachella Valley is situated at the northwestern extent of the Salton Trough which is characterized as a structurally complex transition zone between the Eastern Pacific Rise tectonic spreading center and the right-lateral San Andreas transform fault zone, as displayed in Figure 2-2 (Powell, 1993). 2.2 Climate The climate of the northwestern region of Coachella Valley is classified as a desert, with mild winters and hot summers. Temperatures in the region have varied from ~62F to ~128F in the past 50 years (ACS, 2014). Precipitation on the desert floor varies from year to year, but a weather station in Palm Springs has recorded averages of between 0.03 to 1.14 inches of precipitation per year for the past 50 years (ACS, 2014). Precipitation generally occurs during the December to March months; however, occasional summer month subtropical thunderstorms occur as well. The San Jacinto and San Gorgonio mountains border the valley along the west, creating an effective orographic barrier against coastal storms and reducing the direct precipitation to recharge 9 Legend Study Location Major Freeway Lake N Figure 2-1: Study region (indicated by the red star at 33.920231,-116.508449) is located northwest of the Salton Sea and east of Los Angeles. (Base Layer Credits: ESRI, DeLorme, GEBCO, NOAA NGDC, and other Contributors) 10 Figure 2-2: Study region (indicated by the red star) with respect to the location of the Salton Sea and the transition between the San Andreas Fault and the East Pacific Rise. A detailed map of the study region is shown in Figure 1-2. (Modified figure from Powell, 1993). 11 the Valley’s groundwater basins. Precipitation usually either evaporates, or is consumed by native vegetation, or percolates into the underlying sediments and eventually adds to the subsurface fluid flow. Vegetation in the region is sparse and only xerophytic plants have been identified, including America’s only native palm tree- Washingtonia Filifera (Desert Fan Palms), which are found within the numerous palm tree oases along the San Andreas Fault (Proctor, 1968, Nye, 1994, Guzman, 2010). The Washingtonia Filifera is a relic species of palm trees that has been dated back to the Miocene and Pliocene eras (Nye, 1994). Favorable climatic conditions of the Sonoran Desert (also known as the Colorado Desert) have made it possible for this species to survive in regions with groundwater springs and seeps (referred to as oases). The Desert Fan Palms support a giant radial-root system comprised of hundreds to thousands of tiny rootlets that penetrate to a maximum depth of nearly ~10 feet (Nye, 1994). Thus, the oases are a result of shallow subsurface groundwater resources. 2.3 Water Importation Water supply and management of hydrologic resources is essential to support civilization in desert provinces like the northwestern region of Coachella Valley. Groundwater resources for the cities of Palm Springs, Cathedral, Palm Desert, Desert Hot Springs, and Thousand Palms are managed by three primary agencies: The Coachella Valley Water District (CVWD), Desert Water Agency (DWA), and Mission Springs Water District (MSWD) (MWH, 2013). A decline in groundwater levels due to over pumping of the aquifers promoted the Desert Water Agency and the Coachella Valley Water District to formulate agreements with the State of California to purchase water from the State Water Project (SWP) in 1962 and 1963 (MWH, 2013). However, due to 12 the lack of pre-existing infrastructure to import water to the desired location, the two agencies are held responsible for the construction of a pipeline system for the transportation of water from the State Water Project Canal to the Valley. In order to avoid an estimated cost of $150 Million in the 1970’s (current costs projected at ~ $ 1 Billion), the Desert Water Agency, and Coachella Valley Water District reached an agreement with the Metropolitan Water District of Southern California (MWD) which is valid until January 2035 (MWH, 2013). The agreement was to divert Colorado River water from the Metropolitan’s Colorado River Aqueduct to Coachella Valley in exchange for DWA and CVWD’s water share from the State Water Project (MWH, 2013). Diversion of water was initiated in 1973, when DWA and CVWD started to release Colorado River water into the Whitewater River and implemented artificial recharge within the Whitewater subarea (ED-CVWD, 2013b). The imported water percolates into the Whitewater subarea through nearly 900 acres of percolation ponds located at Windy Point – Whitewater Spreading Facility. However, continuous over-drafting of the aquifer system has required CVWD and DWA to implement artificial recharge activities in the Mission Creek subbasin as well (ED-CVWD, 2013a). In 2002, Colorado River water was diverted to the Mission Creek Spreading Facility to percolate through 200 acres of percolation ponds in efforts to recharge the Mission Creek subbasin (MWH, 2013). Percolation pond locations and groundwater subbasin boundaries are displayed in Figure 1-3. 2.4 Geologic Structure Extending for nearly 1,200 kilometers (740 miles), the San Andreas Fault is a N35-40W trending right-lateral fault, representing the main tectonic boundary between 13 the Pacific and North American plates (i.e. Wallace, 1990). In Southern California, the San Andreas Fault is structurally influenced by a structural knot, often referred to as the San Gorgonio Pass Knot. This is a result of the convergence of Transverse Ranges (San Gabriel Mountains Basement), Peninsular Ranges (San Jacinto Mountains and Santa Rosa Mountains) and the San Bernardino-Mojave Desert block, as displayed in Figure 23 (Langenheim et al., 2005). This convergence adds a great deal of complexity to the San Andreas Fault zone (Langenheim et al., 2005). Because of this convergence zone, the San Andreas Fault splays into a series of strike-slip faults with a dip-slip component creating a non-uniform fault zone (i.e. Proctor, 1968, Wallace, 1990, Yule and Sieh, 2003). The structural formation of the upper Coachella Valley is influenced by the inland shift of the Pacific Plate- North American Plate boundary. This process is described as a two-stage evolution process of the plate boundary that initiated in the late Miocene (9-7 Ma) and continues to the present (Matti and Morton, 1993, Yule, personal communication). During this time, the Salton Trough region experienced oblique extension in the Late Tertiary (6-2 Ma) and oblique convergence in the Quaternary (2.6 Ma-present) (Matti and Morton, 1993, Yule, personal communication). Because of the relative uplift on the valley margins, the Salton Trough is a structurally depressed region with the ability to accumulate sediment. Gravitational and seismic data collected in the region reveals an asymmetrical basin with the deepest portion southwest of the Banning Strand (CDWR, 1964, Catchings et al., 2009); therefore, creating the appearance of a series of stair-stepping half-graben basins between the Little San Bernardino Mountains and the San Jacinto Mountains (Figure 1-2b). 14 Figure 2-3: Various geomorphic regions located in California with the San Andreas Fault (SAF) zone in red. The “Big-Bend” in the San Andrea Fault trace is located at the convergence of the Transverse Ranges, Peninsular Ranges, and the Mojave Desert/Colorado Desert block. Complexity of the SAF is shown by the numerous splays at the convergence zone. The study region is indicated by the location of the star, which encompasses the numerous splays of the San Andreas Fault zone. (Base Layer Credits: ESRI, DeLorme, GEBCO, NOAA NGDC, and other Contributors) 15 The nomenclature of categorizing fault sections, strands, and splays is adopted from previous studies conducted in the region by CSU, Northridge students (Behr et al. 2010, Guzman, 2010). The portion of the San Andreas Fault that is present in the study region is referred to as the Coachella Valley section, which is composed of various strands related to numerous splays. The Coachella Valley section of the San Andreas Fault is characterized by a westerly strike ranging from ~N50-80W, with much of the fault slip distributed along three primary fault strands: San Andreas Fault-Mission Creek Strand (SAF-MCS), San Andreas Fault- Banning Strand (SAF-BS), and San Andreas Fault-Garnet Hill Strand (SAF-GHS). These faults are referenced as the Mission Creek Strand, Banning Strand, and Garnet Hill Strand in this thesis (Figure 2-4). At the surface, the Mission Creek, Banning, and Garnet Hill Strands display several topographic features associated with fault movement (Figure 2-5). The Mission Creek and Banning Strands form the northern and southern boundaries of the uplifted Indio Hills, respectively. The Indio Hills is the largest unit of material that has been displaced due to fault movement. The Indio Hills display characteristics similar to a positive flower structure (also referred to as Palm-Tree structure) where the Mission Creek, Banning, and Garnet Hill Strands merge into a single Coachella Valley section at depth (Catchings et al., 2009). The Indio Hills are cut in an N-S direction by Thousand Palms Canyon (Figure 2-5). The northern portion reaches a maximum elevation of 420 meters (1380 feet) and 530 meters (1740 feet) in the southern portion (CDWR, 1964). The presence of the Garnet Hill Strand near the Indio Hills is unknown due to the lack of surface exposure in the region. However, west of the Indio Hills, evidence for the presence of a Garnet Hill Strand at depth is expressed at the surface through a series of 16 Mission Creek Subbasin Indio Subbasin Whitewater Subarea Garnet Hill Subarea Desert Hot Springs Subbasin Figure 2-4: The three main strands of the San Andreas Fault zone: Mission Creek Strand (SAF-MCS), Banning Strand (SAF-BS), and Garnet Hill Strand (SAF-GHS) are displayed in within the study region, which is outlined in black. The respected subbasins and subareas formed by the faults are also indicated. (Base Layer Credits: ESRI, DeLorme, GEBCO, NOAA NGDC, USGS and other Contributors) 17 Figure 2-5: Vegetation alignments present at the surface along the Banning and Mission Creek Strands. No vegetation alignment was found along the Garnet Hill Strand. Vegetation patches labeled a-g corresponding with Figures A-5a-g (found in Appendix A). (Base Layer Credits: ESRI, DeLorme, GEBCO, NOAA NGDC, USGS and other Contributors) 18 compressive steps between the Garnet Hill and the Banning Strands, as displayed in Figure 2-6 (Yule and Sieh, 2003, 2009). These geomorphic expressions are seen on the surface as active folds (West Whitewater Hill, East Whitewater Hill, Hugo Hill, Devers Hill, Garnet Hill, and Edom Hill) suggesting oblique slip (dextral-reverse slip) with a left step-over along the Garnet Hill Strand (Yule, 2009). The Garnet Hill Strand displays an overall right-lateral movement with a thrust component in the San Gorgonio Pass region and possible northeast dip (~40NE) (Yule and Sieh, 2003, Catchings et al., 2009). East of Edom Hill, the presence of the Garnet Hill Strand is difficult to detect, however, it has been reported that a major oil company conducted a gravitation survey of the region and detected the fault as displayed in Figure 2-6 (Tyley, 1974). Seismic surveys conducted by various agencies have not been able to detect the fault (CDWR, 1964). However, the fault trace was not detected in the geophysical studies conducted by Shawn Biehler in 1964 (Tyley, 1974). The San Andreas Fault- Banning Strand is a right lateral-reverse fault which extends in a southeast direction from the San Bernardino Mountains and crosses the study region at a strike of N70-65W (Proctor, 1968). Seismic surveys conducted along the Banning Strand indicate dips ranging from 50-70 to the northeast along the Banning and Mission Creek Strands (Catchings et al., 2009). Surface traces of the fault can be identified in aerial photography by disruption of alluvial deposits, such as the Mission Creek Upland Wash (Figure 2-7). The presence of vegetation alignments and springs along the north side of the fault strand indicates regions of shallow groundwater accumulation (Figure 2-5). 19 Figure 2-6: Shaded-relief topographic map displaying geomorphic surface expressions between the Garnet Hill Strand and the Banning Strand. Approximate location of the Garnet Hill Strand is based upon probable location detected by gravitation studies (Tyley, 1974). Abbreviations: W-WH, West-Whitewater Hill; E-WH, East-Whitewater Hill; SAF- San Andreas Fault; -?-, Approximate location of fault; , Approximate drainage divide with direction of drainage. (Modified Figure from Yule, 2009) The Mission Creek Strand crosses through the study region at a strike of nearly N50W and is a right lateral reverse fault. Seismic studies conducted in the region have indicated three separate splays associated with the Mission Creek Strand, each dipping at a slightly different angle: Mission Creek splay- A (90NE), Mission Creek splay-B (80SW), and Miracle Hill fault splay (80NE) ( Figure 2-8) (Catchings et al., 2009). Movement along the Mission Creek Strand indicated both lateral and vertical displacement. Geomorphic surface expressions associated with the Mission Creek Strand 20 Figure 2-7: Regions of excessive alluvial deposits within the study region. These regions are referred to as the Dillon Road Piedmont Slope, Mission Creek Upland, and Palm Springs Sand Ridge. The Dillon Road Piedmont Slope and Mission Creek Upland deposits are regions formed by repeated alluvial fan deposits. The Palm Springs Sand Ridge appears to be formed by sand dune deposition. (Modified figure from CDWR, 1964) 21 Figure 2-8: Seismic survey interpretation suggesting three fault splays associated with the Mission Creek Strand (MC splay a, b and the Miracle Hill Fault (sometimes referred to as splay c). Seismic profile was conducted along Long Canyon Creek, located NW of the intersection of Dillon Road and Mountain View Road and east of Sample Location MF. The seismic data interpretation also supports the presence of a fractoconformity. (Modified Figure from Catchings et al., 2009) 22 includes the north face of the Indio Hills, Miracle Hill and the Dillon Road Piedmont Slope (Figure 2-7) (Proctor, 1968). Miracle Hill is an escarpment on the north side of the Mission Creek Strand. Miracle Hill is a continuous scarp with a north-west orientation for nearly 3.5 kilometers (~2.2 miles) and rises nearly 33 meters (110 feet) above the valley floor. Continuous flows from streams in the Little San Bernardino Mountains have formed a series of coalescing alluvial fans which appear to terminate at the Mission Creek Strand and/or the foothills of the Indio Hills. These wash deposits are referred to as the Dillon Road Piedmont Slope (Figure 2-7) (CDWR, 1964). 2.5 Stratigraphy Surrounding and within the study region, the mountains and basement are comprised of highly complex crystalline Precambrian-Cretaceous igneous and metamorphic rocks. Thick deposits of Tertiary and Quaternary sediments of continental and marine origin overlay the basement (CDWR, 1964). Formational names and mapped locations have been adopted from Allen (1957) and Dibblee (1954), see Figure 2-9 for stratigraphic column. Subsurface correlations and surface mapping was conducted by the California Department of Water Resources (CDWR, 1964) in order to determine the potable water supply in the region. Geologic formations are discussed with regard to its ability to store groundwater according to the California Department of Water Resources. 2.5.1 Non-water Bearing Units Units classified under the title of non-water bearing are units that contain little or no water and display poor storage aquifer characteristics. This group of units mainly refers to the Precambrian-Cretaceous crystalline basement rocks, and consolidated Tertiary sediments. 23 Figure 2-9: Stratigraphic column of formations found in the northern Coachella Valley, CA. Formations are categorized to display age and their ability to store groundwater. (Modified Figure from Proctor, 1968). The basement rock and the surrounding mountains (San Bernardino and Little San Bernardino Mountains) are a complex assemblage of gneisses and schists, commonly referred to as San Gorgonio-Igneous-Metamorphic Complex (SGIM Complex). The San Gorgonio-Igneous-Metamorphic Complex is characterized as heterogeneous plutonic rocks with undifferentiable schists and gneissic rocks. The SGIM Complex has many common features resembling the Baldwin gneiss (found on the north side of the San 24 Bernardino Mountains), Chuckwalla Complex, Orocopia Schist, and San Gabriel Formation Igneous-Metamorphic Complex (Allen, 1957). Three distinctive gneissic lithofacies have been distinguished in the foothills of the study region: Flaser Gneiss, Green Schist, and Piedmontite-Bearing gneiss (See Table 2-1 for further description) (Allen, 1957). Overlaying the crystalline rocks is nearly 1,900 meters (6,200 feet) of consolidated tertiary sediments (Proctor, 1968). Two main units are projected throughout the valley floor -Coachella fanglomerate and the Imperial Formation. The Split Mountain Formation is believed to underlie the Coachella fanglomerate, however, there are no surface exposures of the Split Mountain Formation in the upper Coachella Valley (Proctor, 1968). The Coachella fanglomerate is a well-indurated, massive conglomerate creating prominent cliff forming beds exposed in the Upper Whitewater Canyon (Allen, 1957, CDWR, 1964). The exposure on the eastside wall of the Whitewater Canyon is nearly 1,400 meters thick (4,600 feet) and is late Miocene to early Pliocene in age (Allen, 1957). Along the eastern flank of the San Bernardino Mountains, numerous outcrops have been noticed where the formation is interlayered with basalt flows. The upper and lower members of the Coachella fanglomerate are separated by a pale-red-purple basalt flow, which can be found at about 560 meters (850 feet) above its base near the mouth of Mission Creek, (Figure 2-7, Allen, 1957). The number of olivine basalt flows and thickness increases towards the Mission Creek Strand and the basalt appears to have extruded along the fault splays (Allen, 1957, Slade, 1981). The maximum cumulative thickness of these flows is about ~15 meters (50 feet) but wedges out towards the southeast and north (Allen, 1957). 25 Table 2-1: Lithofacies Associated with the San Gorgonio-Igneous-Metamorphic Complex Lithofacies Name Location G1 Flaser Gneiss North of the Banning Fault -Displays distinctive structures developed by cataclastic metamorphism -Ranges from slightly sheared augen gneisses to mylonitzed and lenticularly layered rocks with grading and foliation generally striking east. Green Schist North of the Banning Fault -Localized body of quartz-actinolite-albite-epidote schist. - In hand sample- greenschist is dark blue green with well-developed foliation and contains many porphyroblasts of white albite that gives a “knotty” appearance. -Similar to the Pelona Schist of the San Gabriel Mountains and Orocopia Schist. Three possible explanations for the similarities are: (1) the greenschist represents the parent rock of the migmatitic gneisses and have been upgraded (2) greenschist represents a locally degraded amphilbolite. The environmental setting is of high shear stress and hydrothermal activity. (3) This region is a large fault sliver that was left behind/midway when the Pelona Schist and Orocopia Schist were separated. PiedmontiteBearing Gneiss North of the Banning Fault and West wall of Whitewater Canyon -Piedmontite is a dark-reddish or reddish brown, manganese-bearing mineral of the epidote group. This gneiss facies has been found to contribute fragments to the sedimentary rocks south of the Banning Fault. -The piedmontite-bearing gneiss is associated with pegmatite intrusions in the older occurrence of this gneiss which is on the west wall of Whitewater Canyon by Red Dome and the Trout farm. This occurrence does not display the yellow-green epidote and light-pink mica (alurgite). G2 G3 Description 26 The Imperial Formation outcrops throughout the valley floor and represents the last and only known marine southeasterly incursion dating back to about 6.5-6.3 Ma (Dorsey, 2011). The Imperial formation is characterized by various deep yellowish to brown sandstone, siltstone, and shale beds with thicknesses ranging from 0.15 meters- 30 meters (6 in-100 ft), however, the formation has an overall combined thickness of 480 meters (1,600 feet) (Allen, 1957, Proctor, 1968). Locally abundant fossils and wide variation in grain sizes indicate that the Imperial formation was deposited in littoral or shallow marine environments (Proctor, 1968). Some units consist of fossiliferous sandstone, silty sandstone, and claystones with bentonites (Nye, 1994). Variations in the lithofacies suggest that the northwestern region of Coachella Valley was predominantly a rocky shoreline with moderate relief and deep marine environments towards the southeast. 2.5.2 Semi-Water Bearing Units The category of semi-water bearing units describes units with generally low water-yielding capability. Formations spanning from the Pliocene and early Pleistocene, are categorized as semi-water bearing units (CDWR, 1968). In these formations, water is present in the interstices of rocks, however; due to the highly contorted nature of these formations, it is not readily extracted (CDWR, 1968). Along the western rim of the valley, the Painted Hill Formation conformably overlies the Imperial Formation at Whitewater Canyon. The Painted Hill Formation consists of 1,036 meters (3,400 feet) of pale-brown to light grey conglomerate sandstone with sub-rounded to well-rounded clasts <.5 meters (up to 2 feet in diameter) (Allen, 1957). The semi-consolidated and poorly sorted nature of this formation is indicative of 27 low permeability. The uplifted Indio Hills region exposes thick, compacted beds of the Palm Springs Formation. The Palm Springs Formation is presumed to underlie the Coachella Valley at depth. Previous studies conducted in the region date the Palm Springs Formation as lower Pleistocene in age (Nye, 1994). In the eastern Indio Hills, the Palm Springs Formation lies unconformably beneath the Ocotillo Formation (CDWR, 1964). In this portion of the Indio Hills, the Palm Springs Formation is more than 600 meters (2,000 feet) in thickness with grey to tan arkosic sandstones interbedded with red and green siltstones and claystones (CDWR, 1964). The Canebrake conglomerate, a member of the Palm Springs Formation, is a compacted silty conglomerate and sandstone. The Canebrake conglomerate is believed to be more permeable than other lithofacies associated with the Palm Springs Formation (CDWR, 1964). There is no direct evidence indicating that the Painted Hill Formation, found on the western rim of the valley and identified by Allen (1957), are the same rocks as the Palm Springs Formation which outcrops on the eastern side of the study region and was identified by Dibblee (1954). However, a chrono-stratigraphic correlation may exist between the two formations and deposition may have occurred simultaneously with the regression of the last marine incursion. 2.5.3 Water-Bearing Units The principal groundwater aquifer system lies within undisturbed and unconsolidated recent and Pleistocene alluvial deposits. The study region receives detrital material from the surrounding mountain ranges, which is transported via streams and deposited as heterogeneous alluvial fans. Units classified as water-bearing units are the 28 Ocotillo conglomerate, Cabezon fanglomerate, and Quaternary alluvium and terrace deposits. The Ocotillo conglomerate is the primary water-bearing unit in the upper Coachella Valley. Exposures of the Ocotillo conglomerate in the Indio Hills indicate that the Ocotillo conglomerate unconformably overlies the Palm Springs Formation (CWDR, 1964). It consists of poorly consolidated sandstones and conglomerates interbedded with thin grey-green and red-brown silts and clay like lenses (CWDR, 1964). This unit is at least 730 meters (2,400 feet) thick with the top 30 meters to 60 meters (100 to 200 feet) consisting mainly of lake-deposited sediments, as shown by resistivity breaks in electric logs of water wells (CWDR, 1964). The formation is split into two members- upper and lower- which are distinguished due to differences in lithology. The upper member consists of a poorly sorted grayish conglomerate with discontinuous beds of pebbles and/or boulders with varying thicknesses of a few inches to 1-2 feet (Nye, 1994). The lower member consists of a series of intermixed beds of multi-colored sandstones, conglomerates, and siltstones, as mapped by Popenoe in 1959 near Pushawalla Canyon (Nye, 1994). The exposure is described as a sequence of yellowish buff to pale green, massive, pebbly sandstone beds that range from 1 meter to 9 meters (3 feet to 30 feet) in thickness and alternate with conglomerate beds of similar thickness (Nye, 1994). Stratigraphically correlated to the Ocotillo conglomerate is the Cabezon fanglomerate, which is only found at the base of the San Bernardino Mountains. The Cabezon fanglomerate is a coarse, heterogeneous, and poorly consolidated fanglomerate (CDWR, 1964). The Cabezon fanglomerate reaches a maximum thickness of 300 meters (1,000 feet) and unconformably overlies the Painted Hill Formation (Allen, 1957). The 29 Cabezon fanglomerate is the primary water-bearing unit in the Mission Creek Upland region (CDWR, 1964). The shallow water-bearing zone in the region is contained Pleistocene- Holocene alluvial fan and terrace deposits. These deposits have a maximum thickness of 120 meters (400 feet), as exposed in the Indio Hills (CDWR, 1964). The alluvial fan and terrace deposits found in the Indio Hills are correlated throughout the valley floor to the terrace deposits found the Mission Creek Upland region; correlations are based on lithology exposed at the surface and in water well logs (CDWR, 1964). These undifferentiated terrace deposits consist of grey to tan heterogeneous gravels, sands, and silts (CDWR, 1964). 2.6 Groundwater Basins In 1964, the California Department of Water Resources and United States Geological Survey (USGS) recognized the water-bearing and semi-water bearing formations in the Colorado Desert Region (CDR) as the Coachella Valley Groundwater Basin (CVGB). In the study region, the Coachella Valley Groundwater Basin is divided into appropriate groundwater subbasins and subareas. The boundaries between the subbasins in this region are generally based upon faults which act as effective barriers to lateral flow across the basin (CDWR, 1964). General location of the faults is determined by alignment of geomorphic and phreatophytic vegetation at the surface, which is apparent in aerial photography (Figure 2-5). This thesis examines groundwater flow in four distinct regions, as classified by the CDWR (CDWR, 1964, DWR, 2003) (Table 22). A slight discrepancy in the groundwater subdivisions exists between the classifications system assigned by the DWR and the USGS, where the USGS recognizes 30 Table 2-2: Groundwater basin classification name and assignment number from Department of Water Resources (DWR, 2003). Name Coachella Valley Groundwater Basin Desert Hot Springs Subbasin Miracle Hill Subarea Mission Creek Subbasin Indio Subbasin Garnet Hill Subarea Whitewater River Subarea DWR Basin No. No.7-21 No.7-21.03 No.7-21.02 No.7-21.01 subareas as subbasins. The DWR recognizes 3 subbasins (Desert Hot Springs, Mission Creek, and Indio) with the Indio subbasin subdivided into the Garnet Hill subarea and Whitewater subarea; whereas, the USGS recognizes 4 subbasins: Desert Hot Springs, Mission Creek, Garnet Hill, and Whitewater. However, location of the designated regions is consistent between the two agencies. 2.6.1 Desert Hot Springs subbasin The Desert Hot Springs subbasin covers a surface area of 101,000 acres (158 square miles) (DWR, 2003). Depth of the subbasin varies but seismic data displays at least ~160 feet (50 meters) of saturated unconsolidated sediments in the Desert Hot Springs subbasin region (Catchings et al. 2009). The Desert Hot Springs (DHS) subbasin is an unconfined aquifer system bounded by the Little San Bernardino Mountains and the Mission Creek Strand on the north and south, respectfully. The Little San Bernardino Mountains are considered a no-flow boundary due to their crystalline composition. The main sources of natural groundwater recharge are from the watersheds of Big Morongo Canyon and Morongo Canyon (Figure 2-10). Minor flow paths are present at the base of canyons where runoff from the mountains recharges the subbasin through tributary wash deposits (CDRW, 1964). The Mission Creek Strand is also regarded as a no-flow 31 Figure 2-10: Displays the stream flow paths and directions with the numerous watersheds located in the region. The watershed boundaries provide insight regarding the alluvial depositional patterns in the region. (Base Layer Credits: ESRI, DeLorme, GEBCO, NOAA NGDC, and other Contributors) 32 boundary in computational groundwater models due offsets in the groundwater table found in water wells on opposites sides of the fault strands. The general direction of groundwater flow in the subbasin is towards the southeast (CDWR, 1964). There are two primary water bearing geologic units in the subbasin. They are the late Pleistocene and Holocene coarse-grained and poorly sorted alluvial fan deposits associated with the Dillon Road Piedmont Slop and the Ocotillo Formation (DWR, 2003). The principal exposure of the Ocotillo Formation is at Miracle Hill (CDWR, 1964). Groundwater resources in the vicinity are extracted for their high thermal and mineral properties and are used to supply local hot-spring resorts. This region is referred to as the Miracle Hill subarea due to the unique thermal properties of the groundwater near the Mission Creek Strand northwest of the Indio Hills. The Sky Valley subarea and Fargo Canyon subarea are not discussed in this report because they are not affected by the fault strands being studied. The Desert Hot Springs subbasin is uniquely defined by its hydro-chemical characteristics. The water quality in the subbasin is relatively poor due to high salinity content ranging from 800 mg/L to 1000 mg/L (DWR, 2003). The groundwater contains relatively high levels of fluoride concentration, near 10 mg/L (Slade, 1981). Groundwater resources in the Miracle Hill subarea contain high amounts of sodium and sulfate ions (DWR, 2003). Temperatures in the Miracle Hill subarea range from 82F to 200F (CDWR, 1964, Proctor, 1968). The relatively high temperatures of the water and high concentrated TDS levels is probably a result of emanating gases and hydrothermal activity associated with the Mission Creek Strand. Chemical analysis conducted by the USGS in 1974 concludes that the hydro-chemistry of the groundwater is dominated by 33 sodium-sulfate ions and sulfate levels are generally greater than 250 mg/L (Slade, 1981, DWR, 2003). The high sodium-sulfate content and gypsum deposits are a likely result of hydrothermal activity along the Mission Creek Strand. Gypsum deposits form a thin layer where natural springs have formed; such as in the Thousand Palms region (see Figure page 2-11). Figure 2-11: Displays gypsum (CaSO4·2H2O) deposits resulting from flow of natural springs enriched in sodium sulfate NaSO4. Sulfate content is likely resulting from hydrothermal activity along the Mission Creek Strand. 2.6.2. Mission Creek Subbasin The Mission Creek subbasin is bounded by two faults to the north and south, Mission Creek Strand and Banning Strand, respectively. To the east and west, the 34 subbasin is bounded by the semi-impermeable sediments of the Indio Hills and the impermeable rocks of the San Bernardino Mountains, respectively. The Mission Creek subbasin covers a surface area of 76 square miles (49,000 acres). The depth of the subbasin varies but sediment deposits are expected to be as deep as 2,000 meters (7,000 feet) (Slade, 1981). Geophysical studies conducted by Dr. Shawn Biehler of UC, Riverside suggest that the Mission Creek subbasin contains five layers which can be distinguished by seismic data, see Table 2-3 for layer distinctions (Slade, 1981). However, only the upper 600 meters (2,000 feet) are considered potable water (DWR, 2003). Groundwater is not extracted from deeper depths due to poor water quality and poor hydraulic communication between the unconsolidated and semi-consolidated material. Natural recharge of the subbasin occurs through subsurface flow in intermittent creeks and rivers such as Mission Creek and Little and Big Morongo Washes of the San Bernardino Mountains on the western boundary (see Figure 2-10 for watershed boundaries). Repeated flow events from Mission Creek have deposited a series of alluvial terrace deposits at the base of the San Bernardino Mountains, creating a shallow waterbearing zone known as the Mission Creek Upland (Figure 2-8) (DWR, 2003). The primary water bearing units are the unconsolidated late Pleistocene/ Holocene alluvial deposits and the Ocotillo conglomerate / Cabezon fanglomerate (CDRW, 1964). All water-bearing units are composed of unconsolidated material creating an unconfined aquifer system. The hydro-geochemistry of the Mission Creek subbasin is characterized by calcium-sodium-bicarbonate-chloride ions and groundwater temperature are ambient to 35 Table 2-3: Classification of layers present in the Mission Creek Subbasin distinguished by geophysical studies conducted by Dr. Shawn Biehler in 1979 (Slade, 1981). Layer Depth (ft) 1 - Thickness (ft) (m) 155 47 Velocity (ft/sec) Density (g/cc) Stratigraphic Sequence 2480-3658 1.62 Unconsolidated, active stream deposits and fanglomerate 2 155 564 171 5380-8434 2.16 Older alluviam, terrace deposits, and saturated Ocotillo conglomerate/ Cabazon fanglomerate 3 719 1488 453 9556-10810 2.39 Ocotillo conglomerate, Cabazon fanglomerate 4 2207 3806 1160 11075-12085 2.46 Palm Springs, Painted Hill, and Imperial Formations, and Coachella fanglomerate 5 6013 ? - 16000-16900 2.64 Igneous-metamorphic basement rock Depth to Basement: 6013 feet (1832 meters) 36 surface temperatures, ranging from 73-83F (Slade, 1981, DWR, 2003). Concentration levels of sulfates and TDS (<500 mg/L) are low while fluoride concentrations are moderate (Slade, 1981). A small amount of subsurface flow on the northern boundary has been detected along the Mission Creek Strand (Slade, 1981). Groundwater chemistry and quality similar to Desert Hot Springs subbasin is detected in the Mission Creek subbasin at this location. Leakage across the fault strand is noted in this region due to respectively higher levels of fluoride levels (7-9 mg/L) (Slade, 1981). 2.6.3. Indio Subbain The Indio subbasin is bounded on the north by the Banning Strand and the Indio Hills while impermeable rocks of the San Jacinto and Santa Rosa mountains create the southern boundary of the subbasin. The Indio subbasin extends from the San Gorgonio Pass to the northern extent of the Salton Sea, however, only the region within the study area is described here. In the study region, the Indio subbasin is divided into two subareas: Garnet Hill subarea and Whitewater subarea. A few discrepancies exist in previous studies conducted in the region, such as the reference to the Indio subbasin as the Whitewater subbasin and the Garnet Hill subarea as a distinctive subbasin (CDRW, 1964, DWR, 2003, ED-CVWD, 2013a, ED-CVWD, 2013b, MWH, 2013). In its entirety, the Indio subbasin covers a surface area of nearly 525 square miles (DWR, 2003). The primary water bearing units in the subbasin are heterogeneous Pleistocene and Holocene alluvial deposits, Pliocene alluvial deposits, and the Ocotillo conglomerate. Depth of the basin is unknown in the region; however, it is assumed that the water-bearing units are greater than ~600 meters (2,000 ft) in thickness (DWR, 2003). This assumption is due to decreasing resistivity with increasing depth in geophysical well 37 logs and a gravitation survey revealing gravity anomalies indicating that the deepest subbasin is located south of the Garnet Hill Strand (Proctor, 1968, Tyley, 1974, Reichard, 1992). However, only the upper ~300 meters (1,000 feet) of the aquifer are considered in computational modeling of groundwater resources (Tyley, 1974). Natural recharge of the subarea occurs from watersheds of the Little San Bernardino Mountains and natural subsurface flow from the Whitewater River and runoff from the San Jacinto Mountains (See Figure 2-10 for watershed boundaries). The hydro-geochemistry in the subbasin has been documented prior to the implementation of artificial recharge. The groundwater was dominated by calcium bicarbonate ions and TDS concentrations were relatively low, ~300 mg/L (DWR, 2003). Measurements were taken from the Palm Springs subarea because it dominates most of the subbasin. Groundwater flow in the Palm Springs subarea is generally in a southeasterly direction (Tyley, 1974). However, the gradient steepens at the base of Edom Hill, suggesting a barrier to groundwater flow (Tyley, 1974). The Garnet Hill subarea (also referred to as the Garnet Hill subbasin by CVWD and USGS) is bounded on the northern side by the Banning Strand and the Garnet Hill Strand on the south. The Garnet Hill subarea is regarded as an unconfined aquifer system. The Garnet Hill subarea is hydraulically connected to the Whitewater subarea in the upper 100 feet due to either the lack of faulting or the presence of a conduit fault zone along the Garnet Hill Strand (DWR, 1964, MHW, 2013). The Thermal subarea is a confined region within the Indio subbasin (ED-CVWD, 2013b). The Thermal subarea is comprised of interbedded sands, silts, and clays. This region displays anisotropic permeability where permeability parallel to the bedding is 38 several times greater than the permeability normal to bedding (ED-CVWD, 2013b). These anisotropic conditions suggest an aquitard or perched groundwater conditions. Shallow fine-grained zones have created a series of perched water tables in the Thermal subarea (ED-CVWD, 2013b). Thousand Palms subarea is a small region that is distinguished from other subareas due to its hydro-geochemistry. This region is characterized by sodium sulfate (ED-CVWD, 2013b). There is a sharp hydro-geochemical boundary between the Thermal subarea and the Thousand Palms subarea. An extension of the Garnet Hill Strand to the east would coincide with this hydro-geochemical anomaly. However, gravity measurements and residual gravity profiles do not suggest a subsurface fault in the region; thus, variations in hydro-geochemistry are attributed to variations in permeability within the strata (DWR, 1964). 2.7 Fault Zone Architecture Fault zones are structurally and hydrogeologically complex with heterogeneous zones that influence the geologic framework, as displayed in the San Andreas Fault zone in the northwest Coachella Valley, CA. Such heterogeneities promote anisotropic flow across fault zones influencing a variety of fluid-fault interactions. A simple conceptual model of fault zone architecture has developed over the past 30 years. Architectural components of fault zones are defined as three distinctive regions: main gouge zone (the fault core), damage host rock (damage zone), and undeformed host-rock (protolith rock) (Chester and Logan, 1986, Caine et al., 1996, Faulkner et al., 2010). This conceptual model involves a centralized fault core surrounded by a damage zone, as displayed in Figure 1-1 (Chester and Logan, 1986). The fault core generally 39 consists of gouge, cataclasite, and/or ultracataclasite while the damage zone contains subsidiary structures such as secondary faults, fractures, and folds (i.e. Faulkner et al., 2010). The fault core is defined as the region which accommodates most of the displacement where as the damage zone is mechanically related to the growth of the fault zone (Caine et al., 1996). Fault cores are not the same in every field study. Field based observations suggest that fault cores can be comprised of unconsolidated clay-rich gouge zones, brecciated and geochemically altered zones, or highly indurated cataclasite zones (Caine et al., 1996). However, all fault zones play an important role in controlling fluid flow. Grain size reduction and mineral precipitation along the fault plane generally yield lower porosity and permeability than the primary water units. The reduction in permeability is believed to be the main component that allows the fault core to act as a barrier to fluid flow (Caine et al., 1996). The damage zone consists of a complex network of fractures, faults, and folds. The subsidiary structures present in the damage zone cause anisotropic flow in the damage zone (Caine et al., 1996). The secondary features present can create flow paths for migration of fluids causing an increase in unidirectional permeability or increase tortuosity in the flow path causing a decrease in permeability. The degree of complexity associated with the development of a fault zone is indicative of the number of slip events that have occurred along the fault plane, as an overprinting of damage would have to occur along the fault plane and its associated damage zone. Faults are dynamic systems in which mechanical, geochemical, and hydrogeologic properties are dependent upon variations in lithology, temperature, pressure, and deformation rate; thus, the effectiveness of a fault to impede or enhance 40 fluid flow will vary in respect to time and space (Goodwin et al., 1999). Such dynamic relationships of the fault and fluid interaction system are illustrated in Figure 2-12. In these relationships, it is considered that faulting along a fault plane results in a decrease in fluid pressure (Goodwin et al., 1999). However, subsequent diagenesis and mineralization in the fault zone gradually decreases permeability of the fault zone resulting in an increase in fluid pressure. In return, the high fluid pressure decreases the effectiveness of normal stress across the fault plane, thus, facilitating seismic events (Goodwin et al., 1999). Figure 2-12: Dynamic relationship between faults and fluid systems. (Goodwin et al., 1999). 41 3. Methods For the purpose of this thesis, methods used are divided into field methods and laboratory methods. Laboratory analysis was conducted at CSU, Northridge and Core Laboratories: The Reservoir Optimization Company (referred to as Core Labs) in Bakersfield, California. 3.1 Field Methods Field investigations were primarily carried out in the months of August and September of 2013. Outcrop locations were selected due to the presence of offset in strata as indicated by aerial photography on Google Earth, personal exploration by others (Kimberly Blisniuk, Janice Gillespie, Richard Heermance, Alex Meyer), and analysis of USGS B4 LiDAR imagery. Another factor in selecting sample sites was convenience of access to sample sites. Samples sites had to be accessible either by vehicle or on foot. All samples were collected from fault outcrops along public property. Whenever possible, outcrops were cleaned to provide a relatively fresh surface in order to minimize the weathering effect. Samples were collected from three separate zones within the identified fault outcrop location- fault core, damage zone, and protolith rock, where possible. The protolith rock is regarded as a representative water-bearing unit of the aquifer system in the region. At locations where a consolidated sample was not collectable, unconsolidated sediment was collected for laboratory analysis. The bedding style and attributes (strike and dip) were recorded as well. Specimen colors were identified using the Munsell-Rock Color Chart as issued and standardized by the Geological Society of America. Thickness of fault core and width of damage zone was measured and recorded for comparison 42 among the different sampling locations. Sampling locations were recorded using global position system (GPS) device such as the Garmin etrex-Legend HCx. 3.1.2 Fault Zone architecture Characterized by Numerical Measurements The fault zone architecture and a fault’s ability to act as a barrier and/or conduit to fluid flow can be described by three numerical indices (Caine et al., 1996). The three indices, Fa, Fm, and Fs, are derived from the conceptual model of fault zone architecture (described in Figure 1-1) (Caine et al., 1996). Previous studies involving fault zone architecture have observed a range of fault zone architectures and developed a correlation between the fault zone architecture and permeability structure (Caine et al., 1996). F a is the fault zone architectural index and values can range from 0 to 1. The Fa provides a comparison of the damage zone width with the total fault zone width: (EQ 3-1) Fm= mean of Fa values for a single fault zone (EQ 3-2) (EQ3-3) When Fa is equal to 0, the damage zone is ideally absent causing the theoretically low permeability fault core to dominate the fault zone architecture and cause the fault zone to act as a barrier to fluid flow (Caine et al., 1996). However, when Fa is equal to 1, the fault core is ideally absent causing a theoretically high permeability damage zone to dominate the fault zone architecture and cause the fault zone to act as a conduit to fluid flow (Caine et al., 1996). Fm is an average of the Fa values obtained for a single fault 43 strand (Caine et al., 1996). Fm values incorporate Fa values measured at different transects. Fs is a spatially variable index for a single fault strand where Fa values may vary depending on the fault zone architecture (Caine et al., 1996). 3.2 Laboratory Methods 3.2.1 Core Laboratories: The Reservoir Optimization Company Core Laboratories is a leading reservoir optimization company using technology to conduct analysis of petrophysics and reservoir fluid behavior. Four samples (Samples I1, MF7, P1, and P3) were analyzed at Core Labs in Bakersfield, California for porosity and permeability. Not all samples could be analyzed at Core Labs due to size restrictions and fragility of the samples. Hand specimen samples were cored using a standard Drilling Press Diamond Tool with a 1-inch ID diamond bit (Core Laboratories, 2014). During the core drilling process, samples were cooled using liquid nitrogen in order to prevent samples from falling apart. Cored samples were placed in lead sleeves in order to keep samples intact. Basic measurements like initial weight, texture, grain size, and fluorescence were recorded before exerting 8000 psi of pressure to conform the lead sleeves to the samples. Samples were placed in a humidifying oven for 68 hours and then in a lab oven for 4 hours to dry. After the samples reached equilibrium with room temperature, porosity and permeability was measured. Porosity analysis was conducted using the Ultragrain Grain Volume- UGV200. The core sample is placed in a grain volume cell and helium is isothermally transferred from a reference cell to the grain volume cell, resulting in an equilibrium pressure between the two containers (Core Laboratories, 2014). Through Boyles Law, the grain volume is resolved, thus, providing a porosity measurement (Core Laboratories, 2014). 44 Volume and pressure measurements are recorded and analyzed with the UPore software. Permeability analysis was conducted using the Ultra-Perm 500. The Ultra-Perm 500 is a steady state gas permeameter system applied to plug size core samples (Core Laboratories, 2014). The software UPerm is able to record and plot unitized mercury pressure drop against unitized flow rate. This plotting mechanism allows the software to determine the airflow regime required for validation of Darcy’s Law. This methodology allows measurements to be within 0.0001 mD (millidarcies) of error (Core Laboratories, 2014). 3.2.2 Thin Section Analysis Optical properties of a rock are analyzed using a petrographic microscope. A thin section analysis was conducted on eight samples (B2, B3, MF4, MF7, P1, P3, P4, and P5). Thin sections were not constructed of all samples due to the fragility of the samples. Epoxy injected thin sections were prepared by R.A. Petrographic located in Los Angeles, California. Point counting analysis was conducted on each thin section utilizing the Gazzi-Dickinson method (Folk, 1974, Ingersoll, 1984, Dickinson, 1985). Point counting was conducted with a slide-advancing machine to advance the thin section. The point counting analysis (500-points) was used to distinguish grains, matrix, and void spaces, pore spaces, and epoxy veins. The region being recognized for each point count was identified by the crossing of the ocular micrometer scale present in the ocular portion of the petrographic microscope. Analysis of the thin section was conducted under a magnification of 4x lens. Point counts were recorded in Microsoft-Excel, converted to percentages, and used for grain vs. pore space comparisons. Point counting of thin sections for porosity analysis was selected because by definition a thin section provides a 45 cross sectional view of a rock, thus, if the rock is injected with blue epoxy, the blue epoxy should only reside within the pore space, as shown in Figure 3-1. Visual analysis of thin sections was conducted to examine for fractures promoting fluid migration. Thin sections were photographed using an Ernst Leitz GMBH Wetzlar Germany PRADO-500 thin section projector (CSUN ID # 19871), displayed in Figure 3-2. Figure 3-1: Schematic diagram of a thin section where grains and pore space are differentiated. Figure 3-2: Photograph of the Ernst Leitz PRADO-500 thin section projector used to photograph thin sections. 46 3.2.3 Porosity-Density Analysis Porosity in rocks is defined as the measurement of the void space in a rock sample (i.e. Davis and Dewiest, 1966, Fetter, 2001, Weight et al., 2001). Laboratory processes can be employed to derive porosity measurements from samples. In order to determine porosity of samples, the mass and volume of the sample must be calculated. Volume of an irregular sample is calculated using Archimedes’ Principle. The volume of the original sample solid is calculated as: Volume of water displaced= Final Volume of water- Initial Volume of water Volume of water displaced= Volume of solid (EQ 3-4) This equation assumes that no water enters the pore-space of the rock. For the volume calculations, a known volume of de-ionized water (DI-water) was poured into a graduated cylinder and the intact sample is placed into the graduated cylinder. The initial volume of water can be altered to accommodate the sample sizes, but must always be greater than the sample size. The amount of water displaced by the sample is equal to the volume of the bulk sample, assuming water does not percolate into the pore-space of the sample. This method is described in Figure 3-3. Figure 3-3: Volume of an irregularly shaped object is measured by the water displacement test where a known volume of water is poured into a graduated cylinder and the object is placed to measure the displacement of water. 47 The mass of the sample can be derived using standard laboratory techniques (i.e. Davis et al., 1966, Fetter, 2001, Weight et al., 2001, Kroetsch, et al., 2008). The wet samples are oven dried in a sediment oven at 105C for 24 hours in aluminum foil containers (Fetter, 2001). This expels any moisture in the samples that may be present on the surface of the grains. Samples were dried using a Precision Economy Oven (CSUN ID # 021649). The mass of the oven-dried sample is weighed using a calibrated digital scale (i.e. electronic balance). The density of the material is calculated by dividing the mass of the substance by the volume: (EQ 3-5) Total porosity (n) of a sample can be computed from a porosity relationship (i.e. Fetter, 2001, Weight et al., 2001, Kroetsch, et al., 2008): (EQ 3-6) wheren is the total porosity as a percentage is the bulk density of the material (g/cm3) is the particle density of the material (g/cm3). A standard of 2.65 g/cm3 is used for the particle density. The particle density of quartz is selected because from thin section analysis the grain composition is primarily quartz. 3.2.4 Grain Size Analysis A standard dry grain size analysis was conducted using a mechanical sieve shaker (Ro-Tap Testing Sieve Shaker Model B (CSUN ID #-007130/ 59936). Samples were oven dried for 24 hours at 105C in the sediment oven in aluminum foil containers (i.e. Fetter, 2001, Kalinski, 2011). Each sample was weighed before and after oven drying. 48 Some samples which remained consolidated were loosened with a pestle and mortar. Standard set of sediment sieves were used ranging from 64 mm to 0.045 mm (Table 3-1) Table 3-1: Sieves used in the grain size analysis. All sieves were provided by the Department of Geological Sciences at CSUN and grain size analysis was conducted at CSUN. Sieve # 2 1/2 1 1/4 5/8 5/8 3 1/2 5 6 7 8 10 14 16 18 20 25 30 35 40 45 50 60 80 100 120 170 200 230 325 pan Size (mm) 64.00 32.00 16.00 8.00 5.60 4.00 3.36 2.83 2.38 2.00 1.40 1.18 1.00 0.85 0.71 0.595 0.500 0.420 0.355 0.300 0.250 0.180 0.149 0.125 0.090 0.074 0.0625 0.045 0.000 Size() -6.00 -5.00 -4.00 -3.00 -2.50 -2.00 -1.75 -1.50 -1.25 -1.00 -0.50 0.00 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.50 2.75 3.00 3.50 3.75 4.00 4.50 0.00 49 Mass of Sieve (grams) 544.09 561.25 565.9 546.37 543.37 695.11 674.99 627.48 629.29 621.09 418.52 453.96 401.86 428.56 367.65 566.09 365.93 513.57 335.72 385.07 318.99 307.04 461.77 307.22 307.48 449.32 302.07 365.62 496.03 (Kalinski, 2011). All sieves and the bottom pan were weighed beforehand and cleaned after each sieve shake. The mechanical sieve shaker was operated in 15-minute intervals (Kalinski, 2011). For samples that required longer sieving times, the shaker was operated for an additional 10-minute interval. After the sample was thoroughly sieved, the sieves and samples were weighed. Graphical relationships between the particle diameter size and the percent finer than are used to determine the effective grain size (de) (Fetter, 2001, Kalinski, 2011). 3.2.5 Hydraulic Conductivity Analysis Fluid flow can be enhanced or retarded by the properties of the material in which the fluid is traveling through. The ability of a rock to transmit fluid can be described in terms of hydraulic conductivity and permeability. Permeability is defined as the measurement of the ability of porous material to allow fluid flow through media, whereas, hydraulic conductivity is defined as the measurement of the ease with which water flows through porous material (i.e. Davis et al., 1966, Fetter, 2001, Weight et al., 2001). One of the factors that affects hydraulic conductivity and intrinsic permeability is sorting of grains. From grain-size distributions and previously derived relationships by others, it is possible to relate grain size distribution, intrinsic permeability, and hydraulic conductivity. Empirical methods can be employed to determine the hydraulic conductivity where field methods cannot be implemented. Calculations of hydraulic conductivity of porous media can be derived from grain size analysis (Freeze and Cherry, 1979, Fetter, 2001, Cheng and Chen, 2007). A general relationship between grain size distributions and empirical methods can be summarized as: 50 where: (EQ 3-7) K= Hydraulic conductivity (L/T) g= Acceleration of gravity (L/T2) v= Kinematic coefficient of viscosity (L2/T) C= Dimensionless coefficient (n)= Function of porosity De = Effective grain size There are many variations of calculations for hydraulic conductivity found in literature. Three separate empirical equations are selected to calculate hydraulic conductivity on basis of the parameters required for the calculation. The Beyer relationship is independent of porosity where as the Slichter relationship incorporates the porosity value of the media. The uniformity coefficient (Cu) provides a correlation of measurement of the degree of how well or poorly sorted the sedimentary sample. The uniformity coefficient can be derived from grain size analysis and is equal to the ratio of the grain size that is 60% finer by weight (D60) to the grain size that is 10% finer by weight (D10) (i.e. Fetter, 2001): (EQ 3-8) The uniformity coefficient quantifies the degree of grain sorting within the sample. A sample with a Cu less than 4 is considered to be well sorted where as a sample with a C u of more than 6 is poorly sorted (i.e. Fetter, 2001). The Breyer equation is considered most useful for material that can be characterized with heterogeneous distributions and poor sorting. This equation is only valid for samples with a uniformity coefficient between 1- 20 and an effective grain size 51 between 0.06mm and 0.6mm. The Breyer equation does not factor in porosity but instead assumes a value of 1 for the porosity function (Cheng and Chen, 2007). The Breyer equation was used is (Pliakas, 2011): (EQ 3-9) whereK=Hydraulic Conductivity (m/s) g= Gravitational Constant (9.81 m/s2) v= Kinematic Viscosity (m2/s) CU= Uniformity Coefficient (dimensionless) D10= Effective grain size corresponding to the 10 % grain size (mm) curve The Slichter formula is considered applicable for effective grain sizes between 0.01 mm and 5 mm. The Slichter formula incorporates a porosity value of each sample and the effective grain size of the tenth percentile. The Slichter equation used is (Pliakas, 2011): (EQ 3-10) whereK=Hydraulic Conductivity (m/s) g= Gravitational Constant (9.81 m/s2) v= Kinematic Viscosity (m2/s) n= Porosity value (fraction) D10= Effective grain size corresponding to the 10 % grain size (mm) curve 3.2.6 Permeability Analysis Permeability measurements are dependent upon properties of the porous media. Intrinsic permeability (k) is dependent upon the characteristics of the sediment material, such as the properties associated with clays and sandstones. The relationship between hydraulic conductivity (K) and intrinsic permeability (k) is: 52 (EQ 3-11) where: K= Hydraulic Conductivity (m/s) k= Intrinsic permeability (m2) = Density of the fluid (water density) (1000 kg/m3) = Dynamic viscosity of the fluid (kg m /s) g= Acceleration due to gravity (9.81 m/s2) 53 4. Results The results of this project are presented with respect to the method used for analysis and by field location. Data collected is included in the appendices of this thesis and referenced in the text. 4.1. Field Observations Fault outcrop locations associated with the Mission Creek Strand, Banning Strand, and Garnet Hill Strand were selected due to the recommendation of other scientists who are currently or have in past conducted field studies in the region. Analysis of Google Earth aerial photography, topographic maps, and B4 LiDAR imagery was used to select sampling locations which would display fault outcrops. This analysis assisted in selecting accessible fault outcrop locations. Surface fault expressions associated with Garnet Hill Strand could not be identified within the study region; thus, no samples were collected for this strand. Fault outcrop and sampling locations are displayed in Figure 41. Two fault surface expressions were selected for the Banning Strand: (1) Whitewater Canyon and (2) a river cut near Via Las Palms road, Figures 4-2 and 4-3. Two fault outcrop locations were selected for the Mission Creek Strand: (1) Mt. View Road Cut and (2) Pushawalla Canyon, Figures 4-4 and 4-5. Two samples were collected from the Indio Hills to serve as samples of water bearing units. Field notes and detailed descriptions of field expeditions are provided in Appendix B and sample photographs are available in Appendix C. The Banning Strand crosses perpendicular to Whitewater Canyon and juxtaposes the San Gorgonio-Igneous-Metamorphic Complex with the Cabezon Fanglomerate (Yule 54 Figure 4-1: Locations of collected samples in the study region. Sampling locations a are indicated by a colored star on this map: Green Star: Location B- Whitewater Canyon (Banning Strand), Blue Star: Location MF- Mt. View Road Cut (Mission Creek Strand), White Star: Location I- Indio Hills, Red Star: Location RCB- River Cut near Las Palms Road (Banning Strand), and Orange Star: Location P- Pushawalla Canyon (Mission Creek Strand). (Base Layer Credits: ESRI, DeLorme, GEBCO, NOAA NGDC,USGS, and other Contributors) 55 Figure 4-2: B4 LiDAR data analysis of fault outcrops along the Banning Strand at Whitewater Canyon. Fault outcrop displays fault movement which places the San Gorgonio-Igneous-Metamorphic Complex against Cabezon Fanglomerate. Sampling location is indicated by the yellow star. (Source: OpenTopography Facility, 2014) 56 Figure 4-3: B4 LiDAR data analysis of fault outcrop along the Banning Strand at a river cut located near Las Palms Road. The Banning Strand is mapped as a single strand running along the southern edge of the Indio Hills. Sampling location is indicated by the yellow star. (Source: OpenTopography Facility, 2014) 57 Figure 4-4: B4 LiDAR data analysis of fault outcrop along the Mission Creek Strand. Fault outcrop displays recent movement of water bearing units caused by faulting at Miracle Hill. Sampling location is indicated by the yellow star. (Source: OpenTopography Facility, 2014) 58 Figure 4-5: B4 LiDAR data analysis of fault outcrop along the Mission Creek Strand at Pushawalla Canyon. The river carved canyon provides a cross-sectional view of the fault splays associated with the Mission Creek Strand. Sampling location is indicated by the yellow star. (Source: OpenTopography Facility, 2014) 59 and Sieh, 2003). At this location, the Banning Strand is characterized as a right lateral fault with a dip of 45 NW, as displayed in Figure 4-6 (Yule and Sieh, 2003). Four samples were collected from this location (B1, B2, B3, and B4) at the Whitewater Canyon site (Location B). Detailed field descriptions are available in Table B-1 (see Appendix B). A 13-meter wide fault zone is identified at this location with welldeveloped localized fault gouge, damage zone, and protolith rocks, as displayed in Figure 4-7. Sample B1 is a representative sample of the protolith rock north of the fault plane (hanging wall) and is collected from a range of 0-3.7 meters of the identified fault zone. Sample B1 is characterized as a weak and highly fractured metamorphic rock. Sample B2 is collected from the damage zone associated with the hanging wall at meter marking 8.28.7 meters. Sample B2 is considered a part of the deformed interior damage zone and is banded with a calcareous residue, which range in widths of 2 mm-16 mm. Sample B3 is a representative sample of the localized fault gouge, collected at meter marking 8.7-8.8 meters. The sample is clay-rich in appearance and texture with clay like layering. The interior damage zone associated with the footwall of the fault (located south of the fault) is moderately unconsolidated with cemented angular clasts within the matrix. Sample B4 was collected from the exterior damage zone associated with footwall. Sample B4 is an unconsolidated conglomerate with cobbles ranging in size from 0.5mm to 12 inches. The sample is classified as unconsolidated due the hydrogeologic definition which differentiates consolidated sediments as materials that have been metamorphosed or cemented together such as limestone or sandstone and unconsolidated sediments as sediments ranging from clay to gravel size with pore space connectivity which allows groundwater storage (USGS, 2013). Sample B4 is identified as the Cabezon 60 Figure 4-6: Banning Strand juxtaposing the SGIM Complex against the Cabezon Fanglomerate at Whitewater Canyon 61 Figure 4-7: Identification of localized fault gouge and damage zone associated with the Banning Strand at Whitewater Canyon 62 Fanglomerate, which is also identified as a water-bearing unit in the region An outcrop of a splay associated with the Banning Strand is noted in the Indio Hills near a river cut terrace east of Via Las Palms road. This location is referred to as sampling location: River Cut- Banning Strand (RCB). One sample was collected from this location (RCB 1). An 8.5-meter wide fault zone was identified with a localized fault gouge developed at the 3.7-meter marker, see Figure 4-8. The damage zone is composed of consolidated, very fine-grained sands and silts with popcorn weathering. The fault gouge is weakly developed with a thickness of 5-10 cm. The fault plane appears to be associated with a low angle thrust fault with a strike and dip of 250, 26NW. A representative sample of the fault gouge was collected (sample RCB1). A single strand outcrop of the Mission Creek Strand is not detected in the study region. However, splays associated with the Mission Creek Strand are noted along the northern boundary of the Indio Hills and Miracle Hill. In the city of Desert Hot Springs, California, splays of the Mission Creek Strand can be seen along a road cut of Mt. View Road. This location is referred to as sampling location: Mt. View Road Cut- Mission Creek Strand. A detailed description of field data is available in Table B-2 (see Appendix B). At this location, a 136.6-meter wide fault zone was identified on the hanging wall of the fault, as shown in Figure 4-9. No localized fault gouge was identified at this location; however, eight fault splays were identified. Six samples were collected, 5 samples (MF1, MF2, MF4, MF6, and MF7 are of consolidated precipitate filled damage fault cores, and 1 sample (MF 9) is of the loose matrix sediments. Splays of the Mission Creek Strand are noted at Pushawalla Canyon. At this location, the river carved canyon provides a cross-sectional view of the fault splays 63 Figure 4-8: Fault zone architecture associated with the Banning Strand at river cut near Via Las Palms road. A 8.5 meter wide fault zone was identified with localized fault gouge, ~5-10 centimeters wide. 64 Figure 4-9: Fault zone architecture associated with the Mission Creek Strand at Mt. View Rd. A 136.6-meter wide fault zone was identified as displayed in Figure 4-9a. Localized gouge was indentified is shown in Figure 4-9b and c. 65 associated with the Mission Creek Strand. This location is referred to as Pushawalla Canyon- Mission Creek Strand. A 43-meter wide fault zone was identified on the west sidewall of the canyon and localized fault gouge was identified on the east sidewall, as shown in Figure 4-10. A detailed description of field data is available in Table B-3 (see Appendix B). Five samples (P1-5) were collected from this location. Samples P1 and P2 are collected from the damage zone in the hanging wall from blue-grey siltstones and sandstones of terrace deposits or upper member of the Ocotillo Formation. Sample P3 is collected from a siltstone bed that is affected by fault-drag folding. Sample P4 is collected from the east sidewall. Sample P4 is from a 10 cm thick localized fault gouge which displays clay like layering. The material present in the damage zone associated with Sample P4 is of highly fractured sandstone. Sample P5 is collected from the protolith rock of the footwall associated with the fault zone. This sample does not appear to be effected by the faulting and is a representative sample of the water-bearing units in the region. The water-bearing units in the valley are composed of poorly sorted conglomerates, sandstones, and siltstones of continental origin. Surface deposits in the Indio Hills indicate deposits of lakebeds, fanglomerates, alluvium, and wind-blown sands. The description of these deposits is similar to rock descriptions found in water well and geothermal well logs located in the study region. Thus, samples collected from the Indio Hills are considered representative samples of water bearing units found throughout the valley. Samples I1 and I2 were collected along the north face of the Indio Hills. Detailed description of the samples is provided in Table B-5 (see Appendix B). A numerical evaluation of the Mission Creek and Banning Strands to determine if 66 Figure 4-10: Fault zone architecture associated with the Mission Creek Strand at Pushawalla Canyon. A 43-meter wide fault zone was identified as displayed in Figure 4-10a (western side). Localized gouge was indentified is shown in Figure 4-1-b (eastern side). 67 the strands will behave as a barrier or conduit to fluid flow was conducted according to the conceptual scheme for fault-related fluid flow according to numerical measurements of fault zone architecture and permeability structures developed by Caine and others (Caine et al., 1996). The fault zone indices derived from conceptual modeling of fault zone architecture of the Banning Strand and Mission Creek Strand are presented in Table 4-1 and displayed in Figures 4-11 and 4-12, respectively. 4.2. Results from Core Labs Four samples were processed at Core Labs in Bakersfield, California. Data collected from Core Labs and photographs of cores are presented in Appendix D. Analysis of Sample I1 provides a grain density of 2.70g/cc, porosity of 37.4%, and air permeability of 314.964 md. Analysis of Sample MF7 provides a grain density of 2.60 g/cc, porosity of 48.9%, and air permeability of 47.52 md. Analysis of Sample P1 provides a grain density of 2.68 g/cc, porosity of 26.8%, and air permeability of 8.462 md. Analysis of Sample P3 provides a grain density of 2.68 g/cc, porosity of 46.1% and air permeability of 23.03 md. 4.3 Thin Section Analysis Thin section analysis was conducted on Samples B2, B3, MF4, MF7, P1, P3, P4, and P5. Analysis of thin-sections included conducting a point count to indentify pore space and visual examination of the thin section. Photographs of thin sections under 1.25X magnification are available in Appendix E. Data collected from point count analysis by petrographic microscope is available in Appendix F and data collected from JMicroVision Software is available in Appendix G. 68 Table 4-1: Calculation of fault zone indices from Equations 3-1, 3-2, and 3-3 in accordance to Caine et al (1996). Fault Zone Styles According to Caine et al. (1996) Fault Zone Location Whitewater Canyon (B1) Banning Strand Mission Creek Strand Fault Zone Width Damage Zone Width (meters) (meters) 13 8.3 Fa 8.5 8.4 0.99 Mt. View Road Cut (MF) 136.6 135.78 0.99 Pushawalla Canyon (P) 43 69 Fs 0.64 River Cut near Via Las Palms Rd (RCB) 42.9 Fm (Mean of Fa) 0.99 0.81 0.35 0.99 0.004 Figure 4-11: Fault zone architectural indices for the Banning Strand. Fault zone indices fall between the barrier and conduit index, thus, indicating that the Banning Strand behaves as a barrier-conduit. Figure 4-12: Fault zone architectural indices for the Mission Creek Strand. Fault zone indices fall between the barrier and conduit index, thus, indicating that the Mission Creek Strand behaves as a conduit. 70 4.3.1 Thin Section Analysis Using Petrographic Microscope Results of porosity derived from thin-sections using a petrographic microscope listed in Figure 4-13. Two thin sections (B2 and B3) were created from samples along the Banning Strand at Whitewater Canyon. Sample B2 is from the damage zone associated with the hanging wall and Sample B3 is from the fault gouge. Visual analysis of Sample B2 shows the thin section contains many interconnected pathways which filled in with blue epoxy (see Figure 4-14). Point count analysis of this thin section shows a porosity count of 20.49%. The remaining 79.51% of the thin section is composed of matrix that is too fine to identify. It is presumed that the matrix is composed of silts and clay size particles. Visual analysis of Sample B3 displays three primary epoxy filled pathways propagating in a radial pattern from the epoxy injection point (see Figure 4-15). A number of air bubbles are also present in nearly one quarter of the thin section area. Point Count analysis of the thin section shows a porosity count of 39.04%. The remaining 60.96% is composed of primarily of matrix with rounded grains. Two thin sections (MF4 and MF7) were created from samples along the Mission Creek Strand at Mt. View road cut. Both samples are from the calcareous- precipitate fill found in the fault cores. Visual analysis of Sample MF4 displays angular grains of various sizes suspended in the surrounding matrix (see Figure 4-16). The thin section does not indicate any primary epoxy induced pathways. The point count analysis shows a porosity count of 19.40 %. The remainder of the thin section (80.60 %) is composed of large grains surrounded by matrix. Visual analysis of thin section MF7 shows a large breakage in the sample and other secondary epoxy induced pathways (see Figure 4-17). 71 Figure 4-13: Porosity derived from point counting using a petrographic microscope. 72 Fracture Matrix Matrix Grains Grains Fracture Matrix Fracture Grains Figure 4-14: Thin section of sample B2 in plain light (PPL). The thin section highlights some of the fractures developed in the thin section and regions of matrix and grains. 73 Fracture Grains Grains Fracture Matrix Matrix Air Bubble Figure 4-15: Thin section of Sample B3 in plain light (PPL). Some of the fractures developed, matrix, grains, and air bubbles are indentified in the thin section. 74 Air Bubble Fracture Grains Grains Matrix Grains Grains Matrix Matrix Matrix Grains Figure 4-16: Thin section of sample MF4 in plain light (PPL). Some of the fractures developed, matrix, grains, and air bubbles are indentified in the thin section. 75 Matrix Grains Fracture Grains Grains Matrix Fracture Matrix Matrix Matrix Figure 4-17: Thin section of Sample MF7 in plain light (PPL). Some of the fractures developed, matrix, and grains are indentified in the thin section. 76 The thin section contains numerous angular grains suspended in dense matrix. The point count analysis shows a porosity count of 16.67 %. The remainder of the thin section (83.33%) is composed of dense matrix with small grains. Four thin sections (P1, P3, P4, and P5) were created from samples along the Mission Creek Strand at Pushawalla Canyon. Visual analysis of thin section Sample P1 displays a high quantity of small angular grains with variations in the denseness of the matrix (see Figure 4-18). Regions with high density of matrix also display low quantity of grains and regions with low density of matrix display higher quantity of grains. Point counting analysis shows a porosity count of 14.60%. The remaining 85.40% is composed of angular grains surrounded by weak matrix and patches of dense matrix. Visual analysis of thin section Sample P3 displays a high quantity of sparsely connected weak matrix (see Figure 4-19). The thin section does not display a large quantity of whole grains. Point counting analysis indicates a porosity count of 31.67%. The remainder of the thin section (68.33%) is composed of sparsely interconnected-consolidated matrix. Visual analysis of thin section Sample P4 displays numerous epoxy-induced pathways (see Figure 4-20). The thin section displays a large quantity of angular grains suspended in matrix. Point counting analysis of the thin section indicates porosity count of 14.80%. The remainder of the thin section (85.20%) is composed of matrix and whole grains. Visual analysis of thin section Sample P5 displays a high quantity of sub-rounded grains with very little matrix development (see Figure 4-21). Point counting analysis of the thin section indicates a porosity of 13.92%. The remainder of the thin section (86.08%) is composed of whole grains. 77 Matrix Grains Grains Grains Matrix Fracture Matrix Fracture Figure 4-18: Thin section of Sample P1 in plain light (PPL). ). Some of the fractures developed, matrix, and grains are indentified in the thin section. 78 Matrix Matrix Matrix Grains Fracture Matrix Figure 4-19: Thin section of P3 in plain light (PPL). Some of the fractures developed, matrix, and grains are indentified in the thin section. 79 Grains Fractures Matrix Grains Matrix Fracture Matrix Matrix Grains Figure 4-20: Thin section of Sample P4 in plain light (PPL). Some of the fractures developed, matrix, and grains are indentified in the thin section. 80 Grains Grains Grains Matrix Grains Grains Figure 4-21: Thin Section of Sample P5 in plain light (PPL). Some of the fractures developed, matrix, and grains are indentified in the thin section. 81 4.3.2. JMicroVision Software Point counting using JMicroVision was conducted using photographs of thin sections. JMicroVision uses a random grid system fixated upon the pixel. Three separate categories were created to classify pixels- Pore Space, Grain, and Void. The void category applies to the regions where the point counter landed on a portion of the photograph which is not a part of the thin section or regions where epoxy-filled pore space resulted from thin section construction. Void spaces within the thin section and opaque grains could not be differentiated in the photographs. Thin section diagrams with point counts are provided in Appendix G. Porosity calculations derived from JMicroVision ranges from ~7% to ~46% (see Figure 4-22). 4.4 Porosity-Density Data Porosity data for all the samples was derived mathematically using a porosity and sample density relationship, displayed in EQ 3-6. Each sample was processed three times in order to calculate an average. Porosity calculations for each sample and trial run are included in Appendix H. Porosity calculations used for analysis are displayed in Figure 423. The porosity data is categorized by the sampling location within the fault zone. The protolith samples (I1, I2, B1, and P5) have porosities ranging from ~9% to ~44%. The damage zone samples (B2, P1, P2, P3, and MF9) have porosities ranging from ~10% to 68%. The fault core samples (B3, MF1, MF2, MF4, MF6, MF7, P4, and RCB1) have porosities ranging from ~17% to ~52%. A comparison of all porosity values calculated is presented in Table 4-2. A porosity calculation from bulk density was not calculated for sample B4 because no consolidated material was found in the sample. The material collected for sample B4 was 82 Figure 4-22: Porosity derived from point counting using a JMicroVision software. 83 Figure 4-23: Porosity derived from porosity and sample density relationship. 84 Table 4-2: Porosity comparison of the different methods employed to derive porosity from unconsolidated sediments. The porosity values derived from the Density Porosity method is the only method with a standard deviation error (1) calculation. 85 all loose material. 4.5 Grain Size Analysis A grain size analysis was performed on all the samples using dry sieving mechanisms. Grain analysis and distribution curves are included in Appendix I. Grain size distribution curves were plotted using DPlot software. Four effective grain size diameters were recorded from the grain-size distribution curves. The effective grain sizes are D10, D20, D50, and D60 (mm), see Table 4-3 for values. The D10 effective grain sizes range from 0.03 mm – 0.21 mm. The D20 effective grain sizes range from 0.05 mm0.55mm. The D50 effective grain sizes range from 0.07 mm- 1.53 mm. The D60 effective grain sizes range from 0.06 mm – 2.36 mm. The effective grain-sizes of D10 and D60 are used to calculate the uniformity coefficient of the each sample, as provided in Table 4-4. Sample ranges from well sorted to poorly sorted. 4.6 Hydraulic Conductivity Hydraulic conductivity of each sample was calculated using two separate empirical relationships. The Breyer equation (EQ 3-9) relates effective grain size of D10 from grain size distribution and uniformity coefficient to hydraulic conductivity. Derived hydraulic conductivity values using the Breyer equation are presented in Table 4-5. Breyer hydraulic conductivity values range from ~13 m/s to ~1875 m/s. The Slichter equation (EQ 3-10) relates effective grain size of D10 from grain size distribution and porosity of the sample to the hydraulic conductivity. Hydraulic conductivity values using the Slichter equation are present in Table 4-6. Slichter hydraulic conductivity values range from ~0.17 m/s to ~151 m/s. The difference in calculated hydraulic conductivity values of the two equations is due to different variables used in the calculation. 86 Table 4-3: Effective grain size measurements from grain size distribution curves. D 10 correlates to the 10th %, D20 correlates to the 20th %, and D50 correlates to the 50th % and D60 correlates to the 60th % of sediments finer than. Millimeter P5 Fault Zone Location Protolith B1 Protolith 9.34 0.06 0.11 0.44 0.64 B4 I1 I2 Protolith Protolith Protolith NA 44.78 27.52 0.19 0.07 0.21 0.45 0.09 0.41 2.59 0.14 1.53 5.33 0.17 2.36 MF1 Fault Core 17.62 0.09 0.26 1.29 1.96 MF2 MF4 MF6 Fault Core Fault Core Fault Core 29.09 18.95 47.69 0.10 0.09 0.12 0.27 0.22 0.29 1.19 0.73 1.19 1.83 0.75 1.71 MF7 P4 Fault Core Fault Core 30.52 52.82 0.20 0.06 0.38 0.13 1.38 0.54 2.03 0.71 B3 RCB1 Fault Core Fault Core 23.58 29.20 0.13 0.06 0.26 0.18 0.69 1.51 0.73 2.25 MF9 P1 Damage Zone Damage Zone 10.81 24.81 0.08 0.08 0.14 0.18 0.45 0.69 0.63 0.73 P2 P3 Damage Zone Damage Zone 68.88 32.83 0.03 0.03 0.05 0.04 0.19 0.07 0.21 0.08 B2 Damage Zone 18.52 0.33 0.55 0.75 0.78 Sample Average Porosity D10 D20 D50 D60 10.73 0.07 0.15 0.63 0.92 87 Table 4-4: Uniformity coefficient (Cu) for samples organized by their location within the fault zone architecture. Cu values less than 4 are considered to be well sorted and Cu values greater than 6 are considered to be poorly sorted. Sample P5 B1 B4 I1 I2 MF1 MF2 MF4 MF6 MF7 P4 B3 RCB1 MF9 P1 P2 P3 B2 Fault Zone Location Protolith Protolith Protolith Protolith Protolith Fault Core Fault Core Fault Core Fault Core Fault Core Fault Core Fault Core Fault Core Damage Zone Damage Zone Damage Zone Damage Zone Damage Zone Uniformity Coefficient Cu= D60/D10 12.43 11.20 28.07 2.43 11.41 20.76 18.00 7.89 14.38 9.95 10.97 5.70 39.43 7.91 9.67 6.15 2.64 2.35 88 Sorting Poorly Sorted Poorly Sorted Poorly Sorted Well Sorted Poorly Sorted Poorly Sorted Poorly Sorted Poorly Sorted Poorly Sorted Poorly Sorted Poorly Sorted Poorly Sorted Poorly Sorted Poorly Sorted Poorly Sorted Poorly Sorted Well Sorted Well Sorted Table 4-5: Hydraulic conductivity (K) and intrinsic permeability (ki) values calculated using the Breyer equation. Breyer Hydraulic Conductivity Permeability Sample Fault Zone Location KBreyer ki-Breyer P5 Protolith 65.04 m/s 5.29592E-06 m2 B1 Protolith 40.00 m/s 3.25688E-06 m2 B4 Protolith 331.50 m/s 2.69938E-05 m2 I1 Protolith 79.58 m/s 6.47991E-06 m2 I2 Protolith 516.04 m/s 4.20205E-05 m2 MF1 Fault Core 222.97 m/s 7.16703E-06 m2 MF2 Fault Core 270.15 m/s 8.68326E-06 m2 MF4 Fault Core 293.22 m/s 9.42495E-06 m2 MF6 Fault Core 391.91 m/s 1.25972E-05 m2 MF7 Fault Core 1279.19 m/s 4.11169E-05 m2 P4 Fault Core 51.45 m/s 4.18935E-06 m2 B3 Fault Core 236.21 m/s 1.92342E-05 m2 RCB1 Fault Core 26.49 m/s 2.15686E-06 m2 MF9 Damage Zone 207.12 m/s 6.65751E-06 m2 P1 Damage Zone 72.42 m/s 5.89696E-06 m2 P2 Damage Zone 16.70 m/s 1.35948E-06 m2 P3 Damage Zone 13.78 m/s 1.12218E-06 m2 B2 Damage Zone 1875.66 m/s 0.000152733 m2 Density of water Dynamic Viscosity Gravitational Kinematic Constant Viscosity Temperature 30°C 90°C g (m/s2) v (m2/s) r (kg/m3) m (kg m/s) 9.8 8.01E-07 3.26E-07 1000 0.000798 0.000315 89 Table 4-6: Hydraulic conductivity (K) and intrinsic permeability (k i) values calculated using the Slichter equation. Slichter Hydraulic Conductivity Permeability KSlichter ki-Slichter P5 Fault Zone Location Protolith 0.44 m/s 3.578E-08 m2 B1 Protolith 0.17 m/s 1.36E-08 m2 B4 Protolith NA NA I1 Protolith 40.89 m/s 3.33E-06 m2 I2 Protolith 75.36 m/s 6.136E-06 m2 MF1 Fault Core 3.64 m/s 1.17E-07 m2 MF2 Fault Core 21.92 m/s 7.047E-07 m2 MF4 Fault Core 4.66 m/s 1.498E-07 m2 MF6 Fault Core 151.28 m/s 4.863E-06 m2 MF7 Fault Core 103.17 m/s 3.316E-06 m2 P4 Fault Core 63.44 m/s 5.166E-06 m2 B3 Fault Core 17.56 m/s 1.43E-06 m2 RCB1 Fault Core 7.00 m/s 5.7E-07 m2 MF9 Damage Zone 0.52 m/s 1.673E-08 m2 P1 Damage Zone 7.21 m/s 5.869E-07 m2 P2 Damage Zone 42.78 m/s 3.483E-06 m2 P3 Damage Zone 2.59 m/s 2.111E-07 m2 B2 Damage Zone 52.59 m/s 4.282E-06 m2 Sample Gravitational Constant Kinematic Viscosity Density of water Temperature g (m/s2) v (m2/s) r (kg/m3) 30°C 90°C 9.8 8.01E-07 3.26E-07 1000 90 Dynamic Viscosity m (kg m/s) 0.000798 0.000315 4.7 Intrinsic Permeability Permeability data was derived from an empirical relationship between the hydraulic conductivity and intrinsic permeability. This relationship relates the hydraulic conductivity to properties of the fluid. Intrinsic permeability calculations are derived from the Breyer and Slichter formulas. Intrinsic permeability calculated by using the hydraulic conductivity values from the Breyer equation are presented in Table 4-5. Intrinsic permeability values range from ~9.42E-6 m2 to ~1.52E-4 m2. Calculation of permeability values using the hydraulic conductivity values from the Slichter equation are presented in Table 4-6. Intrinsic permeability values range from ~3.6E-8 m2 to ~6.1E6 m 2. 91 5. Discussion The results of this study are discussed with respect to the method used for analysis. 5.1 Field Data Numerical measurements of fault zone architecture and permeability structure derived from conceptual modeling developed by Caine et al (1996) of fault zones indicates that the Mission Creek Strand behaves as a conduit to fluid flow and the Banning Strand behaves as conduit-barrier fluid flow systems (Figures 4-10 and 4-11). The Mission Creek Strand is characterized as a conduit because the theoretically calculated values associated with the fault zone measurements fall on the conduit extent of the indices created by Caine et al. (1996). The Banning Strand is categorized as conduit-barriers because the theoretical values associated with the fault zone measurements fall in between the barrier and conduit extents set by Caine et. al (1996). In Caine et al. (1996)’s classification the Fa index numbers of 0 indicating a barrier forming fault and 1 indicating a conduit forming fault. Classification and measurement of the fault zone architecture was conducted in the field, thus some discrepancies may exist in the assignment of a damage zone and fault gouge, especially along the Mission Creek Strand at the Mt. View Road Cut location. However, these discrepancies should not alter the identification of zones present in the fault zone architecture. The major discrepancy is expected to lie in the identification of the extent of the damage zone because it is noticed that the width of the damage zone varies along strike. For instance, at Location B1-Whitewater Canyon the damage zone associated with the footwall maybe greater than measured; however, the sampling location is disrupted due to erosion. Another discrepancy may lie in the number of fault 92 cores indentified because some fault planes may not be visible due to deposition of material post-fault movement causing some splays to not be exposed at the surface. Even if some portions of a damage zone or fault cores were not measured in this study, I believe that it will not have drastic impacts on the conceptual classification of these faults. However, additional fault outcrop measurements will aide to provide a more robust data set to calculate the fault zone architectural indices. Sample collection methods can be improved in future fault zone studies by collecting additional samples using a coring device. However, the applicability of a coring device to collect samples of a known size is unknown due to the frailness and unconsolidated stages of some sampling locations. 5.2 Core Labs Data The data analysis conducted by Core Labs provides a standardized data set for comparison. Analysis of the four samples indicates that the protolith rock (Sample I1) has the highest air permeability (314.96 millidarcies (md)), as expected. However, the discrepancy in the data exists between the fault core sample (MF7) and samples from the damage zone (P1 and P3). Sample MF7 indicates an air permeability of 47.52 millidarcies (md) where as Samples P1 and P3 indicate air permeabilities of 8.462 md and 23.03 md, respectively (results are provided in Appendix D- Figure D-1). The four samples processed were associated with the Mission Creek Strand; however, they are associated with different regions of the fault strand. Thus, a correlation of permeability changes across the Mission Creek fault zone cannot be concluded from the samples processed at Core Labs. If the samples are to be applied to the Mission Creek Strand, the results would indicate a permeable fault gouge but an impermeable damage zone. This 93 data is contrary to the conceptual scheme for fault-related fluid flow developed by Caine et al. (1996) where the damage zone is believed to be more permeable than the fault gouge. Samples P1 and P3 are from the same locality, Pushawalla Canyon, however, they indicate varied permeabilities (Figure D-1 located in Appendix D). This difference in calculation can be a result of variation in lithology of the siltstones and sandstones indicating that facies variations during deposition maybe evident, as identified in Figure 5-1. 5.3 Thin Section Data Two different methods were employed to conduct a point count analysis on thin sections in order to calculate pore space relative to grain/matrix space in a thin section. One method involved using a petrographic microscope with a 4X lens and slideadvancing machine to maneuver the thin section, while the second method involved computer software which randomly picked a point based on the pixel location. A discrepancy exists between the two methods used for point counting. The difference in porosity calculation is attributed to the fact that thin-section maneuvering mechanisms employed differ. The slide-advancing machine maneuvered in a lateral format, right to left. Thus, point counts were conducted in a linear format across the thin section. In some cases, only half of the thin section may have been examined for pore space because only a 500-point count was conducted. Point counting on the JMicroVision software was conducted randomly with the point counter moving to various locations of the thin section. Thus, with this method, the entire thin section had the same probability of being examined for pore space under a 500-point count. 94 Figure 5-1: Lithofacies identification at Pushawalla Canyon (west-side wall). Yellow outlines indicate cobble/coarse-gravel deposits. Red outlines indicate fine silt to clay deposits. The remaining material is identified as sand ranging from coarse sand to fine sand. Person provided for scale and photographs correlated with regions A, B, and C are provided below. 95 96 One of the caveats associated with conducting a point-count using JMicroVision, is that it is not possible to distinguish between void spaces and opaque grain orientations. On the other hand, the petrographic microscope provides a higher resolution for accurate determination of pore space vs. grain/matrix space. Due to the randomization of the point counting associated with using JMicroVision, these porosity values are believed to be more accurate than the porosity values calculated using the petrographic microscope. Randomization of point counting across the entire thin section is important for these samples do to the deformation and /or widening of permeable pathways which may have formed due to injection of epoxy. Porosity calculations derived from thin sections are not common due to the deformation that can occur in the thin section making process. From a visual analysis of the thin sections, it is noted that most of the deformation may have occurred due to epoxy injection. Thus, I suggest conducting thin section analysis without epoxy injection. Another method that can be used for pore space analysis using thin section is injecting the thin section with a fluorescing element and using a fluorescence microscope to determine pore space. The fluorescence microscope would also be able to indicate micro pore spaces and pore spaces less than 30 microns (thickness of a thin section). Thin sections should also be examined for degree of cementation that exists in the sample. Thus, a mechanism to measure and indicate cementation in a sample must be devised. The porosity data derived from the thin sections is not accounted for in the hydraulic conductivity and intrinsic permeability calculations because values could not be derived for all samples. The porosity calculations derived from thin-sections indicate high porosity values ranging from 13.91%- 39.04% for thin sections examined via microscope 97 and 7.89%- 46.64% for thin sections examined via JMicroVision. These high porosity values are indicative of high permeability values. However, the expected permeability values and recorded porosity values are not reproducible by standardized testing at Core Labs. Thus, the data derived from thin sections is disregarded and the thin sections are used for visual analysis. Visual analysis of the thin sections provides a cross sectional view of the samples. From the visual analysis, it can be noted that many of the samples developed fracture networks when injected with epoxy. 5.4 Porosity-Density Relationship Porosity for all the samples was derived from an empirical relationship between bulk density of the sample and the grain density. Three separate sample runs were performed on each sample to calculate porosity and average of the values is considered the true porosity of the sample. Data from each sample run is included in Appendix H. In some of the porosity runs, a negative porosity value was calculated. The negative porosity calculations are not considered representative of the sample because a negative porosity value is not realistic, thus, these values were not included in the average calculation. The negative porosity values could have been due to heterogeneity in the sample used in the sample run for volume calculation because negative values were not calculated for every trial run for a sample. Errors associated with the porosity calculation may occur when determining the volume of the sample. The porosity calculations may be higher than the actual porosity values because the bulk volume of the sample was not properly calculated due to infiltration of water into the air-filled pore spaces during the water displacement test. Infiltration of water into the pore spaces was noted due to the presence of air bubbles in the beaker and the disintegration of the sample in the beaker. 98 Thus, it may be possible that for some samples only the volume of the grains was recorded and not the volume of the sample. However, a true sample volume was calculated for samples where the cementation blocked the infiltration of the water. The infiltration of water into the sample would result in a lower volume calculations and resulting in higher bulk density calculations. The miscalculation of the bulk density of the material had direct impact on the porosity calculations because of the selected porosity calculation equation (EQ 3-6). From the various methods employed to calculate porosity, it is noted that the calculations show high variations in porosity for samples from the protolith rock, damage zones, and fault cores. The porosity values did not display a significant trend across the fault zones identified, implying that the faults do not form barriers or conduits in this region. The high variation in the porosity values across the fault zones can be a result of facies variations during deposition. It is noted that not all samples were collected in a lateral form across the fault zone. Thus, not all samples included in the damage zone, fault core, and protolith rock can be stratigraphically correlated. Deposition of various facies is evident in the region due to the convergence of various stream systems, eolian deposits, possible debris flows, and/or variations in alluvial fill material. The various lithologies present in the region are displayed in the geologic map of the study area (Figure 5-2). 5.5 Grain Size analysis A dry sieving grain size analysis was conducted on all the samples. The grain size analysis displays a high degree of heterogeneity within the samples (see Table 4-4 for Uniformity Coefficient values). The sieving process provided an effective grain size (D 10) 99 Figure 5-2: Geologic map of quaternary surficial deposits in the study region. (modified map of Palm Springs 30’X60’ Quadrangle by California Geological Survey, 2012) 100 and analysis of grain size distributions (provided in Appendix I). From the grain-size distribution curves, it is noted that the samples should be processed with a hydrometer or laser particle analyzer to allow the measurement of particle sizes smaller than 0.045mm. Processing samples with clay size particles using a hydrometer or laser particle analyzer will provide a more thorough analysis of the grain size distribution. Comparison of the uniformity coefficients associated with the fault strands does not indicate any correlation between grain-size distributions and fault strand, as indicated in Table 5-1. A comparison of the uniformity coefficients at each sampling location does not indicate that samples associated with a single fault strand are more uniform than the other strand. Some samples (P2, P5 MF1, MF2, MF4, MF7, and B1) needed to be disaggregated using a pestle and mortar. Thus, these samples may contain grains that were crushed during the process due to the pestle hitting a weak plane in the grain. This process should not add a high degree of error as the samples are highly heterogeneous in the first place. The effective grain size calculated from the grain-size distribution charts is D10. However, for some samples, the distribution did not extend to the grain diameter that is correlated to 10% finer by weight (90% coarser by weight). For these samples (MF2, P2, and P3), the grain size curve was extended and the D10 value is approximated. This approximation process can be corrected by conducting hydrometer test or laser particle analyzer. 5.6 Hydraulic Conductivity and Permeability Hydraulic conductivity and intrinsic permeability of the samples was derived from empirical relationships relating grain size and porosity. It is noted that hydraulic conductivity values can vary depending on the equation used, thus, two different 101 Table 5-1: Comparison of Uniformity Coefficients (CU) categorized by fault strand. An average of CU is provided for each sampling location. Protolith Banning Strand Mission Creek Strand Sample Uniformity Coefficient Average P1 P2 P3 P4 P5 MF1 MF2 MF4 MF6 MF7 MF9 9.67 6.15 2.64 10.97 12.43 20.76 18 7.89 14.38 9.95 7.91 B1 11.2 B2 2.35 B3 5.7 B4 28.07 RCB1 39.43 I1 2.43 I2 11.41 8.372 13.1483 11.83 39.43 6.92 equations are used in this analysis (EQ 3-9 and EQ 3-10). An examination of the hydraulic conductivity values across the Banning Strand at Whitewater Canyon (Location- B) indicates lower hydraulic conductivity in the protolith rock and the fault core than in the damage zone (see Figure 5-3). Analysis of the hydraulic conductivity rates and permeability values indicate that the damage zone is more permeable than the fault core and protolith rock. Thus, the fault core impedes fluid flow while the damage zone enhances flow. Analysis of values calculated for the Banning Strand at river cut near Via Las Palms Road (Location RCB) indicates lower hydraulic conductivity rates for the fault 102 Figure 5-3: Schematic diagram of the fault zone architecture at sampling location B (Whitewater Canyon – Banning Strand). Calculated hydraulic conductivities are displayed in their respective location within the architecture. Fluid flow at this location is in the north to south direction. Sample B1 is not considered in the analysis because it is from the SGIM Complex, which is not a water-bearing unit. core than the protolith rock. The intrinsic permeability values indicate that the fault core is less permeable than the protolith rock. See Figure 5-4 for distribution of hydraulic conductivity and intrinsic permeability values across the fault zone architecture. An average of hydraulic conductivity and intrinsic permeability values was used for the protolith rock region of the architectural fault zone because more than one sample was collected. Analysis of hydraulic conductivity rates and intrinsic permeability values indicate that the fault core is less permeable than the surrounding protolith rock. Thus, the fault core impedes fluid flow at this sampling location. 103 Figure 5-4: Schematic diagram of the fault zone architecture identified at a river cut near Via Las Palms Rd (location RCB-Banning Strand). Calculated hydraulic conductivities and permeabilities are displayed in their respective location within the architecture. Fluid flow at this location is believed to be north/northwest to south/southeast direction. Analysis of values calculated for the Mission Creek Strand at Mt. View Road Cut (Location- MF) indicates higher hydraulic conductivity rates and intrinsic permeability values for the fault cores than the damage zone and protolith rocks (See Figure 5-5). An average of hydraulic conductivity and an average of intrinsic permeability values were used for the fault core samples and the protolith rock samples because more than one sample was collected related to the respected region of the architectural fault zone. At this location, the damage zone behaves as a barrier rather than the fault core. The damage zone associated with the fault zone is less permeable than the fault core. Analysis of values calculated for the Mission Creek Strand at Pushawalla Canyon (location P) indicates a higher hydraulic conductivity and intrinsic permeability for the fault core than 104 Figure 5-5: Schematic diagram of fault zone architecture identified at the Mt. View Road Cut (Location MF-Mission Creek Strand). Calculated hydraulic conductivities and permeabilities are displayed in their respective location within the architecture. An average was calculated where more than one sample was collected for the defined zone. Fluid flow at this location is believed to be north/northwest to south/southeast direction. the damage zone and protolith rocks. Figure 5-6 displays a schematic diagram of the fault zone with the hydraulic conductivity and intrinsic permeability values with respect to their locations within the fault zone architecture. An average of hydraulic conductivity and intrinsic permeability values were calculated for the damage zone because multiple samples were collected for this region of the fault zone architecture. The hydraulic conductivity and intrinsic permeability values calculated for this location indicate that the fault core is more permeable than the damage zone. Each sample location is characterized to contain the three fault zone architectural elements discussed in Figure 1-1. Hydraulic conductivity and intrinsic permeability 105 Figure 5-6: Schematic diagram of the fault zone architecture identified at Pushawalla Canyon (Location P-Mission Creek Strand). Calculated hydraulic conductivities and permeabilities are displayed in their respective location within the architecture. An average was calculated where more than one sample was collected for the defined zone. Fluid flow at this location is believed to be a north to south direction. values for the distinguished fault zone architectural zones are provided in Table 5-2. Graphical representation of the hydraulic conductivity values for the damage zone and fault core with respect to the protolith rock provides a visualization of the dominating hydraulic conductivity associated with the architectural zone. Figure 5-7 indicates that samples associated with the Mission Creek Strand (Locations MF and P) are more influenced by the hydraulic conductivity within the fault core than in the damage zone, whereas samples associated with the Banning Strand (Location B) are more dominated by the hydraulic conductivity within the damage zone. Samples associated with the fault zone identified at RCB are not accounted for because samples associated with the damage zone were not collected and processed. Thus, these hydraulic conductivity values associated with the damage zone at RCB cannot be calculated or assumed. A comparison 106 Table 5-2: Hydraulic conductivity and intrinsic permeability values for distinguished zones within the fault zone architecture. Sample Location Fault Zone Location B RCB MF KBreyer Slichter KSlichter ki-Slichter m/s 3.25688E-06 m2 0.17 m/s 1.4E-08 m2 1875.66 m/s 0.000152733 m2 52.59 m/s 4.3E-06 m2 236.21 m/s 1.92342E-05 m2 17.56 m/s 1.4E-06 m2 Protolith B1 Damage Zone B2 Fault Core B3 Protolith B4 Protolith I1 and I2 Fault Core RCB1 Protolith I1 and I2 297.81 m/s Damage Zone 207.12 m/s 491.49 m/s 1.57979E-05 m2 56.94 m/s 1.8E-06 m2 Protolith MF9 MF1, MF2, MF4, MF6, MF7 P5 2.42502E-05 m2 58.13 m/s 4.7E-06 m2 2 2 6.66E-06 m 0.52 m/s 1.7E-08 m 65.04 m/s Damage Zone P1, P2, P3 34.30 m/s Fault Core P4 51.45 m/s 5.29592E-06 m2 0.44 m/s 3.6E-08 m2 2.79287E-06 m2 17.53 m/s 1.4E-06 m2 4.18935E-06 m2 63.44 m/s 5.2E-06 m2 Fault Core P Sample(s) Breyer ki-Breyer 40.00 NA NA 331.50 m/s 2.69938E-05 m2 297.81 m/s 2.42502E-05 m2 58.13 m/s 4.7E-06 m2 26.49 m/s 2.15686E-06 m2 7.00 m/s 5.7E-07 m2 107 Figure 5-7: Graphical representation of fault core and damage zone hydraulic conductivity values with respect to the hydraulic conductivity values of the protolith rock. The graphical representation indicates that the samples associated with the Mission Creek Strand (locations MF and P) fall in the conduit forming fault core region while samples associated with the Banning Strand (locations B and RCB) fall in the barrier forming fault core region. 108 of hydraulic conductivity values for the fault core and the damage zone indicates distinctions for which zone is more hydraulically conductive than the other regions of the fault zone. Figure 5-8 indicates that samples associated with the Mission Creek Strand (Locations MF and P) have higher hydraulic conductivities for the fault core. Whereas, samples associated with the Banning Strand (Location B) have higher hydraulic conductivities for the damage zone than the fault core. Samples from location RCB are not displayed because samples associated with the damage zone were not collected. 5.7 Implications My analysis of intrinsic permeability and hydraulic conductivity values contradict the analysis done in accordance to Caine et al. (1996), where the Mission Creek Strand is characterized as a conduit-forming fault and the Banning Strand is characterized as barrier-conduit fault. Caine et al.’s research stated that the fault core is the portion of the fault zone which has the ability to act as a barrier and the damage zone has the ability to act as a conduit. The analysis of the intrinsic permeability and hydraulic conductivity values across the fault zone architecture indicates that different portions of the fault zone can impede or enhance fluid flow. Along the Mission Creek Strand, it is noted that the fault core is more permeable relative to the damage zone; whereas, along the Banning Strand the damage zone is more permeable relative to the fault core. Analysis of the data at all four sampling locations indicates that the fault zones are barriers to fluid flow relative to the protolith rock; however, the barrier is not a result of just fault gouge development. The barriers formed by the Mission Creek Strand and the Banning Strand are a result of the damage zone and the fault gouge, respectively. The data implies that Caine et al.’s (1996) theoretical classification scheme of indentifying a barrier, barrier- 109 Figure 5-8: Graphical representation of fault core and damage zone hydraulic conductivity values with respect to each other. The graphical representation indicates that the samples associated with the Mission Creek Strand (locations MF and P) fall in the conduit forming fault core region while samples associated with the Banning Strand (locations B and RCB) fall in the conduit forming damage zone region. 110 conduit, and conduit forming faults does not apply to all faults, such as the active San Andreas Fault zone where unconsolidated sediments are faulted. 5.8 Fractoconformity vs. Fault Zone Controlled Fluid Flow Differentiating whether the groundwater table, in northwestern Coachella Valley, is affected by a fractoconformity or by the fault zone architecture is important in order to develop a better understanding of the dynamic relationship between faults and fluid interaction system along the San Andreas Fault zone. If the groundwater table were offset due to a fractoconformity then permeability variations within the fault zone would not have drastic effects upon the pore pressure and effective normal stress across the fault zone. The fractoconformity would not affect the dynamic relationship between fault and fluid systems because offset in the water table would be an offset resulting from vertical displacement of water-bearing units. However, if the groundwater table were offset due to gouge development in the fault zone then extraction and injection of water would impact the dynamic relationship between the faults and fluid systems by altering the pore pressure and effective stresses across the fault zone. The groundwater table would impact the permeability field of a fault zone, fluid pressure, and the effective normal stress across the fault zone; thus, potentially facilitating seismic events along a fault plane (Figure 2-12). My analysis of intrinsic permeability and hydraulic conductivity indicate that the Mission Creek and Banning Strands behave as barriers to groundwater flow due to development of the fault zone architecture, which contradicts the theoretical measurements of Caine et al., (1996) and Catchings et al., (2009) seismic survey. Distinguishing which portion of a fault zone acts as a barrier to groundwater flow is important in order to measure the effectiveness of a groundwater flow barrier. The 111 fault core is characterized as a region where sediments have undergone diagenesis and cataclasis during the fault movement process, whereas, the damage zone is characterized as a zone of deformation and fracture development that has occurred during the fault movement process. Thus, along faults where the fault core acts as a barrier, like the Banning Strand, groundwater flow is retarded due to gouge development and precipitation of carbonate within fractures. Thus, future seismic events along this strand will further encourage the diagenesis of grains along the fault plane and enhance the fault core’s ability to behave as barrier. Along faults where the damage zone acts as a barrier, like the Mission Creek Strand, groundwater flow is retarded due to deformation occurring in the damage zone. Both fault strands, Mission Creek and Banning Strands, behave as barriers to groundwater flow today as shown by a decrease in intrinsic permeability and hydraulic conductivity in the fault zone, however, the barrier is influenced by not only the fault core but by the damage zone as well. 112 6. Conclusion The investigations carried out and presented in this thesis represent a preliminary effort to characterize fluid flow across the San Andreas Fault zone in Northwest Coachella Valley, California. Outcrop investigations allowed detailed mapping of the fault zone architecture associated with the Mission Creek and Banning Strands (no outcrops of the Garnet Hill Strand were indentified in the study region) revealing measurements of fault zone architecture, hydraulic conductivity, and permeability across the fault zones. Porosity variations across the fault zone were determined using three different methods: thin sections, Core Labs Inc., and density-porosity relationship from a water displacement test. Porosity ranged from 9.34% +/- 7.77% to 68.88% +/- 26.11% in the region according to the water displacement test. Intrinsic permeability was determined empirically from hydraulic conductivity values and grain size distributions. Two different equations were used to calculate hydraulic conductivity in order to display possibility of error from porosity measurements. The Breyer equation was selected because it did not require a porosity measurement and hydraulic conductivity measurements ranged from 13.79 m/s to 1875.66 m/s. The Slichter equation was selected because it did require a porosity measurement and hydraulic conductivity measurements ranged from 0.17 m/s to 151.28 m/s. The variations in the calculated values for porosity, hydraulic conductivity, and intrinsic permeability indicate that values can vary depending on the methodology and empirical formula used to derive values. My hypothesis of a fractoconformity being the main cause of variations in the groundwater table rather than a fault behaving as a barrier to fluid flow was not supported 113 by the measured intrinsic permeability and hydraulic conductivity values. However, my data provides insight to what causes a barrier along Mission Creek Strand and the Banning Strand. My analysis of the intrinsic permeability and hydraulic conductivity values across the fault zone architecture associated with the Mission Creek and Banning Strands provide insight as to why these fault cause the water table to be offset. My data indicates that the Mission Creek Strand behaves as a barrier to fluid flow due the presence of an impermeable damage zone; whereas, the Banning Strand behaves as a barrier to fluid flow due to the presence of an impermeable fault core (see Figure 5-7). This analysis indicates that different components within a fault zone can impede or enhance fluid flow with respect to the other components, thus, a general classification system of classifying a fault’s ability to behave as a barrier or conduit in unconsolidated sediments does not exist. 114 References Allen, C.R., 1957, San Andreas Fault Zone in San Gorgonio Pass, Southern California, Bulletin of the Geological Society of America, v.68, p.315-360. Annual Climatologically Summary (ACS), 2014, U.S. Department of CommerceNational Climatic Summary for Palm Springs, CA, US (Station COOP: 046635): www.ncdc.noaa.gov (Accessed January 2014). Behr, W.M., Rood, D.H., Fletcher, K.E., Guzman, N., Finkel, R., Hanks, T.C., Hudnut, K.W., Kendrick, K.J., Platt, J.P., Sharp, W.D., Weldon, R.J., Yule, J.D., 2010, Uncertainties in slip-rate estimates for the Mission Creek Strand of the southern Sand Andreas fault at Biskra Palms Oasis, southern California: Geological Society of American Bulletin, v. 122, no.9/10, p. 1360-1377. Biehler, S., 1964, Geophysical Study of the Salton Trough of Southern California [Masters Thesis]: California Institute of Technology. Boggs, S., 2006, Principles of Sedimentology and Stratigraphy, 4th edition, New Jersey, Prentice-Hall, Inc. CDWR (California Department of Water Resources), 1964, Coachella Valley Investigation: Bulletin No. 108, p. 1-102. Caine, J.S., Evans, J.P., and Forster, C.B., 1996, Fault Zone Architecture and Permeability Structure, Geology, v.24, no. 11, p.1025-1028. Catchings, R.D., Rymer, M.J., Goldman, M.R., and Gandhok, G., 2009, San Andreas Fault Geometry at Desert Hot Springs, California, and Its Effects on Earthquake Hazards and Groundwater, Bulletin of the Seismological Society of America, v. 99, no. 4, p. 2190-2207. Cheng, C., and Chen, X., 2007, Evaluation of Methods for Determination of Hydraulic Properties in an Aquifer- Aquitard System Hydrologically Connected to a River, Hydrogeology Journal, v. 15, p. 669-678. Chester, F.M., and Logan, J.M., 1986, Implications of Mechanical Properties of Brittle Faults from Observations of the Punchbowl Fault Zone, California, Pure Applied Geophysics, v. 124, p.76-106. Core Laboratories, 2014, Core Lab Instruments: http://www.corelab.com (accessed April 2014). Davis. S.N. and DeWiest, R.J.M., 1966, Hydrogeology, New York, John Wiley and Sons, Inc. 115 Dibblee, T. W., Jr., 1954, Geology of the Imperial Valley region, California: in Jahns, R. H., ed.,Geology of southern California: California Division of Mines Bulletin 170, ch. 2, pt. 2, p. 21-28 and Plate 2 (Generalized geologic map of the Imperial Valley region, map scale 1 inch =6 miles). Dickinson, W.R., 1985, Interpreting provenance relations from detrital modes of sandstone, in Zuffa, G. G., ed., NATO ASI Series C:Mathematical and physical sciences, Volume 148: Provenance of arenites: Dordrecht, The Netherlands, D. Reidel, p. 333-361. Dolan, J.F., Bowman, D.D., Sammis, C.G., 2007, Long-range and long-term fault interactions in Southern California, Geology, v.35, p. 855-858. Dorsey, R., 2013, Late Cenozoic Evolution of the Southern San Andreas Fault System: Insights from Stratigraphy and Basin Analysis: USGS Earthquake Science Center Weekly Seminar (accessed January 29, 2014). Department of Water Resources (DWR), 2003, California’s Groundwater: Bulletin 118, P. 1-265. Engineering Department- Coachella Valley Water District, 2013a, Engineer’s Report on Water Supply and Replenishment Assessment- Mission Creek Subbasin Area of Benefit 2013-2014: Coachella Valley Water District. Engineering Department- Coachella Valley Water District, 2013b, Engineer’s Report on Water Supply and Replenishment Assessment- Upper Whitewater River Subbasin Area of Benefit 2013-2014: Coachella Valley Water District. ESRI, DeLorme, GEBCO, NOAA NGDC, USGS, i-Cubed, USDA, FSA, AEX, GeoEye, Getmapping, Aerogrid, and IGP, 2014, BaseMaps for ArcGIS 10.2. Faulkner, D.R., Jackson, C.A.L., Lunn, R.J., Schlische, R.W., Shipton, Z.K., Wibberley, C.A.J., Withjack, M.O., 2010, A review of recent developments concerning the structure, mechanics, and fluid flow properties of fault zones, Journal of Structural Geology, v. 32, p. 1557-1575. Fetter, C.W., 1988, Applied Hydrogeology, 4th Edition, New Jersey, Prentice-Hall, Inc. Folk, R.L., 1974, Petrology of Sedimentary Rocks: Austin, Texas, Hemphill Publishing Company, p. 1-190. Freeze, R.A., and Cherry, J.A., 1979, Groundwater, 1st Edition, New Jersey, PrenticeHall, Inc, p. 1-604. Goodwin, L.B., Mozley, P.S., Moore, J.C., and Haneberg, W.C., 1999, Introduction in Faults and Subsurface Fluid Flow in the Shallow Crust, American Geophysical Union, Geophysical Monograph 113, p. 1-6. 116 Guzman, N.E., 2010, Geology and Geomorphology of the Intersecting Mission Creek and Banning Fault Strands, Southern San Andreas Fault [Masters Thesis]: California State University, Northridge. Ingersoll, R.V., Bullard, T.F., Ford, R.L., Grimm, J.P., Pickle, J.D., and Sares, S.W., 1984, The effect of grain size on detrital modes: A test of the Gazzi-Dickinson pointcounting method: Journal of Sedimentary Petrology, v. 54, p. 103-116. Kalinski, M.E., 2011, Soil Mechanics Lab Manual, 2nd Edition, John Wiley and Sons, Inc. Keller, E.A., Bonkowski, M.S., Korsch, R.J., Shilemon, R.J., 1982, Tectonic geomorphology of the San Andreas fault zone in the southern Indio Hills, Coachella Valley, California, Geological Society of America Bulletin, v. 93, p. 45-56. Kroetsch, D., and Wang, C., 2008, Particle Size Distribution in Soil Sampling and Methods of Analysis, 2nd edition, Canadian Society of Soil Science. Langenheim, V.E., Jachens, R.C., Matti, J.C., Hauksson, E., Morton, D.M., Christensen, A., 2005, Geophysical evidence for wedging in the San Gorgonio Pass structural knot, southern San Andreas fault zone, southern California, Geological Society of America Bulletin, v.117, no.11/12, p.1554-1572. Matti, J.C., and Morton, D.M., 1993, Paleogeographic evolution of the San Andreas fault in Southern California: A reconstruction based on a new cross-fault correlation, in Powell, R.E., Weldon, R.J. II, and Matti, J.C., eds., The San Andreas Fault System: Displacement, Palinspastic Reconstruction, and Geologic Evolution: Boulder, Colorado, Geologic Society of America Memoir 178, p. 107-160. MWH, 2013, Mission Creek and Garnet Hill Subbasins Water Management Plan, January 2013: Coachella Valley Water District, Desert Water Agency, and Mission Springs Water District, p. 1-181. National Climatic Summary, 2014, U.S. Department of Commerce- Annual Climatologically Summary for Palm Springs, CA, US, (COOP: 046635): www.ncdc.noaa.gov (Accessed January 2014). Nye, D.P., 1994, A Hydrogeological Model of Willis Palms and Thousand Palms Oases, Upper Coachella Valley, Southern California [Masters Thesis]: California State University, Fullerton. OpenTopography Facility, 2014, B4 Project- Southern San Andreas and San Jacinto Faults, Airborne LiDAR, Survey Date: 05/18/2005- 05/27/2005. Powell, R.E., 1993, Balanced Palinspastic Reconstruction of Pre-Late Cenozoic Paleogeology, Southern California: Geologic and Kinematic Constraints on Evolution of 117 the San Andreas Fault System, in Powell, R.E., Weldon, R.J. II, and Matti, J.C., eds., The San Andreas Fault System: Displacement, Palinspastic Reconstruction, and Geologic Evolution: Boulder, Colorado, Geologic Society of America Memoir 178, p. 1-106. Proctor, R.J., 1968, Geology of the Desert Hot Springs- Upper Coachella Valley area, California: California Division of Mines and Geology, Special Report 94 (paper copy), p. 1-54. Reichard, E.G., and Meadows, J.K., 1992, Evaluation of a Ground-Water Flow and Transport Model of the Upper Coachella Valley, California: U.S. Geological Survey Water- Resources Investigation Report 91-4142 (print), p. 1-107. Slade, Richard C., 1981, Hydrogeologic Conditions in the Mission Creek Subbasin Upper Coachella, California. In Geology of the San Jacinto Mountains. Annual Field Trip Guidebook No. 9. South Coast Geological Society, p.151-163. Tyley, S.J., 1974, Analog Model Study of the Ground-Water Basin of the Upper Coachella Valley, California: U.S. Geological Survey Water Supply Paper 2027 (print), p. 1-81. USGS (United States Geologic Survey), 2013, The Hydrology of Groundwater: http://md.water.usgs.gov/faq/groundwater.html (Accessed: June 2014). Wallace, R.E., 1990, Then San Andreas Fault System, California: U.S. Geological Survey Professional Paper 1515 (paper copy), p. 1-304. Weight, W.D. and Sonderegger, J.L., 2001, Manual of Applied Field Hydrogeology, New York, McGraw-Hill, p. 1-608. Yule, J.D., and Sieh, K., 2003, Complexities of the San Andreas fault near San Gorgonio Pass: Implications for large earthquakes, Journal of Geophysical Research, v. 108, no.B11, p. 2548. Yule, J. D., 2009, The Enigmatic San Gorgonio Pass: Student-Only Field Trip, 2009 SCEC Annual Meeting, Field Trip Guidebook, p. 1-6. Yule, J.D., 2014, Personal Communication. 118 Appendix A-Aerial Photography Figure A-5a: Vegetation alignment along the Banning Strand at Whitewater Canyon. Figure A-5b: Vegetation alignment along the Mission Creek Strand near the city of Desert Hot Springs, CA. 119 Figure A-5c/d: Vegetation alignment along Banning Strand, west of the Indio Hills. This location is often referred to as Seven Palms. Figure A-5e: Vegetation along the Mission Creek Strand, on the north side of the Indio Hills. Location e is referred to as 1000 Palms. 120 Figure A-5g: Vegetation along the Mission Creek Strand, in the Indio Hills. Location g is referred to as Pushawalla Canyon. Figure A-5g: Vegetation along the Banning Strand, along the south side of the Indio Hills. Location f is referred to as Willis Palms Oasis 121 Appendix B-Field Notes Table B-1 8/18/2013 Location: Whitewater Canyon Date: 33.947025, 116.651536 GPS Location: Fault: Banning Strand Weather- Sunny with a high of 104F with light desert rain flashes. Fault outcrop of the Banning Strand where the Pre-Cambrian San GorgonioField DescriptionIgneous-Metamorphic Complex (SGIM Complex) basement rock is faulted against the late-Pleistocene and/or early-Holocene Cabezon Fanglomerate rocks. A 13 meter-wide fault zone (measured North to South) is identified at this location Fault Zone Identificationwith a well developed localized fault gouge, a clay-rich brecciated damage zone, and surrouding protolith rock. Description Width (m) SGIM Complex with gneissic banding. Banding is alternating in color between very pale orange (10YR 8/2) and dusky yellow green (5GY 5/2). Banding thickness varies from slivers of ~10cm to zones of ~1m-1.5m. This 0-3.7m region is highly fractured causing the rock to be weakly consolidated. Sample B1 collected. Two measurements of attitude were recorded (strike,dip): 250°,66NW and 270°,41NW. Majority of the material is from the SGIM Complex, however, it is more deformed and less consolidated. 3.7-4.7m Vertical fracturing which is perpendicular to the bedding plane is present. Fractures are spaced out by ~5cm-10cm. Contact between SGIM and dark grey material. The dark grey material is weakly consolidated and fine grained. This material is considered to be apart of the damage zone. 4.7m Strike and dip were measured from a consolidated plane: 270,50NW. 122 4.7-5.5m 5.5-8.2m 8.2-8.7m 8.7-8.8m 8.8-8.9m 8.9-13m Continued Table B-1 Exterior damage zone with angular to sub-angular clasts (~0.5mm-2.0mm). The clasts are classified as fault zone breccia. Contains very fine grains and is weakly consolidated. Interior damage zone. Dark yellowish brown in color (10Y 4/2). Clay-rich in appearance and in texture. Carbonate rich veins present. The entire region is HCL reactant. Deformed interior damage zone. Moderate yellowish brown in color (10YR 5/4). Material is banded by a calcareous residue with is 2mm-16mm in width. Representative sample of deformation zone is collected at this location (Sample B2). Localized fault gouge. Moderate brown in color (5YR 3/4). Material is clay-rich is apperance, texture and contains clay-like layering. Root traces and burrowing is present in the region. Representative sample of localized fault core-gouge is collected (Sample B3) Moderately consolidated and cemented breccia clasts within in the matrix of the fault gouge. This region is identified as the interior damage zone. Very unconsolidated (weakly) conglomerate of various size cobbles (~0.5mm12inches). Grains are sub-angular to sub-rounded. A representative sample was collected from the surface of the unconsolidated material (Sample B4). Clast count and identification was conducted in the field as well. 123 Continued Table B-1 Width (meters) 9m Description Size (cm) 1 50 8 6 15 Type Granite Granite Granite Granite Matrix Granite Schist -> Gneiss Granite Matrix-silty to coarse sand Matrix- Coarsed sand with silty matrix Diorite Granite Hornblend rich dike or hornblend diorite Granite Granite Matrix Granite Gneiss Matrix Matrix Granite Quartizite (from a pigmatite) Granite Granite Matrix Matrix Matrix Granite Pigmatite Granite Granite Quartizite Matrix Matrix Matrix Granite 15 12 Hornblend diarite granite 2 5 1 4 5 13 15 9 5 50 Clast Count 20 10 1 15 20 6 2 30 1 13m 124 Continued Table B-1 Width (meters) 13m Description Size (cm) 1 3 3 3 1 2 4 1 8 1 10 1 2 5 2 12 1 Clast Count 3 15 2 10 9m Type Matrix Matrix Granite Matrix Basalt Matrix Schist Diorite Granite Granite Metamorpic Diorite-grey meta-sed rock Granite Granite Shale Granite Granite Granite Granite Garnet Schist Granite Granite Matrix Matrix Matrix Granite Matrix Matrix Granite Quartzite Granite Matrix 125 Table B-2 Location: Mt. View Road Cut 33.941113, -116.475172 GPS Location: Weather- Sunny with a high of 104°F with light desert rain flashes. Field Description- Fault Zone Identification- 8/18/2013 Date: Fault: Mission Creek Strand Fault outcrop of splays associated with the Mission Creek Strand. Outcrop is located on the upthrown block of the Mission Creek Strand and bounded on the northern end by the Miracle Hill Fault. The outcrop being examined is known as Miracle Hill. A 136.6 meter-wide fault zone is identified at this location. No localized fault gouge was found at this location, however, numerous minor fault splays and/or fractures were identified. Fault zone was measured across in a South to North direction where fracture and/or splays were identified by meter markings. The minor faults are identified as Splays 1-8. Meter marking Description 0m Start of uplifted block. The 0-meter marking is considered to be the base location of the Mission Creek Strand. Representative sample collected of the loose matrix sediments (MF9). 22.10m 37.40 m 41m Splay 1: a calcareous cemented pebbly consolidated sand-precipitate filled core. South of the splay is a sand lens that is truncated by Splay 1, showing a slip of >3 meters. No clay like fault gouge or damage zone present. However, a calcareous plane (~7cm in width) is present with a strike,dip of 140°, 64°S. Representative sample collected (MF1). Splay 2: calcareous pebbly sand-precipitate filled core. Truncates beds of sand lens against a bed of conglomerate. Unknown displacement. Only calcareous in the splay zone. Calcareous plane varies in width from ~3 cm to ~7 cm and has a strike,dip of 146°,69°S. Representative sample collected (MF2). Splay 3: calcareous unconsolidated precipitate filled core with a thickness of ~3-4 cm. Splay displays either down dropped block to the south or up thrown block to the north, however, no sense of displacement is identified. Fault plane has a strike,dip of 135°,40°S. No sample collected. 126 Meter marking Continued Table B-2 Description 52.2m Splay 4a: Calcareous pebble sandy- precipitate filled core with a thickness of ~4-5 cm. The fault plane has a strike,dip of 095°,72°S. Splay 4a displays >3 meters of apparent dispalcement. 53.7m Splay 4b: Calcareous pebble sandy- precipitate filled core with a thickness of ~4-5 cm. The fault plane has a strike/dip of 101°,61°S. Representative sample collected (MF4). 74.9m Splay 5a: Very calcareous pebbly sandy-precipitate core with a thickness of ~10-15 cm. Precipitate core contains pebbles upto 3 cm in diameter. Fault splay truncates a channel fill which has pebbly conglomerate at the base. The pebbly conglomerate has calcareous cementing. The fault plane has a strike, dip of 135°,69°S. Fault core is not very consolidated. No sample collected due to the sparcy nature of the fault plane. 77.4m Splay 5b: Moderately consolidated pebbly sandy-precipitate core with a thickness of ~ 5-10 cm. Pebble in fault core are not deformed. Fault core is not planar but is instead wavy and contains pebble to cobble size loose material. Fault plane has a strike,dip of 240°,60°N, dip is in opposite direction than splay 5a. No sample collected. 93.4m Splay 6: Very calcareous pebbly coarse sandy-moderatly consolidated core with a width of ~6-8 cm. Fault truncates pebbly sand lens on the north and conglomerate beds with large boulders on the south. Fault plane has a strike,dip of 094°,69°S and extended up to 1/2 meter from the surface. Representative sample collected from base of the plane because there is a bee-hive at the top (MF6). 107m Splay 7: Very calcareous pebbly coarse sand/precipitate filled core with a thickness of ~15 cm. The fault plane has a strike,dip of 135°,70°S and displays slicken lines with a rake of 15°SW of the strike. Fault plane is consolidated and and a representative sample was collected (MF7). 112m Splay 8a: calcareous pebbly coarse sandy core with a thickness of ~2-3 cm. Fault plane does not have the same white precipitate filled texture and appearance in the same manner as splays 1-7. Fault plane has a strike and dip of 135°,60°S. Sediments surrounding the fault plane are calcareous. No sample was collected. 127 Continued Table B-2 Description Meter marking 118.5m Splay 8b: pebbly coarse sandy core with a thickness of ~2-3 cm. Fault plane does not have the same white precipitate filled core but sediments are calcareous surrounding the fault plane. The fault displays >1 meter of slip and has a strike,dip of 110°,60°S. No sample was collected. 128 W es t Si de W all Table B-3 Location: Pushawalla Canyon Date: 9/8/2013 GPS Location: 33.821139, -116.286875 Fault: Mission Creek Strand Weather- Sunny with a high of 98°F with light cloud coverage. Fault outcrop of splays associated with the Mission Creek Strand. Outcrop location is suggested by Kimberly Field Blisniuk. The river carved canyon provides a cross-sectional view of the fault splays associated with the Mission DescriptionCreek strand. A 43 meter wide fault zone, measured from the South to North was identified on the western side of the canyon. No Fault Zone localized fault gouge was found on the western wall of the canyon, however, a well developed localized fault gouge Identificationwas located on the eastern wall of the canyon. Meter GPS Location Description marking 0 meter marking is marked by a gullie with undeformed beds on the southern side 33.821139, and beds with minor deformation and offset on the northern side. The beds on the -116.286875 southern side of the gullie have an apparent dip to the NE which could be the result of fault dragging. The material north of the 0 meter marking is composed of various beds of sandstone 0m and shale. Colors vary from browns to greens. These deposits maybe associated with the upper member of the Ocotillo Formation or terrace deposits. The material on the south side of the 0 meter marking is composed of conglomerate channels and fine grained sandstone beds. In comparison to the the northern portion, the southern portion has more variations in composition and colors. In the river bed, vegatation is only evident on the north side of the 0 meter marking. 33.821164, 29.4 meter marking is noted because beds start to slope in a nearly verticle 29.4m -116.286583 direction. Sample P3 was collected. 40 meter marking is noted because the beds identified at the 29.4m appear to be 33.821189, 40m horiztonal to deposition. Sample (P1 and P2) were collected slightly north of the 40 -116.286397 meter marking. 129 Continued Table B-3 Meter marking GPS Location Ea st Si d eW al l 33.820767, -116.286456 33.819711 -116.286741 Description The eastern side of the canyon wall was not measured because deformation of rocks was not noticable. Localized fault gouge with clay like layering and a thickness of ~10cm. A representative sample of gouge was collected (P4). The localized gouge is Dark Yellowish Brown (10YR 4/2) in color. The fault plane has a strike, dip of 170The riverbed surrounding the outcrop is densly vegatative. On the northside of the gouge, the material is Yellowish Gray (5Y 7/2) in color. The material is highly fractured with oxidation present in the fractures. Grain size ranges from 0.5mm to clay size. On the southside of the gouge, the material ranges in color from Moderate Yellow (5Y 7/6) to Greyish Orange (10YR 7/4). The material is also fractured. Grain size ranges from 9in-7in to clay size grain grains. The southside is more conglomerate based. Sample P5 collected is from the footwall of the fault zone. The sample is collected to serve as representative sample of water-bearing units away from the fault zone (protolith rock sample). \ 130 Table B-4 Location: River Cut Banning Date: 9/8/2013 GPS Location: 33.842786, -116.370136 Fault: Banning Strand Sunny with a high of 98°F with light cloud coverage. Desert thunder storm clouds are starting to develop. WeatherFault outcrop associated with the Banning Strand. Outcrop location is suggested by Kimberly Blisniuk. Field The river carved canyon provides a cross-sectional view of a fault outcrop associated with the Banning Descriptionstrand. Fault Zone A 8.5 meter wide fault zone was identified (measured North to South) with identifiable fault gouge at 3.7 Identification- meters. Meter marking 0m 3.7m GPS Location Description 33.842786, -116.370136 Start of what appears to be a damage zone associated with faulting. Material is highly affected by popcorn weathering but is consolidated and not brittle. The fault zone is defined by the popcorn weathing. North of the 0 meter marker, the beds are dipping near verticle. This may be associated with down drop/slumping of a block due to faulting or erosion. Localized fault gouge. Fault plane appears to be associated with a low angle thrust fault with a strike,dip of 250, 26NW. Gouge is weakly developed and appears to be 5cm-10cm in thickness. Representative sample collected of gouge (RCB-1). 131 Table B-5 Location: Indio Hills Date: 3/14/2014 GPS Location: 33.833296, -116.3060639 Weather- Sunny with a high of 87°F with light breeze. Field Description- David Nye's thesis from CSU, Fullerton states that water travels through permeable units of the Indio Hills in a NW to S and/or SW direction. Thus, samples are collected from the Indio Hills as representative samples of water bearing units. Sample Description Sample I1 is a very fine grained siltstone. It is weakly consolidated. Homogenous in composition. Sample I2 is a fine grained sandstone with pebbles. Sample contains oxidation in fractures. It is weakly consolidated and hetrogenous in composition. I1 I2 132 Appendix C-Samples Figure C-1: Sample B1 in outcrop (A) and collected sample (B and C). Scale is provided in cm. 133 Figure C-2: Sample B2 in outcrop (A) and collected sample (B). Scale is provided in cm. 134 Figure C-3: Sample B3 in outcrop (A) and collected sample (B and C). Scale is provided in cm. 135 Figure C-4: Sample MF1 collected (A and B). Scale is provided in cm. 136 Figure C-5: Sample MF2 in outcrop (A) and collected sample (B, C, and D). Scale is provided in cm. 137 Figure C-6: Sample MF4 collected. Scale is provided in cm. 138 Figure C-7: Sample MF6 in outcrop (A) and collected sample (B and C). Scale is provided in cm. 139 Figure C-8: Sample MF7 in outcrop (A) and collected sample (B and C). Scale is provided in cm. 140 Figure C-9: Sample MF9 in outcrop. Rock hammer for scale. 141 Figure C-10: Sample P1 collected sample. Scale is provided in cm. 142 Figure C-11: Sample P2 collected sample. Scale is provided in cm. 143 Figure C-12: Sample P3 collected sample. Scale is provided in cm. 144 Figure C-13: Sample P4 in outcrop (A) and collected sample (B and C). Scale is provided in cm. 145 Figure C-14: Sample P5 collected sample. Scale is provided in cm. 146 Figure C-15: Sample RCB1 in outcrop (A) and collected sample (B, C, and D). Scale is provided in cm. 147 Figure C-16: Sample I1 collected sample. Scale is provided in cm. Figure C-17: Sample I2 collected sample. Scale is provided in cm. 148 Appendix D-Core Labs Data Figure D-1: Modified table of data collected at Core Labs showing air permeability and porosity measurements. 149 Figure D-2: Photographs of samples cored at Core Labs in lead sleeves (Samples I1, MF7, P1, and P3). 150 Appendix E- Thin-sections under 1.25X Figure E-1: Sample B2 thin-section under PPL and XP with 1.25x magnification 151 Figure E-2: Sample B3 thin-section under PPL and XP with 1.25x magnification 152 Figure E-3: Sample MF4 thin-section under PPL and XP with 1.25x magnification 153 Figure E-4: Sample MF7 thin-section under PPL and XP with 1.25x magnification 154 Figure E-5: Sample P1 thin-section under PPL and XP with 1.25x magnification 155 Figure E-6: Sample P3 thin-section under PPL and XP with 1.25x magnification 156 Figure E-7: Sample P4 thin-section under PPL and XP with 1.25x magnification 157 Figure E-8: Sample P5 thin-section under PPL and XP with 1.25x magnification 158 Appendix F- Point-Counting under the Petrographic Microscope 159 160 161 162 Appendix G- JMicroVision Data Figure G-1: Point counting (500 valid counts). Blue dots represent epoxy, yellow dots represent grain/matrix, and grey dots represent invalid or undetermined region. Figure G-2: Point counting (500 valid counts). Blue dots represent epoxy, yellow dots represent grain/matrix, and grey dots represent invalid or undetermined region. 163 Figure G-3: Point counting (500 valid counts). Blue dots represent epoxy, yellow dots represent grain/matrix, and grey dots represent invalid or undetermined region. Figure G-4: Point counting (500 valid counts). Blue dots represent epoxy, yellow dots represent grain/matrix, and grey dots represent invalid or undetermined region. 164 Figure G-5: Point counting (500 valid counts). Blue dots represent epoxy, yellow dots represent grain/matrix, and grey dots represent invalid or undetermined region. Figure G-6: Point counting (500 valid counts). Blue dots represent epoxy, yellow dots represent grain/matrix, and grey dots represent invalid or undetermined region. 165 Figure G-7: Point counting (500 valid counts). Blue dots represent epoxy, yellow dots represent grain/matrix, and grey dots represent invalid or undetermined region. Figure G-8: Point counting (500 valid counts). Blue dots represent epoxy, yellow dots represent grain/matrix, and grey dots represent invalid or undetermined region. 166 Appendix H- Porosity- Density Data Table H-1: Porosity data derived from bulk density and volume of samples. Porosity values are organized with respect to their location within the fault zone architecture. Porosity Sample B3 MF1 MF2 MF4 MF6 MF7 P4 RCB1 B2 P1 P2 P3 MF9 I1 I2 B1 P5 Sample Location Fault Core Fault Core Fault Core Fault Core Fault Core Fault Core Fault Core Fault Core Damage Zone Damage Zone Damage Zone Damage Zone Damage Zone Protolith Protolith Protolith Protolith Trail-1 Trail-2 Trail-3 Average 10.12578616 29.71698113 30.90754717 23.58343816 0.824528302 21.49622642 30.54339623 17.62138365 41.21509434 31.10377358 14.95283019 29.09056604 13.88679245 19.50188679 23.46540881 18.95136268 60.49528302 -9.401257862 34.88867925 47.69198113 50.96037736 6.883018868 33.72528302 30.52289308 89.18238994 66.27358491 3.018867925 52.82494759 20.75471698 37.98742138 28.86792453 29.2033543 -8.296855346 16.73962264 20.30691824 18.52327044 17.84528302 -3.091320755 31.76754717 24.80641509 81.41509434 86.36477987 38.86792453 68.88259958 38.11320755 -14.71698113 27.54716981 32.83018868 11.29716981 8.490566038 12.64150943 10.80974843 27.16981132 32.83018868 74.33962264 44.77987421 8.113207547 46.91823899 -29.05660377 27.51572327 0.724528302 15.8327044 11.4745283 9.343920335 5.828092243 1.635220126 24.71698113 10.7267645 167 Standard Deviation 11.66986117 15.23363327 13.24637245 4.812980407 18.10660317 22.21249331 44.62832638 8.621247596 2.522458907 9.844527391 26.11102039 7.471316933 2.117963048 25.7554657 27.43930088 7.776172838 12.29592067 Sample: RCB1 A B C Average RCB1-3 RCB1-2 RCB1-1 Mass of Mass of Sample* Sample Container and Mass* (g) container (g) (g) D E F G H I J Mass of Water Volume of Amount Coffee Sample Level Mass of Original of DI Filter** Mass** and with Container Sample Water and container Sample* (g) 3 (mL) Container and filter (g) (mL=cm ) (mL) (g) Oven Dried Bulk Density Sample (g/cm3 ) Mass (g) 3.45 10.9 7.45 21 24.5 3.5 3.74 5.03 12.38 7.35 3.45 8.47 5.02 10 13 3 3.74 5.03 9.96 4.93 3.45 7.31 3.86 18 20 2 3.74 5.03 8.8 3.77 3.74 5.03 10.38 5.35 3.45 8.89333 5.44333 16.3333 19.1667 2.833333 K 2.1 Porosity: n (%) 20.755 1.64333333 37.987 1.885 28.868 1.87611111 29.203 Standard Deviation: Particle density = 2.65 g/cm3 * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours L 8.6212 n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water 168 Sample: RCB-1 100 90 80 Porosity (%) 70 60 50 40 37.99 30 28.87 20 29.20 20.75 10 0 RCB1-1 RCB1-2 169 RCB1-3 Average Sample: I1 E F G H I J K Mass of Sample Mass of Amount Water Volume of Coffee Mass** Oven Bulk Mass of Sample* Sample Mass of of DI Level with Original Filter** and Dried Density Container and Mass* Container Water Sample* Sample and container Sample (g) container (g) (g) (g/cm3 ) 3 (mL) (mL) Container and filter Mass (g) (mL=cm ) (g) (g) (g) L n (%) I1-1 D 3.45 5.41 1.96 11 12 1 3.74 5.03 6.96 1.93 1.93 27.1698 I1-2 C 3.45 7.04 3.59 10 12 2 3.74 5.03 8.59 3.56 1.78 32.8302 I1-3 B 3.45 8.22 4.77 12 19 7 3.74 5.03 9.79 4.76 0.68 74.3396 Average A 3.45 6.89 3.44 11 3.74 5.03 14.33333 3.333333 8.44667 3.41667 1.4633 44.7799 Standard Deviation: 3 Particle density = 2.65 g/cm * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours Porosity: 170 25.7555 n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water Sample: I1 100 90 80 Porosity (%) 70 74.34 60 50 44.78 40 30 20 32.83 27.17 10 0 I1-1 I1-2 I1-3 171 Average Sample: I2 H I J K Mass of Sample Mass of Volume of Amount Water Coffee Mass** Oven Bulk Mass of Sample* Sample Mass of of DI Level with Original Filter** and Dried Density Container and Mass* Container Sample Water Sample* and container Sample (g) container (g) (g) (g/cm3 ) 3 (mL) (mL) Container and filter Mass (g) (mL=cm ) (g) (g) (g) Average I2-3 a I2-2 I2-1 A B C D E F G n (%) 3.45 8.36 4.91 16 18 2 3.74 5.03 9.9 4.87 2.435 3.45 7.65 4.2 12 15 3 3.74 5.03 9.25 4.22 1.4067 46.918 3.45 10.24 6.79 20 22 2 3.74 5.03 11.87 6.84 3.45 8.005 4.555 14 16.5 2.5 3.74 5.03 9.575 4.545 3.42 3 Particle density = 2.65 g/cm * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours Porosity: sample run was not included in average calculation 172 8.1132 -29.06 1.9208 27.516 Standard Deviation: a L 27.439 n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water Sample: I2 100 90 80 Porosity (%) 70 60 50 46.92 40 30 27.52 20 10 8.11 0 I2-1 I2-2 173 I2-3a Average Sample: P5 H Mass of Mass of Amount Water Volume of Coffee Mass of Sample* Sample Mass of Original of DI Level with Filter** Container and Mass* Container Sample Water Sample* and (g) container (g) (g) 3 (mL) (mL) Container (mL=cm ) (g) (g) Average P5-3 P5-2 P5-1 A B C D E F G I J K Sample Mass** Oven Bulk and Dried Density container Sample 3 (g/cm ) and filter Mass (g) (g) n (%) 3.45 14.75 11.3 22 26.5 4.5 3.74 5.03 16.26 11.23 2.49556 5.8281 3.45 11.28 7.83 27 30 3 3.74 5.03 12.85 7.82 2.60667 1.6352 3.45 7.37 3.92 16 18 2 3.74 5.03 9.02 3.99 3.74 5.03 12.71 7.68 3.45 11.1333 7.6833 21.6667 24.83333 3.166667 1.995 Particle density = 2.65 g/cm3 * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours Porosity: sample run was not included in average calculation 174 24.717 2.36574 10.727 Standard Deviation: a L 12.296 n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water Sample: P5 100 90 80 Porosity (%) 70 60 50 40 30 20 24.72 10 0 10.73 1.64 5.83 P5-1 P5-2 P5-3 175 Average Sample: P4 E F G H I J Mass of Sample Mass of Volume of Amount Water Coffee Mass** Oven Mass of Sample* Sample Mass of of DI Level with Original Filter** and Dried Container and Mass* Container Water Sample* Sample and container Sample (g) container (g) (g) 3 (mL) (mL) Container and filter Mass (g) (mL=cm ) (g) (g) (g) K L Bulk Density n (%) 3 (g/cm ) P4-1 D 3.45 5.27 1.82 10 16 6 3.74 5.03 6.75 1.72 0.28667 89.182 P4-2 C 3.45 10.89 7.44 12 20 8 3.74 5.03 12.18 7.15 0.89375 66.274 P4-3 B 3.45 14.08 10.63 34 38 4 3.74 5.03 15.31 10.28 Average A 3.45 10.08 6.63 6 3.74 5.03 18.667 24.66667 3 Particle density = 2.65 g/cm * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours a 2.57 3.0189 11.41333 6.38333 1.25014 52.825 Standard Deviation: 44.628 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 176 Sample:P4 100 90 89.18 80 Porosity (%) 70 66.27 60 50 52.82 40 30 20 10 3.02 0 P4-1 P4-2 P4-3 177 Average Sample: P3 A B C H I J K Mass of Sample Amount Water Volume of Coffee Mass** Oven Bulk Mass of of DI Level with Original Filter** and Dried Density Container Sample Water Sample* and container Sample 3 (g) (mL) (mL) Container and filter Mass (g) (g/cm ) (mL=cm3 ) (g) (g) L n (%) P3-1 G 3.45 6.33 2.88 10 11.75 1.75 3.74 5.03 7.9 2.87 1.64 38.11 P3-2 a F 3.45 6.49 3.04 24 25 1 3.74 5.03 8.07 3.04 3.04 -14.72 P3-3 E 3.45 6.31 2.86 21 22.5 1.5 3.74 5.03 7.91 2.88 1.92 27.55 Average Mass of Mass of Sample* Sample Container and Mass* (g) container (g) (g) D 3.45 6.32 2.87 15.5 17.125 1.625 3.74 5.03 7.905 2.875 1.78 32.83 Standard Deviation: Particle density = 2.65 g/cm3 * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours a 7.471 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 178 Sample:P3 100 90 80 Porosity (%) 70 60 50 40 38.11 30 32.83 27.55 20 10 0 P3-1 P3-2a 179 P3-3 Average Sample: P2 H Mass of Mass of Amount Water Volume of Coffee Mass of Sample* Sample Mass of of DI Level with Original Filter** Container and Mass* Container Sample Water Sample* and (g) container (g) (g) 3 (mL) (mL) Container (mL=cm ) (g) (g) Average P2-3 P2-2 P2-1 A B C D E F G I Sample Mass** and container and filter (g) J K Oven Bulk Dried Density n (%) Sample (g/cm3 ) Mass (g) 3.45 7.53 4.08 21 29 8 3.74 5.03 8.97 3.94 0.4925 81.415 3.45 6.23 2.78 12 19.5 7.5 3.74 5.03 7.74 2.71 0.3613 86.365 3.45 8.41 4.96 30 33 3 3.74 5.03 9.89 4.86 3.45 7.39 3.94 21 3.74 5.03 8.866667 27.16667 6.166667 1.62 3 Particle density = 2.65 g/cm * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours 38.868 3.83667 0.8246 68.883 Standard Deviation: a L 26.111 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 180 Sample:P2 100 90 86.36 80 81.42 70 Porosity (%) 68.88 60 50 40 30 27.55 20 10 0 P2-1 P2-2 181 P2-3 Average Sample: P1 H I J K Mass of Sample Mass of Amount Water Volume of Coffee Mass** Oven Bulk Mass of Sample* Sample Mass of Original of DI Level with Filter** and Dried Density Container and Mass* Container Water Sample* Sample and container Sample (g) container (g) (g) (g/cm3 ) 3 (mL) (mL) Container and filter Mass (g) (mL=cm ) (g) (g) (g) Average P1-3 P1-2 a P1-1 A B C D E F G n (%) 3.45 7.76 4.31 18 20 2 3.74 5.03 9.3842 4.3542 2.1771 17.845 3.45 10.25 6.8 27 29.5 2.5 3.74 5.03 11.8598 6.8298 2.73192 -3.091 3.45 7.95 4.5 20 22.5 2.5 3.74 5.03 9.5504 4.5204 1.80816 31.768 3.45 7.855 4.405 19 21.25 2.25 3.74 5.03 9.4673 4.4373 1.99263 24.806 Standard Deviation: Particle density = 2.65 g/cm3 * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours a L 9.8445 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 182 Sample:P1 100 90 80 Porosity (%) 70 60 50 40 30 31.77 24.81 20 10 17.85 0 P1-1 P1-2a 183 P1-3 Average Sample: MF9 A B C Average MF9-3 MF9-2 MF9-1 Mass of Mass of Sample* Sample Container and Mass* (g) container (g) (g) D E G H Mass of Water Volume of Amount Coffee Level Mass of Original of DI Filter** with Container Sample Water and Sample* (g) 3 (mL) Container (mL) (mL=cm ) (g) I J Sample Mass** Oven and Dried container Sample and filter Mass (g) (g) K L Bulk Density n (%) 3 (g/cm ) 3.73 41.63 37.9 39 55 16 3.59 5.03 42.64 37.61 3.73 28.22 24.49 36 46 10 3.59 5.03 29.28 24.25 2.425 8.49057 3.73 22.57 18.84 34 42 8 3.59 5.03 23.55 18.52 2.315 12.6415 3.59 5.03 3.73 30.8067 27.0767 36.3333 47.6667 11.3333 Particle density = 2.65 g/cm3 * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours a F 2.35063 11.2972 31.8233 26.7933 2.36354 10.8097 Standard Deviation: 2.11796 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 184 Sample:MF9 100 90 80 Porosity (%) 70 60 50 40 30 20 10 11.30 12.64 8.49 10.81 0 MF9-1 MF9-2 185 MF9-3 Average Sample: MF7 H I J K Mass of Sample Mass of Water Volume of Amount Coffee Mass** Oven Bulk Mass of Sample* Sample Level Mass of Original of DI Filter** and Dried Density Container and Mass* with Container Sample Water and container Sample (g) container (g) Sample* (g) (g/cm3 ) 3 (mL) Container and filter Mass (g) (g) (mL) (mL=cm ) (g) (g) Average MF7-3 MF7-2 MF7-1 A B C D E F G n (%) 3.45 8.72 5.27 19 23 4 3.74 5.03 10.2282 5.1982 1.29955 50.96 3.45 5.9 2.45 14 15 1 3.74 5.03 7.4976 2.4676 3.45 12.5 9.05 20 25 5 3.74 5.03 13.8114 8.7814 1.75628 33.725 3.45 9.04 5.59 17.667 21 3.333333 3.74 5.03 10.5124 5.4824 1.84114 30.523 2.4676 Standard Deviation: Particle density = 2.65 g/cm3 * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours a L 6.883 22.212 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 186 Sample:MF7 100 90 80 Porosity (%) 70 60 50 50.96 40 33.73 30 30.52 20 10 6.88 0 MF7-1 MF7-2 187 MF7-3 Average Sample: MF6 A B C D E F G H I J K Average MF6-3 MF6-2 a MF6-1 Mass of Sample Mass of Amount Water Volume of Coffee Mass** Oven Bulk Mass of Sample* Sample Mass of of DI Level with Original Filter** and Dried Density Container and Mass* Container Water Sample* Sample and container Sample (g) container (g) (g) (g/cm3 ) (mL) (mL) (mL=cm3 ) Container and filter Mass (g) (g) (g) (g) n (%) 3.45 11.84 8.39 18 26 8 3.74 5.03 13.405 8.375 1.0469 60.495 3.45 7.81 4.36 19.5 21 1.5 3.74 5.03 9.3787 4.3487 2.8991 -9.401 3.45 6.95 3.5 15 17 2 3.74 5.03 8.4809 3.4509 1.7255 34.889 3.45 9.395 5.945 16.5 21.5 5 3.74 5.03 10.94295 5.91295 1.3862 47.692 Standard Deviation: Particle density = 2.65 g/cm3 * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours a L 18.107 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 188 Sample:MF6 100 90 80 Porosity (%) 70 60 60.50 50 47.69 40 34.89 30 20 10 0 MF6-1 MF6-2a 189 MF6-3 Average Sample: MF4 H I J K Mass of Sample Mass of Amount Water Volume of Coffee Mass** Oven Bulk Mass of Sample* Sample Mass of Original of DI Level with Filter** and Dried Density Container and Mass* Container Water Sample* Sample and container Sample (g) container (g) (g) (g/cm3 ) 3 (mL) (mL) Container and filter Mass (g) (mL=cm ) (g) (g) (g) Average MF4-3 MF4-2 MF4-1 A B C D E F G n (%) 3.45 5.8 2.35 10 11 1 3.74 5.03 7.312 2.282 2.282 3.45 7.79 4.34 13 15 2 3.74 5.03 9.2964 4.2664 2.1332 19.502 3.45 9.57 6.12 24 27 3 3.74 5.03 11.1145 6.0845 2.0282 23.465 3.45 7.72 4.27 15.667 17.6667 2 3.74 5.03 9.240967 4.21097 2.1478 18.951 Standard Deviation: 3 Particle density = 2.65 g/cm * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours a L 13.887 4.813 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 190 Sample:MF4 100 90 80 Porosity (%) 70 60 50 40 30 20 23.47 19.50 10 18.95 13.89 0 MF4-1 MF4-2 191 MF4-3 Average Sample: MF2 H I J K Mass of Sample Mass of Amount Water Volume of Coffee Mass** Oven Bulk Mass of Sample* Sample Mass of of DI Level with Original Filter** and Dried Density Container and Mass* Container Water Sample* Sample and container Sample (g) container (g) (g) (g/cm3 ) 3 (mL) (mL) (mL=cm ) Container and filter Mass (g) (g) (g) (g) Average MF2-3 MF2-2 MF2-1 A B C D E F G n (%) 3.45 6.6 3.15 15 17 2 3.74 5.03 8.1456 3.1156 1.5578 41.215 3.45 7.2 3.75 14 16 2 3.74 5.03 8.6815 3.6515 1.8258 31.104 3.45 12.49 9.04 43 47 4 3.74 5.03 14.045 9.015 2.2538 14.953 3.74 5.03 10.2907 5.2607 1.8791 29.091 3.45 8.76333 5.3133 24 26.6667 2.666667 Standard Deviation: Particle density = 2.65 g/cm3 * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours a L 13.246 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 192 Sample:MF2 100 90 80 Porosity (%) 70 60 50 40 41.22 30 31.10 29.09 20 14.95 10 0 MF2-1 MF2-2 193 MF2-3 Average Sample: MF1 H I J K Mass of Sample Mass of Amount Water Volume of Coffee Mass** Oven Bulk Mass of Sample* Sample Mass of of DI Level with Original Filter** and Dried Density Container and Mass* Container Water Sample* Sample and container Sample (g) container (g) (g) (g/cm3 ) 3 (mL) (mL) (mL=cm ) Container and filter Mass (g) (g) (g) (g) Average MF1-3 MF1-2 MF1-1 A B C D E F G n (%) 3.45 8.76 5.31 53 55 2 3.74 5.03 10.2863 5.2563 2.6282 0.8245 3.45 7.67 4.22 49 51 2 3.74 5.03 9.1907 4.1607 2.0804 21.496 3.45 6.14 2.69 38 39.5 1.5 3.74 5.03 7.7909 2.7609 1.8406 30.543 48.5 1.833333 3.74 5.03 9.0893 4.0593 2.183 3.45 7.52333 4.0733 46.667 Standard Deviation: 3 Particle density = 2.65 g/cm * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours a L 17.621 15.234 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 194 Sample:MF1 100 90 80 Porosity (%) 70 60 50 40 30 30.54 20 21.50 17.62 10 0.82 0 MF1-1 MF1-2 MF1-3 195 Average Sample: B3 H I J K Mass of Sample Mass of Amount Water Volume of Coffee Mass** Oven Bulk Mass of Sample* Sample Mass of of DI Level with Original Filter** and Dried Density Container and Mass* Container Water Sample* Sample and container Sample 3 (g) container (g) (g) (mL) (mL) (mL=cm3 ) Container and filter Mass (g) (g/cm ) (g) (g) (g) Average B3-3 B3-2 B3-1 A B C D E F G n (%) 3.45 11.17 7.72 57 60 3 3.74 5.03 12.175 7.145 2.3817 10.126 3.45 7.41 3.96 40 42 2 3.74 5.03 8.755 3.725 1.8625 29.717 3.45 7.39 3.94 55 57 2 3.74 5.03 8.6919 3.6619 1.831 30.908 53 2.333333 3.74 5.03 9.873967 4.84397 2.025 23.583 3.45 8.65667 5.2067 50.667 Standard Deviation: Particle density = 2.65 g/cm3 * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours a L 11.67 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 196 Sample:B3 100 90 80 Porosity (%) 70 60 50 40 30 30.91 29.72 20 23.58 10 10.13 0 B3-1 B3-2 197 B3-3 Average Sample: B2 A B C D E F G H I J K Average B2-3 B2-2 B2-1a Mass of Sample Mass of Amount Water Volume of Coffee Mass** Oven Bulk Mass of Sample* Sample Mass of of DI Level with Original Filter** and Dried Density Container and Mass* Container Water Sample* Sample and container Sample 3 (g) container (g) (g) (mL) (mL) (mL=cm3 ) Container and filter Mass (g) (g/cm ) (g) (g) (g) n (%) 3.45 7.49 4.04 53 54.5 1.5 3.74 5.03 9.3348 4.3048 2.8699 -8.297 3.45 5.78 2.33 53 54 1 3.74 5.03 7.2364 2.2064 2.2064 3.45 5.05 1.6 23 23.75 0.75 3.74 5.03 6.6139 1.5839 2.1119 20.307 3.45 5.415 1.965 38 38.875 0.875 3.74 5.03 6.92515 1.89515 2.1591 18.523 Standard Deviation: Particle density = 2.65 g/cm3 * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours a L 16.74 2.5225 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 198 Sample:B2 100 90 80 Porosity (%) 70 60 50 40 30 20 20.31 16.74 18.52 10 0 B2-1a B2-2 199 B2-3 Average Sample: B1 H I J K Mass of Sample Mass of Volume of Amount Water Coffee Mass** Oven Bulk Mass of Sample* Sample Mass of Original of DI Level with Filter** and Dried Density Container and Mass* Container Water Sample* Sample and container Sample 3 (g) container (g) (g) 3 (mL) (mL) (mL=cm ) Container and filter Mass (g) (g/cm ) (g) (g) (g) Average B1-3 B1-2 B1-1 A B C D E F G n (%) 3.45 11.49 8.04 61 64 3 3.74 5.03 12.9224 7.8924 2.6308 0.7245 3.45 10.19 6.74 63 66 3 3.74 5.03 11.7213 6.6913 2.2304 15.833 3.45 12.89 9.44 48 52 4 3.74 5.03 14.4137 9.3837 2.3459 11.475 3.74 5.03 13.01913 7.98913 2.4024 9.3439 3.45 11.5233 8.0733 57.333 60.6667 3.333333 Standard Deviation: Particle density = 2.65 g/cm3 * Before Oven Drying at 104C for 24 Hours ** After Oven Drying at 104C for 24 Hours a L 7.7762 Porosity: n=100 [1-(bulk density/particle density)] = porosity % Bulk density= Oven dried sample mass/original sample volume Volume of Sample= water with sample -original water sample run was not included in average calculation 200 Sample:B1 100 90 80 Porosity (%) 70 60 50 40 30 20 15.83 10 11.47 0.72 9.34 0 B1-1 B1-2 B1-3 201 Average Appendix I- Grain Size Distribution Charts 202 203 204 205 206 207 208 209 210 211 212 213 214 5.72" 215 216 217 218 219