Abstract

The White River altered area, Washington, and the Goldfield mining district, Nevada, are nearly contemporaneous Tertiary (ca. 20 Ma) calc-alkaline igneous centers with large exposures of shallow (<1 km depth) magmatic-hydrothermal, acid-sulfate alteration. Goldfield is the largest known high-sulfidation gold deposit in North America. At White River, silica is the only commodity exploited to date, but, based on its similarities with Goldfield, White River may have potential for concealed precious and/or base metal deposits at shallow depth. Both areas are products of the ancestral Cascade arc. Goldfield lies within the Great Basin physiographic province in an area of middle Miocene and younger Basin and Range and Walker Lane faulting, whereas White River is largely unaffected by young faults. However, west-northwest–striking magnetic anomalies at White River do correspond with mapped faults synchronous with magmatism, and other linear anomalies may reflect contemporaneous concealed faults. The White River altered area lies immediately south of the west-northwest–striking White River fault zone and north of a postulated fault with similar orientation. Structural data from the White River altered area indicate that alteration developed synchronously with an anomalous stress field conducive to left-lateral, strike-slip displacement on west-northwest–striking faults. Thus, the White River alteration may have developed in a transient transtensional region between the two strike-slip faults, analogous to models proposed for Goldfield and other mineral deposits in transverse deformational zones. Gravity and magnetic anomalies provide evidence for a pluton beneath the White River altered area that may have provided heat and fluids to overlying volcanic rocks. East– to east-northeast–striking extensional faults and/or fracture zones in the step-over region, also expressed in magnetic anomalies, may have tapped this intrusion and provided vertical and lateral transport of fluids to now silicified areas. By analogy to Goldfield, geophysical anomalies at the White River altered area may serve as proxies for geologic mapping in identifying faults, fractures, and intrusions relevant to hydrothermal alteration and ore formation in areas of poor exposure.

INTRODUCTION

Crustal faults play important roles in the formation of epithermal-style mineral deposits in arc settings (e.g., Henley and Hughes, 2000; Berger et al., 2003). Faults themselves may serve as conduits for hydrothermal fluids, and systems of faults may establish crustal stress regimes that promote fracturing and modify permeability in intervening regions, thus facilitating circulation of mineralizing fluids at shallow crustal levels. Knowing where faults are located and how they interact are important considerations in mineral exploration of arc and forearc regions.

The White River altered area, western Washington, and the Goldfield mining district, south-central Nevada (Fig. 1), provide good examples of tectonically controlled hydrothermal alteration. Both areas lie within continental-margin arc-related calc-alkaline volcanic rocks that erupted in late Oligocene to early Miocene time. Both display large exposures of intense hydrothermal alteration generated by condensation of magmatic vapor plumes. Both reflect shallow parts of large magmatic-hydrothermal systems, with present surface exposures representing paleodepths of 500 m or less. Both exhibit regional faults that appear related, at least spatially, to the evolution of their hydrothermal systems. Unlike Goldfield, White River is located outside the Great Basin in an area unaffected by Basin and Range block faulting and Walker Lane transtension (Fig. 1).

Gold was discovered at Goldfield in 1902, and peak annual production occurred in 1910. Total recorded production was 130 metric tons (4.2 Moz) of gold, 45 metric tons (1.5 Moz) of silver, and 3240 metric tons of copper (Ashley, 1990). Goldfield is the largest known high-sulfidation (quartz-alunite) gold deposit in North America (Ashley, 1990), whereas at White River, silica is the only commodity that has been exploited to date. This contrast in known economic value may only be apparent, however, owing to the dense vegetation, young surficial deposits, and deep weathering common in western Washington. Indeed, similarities to Goldfield's tectonic and magmatic evolution suggest that the White River altered area may have potential for concealed precious and/or base metal deposits at shallow depth.

Geophysical studies at White River and Goldfield may illuminate structural controls on magmatism and hydrothermal alteration. At Goldfield, for example, magnetic anomalies delineate crustal structures associated with acid-sulfate alteration and precious metal mineralization (Berger et al., 2005), and magnetic anomalies at the more poorly exposed White River altered area may have similar associations. Gravity and magnetic anomalies at both sites may reflect shallow plutons that provided heat and fluids to the hydrothermal system.

In this paper, we examine the geologic, structural, and geophysical setting of the White River altered area and compare this framework to the better-known Goldfield mining district. Specifically, we utilize new geophysical and structural data to investigate the role that crustal stress fields, transverse structures, and shallow plutons may have played in localizing hydrothermal fluid flow at both areas.

GEOLOGIC SETTING

White River

The Cascade arc in northern California, Oregon, Washington, and British Columbia is the long-term magmatic product of subduction of the oceanic Juan de Fuca, Explorer, and Gorda plates beneath North America (Atwater, 1970; Christiansen and Yeats, 1992). The Cascade arc is generally divided into two arc-parallel components on the basis of age and volcanic style. From Eocene to Pliocene time, calc-alkaline volcanic rocks erupted in a broad zone across much of western Oregon and Washington. These rocks are exposed today in the Western Cascades and are located mostly west of the Cascade crest. During Pliocene and Quaternary time, the arc narrowed markedly to form the High Cascades (Fig. 1), a north-south swath generally no wider than 75 km that extends 1100 km from Lassen Peak in northern California to Mount Garibaldi in British Columbia. Both the Western and High Cascade arcs are dominantly calc-alkaline in nature and characterized by voluminous mafic lava flows and scattered stratovolcanoes of mafic andesite to dacite compositions (Christiansen and Yeats, 1992). The volume of silicic volcanism is subordinate to mafic-intermediate volcanism, but quartz dioritic to granitic plutons are exposed in more deeply eroded parts of the Western Cascades of Washington and Oregon.

Hydrothermal alteration is widespread in both the Western and High Cascade arcs. In Tertiary volcanic rocks, most hydrothermal alteration is spatially associated with small intrusions and plutons, some of which formed porphyry copper and related deposits, including copper-rich breccia pipes, polymetallic veins, and epithermal gold-silver deposits (Fig. 1; John et al., 2003).

The White River altered area is located in the Western Cascades of western Washington, ∼13 km east of the city of Enumclaw and 35 km north of Mount Rainier (Fig. 1). The area is underlain by early Miocene Fifes Peak Formation, which consists mainly of porphyritic basaltic andesite, andesite, and dacite lava flows and breccias (Tabor et al., 2000; Fig. 2A). In the altered area, some exposures of the Fifes Peak Formation include interbedded silicic tuffs and tuff breccias, and McCulla (1986) interpreted these rocks to be outflow from a caldera located ∼10 km southeast of the altered area. Sparse silicic dikes and possibly a small dacite flow dome also are exposed in the altered area (John et al., 2003). New 40Ar-39Ar ages of unaltered lava flows suggest that the age of the Fifes Peak Formation is ca. 20.6–21.1 Ma within the altered area (101Tables 1A and 1021B103).

Hydrothermal Alteration and Mineralization

Hydrothermal alteration and mineralization in the White River area have been described in detail by McCulla (1986) and John et al. (2003). Hydrothermally altered rocks at White River are exposed over ∼25 km2 in two adjacent areas bifurcated by the White River and covered by extensive surficial deposits (Fig. 2B). Silicified rocks in the White River area were first explored for gold in the late nineteenth century (Huntting, 1955). Subsequently, alunite was discovered and prospected in the 1930s and 1940s, and areas of intense advanced argillic alteration and silicification have been explored recently for base and precious metals. The only commodity exploited to date is silica, which was first mined some time before 1936; silica currently is being mined from the Scatter Creek Quarry and was formerly mined from the Superior Quarry (Fig. 2; John et al., 2003).

Hydrothermal alteration at White River consists of numerous prominent outcrops of intensely leached, residual (vuggy) silica surrounded by broad zones of poorly exposed advanced argillic (alunite, pyrophyllite, dickite, kaolinite, topaz), argillic, and propylitic alteration (John et al., 2003). Although oxidation of sulfide minerals to hematite or iron oxyhydroxides commonly extends to depths of tens of meters below the surface (John et al., 2003), unoxidized alteration assemblages generally contain 5–20 vol% sulfide minerals, dominantly fine-grained pyrite. The vuggy silica alteration commonly is cut by fine-grained quartz and chalcedony veins, hydrothermal breccias, and massive silica flooding that infill the vuggy texture. This secondary silica locally contains enargite in addition to pyrite. East-west to northeast-southwest elongation of the silica bodies (Fig. 2B) suggests important structural controls on hydrothermal alteration, although these orientations do not correspond to any regionally significant faults mapped in the area.

Hydrothermal alteration at White River probably occurred through condensation of a sulfur-rich magmatic vapor plume in groundwater (e.g., Rye et al., 1992; Rye, 2005). The acidic condensate leached most major and minor elements from the host rock, leaving residual silica and forming zones of progressive alteration outward from the leached silica core. The magmatic-hydrothermal acid-sulfate alteration in the White River area formed at shallow depths (mostly ≤250 m), and exposures reflect the shallowest parts of the hydrothermal system (John et al., 2003). The 40Ar-39Ar ages of alunite from the advanced argillic alteration range from 19.0 to 20.3 Ma (recalculated from McCulla, 1986; 101Tables 1A and 1021B103), and it is likely that the entire area of advanced argillic alteration formed from the same large hydrothermal system (John et al., 2003).

Faults and Structural Considerations

The White River altered area lies amid several regionally important faults (Fig. 2). The White River fault (Frizzell et al., 1984; Tabor et al., 2000) extends west-northwestward, from east of the Cascade Range to the Puget Lowland, and it passes ∼2–3 km north of the White River altered area, where it juxtaposes the early Miocene Fifes Peak Formation against the Oligocene Ohanapecosh Formation (Tabor et al., 2000). The Ohanapecosh Formation consists of andesite to dacite breccia, basalt and andesite flows, and volcaniclastic sedimentary rocks. Tabor et al. (2000) interpreted the White River fault as a subvertical to south-dipping normal fault, but new field data suggest a steeply dipping, south-verging reverse fault (Box et al., 2003; S.E. Box, written commun., 2006). The westward extension of the White River fault is on strike with the Tacoma fault, an active south-verging reverse fault with important earthquake hazard implications for densely populated regions of the Puget Sound (Brocher et al., 2004; Johnson et al., 2004).

Reidel and Campbell (1989) considered the White River fault to be the northwestern portion of a much larger fault zone, the White River–Naches River fault zone, which extends 90 km from Enumclaw to Naches, Washington (Fig. 1). The White River–Naches River fault zone extends eastward from the Cascade Range into the Yakima fold belt, where a series of reverse-fault–related folds affect the middle Miocene Columbia River basalt. These folds postdate 15 Ma, and thus faulting is younger than alteration in the White River altered area, although the younger offset may have been guided by a pre-existing structure. In the Naches River area, the White River–Naches River fault zone separates two terranes: to the northeast, faults and folds in pre-Tertiary to Pliocene rocks trend west-northwest, whereas to the southwest, structures in pre-Tertiary rocks trend more northerly (Reidel and Campbell, 1989). A similar change in trend is seen across the White River fault within our study area (Fig. 2A): north of the White River fault, west-northwest–trending fold axes deform Oligocene to Miocene rocks, whereas south of the White River fault, folds strike north-north-westward (Tabor et al., 2000).

A second fault within our study area is mapped along the Clearwater River valley and projects north-northwestward directly beneath the White River altered zone (Frizzell et al., 1984; McCulla, 1986; Tabor et al., 2000; Fig. 2). The Clearwater River fault is mapped as a west-dipping normal fault, although offsets south of the altered area have apparent down-to-the-east displacement (Tabor et al., 2000). This apparently contrary displacement may have resulted from scissor motion or right-lateral offset on the fault. Detailed mapping of the White River altered area (Fig. 2B) suggests that some alteration was guided and elongated locally along the approximate projection of the Clearwater River fault. Silicification and advanced argillic alteration extend north along the projection of this fault (Fig. 8 of John et al., 2003; Fig. 2B). Silicic alteration is not significantly offset by the Clearwater River fault, however, which suggests that any major movement on the fault predates alteration.

The White River altered area apparently formed in a stress field contrary to the long-term stress regime of western Washington. Long-term stress in western Washington, from Tertiary to modern time, is characterized by north-south to north-northeast–south-southwest compression and east-west to east-southeast–west-northwest extension (Wellsetal., 1998; Zoback, 1992). Several narrow pre- and synalteration dikes exposed in the Superior Quarry and elsewhere (Fig. 2B; Figures 19 and 25 of John et al., 2003) are oriented approximately north-northeast, suggesting that the White River altered area was undergoing east-southeast–west-northwest extension during late stages of magmatism. During later silica deposition, however, ca. 20–19 Ma, fault movement occurred under approximately north-south extension and east-west compression, as evidenced by orientation of striations on fault surfaces and orientation of quartz-pyrite veinlets, both cutting silicified rocks (Box et al., 2003; S.E. Box, written commun., 2006). It is important to note that deformation on the west-northwest–striking White River fault favors left-lateral strike-slip displacement during this time of anomalous east-west compression.

Thus, the White River altered area formed during a time when local stress was rotated roughly 90° from the long-term regional stress regime. The time of onset of this anomalous stress field and its duration are unknown, but it apparently ended shortly after silicification at White River. Gold-bearing hydrothermal clay-filled extensional fractures with north to north-northeast strike and near vertical dip now cut the silicified rocks (Box et al., 2003; S.E. Box, written commun., 2006), suggesting that the stress field at the White River altered area returned to east-west to east-southeast–west-northwest extension during the waning stages of the hydrothermal system.

Goldfield

The Goldfield mining district is located in calc-alkaline volcanic rocks of the Basin and Range Province of southwestern Nevada (Fig. 1). At the time these rocks were emplaced, the Pacific–Farallon–North American triple junction was still south of southwestern Nevada, the Juan de Fuca plate was subducting beneath North America at this latitude, and Goldfield was part of an ancestral Cascade magmatic arc much larger than the present-day active arc (Christiansen and Yeats, 1992).

Host rocks at Goldfield are primarily Oligocene and early Miocene in age and include porphyritic trachyandesite, rhyodacite, quartz latite, and rhyolite (Ashley, 1990; Ashley and Keith, 1976; Fig. 3A). According to Ashley (1990), two episodes of volcanism were important in the Goldfield area: a caldera, ∼6 km in diameter, developed in late Oligocene time along a ring-fracture system formed by eruption of silicic flows and ash-flow tuffs and emplacement of a quartz latite dome. A second phase of volcanism, initiated in the early Miocene, extruded trachyandesite and rhyodacite flows, tuffs, and breccias.

Although hydrothermal systems in both the Goldfield mining district and White River altered area formed in the ancestral Cascade arc at similar times, subsequent northward passage of the triple junction has now placed Goldfield in the Walker Lane belt, a broad region of dextral shear and transtension along the west side of the Great Basin (Fig. 1; Atwater, 1970; Christiansen and Yeats, 1992; Oldow et al., 2001). The Walker Lane belt has accommodated part of the strain between the Pacific and North American plates since passage of the triple junction, but that tectonic overprint is not obvious at Goldfield, probably because the Goldfield area has not undergone large-magnitude extension typical of many other parts of the Great Basin and Walker Lane belt (Seedorff, 1991; John, 2001). In any case, strike-slip faulting related to the Walker Lane belt began after formation of the transform margin and after cessation of the subduction-related magmatism that formed Goldfield. Moreover, the Basin and Range normal faults that formed the Great Basin are younger than 17 Ma. Therefore, we infer that faults active during formation of the Goldfield deposits were related to strain in the ancestral Cascade arc.

Berger et al. (2003, 2005) recently emphasized the importance that transverse deformation zones may play in localizing mineralization at Goldfield and in other mining districts along subduction-zone margins. Permeable zones, caused by faults, fractures, and permeable beds, were important physical controls on ore formation at Goldfield because they provided channels for hydrothermal fluids (Ashley, 1990; Berger et al., 2005). Alteration at Goldfield is concentrated in a broad zone of northwest-striking strike-slip faults and north– to northeast-striking normal faults (Berger et al., 2005; Fig. 3A). Within this broad zone, a prominent, narrow, west-northwest–striking zone of faults and fractures corresponds to a discontinuous magnetic anomaly low (Fig. 3A, label D) and extends east-southeast from the main ore body for a distance of at least 8–10 km. Within the transverse zone, there is a series of en echelon, east-west–striking silica ridges, which were referred to as “ledges” in Ransome (1909). The silica ledges connect with northwest-striking hydrothermal chalced-ony veins, indicating hydraulic linkage between west-northwest–striking faults and fractures within the transverse zone during hydrothermal alteration (Berger et al., 2005). The transverse zone accommodated little displacement, but it apparently was rooted in the prevolcanic basement, influenced the local strain field, and thereby focused hydrothermal fluid flow and formation of the ore bodies (Berger et al., 2005).

Hydrothermal alteration at Goldfield covers an area of 40 km2, nearly twice the area exposed at White River. Hydrothermal alteration consists of ledges of residual vuggy silica surrounded by zones of advanced argillic, phyllic-argillic, argillic, and propylitic alteration; ore zones at Goldfield occur in areas of vuggy silica alteration and were formed at ca. 20.3–19.8 Ma (Vikre et al., 2005). The geologic setting of Goldfield is typical of high-sulfidation epithermal gold-silver deposits worldwide (White and Hedenquist, 1990; Sillitoe and Hedenquist, 2003), which are characterized by calc-alkaline volcanic centers containing prominent areas of magmatic-hydrothermal acid-sulfate alteration. The spatial association between faults and intensely altered areas (Fig. 3A), especially the west-northwest–striking fault zone, suggests that faults and fractures facilitated the flow of hydrothermal fluids and promoted the alteration.

GEOPHYSICAL INVESTIGATIONS

Volcanic rocks have distinctive physical properties (e.g., density, magnetic susceptibility, remanent magnetization, and resistivity) that often produce characteristic geophysical anomalies. Intense hydrothermal alteration of sufficient volume may superimpose other anomalies on these characteristic geophysical patterns. Induced and remanent magnetization, for example, vary dramatically from place to place in unaltered volcanic rocks at all scales, often producing a distinctive pattern of high and low magnetic anomalies. Magnetite is the primary magnetic mineral in volcanic rocks. A hydrothermal plume in such rocks may destroy magnetic minerals and deposit nonmagnetic phases, such as pyrite. Acid leaching resulting in advanced argillic alteration will remove all magnetic minerals from volcanic rocks, leaving behind a residual, nonmagnetic silica core. Zones of alteration of sufficient volumes and intensity should be detectable in magnetic fields measured at low altitudes.

Regional electrical resistivity is generally low for weathered and altered andesitic to rhyolitic volcanic rocks, as compared to high resistivities typical of buried intrusions (Klein and Bankey, 1992). Resistivity lows may be associated with sulfide concentrations, clay minerals (argillic alteration), and increased porosity, as may occur in fractured or brecciated zones. Resistivity highs may be associated with zones of silicification or with intrusions (Klein and Bankey, 1992).

In addition to detection of actual alteration zones, geophysical anomalies also help to delineate near-surface crustal structures and magmatic centers that may be associated with the alteration. Faulted volcanic terranes often produce linear magnetic anomalies (e.g., Blakely et al., 2000), and shallow intrusions are sometimes detectable in aeromagnetic and gravity data (e.g., Blakely, 1994).

Geophysical Data

A low-altitude aeromagnetic survey, suitable for regional-scale interpretations, is available for the White River altered area (Blakely et al., 1999). The survey, flown in 1997, covers the entire Puget Lowland and was conducted primarily for the purpose of assessing earthquake hazards in populated areas of the Puget Lowland. Flight altitude was nominally 250 m above terrain. Flight-line spacing was 400 m over the Puget Lowland but expanded to 800 m over outlying areas, which included the White River altered area (Fig. 4A). Sparse gravity data are also available from the White River altered area (Fig. 5A).

Figure 4B shows a filtered version of aero-magnetic anomalies from the White River study area. The original aeromagnetic data (Fig. 4A) were analytically continued upward a short distance (50 m), and then subtracted from the original data. This two-step procedure is equivalent to a discrete vertical derivative of Earth's magnetic field, a method useful for emphasizing magnetic anomalies caused by shallow crustal sources (Blakely, 1995).

We conducted several site-specific geophysical surveys at the White River altered area as part of the present investigation. (1) Three ground-magnetic transects (Fig. 6B) were established across the altered zones using a portable cesium-vapor magnetometer and global positioning system (GPS). A stationary proton-precession magnetometer in the immediate vicinity served as a base station to record and subsequently correct for time-varying magnetic fields. (2) Magnetic susceptibility measurements were conducted at 27 sites (Fig. 6A; 02Table 2) using a hand-held susceptibility meter. Most susceptibility sites consisted of nine individual measurements spaced over several meters of exposure. (3) Gravity measurements were established at 59 stations in the altered area and surrounding region (Fig. 5A). Appendix A, 03Table 3, and Figure 5A describe these new gravity data. Figure 5B shows the aeromagnetic anomalies of the White River study area transformed to pseudogravity anomalies. This mathematical transform converts magnetic anomalies into the mathematical equivalent of gravity anomalies, which is useful for the direct comparison of measured gravity anomalies with measured magnetic anomalies (Blakely, 1995).

In addition to these site surveys, Weyer-haeuser Company conducted several induced-polarization (IP) transects across the altered area (John et al., 2003).

A high-resolution aeromagnetic survey was conducted over the Goldfield mining district in 2002 (U.S. Geological Survey, 2002; Berger et al., 2005; Fig. 3B). Flight altitude was 250 m above terrain, and flight lines were spaced 250 m apart, providing three times the resolution of the White River aeromagnetic survey. Sparse gravity measurements are also available from the Goldfield area (Ponce, 1997).

To facilitate identification of subsurface faults and other features related to hydrothermal alteration, we have calculated the location of magnetization boundaries (white dots, Figs. 3 and 4) using a method described by Blakely and Simpson (1986). Two sets of magnetization boundaries are shown in both Figure 4A and Figure 4B: those calculated from the original aeromagnetic data (Fig. 4A) and those calculated from the filtered version of the aeromagnetic data (Fig. 4B).

White River Observations

Aeromagnetic anomalies in the Western Cascades near the White River altered area (Fig. 4) display high-amplitude, short-wavelength anomalies characteristic of Tertiary volcanic rocks (Blakely et al., 2000). Superimposed on this pattern are two prominent sets of magnetic lineaments, one set striking west-northwest and a second set striking north-northeast to east-northeast.

The west-northwest–striking anomalies are subparallel to several faults mapped in the area (Tabor et al., 2000). Most notable of these is the west-northwest–striking White River fault, which, within the study area, produces a linear magnetic anomaly along its entire mapped length (Fig. 4, label WR). Along most of the White River fault, specifically that portion east of longitude 121°47.5′W, magnetic anomaly values decrease from north to south across the fault. Assuming uniform magnetic properties and normal magnetization for rocks on each side of the fault, this decrease in anomaly values is consistent with the interpretation that the White River fault is either a south-side-down normal fault or south-verging reverse fault. West of longitude 121°47.5′W, the White River fault, as mapped by Tabor et al. (2000), follows a magnetic ridge, but we suggest that the White River fault actually bifurcates into two strands on opposite sides of this magnetic anomaly (Fig. 4, label WR). The anomaly (and the White River fault) separates two offset magnetic lows (Fig. 4, labels L1 and L2) discussed in the following. Several other west-northwest–striking gradients appear in the northeast part of the study area that correspond approximately to mapped faults (Tabor et al., 2000).

Other west-northwest–striking magnetic lineaments appear where faults have not been recognized by geologic mapping. In particular, a pronounced west-northwest gradient passes just south of the altered area (Fig. 4, label A) and forms the southern margin of the southern magnetic low (Fig. 4, label L2).

North-northeast– and east-northeast–striking magnetic lineaments are also evident in Figure 4. Although no faults with these trends are mapped in the study area, normal faults with this orientation would be compatible with the Tertiary stress regime of the Washington Cascades, with extensional directions ranging from east-west to east-southeast–west-northwest (Box et al., 2003). Indeed, small dikes with north-northeast strike are exposed in the study area (Fig. 2B).

The north-northwest–striking Clearwater River fault is expressed in aeromagnetic anomalies south of the altered area as a broad discontinuous magnetic low (Fig. 4, label CW). If continued northward to the White River fault, the Clearwater River fault would pass directly through the altered zone and bound the eastern margin of a prominent magnetic low (Fig. 4, label L2). Detailed geologic mapping (Fig. 2B) suggests that, although the Clearwater River fault may have guided some of the alteration, it has not subsequently offset the lateral extent of alteration. It appears, therefore, that the Clearwater River fault has not had significant offset this far north since the time of alteration, ca. 20 Ma.

As expected, magnetic susceptibility measurements (Fig. 6A) show a clear spatial correlation with the intensely altered zone. Altered rocks exposed at the surface are essentially nonmagnetic. Given this intense alteration and extreme reduction in magnetic susceptibility, we might expect to see a negative aeromagnetic anomaly corresponding to the White River altered area. Indeed, the main altered area lies entirely within a relatively low magnetic region and extends westward into magnetic low L2 (Fig. 4A). Also as expected, the long-wavelength aspects of the ground-magnetic profile, as it crosses anomalies L2 and B (Fig. 6B), generally agree with the aeromagnetic anomalies.

A small aeromagnetic and ground-magnetic high in the immediate vicinity of the altered area (labels B in Figs. 4, 5B, and 6) correlates with high resistivity, high chargeability (John et al., 2003), and a pronounced positive gravity anomaly (Fig. 5A, label B). The correspondence of these disparate geophysical data and the spatial association to intense exposed alteration suggest that these geophysical anomalies are caused by a shallow intrusion. Using a graphical approach applied to the ground-magnetic profile, we estimated that the source of this anomaly lies between 230 and 390 m depth (John et al., 2003). These depths are consistent with resistivity and chargeability estimates calculated from an induced-polarization survey (John et al., 2003).

The gravity anomaly (Fig. 5A, label B) and thus the pluton appear to be offset in a right-lateral sense by the north-northwest–trending Clearwater River fault, although this gravitational pattern may be influenced in part by the distribution of gravity stations. On the other hand, the lateral extent of alteration (Fig. 2B) shows no discernible evidence of lateral offset by the Clearwater River fault, suggesting that the fault has not had significant offset this far north since the time of alteration. An unmapped west-northwest–striking fault (Fig. 5A, label A), discussed in the following, may truncate the pluton along its southwestern margin.

Goldfield Observations

Aeromagnetic anomalies from the Goldfield mining district show clear correlations with mapped faults and fault zones. Several examples are noted in Figure 3 (labels A, B, C, and D). The dominant magnetic trend is northwest, and individual magnetic gradients with this trend correspond closely with mapped northwest-striking, right-lateral faults (Berger et al., 2005), notably the magnetic gradients that pass immediately south (Fig. 3B, label A) and north (Fig. 3B, label B) of the main zones of alteration. North- to northeast-trending magnetic gradients also correspond to mapped faults of similar trend, e.g., the gradient that follows the arcuate north- to northeast-striking fault interpreted as a ring fracture by Ashley (1990) (Fig. 3B, label C). Superimposed on these two dominant trends is a discontinuous west-northwest–striking magnetic trough (Fig. 3B, label D), which was interpreted by Berger et al. (2005) as a zone of deformation rooted in subvolcanic basement and oriented transverse to the dominant northwest-trending magnetic anomalies and faults.

A distinct set of arcuate faults is mapped in the northwestern part of the Goldfield district (Fig. 3A), which, together with an elliptical pattern in the distribution of argillized and silicified rocks around the district as a whole (Ashley, 1990, their Fig. H3), led Albers and Cornwall (1968), Albers and Kleinhampl (1970), and Ashley (1974) to hypothesize that the underlying structural control on the elliptical patterns was a caldera or ring-fracture zone with little associated caldera collapse (Ashley, 1974). The western and northwestern edges of a prominent arcuate magnetic anomaly (Fig. 3B, label C) closely follow the arcuate part of a major north-northeast–striking fault zone (Fig. 3A, label C). We interpret the broad positive gravity and magnetic anomalies across the district as a whole (Fig. 3B, label E) as evidence for a shallow pluton associated with the transverse deformation zones, hydrothermal alteration, and ore mineralization.

DISCUSSION

The dramatic destruction of magnetic minerals at the White River altered area, so evident in magnetic susceptibility measurements (Fig. 6A) and geologic mapping (Fig. 2), should be reflected in aeromagnetic or ground-magnetic data measured across the altered area. In particular, we should be able to use the wavelengths of magnetic anomalies to estimate the depth to magnetic sources, which in turn is related to the depth extent of alteration. As discussed earlier, we estimated the depth to magnetic sources beneath the White River altered area to be ∼230–390 m. Assuming the lateral extent of alteration exposed at the surface, ∼25 km2, is representative of subsurface rocks, the volume of intensely altered rock is no greater than ∼10 km3. This estimate is appropriate for degrees of alteration sufficiently intense to destroy essentially all magnetite; levels of alteration of lesser intensity may extend to larger volumes.

Aeromagnetic anomalies also constrain the depth extent of intense alteration in the Goldfield mining district. We see little correlation between magnetic lows and mapped zones of intense alteration, suggesting that intense alteration is confined to shallow depths. One exception may be the west-northwest–striking zone of magnetic lows (Fig. 3B, label D) corresponding to a narrow zone of faults and fractures. Berger et al. (2005) interpreted this magnetic anomaly to reflect a transverse deformation zone rooted in subvolcanic basement. Argillic alteration is well developed along this zone (Fig. 3A), suggesting that destruction of magnetic minerals at shallow depths may contribute to the magnetic lows.

Aeromagnetic anomalies in the Goldfield mining district (Berger et al., 2005; Fig. 3B) are associated with faults and fractures intimately tied to the circulation of hydrothermal fluids, and we infer that similar associations are present at White River. The west-northwest–striking White River fault, which passes a few kilometers north of the altered area, is clearly expressed in aeromagnetic anomalies (Fig. 4, label WR). The Tertiary to present-day regional stress field in western Washington is characterized by north-south to north-northeast–south-southwest compression and east-west to east-southeast–west-northwest extension, compatible with other observations that describe the White River fault as a south-verging reverse fault (Box et al., 2003; S.E. Box, written commun., 2006). Detailed field observations within the White River altered area (Box et al., 2003; S.E. Box, written commun., 2006), however, indicate that the long-term stress field was interrupted during a period of time that included formation of the White River altered area. During this time, the anomalous stress field was rotated ∼90°, to east-west compression and north-south extension. Thus, during development of the White River alteration, the White River fault zone was in a stress field conducive to left-lateral displacement.

Other magnetic anomalies with similar west-northwest strike are evident in Figure 4, some of which correspond to mapped faults. Of particular interest is the west-northwest–striking anomaly immediately south of the White River altered area (Fig. 4, label A). The anomaly extends for 15 km and lies subparallel to the White River fault zone north of the altered area. Given its linear nature and orientation parallel to the White River fault, we interpret this anomaly as the result of a concealed fault, subsequently referred to here as fault A. During the time of alteration, ca. 19.0–20.3 Ma, both the White River fault and fault A may have experienced left-lateral displacement, thereby establishing a transverse zone of extension in the overlap region (Fig. 7). We emphasize that fault A has not been identified in geologic mapping, even though the fault would cut exposed Miocene volcanic rocks (Fig. 4). Geologic mapping is hampered in this region, however, by dense vegetation and late Quaternary surficial deposits.

Several magnetic gradients with east to east-northeast strike are present in the step-over region between the White River fault and fault A (Fig. 4, label D), and these anomalies may reflect transverse faults or fractures activated during the transient east-west compressional period (Fig. 7). Other field evidence suggests that extension was important in this region during alteration. Silicic dikes with north-northeast strike and shear zones with northeast strike have been mapped within the altered area (John et al., 2003; Fig. 2B), and the silicified zone itself (Fig. 2) has a general northeastward elongation. We suggest that some of the east– to east-northeast–striking magnetic anomalies through the White River altered area reflect extensional faults or fractures that focused the flow of hydrothermal fluids. One of the east-northeast–striking magnetic lineaments (Fig. 4, label D, passing through the altered area) lies along a fault hypothesized from geologic mapping (Fig. 2B).

The White River altered area extends into a pronounced magnetic low (Fig. 4, label L2), and it is tempting to interpret the entire magnetic low as an unexposed zone of hydrothermal alteration. Other explanations are possible. The low may be caused by a pull-apart basin that opened during the period of anomalous stress and now is filled with glacial deposits or other weakly magnetic deposits. We think this interpretation is unlikely, however, because the magnetic low does not correspond to a gravity depression. Alternatively, the magnetic low may reflect reversely magnetized volcanic rocks erupted during a period of reverse polarity, at least three of which occurred ca. 21–19 Ma (Cande and Kent, 1995).

Geophysical evidence also supports the existence of shallow intrusions at both Goldfield and White River. It has been suggested that a caldera complex formed in late Oligocene time in the Goldfield area, evidenced today by arcuate fault patterns and syn- and postcaldera domes (Ashley, 1974). The presence of broad positive gravity and magnetic anomalies (Fig. 3, label C) nestled in the arcuate faults, however, indicates that the low-density, weakly magnetic volcanic deposits usually associated with calderas are missing or are very thin. We interpret the positive gravity and magnetic anomalies at Goldfield as indications of a near-surface pluton intruded ca. 20 Ma. The intrusion may have provided heat and magmatic volatiles and mobilized fluids that were transported through shallow faults to ore-bearing areas. The arcuate fault and associated magnetic anomaly (Fig. 3, label C) may be a ring fracture associated with this intrusion.

At the White River altered area, a pronounced positive gravity anomaly (Fig. 5A, label B) partially coincides with a subdued positive magnetic anomaly (Fig. 4, label B; Fig. 5B, label B). These anomalies lie over areas of intense alteration and between the White River fault zone and fault A. In analogy with Goldfield, we suggest that the positive gravity and magnetic anomalies reflect a small pluton or stock intruded into the step-over region established between the two west-northwest–striking, sinistral strike-slip faults during the time of anomalous stress regime, and thus the pluton was emplaced synchronously with hydrothermal alteration around 19.0–20.3 Ma. The magnetic anomaly may be caused by magnetite within the pluton itself, or by hydrothermal magnetite associated with that intrusion. Assuming the lateral extent of the pluton is defined by the gravity anomaly, only part of the pluton is significantly magnetic, as has been observed at other plutons and batholiths (e.g., Langenheim et al., 2004). The magnetic part of the pluton may define an early stage of the intrusion event. Fault A, postulated from aeromagnetic anomalies (Fig. 4, label A), lies along the southwestern margin of the gravity anomaly, suggesting that fault A influenced emplacement of the pluton. The abrupt southwestern margin of the gravity anomaly also may be controlled in part by the uneven distribution of gravity stations (Fig. 5A).

One or more of the east– to east-northeast–striking faults postulated from aeromagnetic anomalies (Fig. 4, label D) may have intersected this intrusion and provided fluid pathways to now silicified areas. Indeed, one of the east-northeast–striking faults interpreted from magnetic data (Fig. 4, label D; Fig. 7) appears to segment the gravity anomaly, suggesting that the fault influenced emplacement of the pluton or deformed the pluton after intrusion. This hypothesized fault runs through the center of the northeast-elongated zone of alteration, passes near the two main silica quarries, and coincides with a mapped fault (Fig. 7; Weyerhaeuser Company, written commun., 2004).

The fault in the Clearwater River valley (Tabor et al., 2000) is expressed in both magnetic and gravity anomalies (Fig. 4, label CW; Fig. 5). Although the distribution of gravity stations is uneven, the gravity anomaly that we interpret to be caused by a shallow intrusion appears to be offset several kilometers along the Clearwater River fault in a right-lateral sense, suggesting that the fault was active subsequent to alteration at White River. On the other hand, the mapped extent of the altered zone itself shows no evidence of this fault activity (Fig. 2B), and the aeromagnetic anomalies offer no compelling evidence to extend the Clearwater River fault all the way through the altered zone. We postulate that the Clearwater River fault terminates within the pluton, as shown in Figure 7, where it transfers strain to fault A. We note that the magnetic portion of the pluton abuts the northernmost part of the Clearwater River fault. The apparent offset of the gravity anomaly and the truncation of the magnetic part of the pluton may reflect structural influence on emplacement of the intrusion rather than subsequent offset.

What might have caused the anomalous stress regime at the White River altered area ca. 20–19 Ma? The stress perturbation observed at White River may reflect changes in the far-field stress regime, possibly introduced by slight adjustments in the partitioning of strain between the subducting Juan de Fuca plate, North America craton, and clockwise-rotating, northward-migrating Cascadia forearc (Wells et al., 1998). If the stress perturbation observed at the White River altered area was regional in scope, similar temporal changes should be manifested elsewhere in western Washington ca. 19–20 Ma.

Regional folds with north to north-northwest strike are mapped both north and south of Mount Rainier and extend northward nearly to the White River fault zone (e.g., Walsh et al., 1987). One of these northerly trending anticlines extends to within ∼20 km of the White River altered area (Tabor et al., 2000; Fig. 2A). The northerly strike of these folds is indicative of generally east-west compression, and thus the age of folding bears on the formation of the White River altered area. The Kidd Creek intrusive suite, which is 50 km south of Mount Rainier, may provide this critical timing. Paleomagnetic studies and ages determined with 40Ar/39Ar and zircon fission-track methods indicate that the Kidd Creek intrusive suite was emplaced at ca. 12.7 Ma and has not been subsequently folded or tilted (Hagstrum et al., 1998). Tilted wall rocks around the intrusive rocks, however, have a zircon fission-track age of ca. 18.7 ± 1.9 Ma. Thus, regional folding, consistent with regional east-west compression, apparently occurred between ca. 21 and 13 Ma (Hagstrum et al., 1998), permissibly overlapping the age of White River alteration.

In addition, a west-northwest–striking and esitic dike swarmm was emplaced northeast of Mount St. Helens following intrusion of the Spirit Lake pluton at 24–21 Ma but prior to intrusion of the Kidd Creek pluton at 12.7 Ma, and each of these plutons records normal west-northwest–east-southeast extension (Swanson, 1989; Evarts and Swanson, 1994; Box et al., 2003). The temporal overlap of emplacement of the Mount St. Helens dike swarm and the White River alteration also suggests that the transient stress regime at White River was regional in scope.

It is also possible that the anomalous stress regime at the White River altered area was local in scope, caused by a local event that perturbed the regional stress regime. Intrusion of the shallow pluton beneath the altered area, hypothesized from gravity and magnetic anomalies, may have been such an event, as observed elsewhere (e.g., Tosdal and Richards, 2001).

CONCLUSIONS

Both the Goldfield mining district and the White River altered area are products of Tertiary Cascade arc magmatism and crustal deformation transverse to the subduction margin. By analogy with Goldfield, magnetic anomalies at the White River altered area serve as proxies for geologic mapping in the identification of transverse faults associated with hydrothermal alteration. The White River altered area lies midway between two such faults, both of which have west-northwest strike. Structural data suggest that alteration developed in an anomalous stress field that established a transient zone of extension between the two faults. Gravity and magnetic anomalies indicate a small pluton or stock at shallow depth beneath the White River alteration that may have provided heat and fluids and focused mineralization.

APPENDIX A

A total of 59 gravity stations was acquired throughout the White River altered area and surrounding regions during the summer of 2005 using a LaCoste Romberg gravity meter and differentially corrected GPS navigation. Most stations were acquired along logging roads on land owned and maintained by Weyerhaeuser Company. Elevations were determined via differentially corrected GPS navigation, with estimated accuracy of 5 m. Observed gravity values were adjusted for latitude, Earth tides, elevation, and terrain effects using procedures established by the U.S. Geological Survey. These corrections provided free-air and complete Bouguer anomaly values 03(Table 3). Bouguer anomaly values were additionally corrected for masses that isostatically support topography (Simpson et al., 1986). Although the gravity meter was accurate to within 0.01 mGal, estimates of elevation and terrain effects limit anomaly accuracy to ∼0.5 mGal.

We thank Weyerhaeuser Company for access to the White River altered area and surrounding properties and for allowing us to view proprietary geophysical data. We are grateful to Allegra Hosford-Scheirer, Ray Wells, Al Hofstra, and two anonymous reviewers for comments that greatly improved our manuscript.