Geodynamics and Consequences of Lithospheric Removal in the , themed issue

Foundering lithosphere imaged with magnetotelluric data beneath , California

Linda Ostos and Stephen K. Park Department of Earth Sciences, University of California, Riverside, 2258 Geology, Riverside, California 92521, USA

ABSTRACT xenoliths that equilibrated at a depth of 70 km and magnetic fi elds, but period (the inverse and were found in Miocene volcanic fl ows of frequency) is conventionally used for data A magnetotelluric profi le consisting of dated as 11 Ma (Ducea and Saleeby, 1998). presentation and modeling (Vozoff, 1991). A 14 sites and spanning 240 km from the San Ducea and Saleeby (1998) inferred, from the complex, frequency-dependent, 2 × 2 imped- Joaquin Valley to the Nevada border was loss of garnet pyroxenites and the increase in ance tensor (Z) relates horizontal vector electric installed across the central Sierra Nevada, temperatures estimated from geothermometry at fi elds (Ex, Ey) to horizontal vector magnetic through Yosemite National Park, in an effort 3.5 Ma, that the lithosphere beneath the Sierra fi elds (Hx, Hy). Such a tensor has two princi- to constrain the northward extent of litho- Nevada batholith was replaced by astheno- pal axes that are usually orthogonal. Theoretical spheric removal beneath the range. Broad- sphere between 11 Ma and 3.5 Ma. analysis shows that if fi elds are measured over band and long-period instruments from the The removal mechanism responsible for two-dimensional (2-D) structures, the tensor EMSOC (Electromagnetic Studies of the Con- lithospheric foundering is poorly understood may be decomposed into two orthogonal, inde- tinents) consortium were used to record data and the fairly recent loss of lithosphere beneath pendent modes through rotation to a direction with periods ranging from 0.01 to 20,000 s, the southern Sierra Nevada could indicate either parallel or perpendicular to strike. The allowing the conductivity structure beneath that the phenomenon is still occurring elsewhere tranverse magnetic (TM) mode couples E per- the Sierra to be imaged to a depth of 120 km. in the range. Given that seismic studies (Fliedner pendicular to strike (Ey), H parallel to strike Two-dimensional models reveal that the et al., 1996) showed that the root thickened (Hx), and a vertical magnetic fi eld (Hz). This batholith’s resistive root extends to a depth to the north, we shifted our investigation of mode has a principal axis that is perpendicular of just 30 km beneath the eastern Sierra and the Sierra Nevada’s deep structure to a profi le to strike. The transverse electric (TE) mode cou- 45 km beneath the western Sierra. The batho- through the central Sierra Nevada, specifi cally, ples Ex, Hy, and Hz, and has a principal axis that lith is separated by a thin conductive zone through Yosemite National Park. is parallel to strike. Conventionally, tensors are that coincides with the Moho from a resistive rotated to a principal direction that maximizes mantle structure at a depth of ~55 km. We MT METHOD AND DATA the amplitudes of these two modes and mini- propose that this resistive structure is resid- mizes coupling terms between them. There is ual root. Deeper, a broad conductive feature A 240-km-long MT profi le extended from the still a 90° ambiguity in the rotation because Zxy dipping eastward at depths of 65–100 km San Joaquin Valley to the Nevada border, across (=Ex/Hy) could be larger than Zyx (=Ey/Hx), below the range is upwelling asthenosphere the Sierra Nevada through Yosemite National or vice versa. containing <1% melt that originates from the Park (Fig. 1); 14 stations were recorded with col- We found in this study that the directions of extensional tectonic regime of the Basin and located short-period (0.001–300 s) MT-24 and principal axes were generally aligned either par- Range to the east. long-period (30–10,000 s) NIMS (Narod Intel- allel or perpendicular to the geologic strike of ligent Magnetotelluric Systems) instruments. the major features in California (Figs. 1 and 2). INTRODUCTION The MT method relies on natural electro- Rose diagrams for tensor rotations at periods magnetic signals in the Earth (Vozoff, 1991). longer than 30 s (Fig. 2) show that one of the Magnetotelluric (MT) and seismic studies Horizontal electric fi elds induced by the Earth’s two principal directions at each site is approxi- of the southern Sierra Nevada (California and naturally occurring oscillatory magnetic fi eld mately parallel to the regional geologic strike. Nevada, USA; Fig. 1) revealed the absence are measured at the Earth’s surface with 100 m In some cases, this is the direction of Zxy and of a thick crustal root beneath the high Sierra long dipoles (essentially a digital voltmeter with in others it is Zyx. Each rose diagram shows (Wernicke et al., 1996). Fliedner et al. (1996), long wire leads). Horizontal and vertical mag- directions for several sites that are in the same using seismic refraction, mapped the Moho as netic fi elds are measured with induction coils geologic province (San Joaquin Valley, western varying between 30 and 35 km depth, consis- to record both source and secondary induced foothills metamorphic rocks, batholith, Basin tent with the fi ndings in Park et al. (1996), who horizontal fi elds. Data are acquired at two sites and Range). The principal axes rotate gradu- imaged the conductive mantle asthenosphere simultaneously in order to use a process called ally from N30W in the San Joaquin Valley and at depths as shallow as 35 km. Xenoliths from remote referencing to separate noise from western Sierra Nevada to due north in the cen- Pliocene volcanic fl ows dated as 3.5 Ma are signal. We used the Larsen et al (1996) robust tral Sierra and Basin and Range (Fig. 2). This predominantly peridotitic and marked by the processing technique to do this analysis. is similar to the regional variation seen in the absence of lower crustal eclogite (garnet pyrox- MT transfer functions are computed in the SSCD (Southern Sierra Continental Dynamics) enite). However, garnet pyroxenite is present in frequency domain between observed electric MT line that Park et al. (1996) attributed to the

Geosphere; February 2012; v. 8; no. 1; p. 98–104; doi:10.1130/GES00657.1; 5 fi gures.

98 For permission to copy, contact [email protected] © 2012 Geological Society of America

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/8/1/98/3342537/98.pdf by guest on 01 October 2021 Yosemite MT

Quaternary the TE mode because that is the mode used by sediments the rotation as the fi rst principal axis. The TM Tertiary sedimentary rocks mode was the Zyx component. Cenozoic The impedance and magnetic transfer func- volcanic rocks tions are complex, so they have both magnitude 40°N Subduction and phase. We convert the impedance magni- complex rocks tude into an apparent resistivity, which would Batholithic rocks be the true resistivity if the structure were homo- Prebatholithic metamorphic rocks geneous. Each impedance tensor results in two

125°W MESOZOIC CENOZOIC Precambrian-Paleozoic apparent resistivities and two phases because prebatholithic rocks ! ! !! it has two principal axes (Zxy and Zyx). These ! ! ! ! ! ! ! ! ! ! Precambrian-Cenozoic were plotted against the period in a sounding basement rocks (Mojave YNP Desert) curve, and the sounding curves from each site N were compiled to produce pseudosections of ! ! ! ! ! ! ! ! ! ! ! ! apparent resistivity and phase (Fig. 3). A pseudo- j SSCD Line section shows the distribution of apparent resis- tivities or phases versus spatial location and 35°N period. Because longer periods penetrate deeper into the Earth, the lower portions of pseudo- sections image deeper structure than do the upper portions (Fig. 3). 120°W While a pseudosection is a qualitative view of the resistivity structure with increasing depth toward the bottom, simple conversion to a true 115°W depth section is not possible because the depth of penetration of the electromagnetic wave depends Figure 1. Generalized geologic map of California showing locations on both the period and the average resistivity of of the SSCD (Southern Sierra Continental Dynamics Project) line the material it passes through. A period of 1 s at and this survey through Yosemite National Park (YNP). Geology is site 201 penetrates ~3 km while the same period adapted from Luddington et al. (2005). at site 205 penetrates over 70 km, e.g., penetra- tion depth ≈ √[(apparent resistivity) × (period)]. In addition, MT fi elds are sensitive to lateral change in structural orientation from N30W in While the most thorough analysis of MT data variations as well as vertical ones. In Park et al. California to due north in the Basin and Range. would involve 3-D inversion, studies usually (1991), it was shown that fi elds at periods longer In addition to the MT impedance tensor, we use 2-D inversion because fi eld behavior from than 20 s in the Sierra Nevada were affected by can compute a magnetic transfer function that 3-D bodies is often suffi ciently similar to that the electrical conductivity of the Pacifi c Ocean relates vertical and horizontal magnetic fi elds from 2-D structures (Wannamaker, 1999). In and the structure in between. A rise in apparent for a given frequency. The direction of great- this study, there are three reasons why 2-D mod- resistivity with increasing period may be due to est coherency between the vertical and hori- eling is appropriate. First, the principal axes of a more resistive layer or a conductor off to one zontal fi eld is represented by a set of induction the impedance tensors are aligned either paral- side. For these reasons, qualitative discussion of arrows for each site in the profi le (Fig. 2). An lel or perpendicular to overall geologic strike. pseudosections is appropriate, but inversion induction arrow of unit length means that the Second, induction arrows are small and ones at of data is required to produce a depth section of vertical induced magnetic fi eld has a strength the western sites are perpendicular to geologic electrical resistivity. comparable to that of the horizontal fi eld. Typi- strike. This means that the effect of conductors Both the TE and TM apparent resistivity sec- cally, induction arrows have lengths of <1. The north of the profi le will not dominate the MT tions have higher resistivities at periods <10 s induction arrows in our analysis follow the responses at the eastern sites. Last, the overall for stations across the Sierran batholith, a result convention that real components point away geologic structure is ~2-D in the vicinity of the consistent with the high resistivity of unfrac- from good conductors (Wiese, 1962). If the profi le. An added advantage of 2-D inversion is tured granitic rock (Fig. 3). Conductive sedi- structure of the Sierra Nevada were perfectly that it is fast, allowing many more models to be ments to the west in the San Joaquin Valley are 2-D, then induction arrows would be aligned run. As shown herein, running many models is responsible for the low resistivities at intermedi- perpendicular to geologic strike and real com- important to assessing the reliability of features ate periods (0.1–10 s), and the volcanic terrain ponents would point away from good conduc- in the preferred electrical cross section. in the western Basin and Range is responsible tors. Such behavior is seen at the western end One complication in using 2-D modeling is for intermediate resistivities in both the TE and of the profi le, where the effects of the conduc- that tensors and transfer functions need to be TM pseudosections (Fig. 3). Both sections show tive Pacifi c Ocean are seen at periods of 1000 s rotated to a common strike. We used N10W as a distinctive decrease in resistivity at periods and longer (Fig. 2). However, the real compo- an average between sites on the west and on the longer than 10 s, suggesting increased con- nents of the induction arrows in the Sierras and east of our profi le. Both impedance tensors and ductivity of the lower crust and upper mantle western Basin and Range are pointing south- magnetic transfer functions were rotated to this across the entire section. Complex structure is ward, suggesting electrical conductors to the common direction, and then modes had to be indicated in several portions of the profi le. For north (Fig. 2). identifi ed. The Zxy component was chosen as example, site 201 has multiple increases and

Geosphere, February 2012 99

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/8/1/98/3342537/98.pdf by guest on 01 October 2021 Ostos and Park

N N N N

Basin and Range

NV SR 49 Pmm CA YNP ML 201 38°00′N Mmm 205 216 SR120 212 203 M 209 208

San Joaquin Sierra B Valley SR 99 Nevada SNB 37°7.5′N 121°00′W 118°00′W 0 20 40 60 80 100 km Induction arrows Re at T=1000 s Im 0 1

Figure 2. Location map for magnetotelluric (MT) survey. In topographic map, white denotes elevations above 2500 m, dark gray symbol- izes elevations between 1500 and 2500 m, and light gray is for elevations <1500 m. The profi le extends from station 201 to station 216 (not all are numbered). CA-NV—the California-Nevada border; B—Bishop; M—Modesto; ML— (light blue); YNP—Yosemite National Park (outlined in yellow). SR49, SR99, and SR120 are state routes. Brown lines are geologic contacts from Figure 1. Geologic units labeled in the Sierra Nevada: Mmm—Mesozoic metamorphic rocks; Pmm—Paleozoic (mostly metamorphic) rocks; SNB—Sierra Nevada batholith. The real (red) and imaginary (blue) induction arrows at a period of 1000 s are shown at each station and scaled using the key shown below the map. Rose diagrams of principal axes of MT impedance tensors are shown across top. From left to right, the diagrams show directions of aggregates of sites 213–216, 209–212, 205–208, and 201–204. Colors in rose diagrams represent different stations, all for periods longer than 30 s. See text for discussion of MT parameters.

decreases of apparent resistivity (Fig. 3). While more than a factor of 1.5. The fi rst 18 km were sections in Figure 3 show the data included in it may be tempting to interpret this as alternat- set to 100 ohm-m followed by 10,000 ohm-m the inversion (8–12 points per decade of period). ing conductive and resistive layers, the values from 18 km to 40 km. A layer from 40 km to In addition, data types can be weighted differ- at periods longer than 3 s may be the result of 210 km had a resistivity of 100 ohm-m, and the ently. While we attempted initially to weight all sensing the resistive Sierran batholith and its half-space below that was set to 10 ohm-m. data equally, these converged with unacceptably underlying structure. The Pacifi c Ocean was included in the model high misfi ts. We fi nally downweighted the TE to the west of the profi le as a 0.3 ohm-m band mode apparent resistivity suffi ciently to achieve MT MODELING from 0 to 5km depth. Prior work in California acceptable models; this can be justifi ed because (Park et al., 1991; Mackie et al., 1996) showed TE mode apparent resistivities are particularly An initial model for the 2-D inversion is that electromagnetic induction in the conductive sensitive to truncation of 2-D bodies along needed, and our strategy was to pick one that water of the Pacifi c Ocean affects MT fi elds as strike. Previous work (e.g., Wannamaker, 1999) is layered so that any lateral contrasts seen in the far inland as the western Basin and Range, mak- has shown that use of the TM mode data, the output have been added by the inversion. A 1-D ing the ocean a necessary feature in any model induction arrows, and the TE phase results in a layered structure determined for a single repre- of California. good 2-D approximation to 3-D structure. Error sentative or average sounding curve was chosen All of the TE and TM apparent resistivity and fl oors of 5% for TM apparent resistivity, TM as a basis for the starting model for the inver- phase data and magnetic transfer functions would phase, and TE phase were used while the error sion. The model consisted of a 65 row × 81 col- ideally be used to generate the model; how- fl oor for TE apparent resistivity was 10,000%. umn grid, with 24 rows devoted to topography ever, some data were missing and others were Components of the induction arrows aligned and widths of adjacent columns varying by no of poor quality. The black dots on the pseudo- along the profi le had error fl oors of 0.03.

100 Geosphere, February 2012

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/8/1/98/3342537/98.pdf by guest on 01 October 2021 Yosemite MT

SJV SNB ML W E Ohm-m W E 215 213 212 214 216 205 201 209 203 207 208 206 202 204 215 213 212 214 216 205 201 209 207 203 208 206 202 10000 204 3162 –2 1000 –2 –1 316 –1 0 100 0 1 32 1 2 10 2 3 3.2 OBSV TE Rhoa 1.0 3 Log10(Period, s) CALC TE Rhoa 4 Log10(Period, s) 4 Degrees –2 –2 –1 90 –1 0 80 0 1 70 60 1 2 2 3 50 OBSV TE Phase 40 3 4 30 4 CALC TE Phase Log10(Period, s) 0 100 200 20 Log10(Period, s) 10 0 100 200 Distance (km) Distance (km)

W E Ohm-m W E 215 213 212 214 216 205 201 209 207 203 208 206 202 204 215 213 212 214 216 205 201 209 207 203 208 206 202 10000 204 3162 –2 1000 –2 –1 316 –1 0 100 0 1 32 1 2 10 2 3 3.2 1.0 3 Log10(Period, s) OBSV TM Rhoa 4 Log10(Period, s) 4 CALC TM Rhoa Degrees –2 –2 –1 90 –1 0 80 0 1 70 60 1 2 2 3 50 OBSV TM Phase 40 3 4 30 4 CALC TM Phase Log10(Period, s) 0 100 200 20 Log10(Period, s) 10 0 100 200 Distance (km) Distance (km)

Figure 3. Pseudosections of TE (transverse electric) and TM (tranverse magnetic) mode apparent resistivities (Rhoa) and phases. Observed (OBSV) data are shown on the left and responses calculated (CALC) from the best fi t model are shown on the right. Pseudosections are plotted horizontally with station position from west to east along profi le in Figure 2. The common logarithm of the period is used for the vertical scale from a period of 10–2 s at top to 104 s at bottom. Note the similarities between the observed and calculated pseudosections except for the TE apparent resistivities. Black dots on pseudosections show data values used in the inversion. For display pur- poses, both the observed and calculated responses were contoured in the same way. However, observed or mod- eled data values at black dots were used in all modeling, inversion, and calculation of misfi ts. ML—Mono Lake; SJV—San Joaquin Valley; SNB—Sierra Nevada batholith.

Rodi and Mackie’s (2001) inversion was 1.70 times the data error (Fig. 3). The TE mode through a conductive region spanning depths of used to obtain a model that produced synthetic apparent resistivities were not fi t as well as the 40–120 km (C in Fig. 4). However, this is also responses matching our observed ones. The TM mode values due to the large apparent resis- the region where Frassetto et al. (2011) inter- inversion minimizes a weighted combination tivity error fl oor for the TE mode (Fig. 3). preted the Moho as a fossil structure within of data misfi t and model roughness with the The fi nal model (Fig. 4) reveals the absence high-wave-speed material imaged by Reeg weighting factor, τ, being applied to the model of resistive material underlying the eastern (2008) with P wave tomography. It is common roughness. A larger value of τ places more Sierra Nevada below 45 km depth. A compari- for electrical and seismic images to be sensitive emphasis on a smooth model and a smaller son of the resistivity section with a receiver to different aspects of the physical and chemical value places more emphasis on minimizing function section (Frassetto et al., 2011) shows state of rocks, and we discuss the comparison data misfi t. Inversions were initially tried with that the Moho coincides with a band of resistiv- between sections after we assess the reliability several different weighting factors in order to ity (A in Fig. 4) that is 1–2 orders of magnitude of our model (Fig. 4). determine which provided the best tradeoff. We lower than the batholith or an underlying resis- Sensitivity tests were conducted to evaluate found that a weighting factor of 0.2 produced tor (B in Fig. 4). The Moho thus separates the the robustness of each model feature in Figure 4. the lowest weighted combination of data mis- resistive batholith from this underlying resistor. Because interactions between different resistivi- fi t and model roughness. The inversion was run West of the range, features in the resistivity sec- ties in a model control the overall MT response, for 400 iterations, resulting in a model with an tion are not as well correlated with the Moho, we could not simply change one model feature overall root mean square of 1.70, or a fi t within which deepens in this region to ~60 km, passing and then argue that changes in responses prove

Geosphere, February 2012 101

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/8/1/98/3342537/98.pdf by guest on 01 October 2021 Ostos and Park

WE Sierra Nevada of magnitude change resulting from fractions of San Joaquin Mono a percent of fl uid (e.g., Waff, 1974). While seis- Valley Yosemite Lake mic properties such as attenuation and Vp/Vs 209 213 215 214 212 208 207 206 205 204 203 202 201 (P and S wave velocity) ratios can indicate fl uid 0 ρ = 10–30 ohm-m content, the variation of these properties with ρ > 100 k ohm-m fl uid fraction is a few percent and not orders of magnitude as with resistivity. Both seismic and A ρ > 300 ohm-m electrical properties are sensitive to temperature; 50 wave speeds and resistivity of the solid matrix Moho B E ρ , Ohm-m decrease as temperature increases. C Upper mantle resistivities of 20–30 ohm-m D 10000 ρ = 10–30 ohm-m require elevated temperatures and some melt- 100 3162 ing. One of us (Park, 2004) found that melt frac- 1000 tions of <1% and temperatures of 1200–1250 °C Depth bsl,km Maximum depth of sensitivity 316 could explain such low resistivities at mantle 100 150 depths of 30–60 km. Assuming a model in 32 which a small percentage of basaltic melt forms 10 an interconnected grain boundary fi lm, solution 3 200 of the Hashin-Shtrikman (1962) equation for the 1 effective conductivity of a two phase material 50 100 150 200 estimates a melt fraction of 0.72% for conductor 121°W 120°W 119°W C (Fig. 4). This is consistent with elevated tem- Distance (km) peratures and partial melting found in the west- Figure 4. Best-fi t model for the MT (magnetotelluric) profi le through ern Basin and Range (e.g., Wang et al., 2002) Yosemite National Park. Depth is in kilometers below sea level (bsl). and beneath the southern Sierra Nevada (Park, Resistivity section is plotted with almost no vertical exaggeration. 2004). Coupled with the fact that the resistivity Heavy line denotes the Moho from receiver function study and is of E need not be >100 ohm-m, we propose that dashed where it is a fossil boundary (Frassetto et al., 2011). Where conductor C is the westernmost extension of determined, resistivity bounds are shown for regions of the model. hot, buoyant asthenosphere beneath the Sierra Areas A–E are discussed in text, and dashed lines show outlines Nevada. This asthenosphere could provide sup- of areas used in sensitivity tests. Note that batholith resistivity is port for high elevations in the absence of a deep >100,000 ohm-m (100 K ohm-m). Longitude lines are shown on the crustal root. section for reference. The P wave speed seismic tomography (Reeg, 2008) shows that conductors A, C, and E mostly are within a broad region of normal sensitivity to that feature. It is possible that there ity for this section is 120 km; below this depth, to slightly slow (<2%) mantle beneath a fast is an alternative model without that feature that changes to structure have little to no effect on the crust (Fig. 5). The correlation between wave fi ts the data equally well. To assess sensitivity, model fi t to the data. Therefore, structure deeper speed tomography and electrical resistivity is we fi xed the resistivity of that feature to some- than 120 km cannot be resolved. A 10:1 contrast not good, however, with part of conductor C thing different from the preferred model (Fig. 4). is required between resistor B and conductors A extending into a fast region west of station 208 We then ran the inversion again and allowed it to or C (Fig. 4), thus supporting the earlier state- and resistor D not extending into the fastest continue until convergence. If an alternate model ment that resistor B is not attached to the Sier- region above 50 km. with a fi xed resistivity fi t the data as well as did ran batholith. Additional tests focusing on just The narrow conductor A that is along or just the fi nal model in Figure 4, then we concluded the western part of A above resistor B showed above the Moho and beneath the Sierra Nevada that the alternate model was permissible and that that this 10:1 contrast was also preferred; misfi ts batholith is also likely due to partial melt from the feature was not constrained well by the data. at sites 207, 208, and 209 were elevated if this hot asthenosphere. If E is conductive, then Multiple inversions with different fi xed values of region were as resistive as B. A could be its western extension and simply resistivity for each feature were run in order to result from decompression melting of shallow establish the bounds on features shown in Figure DISCUSSION asthenosphere. If E is instead resistive, then 4. Alternatives were eliminated if misfi ts at sites conductor A could be melt generated beneath exceeded the errors in the observed data (follow- Electrical resistivity (and its inverse, con- the batholith that is ascending eastward from C ing the procedures in Park et al., 1996). ductivity) and seismic wave speeds can be used along the base of the batholith. In either case, More than 120 inversions were run in order to in a complementary fashion to determine the this melt could be the source of eruptions in the establish bounds on features A–E in the model physical and chemical state of rocks. Bulk wave Inyo Domes between sites 203 and 204 (Fig. 2). (Fig. 4). Based on these tests, both the western- speeds are good indicators of mineralogy (e.g., The melt fraction has to be small because the P most (D in Fig. 4) and easternmost (E in Fig. 4) Christensen and Mooney, 1995), while electrical wave speeds in conductor A range from normal resistive bodies at the edges of the MT section resistivity is almost insensitive to the minerals to 4% fast (Fig. 5). are not required to be resistive. Models with composing the solid portion of a rock (Jones, Resistor B is the most enigmatic portion of the these set to resistivities <100 ohm-m fi t the data 1992). Resistivity is extremely sensitive to fl uid model. It is clearly beneath the Moho (Fig. 4), equally well. The maximum depth of sensitiv- content, chemistry, and connectivity, with orders and yet is the most resistive section of mantle

102 Geosphere, February 2012

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/8/1/98/3342537/98.pdf by guest on 01 October 2021 Yosemite MT

WE Sierra Nevada estimated that the eclogitic root of the Sierra San Joaquin Mono Nevada was 250 kg/m3 denser than the sur- Valley Yosemite Lake rounding peridotite, so delaminated fragments 209 213 215 214 212 208 207 206 205 204 203 202 201 would sink quickly into the asthenosphere. Geo-

0 dynamic modeling of the Sierra Nevada shows that the process of delamination and foundering +8% is rapid, occurring over 1–2 Ma (LePourheit et A +6% al., 2006). Given that resistor B is not attached 50 to the current Sierran root, it should be in the Moho B E +4% process of sinking. Thus, our preferred interpre- tation is that resistor B represents a small piece C +2% D of eclogitic crust that has delaminated from the 100 0 Sierran root and is probably sinking into the asthenosphere.

Depth bsl,km Maximum depth of sensitivity –2% CONCLUSIONS 150 –4%

A 14 site magnetotelluric profile imaged widespread mantle asthenosphere extending 200 beneath the western Sierra Nevada. The absence of a deep crustal root beneath highest eleva-

50 100 150 200 tions and the presence of a conductive body 121°W 120°W 119°W immediately beneath the resistive batholith Distance (km) imply that the Sierra’s high elevations are sup- ported by hot buoyant asthenosphere. The thick Figure 5. P wave speed tomographic section along MT (magneto- eclogitic crust that was present during the Mio- telluric) profi le (adapted from Reeg, 2008). Depth is in kilometers cene Epoch (Ducea, 2002) was not imaged in below sea level (bsl). Location information and features from Fig- this section. A small resistive body at the top of ure 4 are displayed to facilitate comparison. Note that there is little the astheno sphere is inferred to be a remnant correspondence between conductive areas (A, C, E) and low wave of this eclogitic crust that is no longer attached speeds or resistive areas (B, D) and high wave speeds. to the batholith.

ACKNOWLEDGMENTS that is well constrained by data. The P wave lithosphere and would more likely be eclogitic This project was supported by National Science seismic tomography shows that the resistor has than peridotitic. Foundation Continental Dynamics Program grant wave speeds characteristic of mantle just below Features A, B, and C could all be astheno- EAR-0607349. Any opinions, fi ndings, and conclu- the Moho and is not anomalous (Fig. 5). It is sphere with varying degrees of partial melt. If so, sions or recommendations are those of the authors and clearly separated from the Sierran batholith by then the high resistivity of B could be explained do not necessarily refl ect the views of the National Science Foundation. Magnetotelluric instruments conductor A along the Moho. There are several by the absence of melt. This would mean that from the EMSOC (Electromagnetic Studies of the possible interpretations of resistor B: it could either B was warm enough to have partially Continents) instrument pool were used for this study. be an extension of the mafi c root of the batho- melted and the melt was subsequently extracted, We thank the staff of Yosemite National Park and the lith, a region of cooler mantle lithosphere that or that B is cooler than A and C. However, melt- Bishop offi ce of the Bureau of Land Management for has not yet melted, a region of asthenosphere ing in rising asthenosphere usually results from helping us obtain permits quickly. Many private land- owners made sites west of the Sierra Nevada available from which the melt has been extracted, or an decompression, so it is not clear why conduc- to us. This paper was improved by comments from eclogitic fragment of the root that has detached tors A and C would contain melt and resistor B two reviewers and the associate editor. from the batholith and has not yet sunk. If resis- would not. While a small temperature change tor B were simply a mafi c root separated from at the solidus can lead to a transition to partial REFERENCES CITED the rest of the Sierran batholith by a fault along melt and a drop in resistivity, this would suggest Christensen, N.I., and Mooney, W.D., 1995, Seismic velocity conductor A, then we would not expect it to that the asthenosphere is at a critical state and structure and composition of the continental-crust—A have mantle wave speeds. we would expect resistors like B elsewhere in global view: Journal of Geophysical Research, v. 100, p. 9761–9788, doi: 10.1029/95JB00259. Another alternative is that resistor B is mantle the mantle. Ducea, M.N., 2002, Constraints on the bulk composition and lithosphere but simply cooler and melt free. A Resistor B is deep enough that the pressure root foundering rates of continental arcs: A California consequence of this interpretation is that B is (>1.2 GPa) is suffi cient to metamorphose mafi c arc perspective: Journal of Geophysical Research, v. 107, 2304, doi: 10.1029/2001JB000643. lithosphere from the Mesozoic formation of the crustal materials to eclogite (Hacker, 1996), so it Ducea, M.N., and Saleeby, J., 1996, Buoyancy sources for batholith. Ducea and Saleeby (1996) used xeno- is possible that the resistor is delaminated resid- a large, unrooted mountain range, the Sierra Nevada, lith thermobarometry to show that the Miocene ual root. Ducea and Saleeby (1996) estimated California: Evidence from xenolith thermobarometry: Journal of Geophysical Research, v. 101, p. 8229– crustal root extended to a depth of at least 70 km that the root extended to at least 70 km, so resis- 8244, doi: 10.1029/95JB03452. and possibly to 100 km. Xenoliths from these tor B would just be a remnant of the original Ducea, M., and Saleeby, J., 1998, A case for delamination of the deep batholithic crust beneath the Sierra Nevada, depths are in the eclogite facies (Ducea, 2002); root of the Sierra Nevada that is now underlain California: International Geology Review, v. 40, resistor B is well above the transition to mantle by asthenosphere (conductor C). Ducea (2002) p. 78–93, doi: 10.1080/00206819809465199.

Geosphere, February 2012 103

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/8/1/98/3342537/98.pdf by guest on 01 October 2021 Ostos and Park

Fliedner, M.M., Ruppert, S.D., Malin, P.E., Park, S.K., etary Science Letters, v. 251, p. 104–119, doi: 10.1016 Waff, H.S., 1974, Theoretical considerations of electrical con- Jiracek, G.R., Phinney, R.A., Saleeby, J.B., Wernicke, /j.epsl.2006.08.028. ductivity in a partially molten mantle and implications for B.P., Clayton, R.W., Keller, G.R., Miller, K.C., Jones, Luddington, S., Moring, B.C., Miller, R.J., Flynn, K.S., geothermometry: Journal of Geophysical Research, v. 79, C.H., Luetgert, J.H., Mooney, W.D., Oliver, H.L., Stone, P.A., and Bedford, D.R., 2005, Preliminary p. 4003–4010, doi: 10.1029/JB079i026p04003. Klemperer, S.L., and Thompson, G.A., 1996, Three- integrated databases for the western states: California, Wang, K., Plank, T., Walker, J.D., and Smith, E.I., 2002, A dimensional crustal structure of the southern Sierra Nevada, Arizona, and Washington: U.S. Geological mantle melting profi le across the Basin and Range, Nevada from seismic fan profi les and gravity modeling: Survey Open-File Report OFR 2005–1305, http://pubs southwestern USA: Journal of Geophysical Research, Geology, v. 24, p. 367–370, doi: 10.1130/0091-7613 .usgs.gov/of/2005/1305. v. 107, 2017, doi: 10.1029/2001JB000209. (1996)024<0367:TDCSOT>2.3.CO;2. Mackie, R.L., Madden, T.R., and Park, S.K., 1996, A Wannamaker, P.E., 1999, Affordable magnetotellurics: Inter- Frassetto, A., Zandt, G., Gilbert, H., Owens, T.J., and Jones, three-dimensional magnetotelluric investigation of pretation in natural environments, in Oristaglio, M., and C., 2011, Lithospheric structure of the Sierra Nevada the California Basin and Range: Journal of Geophysi- Spies, B., eds., Three-dimensional electro mag netics: from receiver functions and implications for lithospheric cal Research, v. 101, p. 16221–16239, doi: 10.1029 Society of Exploration Geophysicists Geo physics Devel- foundering: Geosphere, v. 7, p. 898–921 doi: 10.1130 /96JB01276. opment Series, v. 7, p. 349–374. /GES00570.1. Park, S.K., 2004, Mantle heterogeneity beneath eastern Cali- Wernicke, B., Clayton, R., Ducea, M., Jones, C.H., Park, S., Hacker, B.R., 1996, Eclogite formation and the rheology, fornia from magnetotelluric measurements: Journal of Ruppert, S., Saleeby, J., Snow, J.K., Squires, L., Flied-

buoyancy, seismicity, and H2O content of oceanic crust, Geophysical Research, v. 109, B09406, doi: 10.1029 ner, M., Jiracek, G., Keller, R., Klemperer, S., Luetgert, in Bebout, G.E., ed., Subduction: Top to bottom: Amer- /2003jb002948. J., Malin, P., Miller, K., Mooney, W., Oliver, H., and ican Geophysical Union Geophysical Monograph 96, Park, S.K., Biasi, G.P., Mackie, R.L., and Madden, T.R., Phinney, R., 1996, Origin of high mountains in the con- p. 337–346. 1991, Magnetotelluric evidence of crustal suture zones tinents: The southern Sierra Nevada: Science, v. 271, Hashin, Z., and Shtrikman, S., 1962, A variational approach bounding the southern Great Valley, California: Jour- p. 190–193, doi: 10.1126/science.271.5246.190. to the theory of effective magnetic permeability of nal of Geophysical Research, v. 96, p. 353–376, doi: Wiese, H., 1962, Geomagnetische tiefensondierung. Teil multi phase materials: Journal of Applied Physics, v. 33, 10.1029/90JB02078. II: Die Streichrichtung der Untergrundstrukturen des p. 3125–3131, doi: 10.1063/1.1728579. Park, S.K., Hirasuna, B., Jiracek, G.R., and Kinn, C.L., electrischen Widerstandes, erschlossen aus geomagne- Jones, A.G., 1992, Electrical conductivity of the continen- 1996, Magnetotelluric evidence of lithospheric mantle tishen variationen: Geofi sica Pura e Applicata, v. 52, tal lower crust, in Fountain, D.M., et al., eds., Conti- thinning beneath the southern Sierra Nevada: Journal p. 83–103, doi: 10.1007/BF01996002. nental lower crust: New York, Elsevier, p. 81–143. of Geophysical Research, v. 101, p. 16241–16255, doi: Vozoff, K., 1991, The magnetotelluric method, in Nabig- Larsen, J.C., Mackie, R.L., Manzella, A., Fiordelisi, A., and 10.1029/96JB01211. hian, M.N., ed., Electromagnetic methods in applied Rieven, S., 1996, Robust smooth magnetotelluric trans- Reeg, H., 2008, Seismic structure of the crust and upper geophysics Volume 2: Society of Exploration Geo- fer functions: Geophysical Journal International, v. 124, mantle of the Sierra Nevada, California [M.S. thesis]: physicists Investigations in Geophysics 3, p. 641–711. p. 801–819, doi: 10.1111/j.1365-246X.1996.tb05639.x. Boulder, University of Colorado, 139 p. LePourheit, L., Gurnis, M., and Saleeby, J., 2006, Mantle Rodi, W.L., and Mackie, R.L., 2001, Nonlinear conjugate MANUSCRIPT RECEIVED 7 JANUARY 2011 instability beneath the Sierra Nevada Mountains in gradients algorithm for 2-D magnetotelluric inversion: REVISED MANUSCRIPT RECEIVED 19 OCTOBER 2011 California and Death Valley extension: Earth and Plan- Geophysics, v. 66, p. 174–187, doi: 10.1190/1.1444893. MANUSCRIPT ACCEPTED 1 NOVEMBER 2011

104 Geosphere, February 2012

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/8/1/98/3342537/98.pdf by guest on 01 October 2021