
==== Front
Proc Natl Acad Sci U S A
Proc Natl Acad Sci U S A
PNAS
Proceedings of the National Academy of Sciences of the United States of America
0027-8424
1091-6490
National Academy of Sciences

38466842
202317809
10.1073/pnas.2317809121
research-articleResearch Articleearth-sciEarth, Atmospheric, and Planetary Sciences413
Physical Sciences
Earth, Atmospheric, and Planetary Sciences
Multilevel transcrustal magmatic system beneath the Geysers-Clear Lake area
Li Tianjue a b https://orcid.org/0000-0003-2033-2793

Wu Shucheng a b https://orcid.org/0000-0002-1402-5669

Tong Ping tongping@ntu.edu.sg
a b c 1
aDivision of Mathematical Sciences, School of Physical and Mathematical Sciences, Nanyang Technological University, Singapore 637371, Singapore
bEarth Observatory of Singapore, Nanyang Technological University, Singapore 639798, Singapore
cAsian School of the Environment, Nanyang Technological University, Singapore 639798, Singapore
1To whom correspondence may be addressed. Email: tongping@ntu.edu.sg.
Edited by Michael Manga, University of California, Berkeley, CA; received October 13, 2023; accepted January 28, 2024

11 3 2024
19 3 2024
11 9 2024
121 12 e231780912113 10 2023
28 1 2024
Copyright © 2024 the Author(s). Published by PNAS.
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).

Significance

Petrological and geochemical studies of the volcanic deposits and crustal xenoliths contained therein generally provide the fundamental knowledge of the underground magmatic systems in the past. Geophysical imaging instead imposes quantitative constraints on the present underground volumetric structures typically at a resolution of tens of kilometers. Challenges remain in revealing the magmatic structures in the lower crust. By jointly inverting seismic first P and PmP traveltimes, we can image the multilevel architecture of the melt-bearing reservoirs beneath the Geysers-Clear Lake area over the entire crustal column. This transcrustal magmatic system produced a wide spectrum of erupted rocks over the past 2 Ma. Essentially, joint interpretation of the seismic images with geological constraints could advance the development of geothermal resources.

Magmatism in the Quaternary Clear Lake volcanic field (CLVF), with its youngest eruption having only occurred c. 10 ka ago, is commonly invoked as the heat source for the world’s largest commercial geothermal reservoir, The Geysers, in northern California. A shallow silicic magma reservoir in the upper-middle crust has been discovered for some time, but the location and mechanism of a potential deep mafic magma reservoir have remained elusive. Here, we present a seismic tomographic model that images the entire crustal column, clearly revealing a multilevel transcrustal magmatic system beneath the Geysers-Clear Lake area. Upwelling melts from the mantle traverse across the crust-mantle boundary and accumulate in the lower crust underneath the southeastern part of Clear Lake, resulting in a hot Moho in between. Mafic melts primarily ascend westward due to the extensional regime in the west and physical barrier effect from the overlying rigid ophiolite fragment, ultimately forming a shallow silicic magma reservoir underlying and heating The Geysers geothermal field. In addition, this study also links compositionally diverse volcanism in a continental setting to differentiation in a multilevel transcrustal magmatic system.

seismic reflection tomography
crustal P-wave velocity
The Geysers geothermal field
transcrustal magmatic system
==== Body
pmcSituated in the central belt of Franciscan Complex in Northern California, The Geysers is the world’s largest commercial geothermal reservoir. Having been exploited for more than 50 y, the field production levels are still on the order of 850 MWe (megawatts of electrical output, ref. 1). The Geysers is believed to be heated by magmatic intrusions associated with volcanism in the Clear Lake volcanic field (CLVF), which has experienced four eruption episodes with a noticeable northward progression during the past 2.1 Ma (Fig. 1; refs. 2, 3). The northward-younging trend of volcanism is thought to be related to the progressive movement of a slab window trailing the Mendocino Triple Junction (MTJ; ref. 4). Hot asthenosphere upwelled through the slab window and partial melts generated therein heated and melted the overlying crustal rocks, which ultimately erupted through the fault zones within the pull-apart basin (5). Although the CLVF consists predominantly of silicic extrusive rocks (e.g., dacite and rhyolite), eruptions of intermediate and mafic lavas (e.g., andesite and basalt) have also occurred (Fig. 1 B and C). Episodic eruptions resulted in a wide compositional variety of extrusive rocks, hinting a complex underground magmatic system (6–8).

Fig. 1. (A) Location of the study region (white/blue rectangles) with geologic features in Northern California south of the MTJ. There are three volcanic groups along the coastal California (3, 9): the Berkeley Hills (12.0 ~ 7.9 Ma), Sonoma volcanics (13.62 ~ 2.6 Ma), and Clear Lake volcanics (CLVF, 2.1 ~ 0.01 Ma). Red and pink shaded areas refer to the locations where extrusive and intrusive igneous rocks are found, respectively (from USGS California Geologic Map). The Geysers is outlined by the white dashed line. The San Andreas Fault (SAF), Maacama Fault (MF), and Bartlett Springs Fault (BSF) are three main NW-SE trending faults. (B) The eruption history of the CLVF. Colored dots denote different types of volcanic rocks at several sampling sites (2, 3). Note that basalt* includes both basalt and basaltic andesite. (C) The volumes of the erupted rocks at the CLVF modified from ref. 2. The Inset shows the accumulated volume of each rock type, including both the volume of existing rocks shown in C and the estimated volume accounting for erosion since deposition.

There is a general consensus that the deep heat source for the CLVF originates from the underlying hot mantle-derived magmas (e.g., refs. 4 and 10). However, the nature of the crustal magmatic system remains a topic of debate. In the early 1980s, based on gravity and teleseismic traveltime observations, a model of a magma chamber with its top sitting at 7 km depth beneath Mt. Hannah was proposed (11, 12). But subsequent local seismic tomography studies and magnetotelleric surveys did not find conclusive evidence to support this model (e.g., refs. 13–16). More recently, the proposed magma chamber in the upper-middle crust beneath the CLVF was imaged as a low-velocity zone (17), and its depth range was determined to be from about 5 to 13 km (18, 19). In the lower crust, the presence of a mafic magma reservoir was inferred based on seismic, gravity, and magnetotelleric observations (14). Further thermobarometry analyses of crustal xenoliths provided a conceptual model deciphering a two-level system of the transcrustal magmatism (8). A recent seismic velocity model using full waveform inversion also provided evidence to support the existence of a mafic magma reservoir (20). However, the resolution of the velocity model in the lower crust is still limited, which hinders our understanding of the transcrustal magmatic system (e.g., the spatial extent, melt fraction, and interplay among different melt zones) beneath the Geysers-Clear Lake area.

In this study, we conducted a joint inversion of seismic first P and Moho-reflected PmP traveltimes of local earthquakes from the past 20 y (Fig. 2 A and B). By combining and utilizing the complementary resolving abilities of the first P and PmP data, we were able to image the complete crustal-scale three-dimensional (3D) velocity structure. Our findings provide a clear tomographic image of a two-level crustal magma reservoir beneath the Geysers-Clear Lake area with high spatial resolution, which together with mantle velocity pattern from former studies enhances our comprehension of the transcrustal magmatic system beneath the CLVF. Furthermore, incorporating previous studies conducted in other well-studied volcanic fields in the western United States, we propose that the existence of a multilevel transcrustal magmatic system may represent a prevalent mechanism among those continental volcanoes which have experienced multiple eruptions with a wide compositional spectrum.

Fig. 2. (A) The source–receiver distribution of the first P dataset. Earthquakes are denoted by red stars, and seismic stations are represented as blue squares. (B) The source–receiver distribution of the PmP dataset. The open black square denotes an example seismic station TA.Q03C, which records the waveforms from earthquakes represented by the cyan stars beneath The Geysers. (C) The Moho geometry used in seismic tomographic inversions, which is constructed through the common conversion point stacking of receiver functions in a 3D velocity model developed by Shen et al. (21). (D) A reference 1D P-wave velocity model. The initial velocity model for seismic tomography incorporates this 1D velocity model and the Moho geometry in C. (E) Example waveforms with clear PmP signals recorded at station TA.Q03C, where the waveforms are normalized and aligned to the first P arrivals. The onset times of the PmP waves are indicated by red bars. Numbers in purple show the event-station distance (in km) and earthquake depth (in km).

Results

We present the obtained seismic velocity model as depth slices (Fig. 3 B–D) and vertical profiles (Fig. 3 E and F). At shallow depths (e.g., 2 km below the sea level, Fig. 3B), high-velocity features dominate the Franciscan Complex in comparison to low velocities beneath the Great Valley sequence, coinciding well with surface geology (Fig. 3A). Specifically, we observe high P-wave velocities of 5.5 km/s or greater beneath The Geysers and extending to its south, which correspond to the reported dry graywacke with low pore pressure (22, 23). Underlying the Great Valley sequence, we observe several discrete low-velocity zones (with Vp ≤ 4.0 km/s), which may represent the early isolated depocenters where the latest Jurassic-Early Cretaceous sediments accumulated (24). In the upper-middle crust, the Great Valley sequence extends from the surface down to approximately 10 km depth and sits atop the Great Valley ophiolite which is imaged as a high-velocity zone (Fig. 3E). These findings are consistent with a previous active-source survey (25).

Fig. 3. (A) Surface geology map of the study region (extracted from USGS California Geologic Map). The two black arrows are used to show the locations of two vertical profiles, P1 and P2. (B–D) Depth slices of the velocity model obtained in this study at 2 km, 7 km, and 17 km depths below the sea level. (E) Velocity structures along the vertical profile P1. Earthquakes with magnitude larger than 1.0 (NCEDC double difference catalog during 1984-2022, ref. 26) within 10 km distance from the profile are projected as open circles whose sizes are logarithmically scaled according to earthquake density. (F) is the same as E but for the velocity structure along the vertical profile P2.

Beneath the Geysers-Clear Lake area, a relatively low-Vp region (≤5.5 km/s) is present centered at the northern part of The Geysers from about 4 to 14 km depth (Fig. 3 C and E). It has a diameter of about 30 km, likely representing the proposed silicic magma reservoir (11, 12, 18, 19). In the middle-lower crust, we identify two obvious low-velocity zones (≤5.75 km/s) that are strengthened and more visible when PmP traveltime data are incorporated into the inversion (SI Appendix, Figs. S13 and S15). One is situated beneath the southeast of Clear Lake, starting from approximately 15 km down to the Moho discontinuity at 23 km depth (Fig. 3 D and E). The other low-velocity zone is located beneath the Berkeley Hills, spanning from approximately 12 to 19 km depth (Fig. 3 D and F), and has been reported before (e.g., ref. 16). It might constitute the remnants of melted materials left behind by volcanic activity around 7.9 Ma (9). Although there are no more hot mantle materials supplied from the bottom (e.g., ref. 10), the melt body could be maintained through thermal blanketing effect (27) as the overlying sedimentary basin evolved (28). In the west of this low-Vp feature, we observe a wide high-Vp zone in the middle-lower crust (at 12.5 km and greater depths), which is thought to have oceanic affinity and could be the stalled crust of the Pacific Plate (29).

Discussion

Multilevel Transcrustal Magmatic System Beneath the Geysers-Clear Lake Area.

Our tomographic model provides a comprehensive picture of the transcrustal magmatic system beneath the CLVF (Fig. 4A). Directly beneath the high-Vp geothermal reservoir rocks (graywacke), there is an abnormally low-Vp zone, which we interpret as a silicic magma reservoir following prior studies (11, 12, 18, 19). This low-velocity feature is robust from both the checkerboard resolution tests (SI Appendix, Figs. S17 and S18) and restoration tests (SI Appendix, Fig. S19). Although its geometrical shape may be influenced by the uneven data converge, as shown in the restoration tests (SI Appendix, Fig. S19), its upper boundary matches the distribution of local seismicity well, with most earthquakes occurring at depths shallower than 5 km above the imaged magma reservoir (Fig. 3E). We used the Gassmann fluid substitutions following (30) to estimate the silicic melt fraction, considering a granite-rhyolite-melt system (see more details in SI Appendix). The low-Vp anomaly was estimated to have a melt fraction of 8.4 vol%, comparable to the melt fraction (>10 vol%) estimated through analyzing density anomalies (19). Multiplying the estimated magma reservoir volume by its melt fraction, we obtained a contemporary silicic melt volume of 243 km3, which matches the estimation (275 km3) from thermal modeling for the scenario c. 1.0 Ma ago based on zircon age analysis (31). According to the thermal evolution model (31), it is likely that the shallow magma reservoir was first emplaced in its current position during the late stage of the first volcanic eruption episode. The present silicic magma reservoir could be an integral of a series of intrusions after absorbing the magma pulses supplied from the lower crust (7, 32, 33), and may have maintained a nearly constant volume with elevated temperature since 1.0 Ma.

Fig. 4. (A) The multilevel transcrustal magmatic system beneath the CLVF. The inverted crustal velocity structure is displayed in a quasi-3D view. We also present the estimated temperature along the Moho discontinuity (34) and the P-wave velocity perturbations in the uppermost mantle (60 km depth, (10)) on the bottom. Note that the topography of the bars (3D cubes) is proportional to the amplitude of P-wave velocity perturbation (dVp) at 60 km depth, and each bar is colored according to the value of dVp at its specific location based on the color bar showing on the left. The two arrows denote the possible main melt ascending pathways, and the ellipses show the speculated trajectories of small-volume melt diapirs/dikes. (B) Summary of four volcanic fields in the western United States: Clear Lake (this study), Coso (35), Mount St. Helens (36, 37), and Yellowstone (38). The volcano symbols with varied sizes approximate the volumes of erupted rocks, silicic (red) versus mafic (purple). The magma reservoirs with the inferred depth range, melt fraction (color-coded with magma colorbar), and averaged seismic velocity (solid edge for Vp and dashed edge for Vs) are shown as colored boxes.

In the lower crust, we observe an obvious low-Vp feature beneath the southeast of Clear Lake. Such a low-Vp feature is robust based on the checkerboard resolution tests (SI Appendix, Fig. S18), restoration tests (SI Appendix, Fig. S19), and inversions starting from other Moho geometries (SI Appendix, Fig. S21). We interpret it as a lower crustal mafic magma reservoir according to previous studies (8, 20). An approximate 25-km horizontal offset between the centers of the two imaged crustal magma reservoirs is observed, where the deep mafic magma reservoir is situated to the east of the shallow one. This phenomenon may result from the combined effect of an extensional regime around the Geysers-Clear Lake area (5) and the impediment from the overlying rigid high-Vp rocks (39) slightly west of the Bartlett Springs Fault (Figs. 3C and 4A). We interpret the high-Vp rocks as an ophiolite fragment, given that the dismembered ophiolite outcrops have been widely observed nearby (40). The Moho temperature underlying the inferred mafic magma reservoir is higher than surrounding regions (Fig. 4A and SI Appendix, Fig. S22; ref. 34), which not only supports our interpretation but also suggests the presence of an active pathway through which magma migrates from the upper mantle to the lower crust.

We estimated the mafic melt fraction using the same method aforementioned but with a mafic-granulite-basalt-melt system. The inferred melt fraction of 7.4 vol% is higher than other volcanoes in the western United States (Fig. 4B), e.g., Yellowstone (2 vol%, ref. 38), Coso (2.6 vol%, ref. 35), and Mount St. Helens (6 vol%, ref. 36), which may arise from variations in physical models and/or velocity contours used to estimate the melt fraction. However, previous petrologic studies and modeling experiments (e.g., refs. 6 and 8) suggested that the mafic magma reservoir is regularly replenished by hot mantle materials and thus could exhibit a high proportion of molten rocks. In fact, the melt fraction could be even higher, given that the inverted P-wave velocity in the deep crust may be underestimated as seen in the restoration tests (SI Appendix, Fig. S19). To reconcile the constraint that the intermediate-felsic magma reservoir is fed by an ~5-times larger mafic magma reservoir underneath (7, 41), we speculate that the lower crustal magma reservoir may have migrated northward over the past 2 Ma in response to the progressive movement of the MTJ (4), and eventually settled in its current position with a melt volume of 693 km3 nowadays. During the evolution of the CLVF, mafic melts from the lower crust may have stalled within the upper-middle crust and formed evolved magma reservoirs through a series of magma pulses (7, 33). Only a small portion of the mafic magma was able to reach the surface via diapirs/dikes (39), which could be responsible for the episodic mafic eruptions throughout the lifetime of the CLVF. Collectively, this resulted in a wide range of erupted rock compositions spanning from basalt to rhyolite over the past 2 Ma (2).

In the upper mantle, a broad area with low Vp, low Vs, and high Vp/Vs sits between 60 and 120 km depth beneath the CLVF (Figs. 3F and 4A and SI Appendix, Fig. S22; (ref. 10), outlining the spatial extent of the hot upwelling asthenosphere induced by the northward passage of a slab window trailing the MTJ (4). The active transcrustal magmatic system beneath the Geysers-Clear Lake area may have been regularly recharged/rejuvenated through ascending partial melts derived from the hot upwelling asthenosphere (6, 8).

Direct Heat Source for the Geothermal Field.

The Geysers is a vapor-dominated geothermal field which has been active since at least c. 0.26 Ma (e.g., ref. 32). It is primarily heated by a series of shallow intrusions (e.g., refs. 15 and 42), among which the Geysers Plutonic Complex is the largest one with a present volume of about 75 km3 (31). Zircon geochronology data have shown that the drilled composite plutonic complex formed between c. 1.1 and 1.8 Ma (43), suggesting that they should have cooled by now if there was no thermal recharge from below (8, 32). To explain the present-day elevated surface heat flow of 167.2 mW/m2 and higher (44), we invoke an active transcrustal magmatic system imaged in this study as the heat source, where the silicic magma reservoir regularly recharges and heats the shallow intrusive bodies (3, 42). The presently hottest point is located beneath the northern part of The Geysers, where the temperature can reach 400 °C at a depth of no more than 3 km (1). This spatially coincides well with the center of the inferred silicic magma reservoir in our tomographic image (Fig. 3C). Reservoir rocks that exhibit lower levels of fracturing, along with the presence of existing intrusive bodies, may have impeded the formation of new intrusions (39). Alternatively, the newly formed intrusions could have been constrained, occurring primarily along limited fractures and manifesting in small volumes (45). Although new intrusions with an age younger than 1.1 Ma have not been found to date by drill holes, their existence is manifested by the stable isotopes in the geothermal vapor which are indicative of a magmatic component (46). Besides, the heat transferred from the silicic magma reservoir to the shallow existing intrusions through thermal conduction could have played a more vital role in keeping the geothermal field hot (e.g., ref. 32). Thermal modeling based on our imaged magmatic system together with the previously obtained emplacement history of the shallow intrusions (e.g., refs. 32, 33, and 45) could provide more details about the thermal evolution of The Geysers, and give valuable instructions for the further exploitation of geothermal resources, e.g., denoting the most likely region where the young intrusion exists and outlining the most productive region where the enhanced geothermal system is suitable to construct (1).

Multilevel Structure as a Prevalent Feature for Continental Transcrustal Magmatic Systems.

In this study, we demonstrated through tomographic images that episodic volcanic eruptions with a wide compositional spectrum are strong indicators of an underlying complex magmatic system. In addition, we compiled the depth range, melt fraction estimate, and representative seismic velocities of the imaged magma reservoirs at four well-studied volcanic fields in the western United States (Fig. 4B). Despite the varying estimates of magma volume, depth range, and melt fraction from one volcano to another, we find that all these volcanic fields present a strong causality between a multilevel transcrustal volcano-related structure at the scale of tens of kilometers and a wide compositional variety of the overlying volcanic rocks formed through multiple eruptions. The recently proposed term “transcrustal magmatic system” by ref. 47 reflects the expansion of the research scope from the upper crustal magma chambers to the magmatic processes over the entire crustal column and beyond. Our seismic images beneath the CLVF together with the geophysical studies at other volcanic fields provide alternative evidence apart from the geochemistry and petrology studies toward the transcrustal-scale view, revealing additional details regarding the spatial extent, melt fraction, and interplay among different melt storages of the complex magmatic system. The multilevel transcrustal magmatic system model provides a more complete picture of the plumbing system of continental volcanic fields, which helps to address the important questions regarding magma/melt transports through multiple reservoirs at different depths, the interaction between deep magma reservoirs and overlying volcanic systems, and the relationship with the diverse volcanic products on the surface.

Methods

Initial Model.

We first scaled an accurate regional three-dimensional (3D) S-wave velocity model (21) into a P-wave velocity model using an empirical relationship (21). Then, a 1D P-wave velocity model was calculated by laterally averaging the 3D P-wave velocity model in the study region. We further updated the 1D model in the 3D space by inverting first P traveltimes to obtain a reference 1D P-wave velocity model (Fig. 2D). In addition, an accurate Moho model is required to mitigate the tradeoff between crustal velocity and Moho discontinuity when inverting PmP traveltimes. We employed the common conversion point stacking of receiver functions (RFs) to image the Moho discontinuity. The time-to-depth mapping of RFs was conducted in a 3D velocity model (21). We built the Moho geometry model following the procedures outlined in (48). More details related to the RFs data preparation and Moho topography construction can be found in SI Appendix. The resulting Moho model agrees well with the contiguous U.S. Moho model (49) and Crust 1.0 (50). In the study region, the Moho is shallow (<20 km) offshore, reaches an intermediate depth (~25 km) beneath the inland Coast Ranges, and deepens under the Great Valley (Fig. 2C and SI Appendix, Fig. S10). The P-wave velocity model incorporating the reference 1D velocity model (Fig. 2D) and the Moho geometry determined via receiver function analysis (Fig. 2C) serves as the initial model for the tomographic inversions of this study.

Tomographic Inversions.

A total of 174,991 first P traveltimes covering the period from January 2000 to December 2021 (Fig. 2A) were selected to image the upper and middle crust from the Northern California Earthquake Data Center (NCEDC, (51)). To illuminate the lower crust, 9,247 high-quality PmP arrivals were picked by applying a reliable seismic reflection identification workflow (48) to the waveforms archived by the Incorporated Research Institutions for Seismology (IRIS, (52)) for the same time span as first P. We employed the seismic reflection tomography technique outlined in ref. 12 to determine the velocity structure of the entire crustal column. In summary, we solved eikonal equations on a uniformly spaced forward grid, followed by ray tracing. Five sets of inversion grids were used to parameterize the model space. A sequential inversion strategy was adopted to fully explore the complementary resolving abilities of the first P and PmP traveltimes. Specifically, we began with inversions that only use first P traveltimes, followed by inversions that only utilize PmP traveltimes. Finally, we simultaneously inverted both the first P and PmP traveltimes to refine the model. Further details regarding the selection of first P traveltimes, the workflow used to pick PmP waves, and inversion process are provided in SI Appendix.

Supplementary Material

Appendix 01 (PDF)

We thank Dr. May R. Berenbaum, Dr. Michael Manga, and two anonymous reviewers for their constructive comments that have greatly improved the paper. We also thank Michael A. Mitchell, Jared R. Peacock, Axel K. Schmitt, and Robert S. White for their constructive discussions. This work is funded by Minister of Education, Singapore, under its MOE AcRF Tier-1 Thematic Grant (RT12/22). This research is partly funded by the National Research Foundation Singapore and the Singapore Ministry of Education under the Research Centers of Excellence Initiative (Project Code Number: 04MNS001953A620).

Author contributions

P.T. designed research; T.L. performed research; T.L. and P.T. contributed new reagents/analytic tools; T.L., S.W., and P.T. analyzed data; and T.L., S.W., and P.T. wrote the paper.

Competing interests

The authors declare no competing interest.

Data, Materials, and Software Availability

Seismic phase data (first P traveltimes) and local earthquake catalog are requested from Northern California Earthquake Data Center (https://ncedc.org/ (51)). Event waveforms are downloaded from the Incorporated Research Institutions for Seismology (https://www.iris.edu/hq/ (52)). PmP traveltimes and the 3D P-wave velocity model obtained in this study are available at https://doi.org/10.21979/N9/JPCTF9 (53).

Supporting Information

This article is a PNAS Direct Submission.

Although PNAS asks authors to adhere to United Nations naming conventions for maps (https://www.un.org/geospatial/mapsgeo), our policy is to publish maps as provided by the authors.
==== Refs
1 J. Stimac, M. Wilmarth, P. E. Mandeno, P. Dobson, J. Winick, Review of exploitable supercritical geothermal resources to 5 km at geysers-Clear Lake, Salton Sea, and Coso. GRC Trans. 41 , 806–835 (2017).
2 J. Donnelly, B. Hearn Jr., G. Curtis, R. Drake, Geochronology and evolution of the Clear Lake Volcanics, California. U.S. Geological Survey Professional Paper 1141 , 47–60 (1981).
3 J. Donnelly-Nolan, M. Burns, F. Goff, E. Peters, J. Thompson, The Geysers-Clear Lake area, California; thermal waters, mineralization, volcanism, and geothermal potential. Econ. Geol. 88 , 301–316 (1993).
4 K. P. Furlong, S. Y. Schwartz, Influence of the Mendocino triple junction on the tectonics of coastal California. Annu. Rev. Earth Planet. Sci. 32 , 403–433 (2004).
5 R. McLaughlin, Tectonic setting of pre-Tertiary rocks and its relation to geothermal resources in The Geysers-Clear Lake area. Research in The Geysers-Clear Lake Geothermal Area, Northern California 1141 , 3–23 (1981).
6 L. Hammersley, D. DePaolo, Isotopic and geophysical constraints on the structure and evolution of the Clear Lake volcanic system. J. Volcanol. Geothermal Res. 153 , 331–356 (2006).
7 A. K. Schmitt, R. L. Romer, J. A. Stimac, Geochemistry of volcanic rocks from the Geysers geothermal area, California Coast Ranges. Lithos 87 , 80–103 (2006).
8 J. A. Stimac, F. Goff, K. Wohletz, Thermal modeling of the Clear Lake magmatic-hydrothermal system, California, USA. Geothermics 30 , 349–390 (2001).
9 K. F. Fox Jr., R. J. Fleck, G. H. Curtis, C. E. Meyer, Implications of the northwestwardly younger age of the volcanic rocks of west-central California. Geol. Soc. Am. Bull 96 , 647–654 (1985).
10 B. Schmandt, E. Humphreys, Complex subduction and small-scale convection revealed by body-wave tomography of the western United States upper mantle. Earth Planet. Sci. Lett. 297 , 435–445 (2010).
11 H. M. Iyer, D. H. Oppenheimer, T. Hitchcock, J. N. Roloff, J. M. Coakley, Large teleseismic P-wave delays in The Geysers-Clear Lake geothermal area. Geol. Survey Professional Paper 1141 , 97–116 (1981).
12 W. F. Isherwood, Geophysical overview of the Geysers. Research in the Geysers-Clear Lake geothermal area, Northern California. 1141 , 83–95 (1981).
13 D. Eberhart-Phillips, Three-dimensional velocity structure in northern California Coast Ranges from inversion of local earthquake arrival times. Bull. Seismological Soc. Am. 76 , 1025–1052 (1986).
14 W. D. Stanley, R. J. Blakely, The Geysers-Clear Lake geothermal area, California—An updated geophysical perspective of heat sources. Geothermics 24 , 187–221 (1995).
15 W. D. Stanley, H. M. Benz, M. A. Walters, A. Villaseñor, B. D. Rodriguez, Tectonic controls on magmatism in The Geysers-Clear Lake region: Evidence from new geophysical models. Geological Soc. Am. Bull 110 , 1193–1207 (1998).
16 J. A. Hole , Three-dimensional seismic velocity structure of the San Francisco Bay area. J. Geophys. Res. Solid Earth 105 , 13859–13874 (2000).
17 Y. Liu, P. Tong, Eikonal equation-based P-wave seismic azimuthal anisotropy tomography of the crustal structure beneath northern California. Geophys. J. Intern. 226 , 287–301 (2021).
18 Y. Qi, P. Tong, Structure of the crustal magmatic system in the Geysers-Clear Lake Area (Northern California) imaged by adjoint-state travel-time tomography. Seismolog. Res. Lett. 94 , 414–427 (2022).
19 M. A. Mitchell, J. R. Peacock, S. D. Burgess, Imaging the magmatic plumbing of the Clear Lake Volcanic Field using 3-D gravity inversions. J. Volcanol. Geothermal Res. 435 , 107758 (2023).
20 A. Rodgers , WUS256: An adjoint waveform tomography model of the crust and upper mantle of the Western United States for improved waveform simulations. J. Geophys. Res. Solid Earth 127 , e2022JB024549 (2022).
21 W. Shen, M. H. Ritzwoller, Crustal and uppermost mantle structure beneath the United States. J. Geophys. Res. Solid Earth 121 , 4306–4342 (2016).
22 R. Gritto, S. P. Jarpe, D. L. Alumbaugh, “New large-scale passive seismic monitoring at The Geysers Geothermal Reservoir, CA, USA” in Proceedings, 47th Workshop on Geothermal Research Engineering (2022), pp. 1–11.
23 G. Lin, B. Wu, Seismic velocity structure and characteristics of induced seismicity at the Geysers geothermal field, eastern California. Geothermics 71 , 225–233 (2018).
24 D. A. Orme, K. D. Surpless, The birth of a forearc: The basal Great Valley Group, California, USA. Geology 47 , 757–761 (2019).
25 N. Godfrey, B. Beaudoin, S. Klemperer, Ophiolitic basement to the Great Valley forearc basin, California, from seismic and gravity data: Implications for crustal growth at the North American continental margin. Geol. Soc. Am. Bull. 109 , 1536–1562 (1997).
26 F. Waldhauser, W. L. Ellsworth, A double-difference earthquake location algorithm: Method and application to the northern Hayward fault, California. Bull. Seismological Soc. Am. 90 , 1353–1368 (2000).
27 M. Wangen, The blanketing effect in sedimentary basins. Basin Res. 7 , 283–298 (1994).
28 W. P. Elder, Bedrock geology of the San Francisco Bay Area: A local sediment source for bay and coastal systems. Marine Geol. 345 , 18–30 (2013).
29 T. M. Brocher , Seismic evidence for a lower-crustal detachment beneath San Francisco Bay, California. Science 265 , 1436–1439 (1994).17833818
30 R. Chu, D. V. Helmberger, D. Sun, J. M. Jackson, L. Zhu, Mushy magma beneath Yellowstone. Geophys. Res. Lett. 37 , L01306 (2010).
31 C. A. Angeles-De La Torre , A common magma source for plutonic and volcanic rocks of The Geysers geothermal field, California: Volume and intrusive history derived from zircon. Chem. Geol. 624 , 121414 (2023).
32 G. B. Dalrymple , Age and thermal history of the Geysers plutonic complex (felsite unit), Geysers geothermal field, California: A 40Ar/39Ar and U-Pb study. Earth Planet. Sci. Lett. 173 , 285–298 (1999).
33 A. K. Schmitt , The Geysers-Cobb Mountain Magma System, California (Part 1): U-Pb zircon ages of volcanic rocks, conditions of zircon crystallization and magma residence times. Geochim. Cosmochim. Acta 67 , 3423–3442 (2003).
34 D. L. Schutt, A. R. Lowry, J. S. Buehler, Moho temperature and mobility of lower crust in the western United States. Geology 46 , 219–222 (2018).
35 D. Wang, S. Wu, T. Li, P. Tong, Y. Gao, Elongated Magma Plumbing System beneath the Coso Volcanic Field, California, Constrained by Seismic Reflection Tomography. J. Geophys. Res. Solid Earth 127 , e2021JB023582 (2022).
36 C. Jiang, B. Schmandt, G. A. Abers, E. Kiser, M. S. Miller, Segmentation and radial anisotropy of the deep crustal magmatic system beneath the Cascades arc. Geochem. Geophys. Geosyst. 24 , e2022GC010738 (2023).
37 E. Kiser, A. Levander, C. Zelt, B. Schmandt, S. Hansen, Focusing of melt near the top of the Mount St. Helens (USA) magma reservoir and its relationship to major volcanic eruptions. Geology 46 , 775–778 (2018).
38 H.-H. Huang , The Yellowstone magmatic system from the mantle plume to the upper crust. Science 348 , 773–776 (2015).25908659
39 V. Acocella, Volcano-tectonic Processes (Springer, 2021), vol. 567 .
40 J. W. Shervais , Multi-stage origin of the Coast Range ophiolite, California: Implications for the life cycle of supra-subduction zone ophiolites. Intern. Geol. Rev. 46 , 289–315 (2004).
41 S. J. Barker , What lies beneath? Reconstructing the primitive magmas fueling voluminous silicic volcanism using olivine-hosted melt inclusions. Geology 48 , 504–508 (2020).
42 K. Erkan, D. D. Blackwell, M. Leidig, “Crustal thermal regime at the Geysers/Clear Lake area, California” in Proceedings World Geothermal Congress, Antalaya, Turkey (2005), pp. 24–29.
43 A. K. Schmitt , The Geysers-Cobb Mountain Magma System, California (Part 2): Timescales of pluton emplacement and implications for its thermal history. Geochim. Cosmochim. Acta 67 , 3443–3458 (2003).
44 M. A. Walters, J. Combs, “Heat flow in The Geysers-Clear Lake geothermal area of northern California, USA” in Monograph on the Geysers Geothermal Field (Geothermal Resources Council, 1992), vol. 17 , pp. 43–53.
45 J. R. Peacock , Geophysical characterization of the Northwest Geysers geothermal field, California. J. Volcanol. Geothermal Res. 399 , 106882 (2020).
46 A. Truesdell, M. Walters, M. Kennedy, M. J. Lippmann, An integrated model for the origin of The Geysers geothermal field. Transactions 17 , 273–280 (1993).
47 K. V. Cashman, R. S. J. Sparks, J. D. Blundy, Vertically extensive and unstable magmatic systems: A unified view of igneous processes. Science 355 , eaag3055 (2017).28336610
48 T. Li, J. Yao, S. Wu, M. Xu, P. Tong, Moho complexity in Southern California revealed by Local PmP and Teleseismic Ps waves. J. Geophys. Res. Solid Earth 127 , e2021JB023033 (2022).
49 B. Schmandt, F. C. Lin, K. E. Karlstrom, Distinct crustal isostasy trends east and west of the Rocky Mountain Front. Geophys. Res. Lett. 42 , 10290–210298 (2015).
50 G. Laske, G. Masters, Z. Ma, M. Pasyanos, “Update on CRUST1. 0—A 1-degree global model of Earth’s crust” in Geophysical Research Abstracts (EGU General Assembly Vienna, Austria, 2013), p. 2658.
51 NCEDC, Northern California Earthquake Data Center. UC Berkeley Seismological Laboratory. Dataset. 10.7932/NCEDC. Accessed 18 June 2022.
52 T. Ahern, The FDSN archive at the IRIS data management center. SAGE. https://ds.iris.edu/. Accessed 1 April 2022.
53 P. Tong, “Replication Data for: Multilevel trans-crustal magmatic system beneath the Geysers-Clear Lake area.” DR-NTU (Data). 10.21979/N9/JPCTF9. Deposited 6 January 2024.
