Newsletter_Final2020_021921_red

science nuggets Portfolio of research projects supported by GeoPRISMS 132 Volcanic seismicity beneath Chuginadak Island, Alaska (Cleveland and Tana Volcanoes): Implications for magma dynamics and eruption forecasting 134 From the slab to the surface: Origin, storage, ascent, and eruption of volatile-bearing magmas 136 Water and oxygen fugacity controls on continental signatures in western Aleutian arc magmas 137 Carbon in the mantle lithosphere 138 Exploring the Alaska Peninsula subduction zone with AACSE: The Alaska Amphibious Community Seismic Experiment 140 After field work leaves the field – The AACSE active source supplement 142 Plumes, plate thinning, and magmatism in the East African Rift 144 Paleogene flood basalt stratigraphy in East Africa 146 Deformation of the East African Rift 147 EARThD: a compilation of explosive volcanism in East Africa 148 Displaced cratonic mantle concentrates deep carbon during continental rifting 150 Revealing asthenospheric rheology beneath continental regions 152 Protracted continental breakup along the Eastern North American Margin Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 101 

Science Nuggets Hikurangi Ocean Bottom Investigation of Tremor and Slow Slip (HOBITSS) - Revealing the environment of shallow slow slip Susan Schwartz, Anne Sheehan, and Rachel Abercrombie Figure 1. Hypocenters determined using Bayesloc and onshore and OBS stations The 2014-15 HOBITSS deployment of twenty-four ocean (triangles). Event symbol is scaled by local bottom pressure sensors and fifteen ocean bottom magnitude size and colored by depth. Gray seismometers (OBSs) at the northern Hikurangi margin, shaded area encloses the 2014 slow slip area. New Zealand, captured a M7.0 slow slip event (SSE). The vertical Figure modified from Yarce et al. (2019). deformation data collected were used to image one of the best- Seismicity at the northern Hikurangi Margin, resolved slow slip distributions to date, and indicated slip very close, if not all the way to the trench (Wallace et al., 2016). We used New Zealand, and investigation of the the seismic data collected during this experiment to evaluate the potential spatial and temporal relationships spatiotemporal relationship between seismicity (both earthquakes and tremor) and the slow slip event and the role that seismic with a shallow slow slip event. structure plays in controlling slip behavior. Significant results include: Right page: Rough seas off the shores of New 1. Creation of a catalog of local earthquakes consisting of about Zealand during the HOBITSS deployment in 2300 events ranging in magnitude between 0.5 and 4.7 that reveals 2014. Photo credit: Justin Ball. two NE-SW seismicity bands with a 20-km wide gap between them. This gap, beneath the inner forearc wedge, locates at the downdip edge of the slow slip patch. 2. Most earthquakes are within the subducting slab rather than at the plate interface. 3. Template matching reveals “burst-type” repeating earthquakes occur coincident with tremor on an upper-plate fracture network above a subducted seamount. This activity locates at the edge of the slow slip patch, but begins just after the SSE. We propose that during the large plate-boundary SSE, fluids migrated from downdip overpressurized sediments imaged in seismic reflection data, into the fracture network, diverting aseismic slip to multiple faults in the upper plate. Thus, seamount subduction appears to play a key role in controlling the mechanics of shallow slow slip and microseismicity at the northern Hikurangi margin. 4. Changes in principal stress ratios, obtained from focal mechanism inversions, prior to slow slip events represent the accumulation and release of fluid pressure within overpressured subducting oceanic crust, the episodicity of which may influence the timing of slow slip event occurrence on subduction megathrusts. 5. Overall temporal variations in VP/VS and shear wave splitting delay times observed during slow slip are consistent with fluid pressurization below a permeability barrier and movement of ■fluids during the build-up to and rupture of slow-slip patches. 102 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

References Barker, D.H.N., R. Sutherland, S. Henrys, S.C. Bannister (2009). Geometry of the Hikurangi subduction thrust and upper plate, North Island, New Zealand. Geochem, Geophys, 10, Q02007, doi.org/10.1029/2008GC002153 Bell, R., C. Holden, W. Power, X. Wang, G. Downes (2014). Hikurangi margin tsunami earthquake generated by slow seismic rupture over a subducted seamount. Earth Planet Sci Lett, 397, 1-9, doi.org/10.1016/J.EPSL.2014.04.005 Bell, R., R. Sutherland, D.H.N. Barker, S. Henrys, S.C. Bannister, L.M. Wallace, J. Beavan (2010). Seismic reflection character of the Hikurangi subduction interface, New Zealand, in the region of repeated Gisborne slow slip events. Geophys J Int , 180, 1, 34-48, doi.org/10.1111/j.1365‐246X.2009.04401.x Shaddox, H.R., S.Y. Schwartz (2019). Subducted seamount diverts shallow slow slip to the forearc of the northern Hikurangi subduction zone, New Zealand, Geology, 47,5, doi.org/10.1130/G45810.1 Wallace, L.M., S.C. Webb, Y. Ito, K. Mochizuki, R. Hino, S. Henrys, S.Y. Schwartz, A.F. Sheehan (2016). Slow slip near the trench at the Hikurangi subduction zone, New Zealand. Science, 352, 6286, 701-704, doi.org/10.1126/science.aaf2349 Yarce, J., A.F. Sheehan, J.S. Nakai, S.Y. Schwartz, K. Mochizuki, M.K. Savage, L.M. Wallace, S.A. Henrys, S.C. Webb, Y. Ito, R.E. Abercrombie, B. Fry, H. Shaddox, E.K. Todd (2019). Seismicity at the Northern Hikurangi Margin, New Zealand, and investigation of the potential spatial and temporal relationships with a shallow slow slip event. J Geophys Res, 124, 5, 4751-4766, doi.org/10.1029/2018JB017211 2014 Gisborne SSE Repeating General Earthquakes Seismicity 0 2014 SSE Tremor GrowClust Relocated NW Accretionary Not Relocated Wedge Deformation Tūranganui SE Depth (km) Australian Front Knolls 5 Plate Fracture Zone Subducted Hikurangi Plateau Thrust fault Seamount HRZ Paci c Plate Décollement 10 2014 Gisborne SSE Tremor Duration 304 312 320 328 326 HRZ High-amplitude -50 -40 Julian Day 2014 re ectivity 280 288 296 -30 -20 -10 0 10 20 30 Distance (km) from the deformation front Figure 2. Seismic reflection profile 05CM-04 modified from Barker et al. (2009), Bell et al. (2010), and Bell et al. (2014). Bottom: Depth-converted interpretation. Repeating earthquakes (stars) and all seismicity within 10 km of reflection profile are colored by time. Large stars and circles are relocated with GrowClust. Figure from Shaddox and Schwartz (2019). Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 103 

Science Nuggets Preliminary findings from the NZ3D 3D seismic imaging experiment Nathan Bangs, Greg Moore, Shuoshuo Han, Hannah Tilley, and the NZ3D project team The NZ3D 3D seismic imaging experiment was designed The landslide deposits likely buried BSR2, shifting the gas hydrate to examine structure, tectonics and the role of fluids in stability boundary to BSR1. shallow slow slip earthquakes along the northern Hikurangi We also found four patches with double BSRs within the NZ3D margin of New Zealand. During the initial phase, the University volume that lie within thrust anticlines and are not related to of Texas group has focused on recent mass wasting, fluid flow, and landslides. These imply recent rapid uplift, potentially as a result of tectonics within the accretionary wedge as indicated by bottom seamount collision, or possibly enhanced fluid flow directed into simulating reflections (BSRs). The NZ3D data show a rare double the anticlines. BSR (Fig.  1), which we exploit to examine recent subduction processes. We observe overlapping BSRs where a broadly extensive A primary BSR1 lies above a more localized, weaker secondary BSR2. Double BSRs can indicate significant, rapid changes in pressure- temperature conditions that shift the base of gas hydrate stability zone to shallower depths. By mapping the double BSR near IODP Site 1519, we find that BSR2 follows the paleo seafloor defined by an unconformity that was drilled at Site 1519, but has recently been covered by landslide deposits (Fig. 1B). Figure 1. A. Bathymetry of the B Hikurangi margin offshore Gisborne, C NZ showing the region imaged during the NZ3D experiment (gray shading). Red circles show IODP Exp. 372/375 drill sites. Yellow box shows the region covered in the middle panel B. B. Depth to unconformity mapped from the NZ3D data and drilled at Site U1519. A patch of BSR2 (blue outline) overlaps with the landward edge of the unconformity. Black line is the track of the seismic profile shown in the bottom panel C. C. Seismic image along crossline 2038 showing BSR2 (magenta arrows), BSR1 (yellow arrows), and predicted BSR position (blue). Note BSR2 follows nearly parallel with the unconformity (orange) and is consistent with a paleo BSR that tracks a seafloor position along the unconformity. The large separation between the BSRs and predicted BSR implies both BSR1 and BSR2 are relict BSRs that may have been buried by landslide deposits and are out of equilibrium with current pressure/temperature conditions. 104 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

The University of Hawaii group has integrated regional shallow The NZ3D science party, technical staff and crew of structure with recent drilling results from IODP Expeditions the R/V Langseth during the January 2018 seismic 372/375. Recent tectonic activity in the trench slope basin is also implied from units imaged in the 3D data and drilled at Site U1519. survey off the New Zealand north island. Significant along strike variations of these units are consistent with continued thrusting within the underlying accretionary prism. Strata within the basin have been trapped by uplift of the seaward edge of the basin. Several sub-units are defined by internal stratal onlap and truncation and help to understand the timing and magnitude of thrusting within the prism. Strata on the Pacific plate east of the trench, sampled at Site 1520 represent the inputs to the subduction zone. Our 3D data show that the input sequence is highly variable throughout the basin. The pelagic sequence contains sediment waves that are ~1km in width and ~300-400 m high and extend across the entire width of the volume. They are onlapped by the overlying trench strata that includes the Ruatoria MTD, which ranges in thickness from > 200 m at the NE side of the 3D box to < 100m at the SW side. At the deformation front, most of the sedimentary cover sequence is accreting along the plate boundary décollement located close to the base of the sequence. During the second phase of this project, we hope to document how these lateral variations in the input sequence cause variations in deformation along the frontal thrust and how they might influence slow slip events deeper in the subduction zone. The second phase for both groups, just starting, will target the deep structures directly related to slow slip and ■integrate them with tectonic and fluid flow processes implied by BSRs and shallow structures. The New Zealand primary site exhibits a wide range of fault slip and volcanic phenomena with significant variation along-strike in a small, compact setting. Excellent exhumed exposures of arc crust and accretionary prism, a zone of active subduction initiation, and significant government investments in onshore and offshore scientific infrastructure made New Zealand an exciting site for GeoPRISMS research. From north to south, the New Zealand primary site includes the Puysegur Ridge (subduction initiation), Puysegur Trench (subduction), Fiordland (exhumed arc crust), Hikurangi Trench (subduction), the Taupo Volcanic Zone (arc and rift volcanism), the southern Kermadec Arc (subduction), and the Havre Trough (back-arc rifting). Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 105 

Science Nuggets Unraveling the effects of upper plate lithology and stress on seismic velocities at Hikurangi Andrew Gase, Harm Van Avendonk, Nicola Tisato, Nathan Bangs, Kelly Olsen, Carolyn Bland, Kyle Bland, and Dominic Strogen The Hikurangi margin, which parallels the East Coast of At convergent margin forearcs, seismic velocities typically North Island, New Zealand, is a convergent margin with a increase towards the interior of the overriding plate, indicating a prolonged accretionary history. The margin underwent two variation in compaction and metamorphic grade, or a transition major phases of convergence, from ~120-85 Ma and ~22 Ma to to igneous basement. Active-source seismic tomography present, separated by a period of passive margin sedimentation. and multi-channel seismic imaging results from the SHIRE As a result of this history, the eastern North Island is a 150-200 project indicate distinct structural domains in the northern km wide accretionary wedge, with three major sedimentary units Hikurangi margin prism (Fig. 1; Gase et al., in review). that include (1) an Early Cretaceous metasedimentary basement, The outer ~35 km of the overriding plate is a low-velocity frontal (2) Late Cretaceous-Paleogene passive margin sediments, and (3) prism of off-scraped trench-fill sediments. A network of irregular Miocene-Present sediments that extend to the offshore deformation thrust faults that outcrop at Tuaheni Ridge separates the frontal front (e.g., Bland et al., 2015). Within the overriding plate, where prism from a higher velocity inner prism. This abrupt landward and how transitions between these three units occur remains increase in seismic velocity can be explained by the exhumation of unclear, particularly in the offshore accretionary prism where stronger sediments within the hanging wall of the Tuaheni thrust lithology could control the permeability and elastic stiffness of the and/or a boundary between stronger Late Cretaceous-Paleogene upper plate and have a major impact on subduction earthquakes. sediments and weaker Miocene-Present sediments. Figure 1. a) Map of SHIRE seismic Line 1b/MC10 (red line), high-resolution bathymetry from NIWA. b) Interpreted seismic reflection image of SHIRE seismic Line 1b/MC10 from Gase et al., (in review) showing faults (red dashes), bottom simulation reflector (blue dashes), and slope cover sediments (black dash). c) Overlain seismic velocity model from joint OBS and streamer tomography and seismic reflection image with fault interpretations. 106 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Figure 2. Geologic map of North Island, New Zealand (QMAP, GNS), and measurements on dry samples as a function of confining pressure (Pc). a) Relevant geologic units have the following colors: Early Cretaceous - Turquoise, Late Cretaceous-Paleogene - Green, Miocene-Eocene - Orange. Sample sites for rocks presented in b) (Cretaceous Torlesse graywacke) and c) (Miocene-Eocene mudstone) are indicated by gray inverted triangle and yellow diamond. b) Ultrasonic velocities. c) Relative vertical deformation. In addition, analyses of seismic traveltimes (Bassett et al., 2014) References and ground-motion from the 2017 Kaikoura earthquake (Kaneko et al., 2019) reveal large contrasts along the margin, with faster Bassett, D., R. Sutherland, S. Henrys (2014). Slow wavespeeds and fluid velocities in the southern margin, and slower velocities in the overpressure in a region of shallow geodetic locking and slow slip, northern margin. These observations could result from variations Hikurangi subduction margin, New Zealand. Earth Planet Sci Lett, in fluid controlled effective stress (Bassett et al., 2014), but may 389, 1-13. also indicate differences in the extent of upper plate lithologies. Older, stronger, rocks of the Early Cretaceous accretionary wedge Bland, K.J., C.I. Uruski, M.J. Isaac (2015). Pegasus Basin, eastern New could be more prevalent along the southern margin, whereas less Zealand: A stratigraphic record of subsidence and subduction, ancient consolidated Late-Cretaceous-Paleogene sediments may form the and modern. New Zealand Journal of Geology and Geophysics, 58, 4, interior of the northern margin. 319-343. doi.org/10.1080/00288306.2015.1076862 CRUSH, a new GeoPRISMS funded project, will test the effects of lithology and stress through geomechanical experiments on Gase, A.C., H.J.A. Van Avendonk, N.L. Bangs, D. Bassett, S. Henrys, D.H.N. rocks from across the Hikurangi forearc. Rock samples will be Barker, et al. (in review). Crustal structure of the northern Hikurangi tested under varying confining and fluid pressures, and differential margin, New Zealand: Variable accretion and overthrusting plate stresses, while measuring ultrasonic velocities, permeability, and strength influenced by rough subduction. Submitted to J Geophys strain. Preliminary experiments on dry (1) Cretaceous Torlesse Res: Solid Earth. grackwacke from the western Raukumara Peninsula, and (2) Miocene-Eocene Mudstones from near Tolaga Bay show that Kaneko, Y., Y. Ito, B. Chow, L.M. Wallace, C. Tape, R. Grapenthin, E. under increasing confining pressure, the mudstone deforms more, D'Anastasio, S. Henrys, R. Hino (2019). Ultra-long duration of seismic although the two examples exhibit only minor differences in Vp and ground motion arising from a thick, low-velocity sedimentary Vs (Fig. 2). Results of these ongoing experiments will be integrated wedge. J Geophys Res: Solid Earth, 124, 10, 10347-10359. doi. with seismic tomography and imaging along the Hikurangi margin org/10.1029/2019JB017795 ■to help us understand the hydromechanical controls of lithology and its impact on earthquake hazards. Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 107 

Science Nuggets How are large eruptions different? Reconstructing changes in magma reservoirs in the Taupo Volcanic Zone Kari Cooper, Chad Deering, and Adam Kent The largest volcanic eruptions are rare events but when 2. New high-precision Sr, Nd, and Pb isotopic data for glasses they occur can represent a global catastrophe. This project from the complete caldera cycles show that the OVC and TVC focuses on examining recent caldera-forming eruptions erupted magmas that are largely isotopically distinct from each from the highly active Taupo Volcanic Zone (TVZ) in New other, with temporal changes in composition evident for each Zealand, focusing on the Okataina and Taupo Volcanic Centers center. Furthermore, some samples show Pb isotopic compositions (OVC and TVC). Our approach is primarily a petrological and that are more radiogenic than existing data for the local crustal geochemical one and focuses on studying full caldera cycles. The rocks, suggesting that there is unsampled diversity in crustal overarching project goal is to develop a better understanding of materials contributing to the TVZ magmas, and/or that additional how the compositional variability, temperature and mobility of components (enriched mantle or lower crust) are contributing the magma reservoir below the surface changes before, during, significantly to the melts. and after a major eruption. As such the project contributes to an 3. Polytopic vector analysis, a multivariate statistical analytical tool, emerging understanding of the magmatic processes leading to was used to determine the variability in plagioclase compositions large eruptions, and provides context for interpretation of hazard among eruptions from the OVC spanning the time before, during monitoring. The project also included field research experience for and after the large, caldera-forming event. The results show that two K-12 teachers (one in the US and one in New Zealand) and distinct crystal populations were shared (i.e. recycled) throughout K-12 course content development based on this experience. the eruptive history, but specific end-member compositions Highlights of the results to date are: dominate the crystal cargo following caldera collapse. This suggests 1. Zircon age and trace-element data from unpolished rims that new, distinct magma compositions began accumulating and polished grain interiors provide a record of compositional following caldera collapse with limited recycling of material from changes within the reservoir spanning tens of thousands of years. the previous magmatic system. Comparison of the compositional and thermal diversity of interiors Overall, the picture emerging from these combined data is one of vs. surfaces indicates that magmas drew crystalline material from a complex and compositionally diverse magma system present at a compositionally diverse crystal-rich reservoir, and that more any given time, with rapid assembly of the erupted magmas before than one compositionally distinct magma body was present at eruptions. Although final data collection and synthesis has been the time of the most recent caldera-forming eruption at both the delayed due to covid-19, additional data for the full sample suite OVC and TVC. In addition, the composition and ages of zircon hosted within plagioclase are distinct from those in the whole rock, ■will provide a more nuanced view of the evolving magma reservoir providing insights into the nature of the crystal-rich source region. beneath the TVZ. Right page, top photo. Tephra sequence from Okataina Caldera Center (including Rotorua eruption, ~15 ka) in quarry section. Left to right, Tyler Schlieder, Lydia Harmon, Elizabeth Grant, Damien Cranney, Nicole Rocco. Photo credit: Kari Cooper 108 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

View of Mount Nguarahoe along the Tongariro crossing. Photo credit: Kari Cooper Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 109 

Science Nuggets SISIE: South Island, New Zealand, Subduction Initiation Experiment Michael Gurnis, Brandon Shuck, Erin Hightower, Harm Van Avendonk, Sean P. S. Gulick, Joann Stock, and Rupert Sutherland The South Island Subduction Initiation Experiment (SISIE) SISIE, our understanding of the antecedent tectonics and evolving was successfully completed in February and March 2018 dynamics associated with subduction initiation have improved using the R/V Langseth (as MGL1803). Seismic reflection enormously. and refraction data were acquired across the Puysegur Trench With the new seismic images, much of the crust immediately to the and Ridge and the Solander Basin (Fig. 1A). For the multichannel east of the Puysegur Trench was found to be rifted continental crust seismic (MCS) imaging, the Langseth used a 12.6 km long streamer (Fig. 1D) and not oceanic crust as previously interpreted along for much of the 1,300 km of lines acquired. A group of 28 ocean- the entirety of the MRC. This is an important new observation, bottom seismometers (OBSs) from the University of Texas Institute because the density difference across an ocean-continental margin of Geophysics (UTIG) were used at 43 sites on two refraction lines. is substantially larger than that across an ocean-ocean margin. One combined refraction/MCS line targeted the more juvenile part A plate tectonic reconstruction shows that the density difference of the Puysegur Ridge (SISIE 1) while the other targeted the more across the plate boundary rapidly increased during strike-slip evolved part (SISIE-2). Two lines of onshore seismic receivers motion between 18 and 15 Ma, just before subduction initiation, and several broadband seismometers on islands were deployed supporting the role of compositional differences in the initiation by New Zealand collaborators. Since the experiment, there have of Puysegur subduction. During initiation, a large fault (Tauru been important results from analysis of the seismic reflection and Fault) within the northern Solander Basin, inverted from normal refraction data and modeling subsurface density structure. With to reverse. Figure 1. Regional A and detailed B elevation maps showing major tectonic features of the Puysegur margin and the SISIE survey. Gray and black lines are MCS lines. Yellow circles represent successful OBS deployments used in analysis. The box green delimits the region of the gravity inversion shown in Figure 2. SZ = Snares Zone; PB = Puysegur Bank; PF = Puysegur Fault; TF = Tauru Fault; PaF = Parara Fault; SF = Solander Fault; HF = Hauroko Fault; eBF = eastern Balleny Fault; SI = Solander Island. The combined MCS and seismic tomography results from the OBS are shown for lines SISIE-1 (F) and SISIE-2 (D) along with high resolution multibeam bathymetry (C and E). For D and F the compressional-wave (Vp) seismic velocity model overlain on the Pre-stack depth migrated MCS reflection images. Figure modified from Shuck et al. (2020). 110 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Using sequence stratigraphy with the N-S MCS lines and an existing constraints in a 3D Bayesian gravity inversion, which showed petroleum exploration well, we constrained the compressional that the crust below the Snares zone was thicker compared to the event between 12 and 8 Ma. From our interpretation of the Puysegur Ridge to the south even though it was topographically seismic images, we suspect that subduction may have initiated depressed (Fig. 2). This likely reflects a strong change in the vertical on the western-most, thickest fragment of the thinned Solander force balance along strike. The structural model, constraints on the Basin crust, and through sequence stratigraphy we estimate that evolution of the state of stress, and inferences on the non-isostatic there was an evolution of stress from compression to tension from topography are being used for a high resolution, three-dimensional north to south. We also developed a seismic tomographic model of the crust and lithosphere with OBS-inferred velocities mapped ■geodynamic model, which shows the change from forced to self- to density (Fig. 1). These mapped images were then used as prior sustaining subduction. References Hightower, E., M. Gurnis, H. Van Avendonk (2020). A Bayesian 3D linear gravity inversion for complex density distributions: Application to the Puysegur subduction system, Geophys J Int, 223, 3, 1899-1918, doi: 10.1093/gji/ggaa425 Sandwell D.T., H. Harper, B. Tozer, W.H. Smith (2019). Gravity field recovery from geodetic altimeter missions, Adv. Space Res. doi:10.1016/j. asr.2019.09.011 Shuck, B., H. Van Avendonk, S.P.S. Gulick, M. Gurnis, R. Sutherland, J. Stock, J. Patel, E. Hightower, S. Saustrup, T. Hess (2020). Strike-slip enables subduction initiation beneath a failed rift: New seismic constraints from Puysegur Margin, New Zealand, Tectonics, doi.org/10.1002/essoar.10503735.1 Figure 2. Perspective views looking north for the area shown by the green box in Figure 1. A. Free air gravity anomaly (from Sandwell et al., 2019) for the Puysegur study area used in the Bayesian gravity inversion of crustal thickness. PB, Puysegur Bank; SZ, Snares Zone; CP, Campbell Plateau; PR, Puysegur Ridge and SB, Solander Basin; B. Moho depth interpreted from the 3-D density model; C. Crustal thickness for the Puysegur study area calculated by subtracting the bathymetry from the Moho depth and overlain on the bathymetric surface. The crustal volume is filled to the base of the crust using the Moho surface in panel B. Figure modified from Hightower et al. (2020). SISIE Science Party in front of the R/V Langseth Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 111 

Science Nuggets Slow-Slip and Fluid Flow Response Offshore New Zealand (SAFFRONZ) - Probing the nature of the Hikurangi margin hydrogeochemical system Evan A. Solomon, Irita Aylward, Marta Torres, and Robert Harris Fluid generation, migration, and pore fluid pressure at on the northern margin. SAFFRONZ complements and extends subduction zones are hypothesized to exert a primary these efforts by providing 1) a continuous two-year record of fluid control on the generation of seismicity, low-frequency flow rates and composition through the 2019 slow slip event, 2) earthquakes, and slow slip events (SSEs). The SAFFRONZ (Slow- information on the present background state of fluid flow and slip and fluid flow response offshore New Zealand) project how it relates to inferred pore fluid overpressure along the plate addresses the GeoPRISMS Subduction Cycles and Deformation boundary, and 3) comparative geochemical and hydrologic Initiative Science Plan by testing interrelationships among fluid data between the northern and southern sections of the margin. production, fluid flow, and slow slip at the Hikurangi Margin. The Our field strategy combined ship operations and ROV surveys recognition of dramatic changes in the along-strike distributions in a nested approach to constrain the margin-wide fluid flow of SSEs and their recurrence intervals, interseismic coupling, distribution. Coring, heat flow measurements, and benthic fluid inferred pore pressure, and other subduction-related parameters flow meter deployments targeted fault-hosted seep sites and off- at the Hikurangi margin have resulted in a concerted international fault locations from the deformation front to the shelf-break at effort to acquire seismological, geodetic, other geophysical, and both the southern and northern Hikurangi margin. Continuous geomechanical data both onshore and offshore the Hikurangi fluid flow rate measurements at off-fault locations will quantify margin. This effort includes recent scientific ocean drilling, logging, the fluid flow response to local volumetric strain during slip, and and the deployment of two subseafloor observatories during IODP Expeditions 372 and 375, as well as a 3D seismic reflection survey ■comparative data at fault zones will provide information on the hydrologic responses to slip. Figure 1. SAFFRONZ core and instrument deployment locations. Map shows all sites surveyed (hydroacoustic and ROV Jason dives) during the 2019 SAFFRONZ research expedition, as well as the locations of cores recovered, heat flow measurments, and benthic fluid flow meter deployments (CAT and Mosquito fluid flow meters). 112 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Science Nuggets Deformation and anisotropy of antigorite Charis Horn, Pierre Bouilhol, and Philip Skemer Hydrous silicates such as antigorite have highly anisotropic extremely strong bulk anisotropy (P-wave and S-wave anisotropy seismic and rheological properties. It is thought that these up to ~25% and 31%, respectively). minerals may influence several geologic phenomena of Moreover, these microstructures would be capable of generating subduction zones, including shear localization, intermediate-depth strong trench-parallel shear wave splitting when the sheared interface earthquakes, and shear-wave splitting. To understand the feedbacks is dipping, or otherwise oriented at a high angle to an incident between serpentinization, deformation, and seismic anisotropy, we shear wave. Given the apparent rheological weakness of antigorite, analysed microstructures from a suite of antigorite-bearing samples we hypothesise that progressive deformation will localise in areas from the Kohistan paleo-island arc in Pakistan. Weakly deformed with higher modal percentages of antigorite, thus overprinting any samples display evidence for crystallographically-controlled original topotactic signature in the rocks. These regions of highly growth of antigorite after olivine. Two distinct relationships were deformed antigorite then have the capability to substantially affect found: seismic wave patterns in a subduction zone. Mylonitic antigorite in 1. (010)ant//(100)ol with [100]ant//[001]ol and 2. (010)ant//(100)ol with [100]ant//[010]ol ■a dipping structure could explain some of the complex patterns of However, this topotactic replacement produces a bulk texture with only modest seismic anisotropy. In contrast, highly deformed shear wave wave splitting in subduction zones. samples, in which the serpentine was sheared to produce a For more information, please see: Horn, C., Bouilhol, P., Skemer, strong lattice-preferred orientation (LPO), were found to possess P. (2020) Serpentinization, deformation, and seismic anisotropy in the subduction mantle wedge, Geochem, Geophys, doi: 10.1029/2020GC008950 Figure 1. Photomicrograph of samples D9 (protomylonite) and D3 (mylonite) in crossed‐polarized light. (a&b) and (c&d) display the same fields of view but with the polarizers rotated by 45°. Low birefringence (grey to white) grains are antigorite, and higher birefringence (orange, pink, and purple) grains are olivine. (a,b) In the relatively undeformed protomylonite “haloes” of antigorite with similar orientations surround relict olivine (marked as white dashed lines), suggesting there is a topotactic relationship between the olivine and antigorite grains that formed during hydration. (c,d) Mylonites exhibit a strong degree of crystallographic alignment between antigorite grains, which can be seen in the similar extinction angles visible across the entire thin-section. These mylonites are inferred to be the product of progressive deformation of a more weakly deformed protolith. Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 113 

Science Nuggets Slow slip events in Cascadia and New Zealand Noel Bartlow This project looked at the relationships between slow slip subduction zones, including where they are located relative to the events, also called slow earthquakes, and fault locking locked zone. This will inform how likely these slow slip events are in Cascadia and New Zealand. Slow slip events occur on to eventually trigger a big earthquake. In addition, the results of the subduction megathrust, which is the contact between the this project form the basis for long-term monitoring of slow slip subducting plate and the overriding plate. A slow slip event is slip events in both subduction zones. Monitoring may yield observable on the plate interface, similar to an earthquake, but occurs over a changes in slow slip behavior prior to future large earthquakes. longer time scale ranging from days to months or even longer. In We demonstrated interactions between earthquakes and slow slip both New Zealand and Cascadia, slow slip events occur below the events in New Zealand, including the first documented case of depth of interseismic locking - the region of the fault that is stuck dynamically triggered slow slip. The 2016 magnitude 7.8 Kaikōura due to friction and will eventually rupture in large earthquakes. earthquake on the South Island triggered slow slip events over Slow slip events can trigger earthquakes, potentially including 200 km away on the Hikurangi subduction zone under the North devastating megathrust events with magnitude of 8.5 or larger. Island. This project aimed to better characterize slow slip events in these Figure 1. A. Time-averaged episodic tremor and slip rate (colors) and contours of density of tremor detections (brown lines) on the Cascadia plate interface (modified from Bartlow, 2020); B. Same as A, but with a comparison to the location of the locked zone (red and yellow colors) from Schmalzle et al. (2014). 114 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Figure 2. Time-dependent slip inversion results from TDEFNODE during the time following the Kaikōura M7.8 earthquake. Slip amounts on the interface are shown in centimeters (yellow to red colours), and horizontal displacements for the cGPS sites from the best-fitting model are shown as arrows (with scale). The blue circles are earthquakes from the GeoNet catalogue for each time slice, and the green circles are repeating earthquakes determined from template matching. The focal mechanism on the fifth panel is for the Mw 6.0 thrust event on 22 Nov 2016 from www.geonet.org.nz. Figure from Wallace et al. (2017). We conducted time-dependent modeling of slow slip events in The results of this project were used to motivate funding for a new both subduction zones, allowing us to see slip move around on the plate interface. This allowed us to better illuminate the relationships ■NSF-funded project aimed at detecting the second potential slow between slow slip events and any earthquakes occurring nearby. For the 2012 Gisborne, New Zealand slow slip event, we conducted slip zone at the downdip limit of locking. static and time-dependent models incorporating both onshore GPS data and offshore Absolute Pressure Gauge data. This allows References for the best constraints to date on slow slip in this offshore region with a history of damaging shallow tsunami earthquakes. Bartlow, N.M. (2020). A Long-term view of episodic tremor and slip in In Cascadia, we also measured the cumulative effect of all slow slip Cascadia. Geophys Res Lett, 47, 3, doi.org/10.1029/2019GL085303 events to quantify the role of slow slip in overall plate motion. We found it to be highly variable along strike. We also found a \"gap\" Schmalzle, G.M., R. McCaffrey, K.C. Creager (2014). Central Cascadia between the deep limit of the locked zone and the shallow limit of subduction zone creep. Geochem, Geophys, 15, 1515-1532, doi. the slow slip zone, with hints of a possible second slow slip zone org/10.1002/2013GC005172 at the deep limit of locking. This means that slow slip in the main slow slip zone is not likely to trigger a great earthquake, but slow Wallace, L.M., Y. Kaneko, S. Hreinsdóttir, I. Hamling, Z. Peng, N. Bartlow, slip in the potential second slow slip zone might. E. D’Anastasio, B. Fry (2017). Large-scale dynamic triggering of shallow slow slip enhanced by overlying sedimentary wedge. Nat Geosci, 10, 10, 765-770, doi.org/10.1038/ngeo3021 Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 115 

Science Nuggets The Deschutes Formation: North America's most recent arc-related ignimbrite flare up Bradley Pitcher, Adam Kent, Anita Grunder, and Roberta Duncan A B This project studied the Deschutes formation of the Oregon Cascade range. This formation is located immediately to the east of the current Cascade arc and contains a 5-7 Ma sequence of distal volcanogenic sediments with a high proportion of silicic ignimbrite and tephra fall deposits. The volcanic units are derived from sources within the region of modern-day Cascade arc volcanic front, and mark a period of unusual high production and eruption of silicic magmas immediately following arc rearrangement. Extensive new geochronology, tephra correlations, and field and geochemical data show that volcanism was silicic, highly explosive in characters, and restricted to the period between 6.25 and 5.45 Ma. Overall this period resulted in eruption of at least 400 to 675 km3 (210 to 330 km3 dense-rock equivalent DRE) of magma. The volumetric output (2.7 to 4.1 km3/10 kyr, DRE) and average frequency of eruptions (1.0 to 1.5 eruptions/10 kyr) during the deposition of the Deschutes Formation are more than a factor of ten greater than the Cascades arc activity outside this interval. The Deschutes Formation thus represents North America’s most recent arc-sourced ignimbrite flare-up. We hypothesize that a heightened flux of basalt, induced by slab-rollback following arc reorganization, was focused beneath the arc into the shallow crust by crustal extension. Storage of basalt at shallow levels beneath a new arc locus within fertile crust produced this unusually silicic and ■explosive period of Cascade arc history. References Pitcher, B.W., A.J.R. Kent. A.L. Grunder, R.A. Duncan (2017). Frequency and volumes of ignimbrite eruptions following the Late Neogene initiation of the Central Oregon High Cascades. JVGR, 339, 1-22 Priest, G.R. (1990). Volcanic and tectonic evolution of the Cascade Volcanic Arc, central Oregon. J. Geophys. Res.: Solid Earth 95, 19583-19599 Figure 1. A. Schematic cross section of the Deschutes Formation with major ignimbrites showing with stratigraphic profile and north-south extent, and major field features. Dates show 40Ar-39Ar ages from Pitcher et al. (2017) and sources therein; B. Volcanic eruption rates estimated for the Oregon Cascade arc since 40 Ma from Pitcher et al. (2017) and Priest (1990). Dashed lines show the Deschutes formation and letters refer to different volume estimate methods. Eruption rates during the Deschutes Formation time are significantly great than current (Holocene) rates and the largest in the Oregon Cascade arc in the last 15 million years. 116 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Dramatic ignimbrite “hoodoos” produced by caprock weathering of the Balanced Rock ignimbrite. The large dark and banded pumice clasts that are characteristic of this unit are evident in the hoodoos. Photo credit: A. Kent Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 117 

Science Nuggets Small-scale mantle convection in the back-arc modulates volcanic activities along the Cascade arc Haiying Gao and Yang Shen Using continuous ground motion recorded by over one induced by the subduction of the Juan de Fuca plate underneath thousand seismic stations in the Pacific Northwest from the North America plate. The magnitude of the velocity reduction the northern California to Washington State, Prof. Yang in those volumes requires the presence of partial melt. The along- Shen and Dr. Haiying Gao extracted data that represent seismic strike variations in the velocity structure suggest that the large-scale waves that propagate between pairs of the stations along the Earth plate-motion-induced flow in the back-arc mantle is modulated surface. From those data, they constructed a wave speed model for the crust and upper mantle in the region with a tomographic ■by small-scale convection, resulting in a highly 3D process that method that fits the observed waveforms and the synthetics from 3D wave propagation simulation. The model reveals three low- defines the segmentation of volcanism along the Cascade arc. velocity volumes in the mantle east of the Cascadia volcanic arc, which are spatially correlated with the three arc-volcano clusters. These low-velocity volumes indicate upwelling of hot mantle Segmented low-velocity anomalies along the Cascade back-arc. (a) Horizontal slice at depth of 94 km (Vs in km/s). The black dashed lines outline the amplitude of largest negative Sp phase from receiver functions in the back-arc (Hopper et al., 2014). The magenta lines mark the profile locations in (b), (c), (d) and (e), respectively. All the panels share the same color bar. (b-d) W-E profiles across the back-arc anomalies. The y-axis has the approximate same length scale as the x-axis. The triangles mark the volcano centers. The Juan de Fuca plate interface at depths of 20-100 km from the model of McCrory et al. (2006) is projected. At greater depth, the plate interface is poorly defined. (e) S-N profile along the back-arc low- velocity anomalies, which spatially correlate with the three volcano clusters along the Cascadia. The length scale of y-axis is exaggerated two times of the x-axis. Gao and Shen (2014). References Gao, H., Y. Shen (2014). Upper mantle structure of the Cascades from full-wave ambient noise tomography: Evidence for 3D mantle upwelling in the back-arc. Earth Planet Sci Lett, 390, 222-233, doi.org/10.1016/j.epsl.2014.01.012 Hopper, E., H.A. Ford, K.M. Fischer, V. Lekic, M.J. Fouch (2014). The lithosphere–asthenosphere boundary and the tectonic and magmatic history of the northwestern United States. Earth Planet Sci Lett, 10.1016/j.epsl.2013.12.016 McCrory P.A., J.L. Blair, D.H. Oppenheimer, S.R. Walter (2006). Depth to the Juan de Fuca slab beneath the Cascadia subduction margin: A 3D model for sorting earthquakes. US Geol. Surv. Data Ser. DS, 91 118 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Science Nuggets Unravelling monogenetic volcanism in the Oregon Cascades Fiona Couperthwaite and Adam Kent We have studied tephra and scoria deposits from several weeks or several months. This indicates the olivine resided for volcanic cones amongst the numerous basaltic and some time in the sub-surface diffusively re-equilibrating before basaltic-andesite eruptions across the Central Oregon eruption, suggesting shallow storage beneath Belknap Crater. Cascades. This type of volcanism is under studied in the region. At Four in One Cone we studied material from the entire eruptive Gaining an understanding of the eruption timescales associated sequence, to study pre-eruptive processes at a higher resolution. with this type of volcanism is important in the event of future The earliest tephra layer contains weakly zoned (both normal and eruptions, due to their spatial and temporal distribution. We used reverse) or unzoned olivine, with core compositions of Fo83-82. Fe-Mg diffusion modeling techniques in olivine together with Later tephra layers contain olivine with mostly the same or similar crystal textures to calculate eruption timescales and to consider forsterite cores (Fo84-81) with the exception of one reversely the presence (or not) of a transport and storage network in the zoned olivine with a core composition of Fo73. Olivine within subsurface and how this may or may not differ between the various the later tephra layers exhibit stronger zoning patterns with rim cones. Our early work has focused on two volcanic centers within compositions down to Fo71 for normally zoned crystals. Together 9 km of each other - Four in One Cone, a ~2000 year old fissure with the weaker zoning, olivine within the initial tephra layer eruption made up of six small cones and Belknap Crater, a ~3000 are mostly euhedral, however olivine from the later tephra layers year old small shield volcano. exhibit dendritic overgrowths. The indicates the more evolved, The latest tephra from Belknap contains at least two olivine dendritic rims grew under conditions favourable for rapid crystal populations – a more dominant population with core compositions growth, potentially associated with degassing processes. Maximum of Fo81-80 and a second, smaller olivine population with core compositions of Fo84. Both olivine populations contain normally ■diffusion timescales range from days to months suggesting these zoned olivine that give maximum Fe-Mg diffusion timescales of magmas did not travel straight to the surface unimpeded. Four-in-One Cone, of the holocene monogenetic centers in Central Oregon studied for this project. Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 119 

Science Nuggets Physical properties of Cascadia incoming sediments Juan Pablo Canales, Shuoshuo Han, Suzanne M. Carbotte, Adrien Arnulf, Jian Zhu, Bridgit Boulahanis, and Mladen .R. Nedimović Seismic data from the Ridge2Trench (R2T) experiment The FWI-derived fine-scale structure of the incoming sediments combined with ocean drilling data indicate that offshore offshore central Oregon shows that a ~400-m-thick low-velocity Washington where the megathrust is inferred to be strongly interval initiates ~7 km seaward of the deformation front likely locked, over-consolidated sediments near the deformation front associated with anomalously high porosity that developed due to are incorporated into a strong outer wedge, with little sediment poor drainage beneath a thin layer of low permeability (Fig.  1). subducted (Han et al., 2017). These conditions are favorable for Further landward, décollement develops within this interval strain accumulation on the megathrust and potential earthquake with along-strike variations in depth of a few hundred meters. rupture close to the trench. In contrast, offshore central Oregon, In contrast, offshore Washington, we do not observe low velocity a thick under-consolidated sediment sequence is subducting, and intervals in the incoming sediment section near the deformation is probably associated with elevated pore fluid pressures on the front and the décollement is only ~200 m above the basement. megathrust in a region where reduced locking is inferred. These The average Vp/Vs structure of Cascadia basin sediments derived results suggest that the consolidation state of the sediments near from analyses of P-to-S converted waves (Fig. 2A) can be well the deformation front is a key factor contributing to megathrust described by a compaction trend decaying exponentially with slip behavior and its along-strike variation, and it may also have a depth. On a finer scale, Vp/Vs structure of incoming sediments near significant role in the deformation style of the accretionary wedge. the deformation front offshore northern Oregon and Washington Using advanced methodologies - full waveform inversion or FWI shows little variability along strike, while the structure of incoming and analyses of seismic reflection attributes and P-to-S converted sediments offshore central Oregon is more heterogeneous and waves - applied to the R2T multichannel seismic (MCS) and includes intermediate-to-deep sediment layers of anomalously ocean bottom seismometer (OBS) datasets, we map out and infer high Vp/Vs likely due to elevated pore pressures (Fig. 2). the physical properties of the incoming sediment section and the megathrust, and their along-strike variations. Figure 1. FWI velocity gradient model superimposed on pre-stack depth migrated image of the Oregon transect. LVZ: low velocity zone. (Han et al., 2019). 120 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Figure 2. Modified from Zhu et al. (2020). A. OBS receiver gather (radial) after downward continuation of sources to the seafloor. Yellow arrows PS reflection from the basement (PBasementS), and from the proto-décollement (PPdS); B. Contoured semblance Vp/Vs spectrum of OBS gather. Black crosses connected with solid line indicate picked PS arrivals with high semblance; C. Calculated interval Vp/Vs ratio as a function of P-wave two-way travel time; D. MCS section in the vicinity of the OBS showing sediment layering and oceanic basement, with calculated Vp/Vs function overlaid. We infer a step down in the proto-décollement from shallow depths References of ~1500 m to ~300 m above basement at 44°46'N from analyses of seismic attributes. Further north along the margin a horizon Han, S., N.L. Bangs, S.M. Carbotte, D. Saffer, J.C. Gibson (2017). interpreted as a mid-level detachment in the accretionary wedge Variations in sediment consolidation along the Cascadia margin and can be identified above a layer of distinct sediment properties that contribution to forearc deformation and megathrust slip behavior. may reflect diagenetic alteration. Nat Geosci, 10, 954-950, doi:10.1038/s41561-017-0007-2 Collectively, our results indicate that the incoming sediment Han, S., A. Arnulf, J.P. Canales, S.M. Carbotte, M. Nedimović (2019). section offshore Washington where landward vergence dominates Décollement initiation at Cascadia Subduction Zone from Full- is well drained and characterized by normal fluid pore pressures. Waveform Inversion, Seismol Res Lett, 90, 2B, 791-1069 In contrast, incoming sediments with the highest estimated fluid overpressures occur offshore central Oregon where deformation of Zhu, J., J. P. Canales, S. Han, S. M. Carbotte, A. Arnulf, M. R. Nedimović the accretionary prism is seaward vergent. Our results also suggest (2020). Vp/Vs ratio of incoming sediments off Cascadia subduction that the presence of a low permeability layer at the base of Astoria zone from analysis of controlled source multicomponent OBS records. Fan sediments with fluid overpressures below it, may play an J Geophys Res, 125, e2019JB019239, doi:10.1029/2019JB019239 ■important role in forming a shallow décollement offshore central Oregon. Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 121 

Science Nuggets Seismic characterization of the Juan de Fuca Plate from Ridge to Trench Suzanne M. Carbotte, Juan Pablo Canales, Shuoshuo Han, Greg Horning, Bridgit Boulahanis, and Mladen Nedimović Plate-scale seismic reflection images and seismic velocity 130˚W 128˚W 126˚W 124˚W 122˚W characterization of the sediments, crust, and shallowest 48˚N mantle along two transects crossing the Juan de Fuca Endeavor Washington Transect WASHINGTON 47˚N (JdF) plate and along a margin parallel transect seaward of the Cascadia deformation front are used to characterize the evolution N. Symmetric Along-Margin Transect and hydration of this young plate prior to subduction. Reflection images for the two-plate crossing transects (Han et al., 2016) Coaxial reveals faults in the sediment section beginning 55-70 km from the ridge axis with sparse reflections from presumed hydrated abyssal Axial 46˚N hill faults in the upper crust below. The lower crust is mostly acoustically transparent except for low-angle (20-40°) ridgeward Vance Oregon Transect OREGON dipping reflectors found in same age (6-8  Ma) crust along both transects, possible ductile shear zones in the lower crust due to 45˚N temporal variations in mantle upwelling. Near the deformation front (DF), plate bending due to sediment loading and subduction Cleft begins ~150 and 80 km from the DF offshore Oregon and Washington respectively with greater bending observed along the Blanco Transform Fault Explorer 52˚N Oregon transect. Bright fault plane reflections are imaged on the 48˚N Oregon transect that extend through the crust and 6-7 km into the mantle beginning ~40  km from the DF (Fig. 2A) and attributed 130˚W 128˚W 126˚W Juan de Fuca to hydrated fault zones due to subduction bend faulting. On the Washington transect, bend faults are confined to the sediment 44˚N section and upper crust and more limited plate hydration in this outer trench slope region is inferred. From tomographic analysis −5000 −4000 −3000 −2000 −1000 0 1000 2000 3000 Gorda 40˚N of seismic velocity structure along the plate transects, we infer Bathymetry (m) that the JdF plate acquires a mature structure within one million 132˚W 128˚W 124˚W year after accretion (Fig.2B, Horning et al., 2016; Boulahanis et al., 2020). This mature structure is characterized by a hydrated, porous Figure 1. Map of the seismic experiment. upper crust, and a lower crust and uppermost mantle with seismic velocities consistent with dry gabbro and peridotite, respectively. Along the margin transect, we find that the JdF lower crust and Propagator wakes are associated with velocity and hydration mantle are drier than at any other subducting plate, with most anomalies at lower crustal and upper mantle levels. Along both of the water stored in the sediments and upper crust (Canales et transects, three age intervals (1-3.4/4Ma, 3.5/4-6 Ma, > 6-8 Ma) of al., 2017). Variable but limited bend faulting along the margin distinct crust and upper mantle Vp structure, basement roughness, (Han et al., 2016; 2018) limits slab access to water, and a warm and spreading rate history are identified. These distinct periods are thermal structure resulting from a thick sediment cover and young attributed to differences in the mode of accretion at the paleo-JdF plate age prevents significant serpentinization of the mantle. The Ridge due to plate motion change at 6 Ma and a period of possible implications from these findings for subduction processes are enhanced flux to the ridge from the Cobb hotspot at ~ 4Ma. that fluids that facilitate episodic tremor and slip downdip from the seismogenic zone must be sourced from the subducted upper ■crust,and that decompression rather than hydrous melting must dominate arc magmatism in central Cascadia. 122 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

TWTT (ms)A B Distance from deformation frontDepth (km) Vp (km/s) Distance from deformation front Figure 2. A. Interpretation of Oregon plate transect showing faults within the sediment section (brown), crust and uppermost mantle (purple), ridgeward dipping lower crustal reflections (orange), and other sparse events in crust and near Moho (blue) (from Han et al. submitted); B. Seismic Tomography Model for Oregon transect (Horning et al., in prep). Propagator wakes are indicated with black horizontal bars in (A) and gray shading in (B). References Boulahanis, B., S.M. Carbotte, J.P. Canales, S. Han, M.R. Nedimović (2020). Structure and evolution of Juan de Fuca crust and uppermost mantle over the last 8 Ma from an active source seismic tomography study, for submission J Geophys Res Canales, J.P., S.M. Carbotte, M.R. Nedimovic, H. Carton (2017). Dry Juan de Fuca slab revealed by quantification of water entering Cascadia subduction zone. Nat Geosci, 10, 864-870 Han, S., S.M. Carbotte, J.P. Canales, M. Nedimovic, H. Carton (2018). Along-trench structural variations of the subducting Juan de Fuca Plate from multichannel seismic reflection imaging, J Geophys Res, 123,4, 3122-3146 Han, S., S.M. Carbotte, J.P. Canales, M.R. Nedimović, H, Carton, J.C. Gibson, G.W. Horning (2016). Seismic reflection imaging of the Juan de Fuca plate from ridge to trench: new constraints on the distribution of faulting and evolution of the crust prior to subduction. J Geophys Res, 121, 1849–1872, doi:10.1002/2015JB012416 Horning, G., J.P. Canales, S.M. Carbotte, S. Han, H. Carton, M.R. Nedimović, P.E. van Keken (2016). A 2‐D tomographic model of the Juan de Fuca plate from accretion at axial seamount to subduction at the Cascadia margin from an active source ocean bottom seismometer survey. J Geophys Res, 121, 8, 5859-5879 T C S Zhe ascadia ubduction one primary site offered outstanding opportunities, in particular leveraging onshore and offshore infrastructure associated with EarthScope’s Plate Boundary Observatory (PBO), and deployment of the EarthScope Amphibious Array, all part of the ARRA-funded Cascadia Initiative. Work in this region built upon a broad spectrum of geological and geophyscial data collected over the past decades and engaged a range of US, Canadian, and international scientists. Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 123 

Science Nuggets Constraining the temperature conditions of paleo-subduction plate interfaces Kayleigh M. Harvey, Xin Zhou, Besim Dragovic, Sarah Penniston-Dorland, Ikuko Wada, and Peter van Keken petrological studies. The effects of uncertainties in these parameter values on the temperature estimates are also evaluated. As there Pressure-temperature (P-T) estimates from exhumed are no constraints on the geometry of the subducting slab, we metamorphic rocks are often used to constrain the thermal construct a range of generic model geometries and quantify the conditions of paleo-subduction zone plate interfaces. effect of slab geometry on the subduction interface temperature. However,theexhumedrockrecordonaverageindicatestemperatures For the petrology we use quartz-in-garnet and zircon-in-garnet 200-300°C warmer than those predicted by geodynamic models elastic thermobarometry combined with Zr-in-rutile thermometry for modern subduction zones. To elucidate the difference in to evaluate the P-T history of a block within mélange zone of the paleo and modern subduction zone thermal structures, we the Rio San Juan Complex in order to test assumptions about compare newly acquired P-T estimates from petrologic data to chemical equilibrium commonly made to reconstruct P-T paths. newly constructed geodynamic models of the regional tectonics Previous work suggests that this subduction complex is a warm at selected paleo-subduction localities, including the Franciscan end-member relative to other exhumed terranes. Mélange zones Complex in California, the Raspas Complex in Ecuador, the Rio within the Rio San Juan Complex preserve a near-continuous San Juan Complex in the Dominican Republic, the Sanbagawa Belt record of its P-T history during subduction and exhumation. We in Japan, and the Pam Peninsula in New Caledonia. For this nugget compare the model-predicted subduction thermal structures with we focus on the Rio San Juan Complex in the Dominican Republic. the P-T conditions that are estimated from exhumed rocks in the For the geodynamic models we develop 2-D coupled kinematic- dynamic models, using the paleo-subduction parameters, such ■selected localities and assess the key factors that contributed to the as convergence velocity and plate age, that are constrained by global plate reconstruction models and regional geological and petrologically constrained P-T conditions. Results to date • Sample 25-228 followed a counter-clockwise P-T path consistent with heating during nascent subduction • Validated by petrologic observations, phase equilibria modeling and trace element and elastic thermobarometry • P-T conditions agree with 2D coupled kinematic-dynamic models for early subduction • Elastic thermobarometry methods are still being refined to consistently reproduce results from other methods • Working to reduce uncertainty • May help elucidate parts of the P-T path where phases did not reach chemical equilibrium • Combining trace element thermometry, elastic thermobarometry and pseudosection modeling shows promising results for predicting P-T paths of paleo-subduction rocks • This approach is broadly applicable to a variety of lithologies and metamorphic grades • 2D coupled kinematic-dynamic models for Rio San Juan can reproduce the P-T estimates from the rocks without invoking frictional heating References Angel, R.J., M. Alvaro, R. Miletich, F. Nestola (2017). A simple and generalised P–T–V EoS for continuous phase transitions, implemented in EosFit and applied to quartz. Contrib Mineral Petrol, 172, 29, doi.org/10.1007/s00410-017-1349-x Kohn, M.J. (2020). A refined zirconium-in-rutile thermometer. Am Mineral, 105, 6, 963-971, doi.org/10.2138/am-2020-7091 Krebs, M., H.-P. Schertl, W.V. Maresch, G.Draper (2011). Mass flow in serpentinite-hosted subduction channels: P–T–t path patterns of metamorphic blocks in the Rio San Juan mélange (Dominican Republic). Jour Asian Sci, 42, 4, 569-595, doi.org/10.1016/j.jseaes.2011.01.011 Milani, S., F. Nestola, M.Alvaro, D. Pasqual, M.L. Mazzucchelli, M.C. Domeneghetti, C.A. Geiger (2015). Diamond–garnet geobarometry: The role of garnet compressibility and expansivity. Lithos, 227, 140-147, doi.org/10.1016/j.lithos.2015.03.017 124 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Petrologic Data D 2.5 A Peak Conditions 2.5 B Core 2.0 A Pressure (GPa) 2.0 Zr-in-rutile 1.5 1.5 Quartz-in-garnet Pressure (GPa) 1.0 Zircon-in-garnet 1.0 0.5 500 600 700 B 400 0.0 0.5 350 Temperature (°C) Core Mantle 1 Mantle 2 Rim Matrix 400 0.0 Mantle (All) 800 400 500 600 700 800 Mn 300 Peak Conditions Temperature (°C) Zr (ppm) 250 Mantle Rim 200 2.5 2.5 C D 150 2.0 2.0 100 Pressure (GPa) Pressure (GPa) 50 1.5 1.5 0 20C0a 0 3000 4000 5000 1.0 1.0 0 1000 Distance (μm) C 1000 µm Q7 0.5 0.5 Z10 0.0 0.0 800 400 400 500 600 700 500 600 700 800 Temperature (°C) Temperature (°C) Z7 Q6 M2 Q4 Q5 C M1 Amp + R Figure 1. A. Ca-Amph-bearing eclogite mélange block (sample 25-228) with Amp + Ph + Ep Amp + Ph + Omp + Ph blueschist-facies overprint suggesting counter-clockwise P-T path (Krebs et + Rt + Ttn + Qz Ep + Rt + Qz + Ep + Rt + Amp + Ph + Chl + Rt + al., 2011); B. Concentration of Zr in rutile for both rutile inclusions in garnet + Zrc + Ap Qz + Zrc + Qz + Zrc + Ap and matrix rutile; C. Sketch of garnet showing mineral assemblage and + Zrc + Ap location of rutile, quartz and zircon inclusions. Rutile inclusions are colored 280 by Zr concentration. Zones are drawn based on major element zoning in Q8 Z1 Ap 260 garnet; D. P-T estimates for each zone of the garnet using Zr-in-rutile (red) 240 thermometry paired with quartz-in-garnet (blue) and zircon-in-garnet Q3 Zr-in-rutile (ppm) 220 (yellow) elastic barometry. Isomekes calculated using the equations of Z4 200 state for quartz (Angel et al., 2017), zircon (Ehlers et al., in prep), and Z3 180 pyrope (Milani et al., 2015). Zr-in-rutile isopleths (red) calculated using the 160 combined experimental-empirical thermometer by Kohn (2020). Zircon inclusion Z11 140 Quartz inclusion 120 * Mineral assemblages in parentheses Z16 100 as reported by Krebs et al. (2011) 80 Matrix: Amp + Omp + Czo + Chl + Na-Amp + Ph + Rt + Ttn + Qz + Zrc + Ap + Pl B0 Thermal Models A 120oW 80oW 40oN 40oW 0o120oW 80oW 40oW 0o120oW 80oW 40oW 0o 200 oC 600 oC 20oN 0o Depth (km) 100 1000 oC Study 1400 oC 20oS area 200 (a) 120 Ma 136 Ma 60 Ma 5 Ma slab age 600 oC 120 Ma 0 200 oC Depth (km) C 0 70 140 210 280 4 Age of The Oceanic Lithosphere [Myr] 3 100 1000 oC 2 Figure 2. A. Subduction initiates around 136 Ma at Rio 1 200 (b) 100 Ma 1400 oC 0 60 Ma 25-228 (Krebs et al., 2011) San Juan. Both the subducting and overriding plates are 25 Ma slab age 500 0 80 Ma 25-228 (this study) relatively young; B. Time-dependent models at time (a) 100 Ma 120 Ma, (b) 100 Ma, and (c) 60 Ma for Rio San Juan; 120 Ma 0 200 oC Pressure (GPa) v = 3 cm/yr, R = 600 km. The slab ages as subduction proceeds; C. Slab surface P-T conditions predicted by the 600 oC model in Fig. 11 (solid lines) and another time-dependent Depth (km) model that assumes a 10 Ma slab at 120 Ma (dashed 100 1000 oC lines). Peak P-T estimate shown along with P-T estimates 1400 oC from Krebs et al. (2011) (red dots). The subduction (c) 60 Ma interface is relatively hot given the relatively young age 65 Ma slab age 200 of the subducting plate, and the petrologic data can be explained if the slab age is ~5 Ma without invoking 0 100 200 300 400 200 400 600 800 1000 frictional heating (µ’ = 0). Distance (km) Temperature (oC) Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 125 

Science Nuggets Mélange-peridotite interactions in the source of arc magmas Emmanuel Codillo, Veronique Le Roux, Mark Behn, Gray Bebout, Glenn Gaetani, and Horst Marschall The mechanisms of material transfer from the slab to the We performed a series of piston-cylinder experiments on natural overlying mantle in subduction zones are still highly mélange rocks from several exhumed high-pressure terranes (e.g., debated. Whether material transfer primarily occurs during Syros, Greece, Santa Catalina, USA) that cover the endmember partial melting, fluid percolation, or detachment of solid diapirs compositions of global mélanges over a wide range of pressure has critical implications for the timing of elemental fractionation and temperature conditions. In particular, we aim to constrain the observed in arc magmas, the composition of the overlying arc initiation temperature of mélange melting (solidus) which allows crust, and the geodynamic processes in subduction zones. us to assess the likelihood of mélange melting along the slab-tops It has been postulated that mélange rocks, which are physical at different subduction zones (Fig. 1B). We then determine the mixtures of sediments, oceanic crust, ultramafic rocks formed elemental compositions of melts and mineral residues for various from the mechanical and metasomatic interactions along the bulk starting compositions, and quantify the density evolution of slab-mantle interface, could play a key role in arc magmatism. mélange rocks along the slab-top (Fig. 1C). This information will Field observations of exhumed high-pressure mélange rocks often be combined with numerical modeling to determine the conditions display blocks of crustal rocks embedded in mafic to ultramafic necessary for mélange diapirs to rise from the slab with respect to matrices. Importantly, mélange rocks show compositional the onset of mélange melting. Critical information on the effects of similarities with global arc lavas in terms of trace elements and melting and instantaneous melt extraction on the density evolution isotopes, implying that some characteristic slab signatures of arc of the melting residues will be fed into numerical model calculations lavas may already be imprinted during the formation of mélange. to ensure that the dynamics are consistent with the compositional However, the fate of mélange rocks along the descending slab is evolution of the mélange diapirs. Preliminary results from this study still rather unconstrained. Once formed, these mélange rocks show that melting of mélange rocks along the slab-top is unlikely may either melt along the slab-top, or rise as solid or partially at low pressures (<2.5  GPa) even along a hot subducting slab. In molten diapirs (Fig. 1A). If favored, these diapirs could effectively addition, the intrinsic buoyancy of most mélange rocks relative to deliver the compositional signatures of mélange rocks into the the overlying mantle may promote the ascent of mélange diapirs source of arc magmas. The goal of our project is to investigate the and subsequent interactions with the overlying mantle wedge. phase equilibria, melting properties, and densities of a variety of Using these experiments, we extract critical insights on the physical mélange rocks in order to constrain their physical behavior during subduction and their influence on arc magmatism. ■behavior of mélange rocks during subduction and the geochemical consequences of mélange melting. References Lambert, I.B., P.J. Wyllie (1970). Melting in the deep crust and upper mantle and the nature of the low velocity layer. Phys Earth Planet Int, 3,316-322 Liu, J., S. Bohlen, W. Ernst (1996). Stability of hydrous phases in subducting oceanic crust. Earth Planet Sci Lett, 143, 161-171 Nichols, G.T., P.J. Wyllie, C.R. Stern (1994). Subduction zone melting of pelagic sediments constrained by melting experiments. Nature, 371, 785-788 Skora, S., J. Blundy (2010). High-pressure hydrous phase relations of radiolarian clay and implications for the involvement of subducted sediment in arc magmatism. J Petrol, 51, 2211–2243 Syracuse, E.M., P.E. van Keken, G.A. Abers (2010). The global range of subduction zone thermal models. Phys Earth Planet Inter, 183, 73-90 Till, C.B., T.L. Grove, A.C. Withers (2012). The beginnings of hydrous mantle wedge melting. Contrib Mineral Petrol, 163, 669-688 126 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Figure 1. A. Schematic illustration of three possible end-member scenarios for the fate of mélange rocks along the slab-mantle interface; B. Experimentally-derived solidi of different mélange rocks plotted along with other solidi of different bulk compositions relevant to subduction zone melting (Literature data: Lambert and Wyllie, 1970; Liu et al., 1996; Nichols et al., 1994; Skora and Blundy, 2010; Till et al., 2012), and representative global slab-top P-T paths (Syracuse et al., 2010); C. Calculated bulk density difference between mélange rocks and overlying peridotite mantle at similar P-T conditions. Circle and square symbols represent experiments at 1.5 and 2.5 GPa, respectively. Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 127 

Science Nuggets Uplift and exhumation history of the Central Aleutian Arc Emily H. G. Cooperdock, Claire Bucholz, and Anahi Carrera Acentral question surrounding island arcs is whether This work will place the exhumation history of the Central Aleutians processes that occur deeply in the subduction zone affect in the context of 1) driving forces involving regional tectonics and processes that are observed on the surface, such as uplift, subduction zone processes; and 2) geochemical inputs to the N. exhumation, and erosion. In many locations, the arc exhumation Pacific via erosion of arc material. In order to quantify when and history is difficult to study because it may be overprinted by later by how much the Aleutians experienced surface uplift and erosion, thermal or physical processes. The Aleutian Arc is unique in that it plutonic samples from about ten islands that span 870 miles of arc is includes over fifty islands composed of both active and inactive length will be analyzed for emplacement depth and subsequent volcanoes over 1,900 miles from Alaska to Russia, and has limited uplift rates (including samples collected on the 2015 GeoPRISMS overprinting by secondary events. In addition, it has a volcanic sampling campaign). Samples will be characterized by crystallization history that spans fifty million years, so that the ancient magma age, emplacement depth, and two or more thermochronometers chambers of past volcanoes are now exposed on the islands' with different thermal sensitivities (e.g., apatite and/or zircon, surfaces today. These plutonic rocks hold a record of past island (U-Th)/He and/or fission track). Thermochronology techniques, uplift, exhumation, and erosion that can address key questions particularly the use of multiple chronometers with different regarding: When were the Aleutians uplifted? When and how temperature sensitivities within the same sample, can constrain much has been eroded through time? Was uplift and subsequent exhumation rates and be used to estimate erosion rates over time. erosion constant or cyclical? How does it vary geographically along Coupling the thermochronological data with emplacement depth the length of the arc? Specifically, the answers to these questions will help constrain the amount of material eroded over that time. will shed light on the link between plate-scale processes and uplift; The timing and geographic trends revealed by these data can then the timing and magnitude of erosion, contributing sediment and be related in time and space to previously proposed drivers for geochemical inputs to North Pacific; and the relationship between plutonic exhumation rates, geochemistry, and emplacement depth. ■uplift, including plate rotation, change in convergence angle and The Aleutians have functioned as a natural laboratory to test fundamental principles regarding subduction zone evolution for rate, or the development of an accretionary prism. more than four decades. This study will make a novel contribution to the existing extensive geochemical and geodynamic research. Reference Geist, D.J., J.D. Myers, C.D. Frost (1988). Megacryst-bulk rock isotopic disequilibrium as an indicator of contamination processes: The Edgecumbe Volcanic Field, SE Alaska. Contrib Mineral Petrol, 99, 105-112, doi.org/10.1007/BF00399370 128 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Bering Sea Bowers Ridge 6 Umnak Unalaska 5 Attu 1 Adak Atka 2 Kanaga KiskaAmchitka 4 Kagalaska 3 Ilak Amatignak Paci c Ocean Map of the Central Aleutian Arc (180° to 165°W). The research area encompasses labelled islands. Plate trench shown with dashed line and convergence vectors in black arrows. Proposed rotated tectonic blocks are outlined (from left to right): 1) Near Block, 2) Buldir Block, 3) Rat Block, 4) Delarof Block, 5) Andreanof Block, 6) Cold Bay Segment. (Blocks from Geist et al., 1988 and Kay et al., 1992). Sampling on Unalaska Island during the 2015 GeoPRISMS field campaign. Photo credit: E.H.G. Cooperdock Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 129 

Science Nuggets Tracking carbon from subduction to outgassing along the Aleutian Subduction Zone Taryn Lopez, Tobias Fischer, Terry Plank, Alberto Malinverno, Andrea Rizzo, Daniel Rasmussen, Elizabeth Cottrell, Cynthia Werner, Christoph Kern, Deborah Bergfeld, Tehnuka Ilanko, Janine Andrys, and Katherine Kelley Volatile cycling among Earth’s mantle, crust, and surface can be explained in part by differences in sediment subduction reservoirs is an important process for magma generation, vs. accretion, likely contributions of crustal C to EA volcanic volcanism, and the long-term evolution and habitability emissions, and potentially greater slab AOC devolatilization in of Earth. Volatile migration among Earth’s reservoirs can be the EA and WA segments near slab edges. When volatile source estimated by mass balance, using a volatile tracer such as carbon proportions are averaged over the full arc, we find nearly equal (C), whose chemical composition and flux can be measured or proportions of volcanic C are supplied from mantle (~32%), AOC estimated both for the various source inputs and volcanic outputs. (~30%), and sediment (~38%) sources. When combined with This project aimed to use new and existing measurements of published estimates of an Aleutian Arc volatile flux, we find that volcanic gases from remote volcanoes within the Alaska-Aleutian on an arc-wide scale only ~15% of trench sediments are recycled arc, along with constraints on the composition and thickness of back to the atmosphere through volcanism. This may support sediments presumed to be subducted into the Aleutian trench, to previous studies of arc systems that indicate that the majority of compare subduction inputs with volcanic outputs and ultimately C found in trench sediments is accreted to the overriding plate, characterize volatile cycling within this region. The composition subducted to the deep mantle, and/or stored in the overriding of volcanic gas outputs and sediment inputs were used with two crust. Alternatively, this may reflect the previous subduction cycle mixing models to constrain the proportions of C supplied from before the Pliocene glaciation, when sedimentation rates were the three main subduction-zone C sources: subducted (altered) lower. This lower trench sedimentation input may have manifested oceanic crust (AOC), subducted sediment, and mantle. These as low sediment signals in present-day volcanic C output. These results were then combined with published constraints on the findings show that C contribution to the atmosphere from the C flux from Aleutian volcanoes to estimate the amount of C Aleutians includes lower proportions of subducted sediment and supplied to volcanic outputs from each source within the Aleutian crustal C, and higher proportions of AOC and mantle C, relative subduction zone. to other arcs. The dominance of AOC suggests that it may be a Through this work we find that C source proportions vary more globally significant input to atmospheric C than previously significantly along the length of the Aleutian Arc, with Western thought, especially in arcs where crustal C sources are minor. Aleutian (WA) volcanoes having primarily mantle and AOC derived These combined findings have implications for the global C budget C, Central Aleutian (CA) volcanoes having primarily subducted sediment derived C, and Eastern Aleutian (EA) volcanoes having ■over Earth’s history, and in turn the evolution of Earth’s climate and variable C inputs from all three sources. These along-arc variations suitability for life on Earth. Reference Sano, Y., B. Marty (1995). Origin of carbon in fumarolic gas from island arcs. Chem Geol, 119, 265-274 130 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

AB Figure 1. A. C-He three component mixing model results for the Aleutian Arc, following the methods of Sano and Marty (1995). This figure shows the volcanic gas compositions relative to those of end-member C sources of bulk sediment (S), carbonate (C) and Mid Ocean Ridge Basalt (MORB; M), used here as a proxy for upper mantle. Volcanoes within the different arc segments are colored, where blue, green, and red correspond with the eastern, central, and western Aleutian Arc segments, respectively. The M-S black line shows the mantle- sediment mixing line for the arc-minimum bulk sediment δ13C value (-18.7), while the blue line shows the minimum bulk sediment δ13C value (-12.6) seen in the Central Aleutians; B. Normalized mean proportions of sediment, carbonate and mantle C sources based on calculation results shown at left for Aleutian Arc volcanoes from West to East: Little Sitkin (LS), Kanaga (KAN), Okmok (OK), Makushin (MAK), Akutan (AK), Mageik (MGK), Griggs (GR), Trident (TRI) and Augustine (AUG). Mixing proportions were calculated using the minimum bulk sediment δ13C value observed for each arc segment (-18.4 for EA, -12.6 for CA, and -18.7 for WA). The Alaska & Aleutian primary site defines the most tectonically active region in North America. It is the ideal location to study arc magmatism, structure and the contributions of arc volcanism to the development of continental crust. The Alaska and Aleutian subduction zone is also ideal for the study of earthquake processes and seismic cycle. The Alaska and Aleutian subduction zone offered important opportunities to leverage onshore and offshore infrastructure associated with the EarthScope's Plate Boundary Observatory and the deployment of the US Transportable Array. GeoPRISMS investigations in Alaska faced logistical challenges due to remote locations of field work and required significant advance planning. Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 131 

Science Nuggets Volcanic seismicity beneath Chuginadak Island, Alaska (Cleveland and Tana Volcanoes): Implications for magma dynamics and eruption forecasting John Power, Diana Roman, John Lyons, Matthew M. Haney, Dan Rasmussen, Terry Plank, Kirsten Nicolaysen, Pavel Izbekov, Cynthia Werner, and Max Kaufman Cleveland and Tana are remote volcanoes located in the beneath the active crater of Mount Cleveland and almost all of the central Aleutian volcanic arc on the eastern end of the explosions occur without identifiable short-term (hours to days) Islands of Four Mountains. The persistently active Mount seismic precursors. VT earthquakes beneath Mount Cleveland Cleveland volcano, located on the western side of Chuginadak occur at depths of 2 to 8 km BSL and range in magnitude from -0.2 Island, is surrounded by several closley spaced Quaternary volcanic to 1.8. VT focal mechanisms indicate have horizontal P-axes that centers including Carlisle, Herbert, Kagamil, Tana, and Uliaga, and align with the regional axis of maximum stress. These observations, numerous small satellite vents on Chiginadak, between Cleveland and a relatively slow one-dimentional seismic velocity model, and Tana (Fig. 1). The Alaska Volcano Observatory (AVO) suggest are consistent with a shallow body of magma that is installed two permanent broadband seismometers on Chuginadak fed through a deeper conduit system. The time-history of VT Island near Cleveland in 2014, and we operated a temporary earthquakes and shallow LP events suggest their occurrence may broadband network on the western side of the island in 2015-2016. track the transfer of magma and fluids from the mid-crust to the Collectively, these stations provided the first seismic observations shallow portions of the conduit system and may provide a means of this frequently active volcano and the surounding Holocene- to anticipate future explosions and periods of dome growth. VT aged volcanic vents. During the study period (July 2014-January hypocenters also occur ~7 km to Cleveland's northeast at depths 2019), eruptive activity at Cleveland was characterized by small of 5 to 10 km BSL, below a group of Holocene-aged vents between explosions separated by periods of lava effusion that fomed small Mount Cleveland and Tana. These earthquakes have vertically- domes in the volcano's summit crater. We characterize seismicity oriented P-axes and a greater percentage occur in families. These beneath Chuginadak Island through automated analysis of event observations, combined with observations of vent orientation and waveform frequency content, development of a one-dimensional morphology and gas flux, suggest the area between Cleveland and P-wave velocity model, calculation of earthquake hypocenters, Tana represents a zone of complicated volcano-tectonic interaction, magnitudes, focal mechanisms, and identification of earthquake similar to calderas elsewhere in the Aleutian arc. The presence of families. This analysis reveals the full range of seismic event types a larger volcanic system in the IFM could influence magmatism expected in a highly active volcanic environment and includes Volcano-Tectonic (VT) earthquakes, Long-Period (LP) events, ■and account for the multiple closely-spaced volcanic centers in this and explosion signals. LP events appear to cluster at shallow depth region. References Haney, M., J.J. Lyons, J.A. Power, D.C. Roman (2019). Moment tensors of small vulcanian explosions at Mount Cleveland, V41A–V05A, Alaska: American Geophysical Union, San Fransisco. Iezzi, G., G. Lanzafame, L. Mancini, H. Behrens, S. Tamburrino, M. Vallefuoco, S. Passaro, P. Signanini, G. Ventura (2020). Deep sea explosive eruptions may be not so different from subaerial eruptions. Scientific Reports, 10, 6709, doi.org/10.1038/s41598-020-63737-7 Janiszewski, H. A., L.S. Wagner, D.C. Roman (2020). Aseismic mid-crustal magma reservoir at Cleveland Volcano imaged through novel receiver function analyses. Scientific Reports. 10, 1, doi.org/10.1038/s41598-020-58589-0C Werner, C., D.J. Rasmussen, T. Plank, P.J. Kelly, C. Kern, T. Lopez, J. Gliss, J.A. Power, D.C. Roman, P. Izbekov, J. Lyons (2020). Linking subsurface to surface using gas emission and melt inclusion data at Mount Cleveland volcano, Alaska. Geochem, Geophys, doi.org/10.1029/2019GC008882 132 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Figure 2 (right). Sketch of the major components of the Chuginadak Island magmatic system defined by observed seismic activity after the conceptual model proposed by Werner et al. (2020). The principal seismic components of this model include the inferred source of LP seismicity likely within the Cleveland cone, the VLP explosion source zone at 300 - 600 m ASL (Haney et al., 2019), a modeled seismo-acoustic source zone that extends from the summit to sea level (Iezzi et al., 2020), VT earthquake hypocenters that range in depth from 2 to 8 km BSL, The zone of VT hypocenters under the Isthmus cones that range from 5 to 10 km depth BSL, and the deeper low velocity anomaly from 10 to 20 km depth identified by seismic receiver functions (Janiszewski et al., 2020). Figure 1. (bottom) Location of Mount Cleveland volcano, ~70 km west of the settlement of Nikolski, Alaska, in the Central Aleutian arc. Volcanoes are shown with triangles and settlements with plus symbols. Modified after Werner et al. (2020). Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 133 

Science Nuggets From the slab to the surface: Origin, storage, ascent, and eruption of volatile-bearing magmas Terry Plank, Daniel Rasmussen, Diana Roman, and John Power Does the volcano know about the slab? Our work in the Our own analysis of seismic S-P times for earthquakes measured central-eastern Aleutian was aimed at this question using local stations confirms the variation in slab depth published (Fig.  1). Spanning from Seguam volcano (west) to previously using global catalogs. We investigate new major, trace, Shishaldin volcano (east), our corridor is marked by significant and volatile elements in melt inclusions and bulk rock samples variations in magmatic water contents, seismicity, deformation, from eight volcanoes, which exhibit systematic trends with slab and style and frequency of eruption. By contrast, most subduction depth (Fig. 2). Maximum water contents of inclusions vary from parameters, such as slab age and velocity, remain constant. One ~2 wt.% (Fisher) to ~5 wt. % (Akutan), spanning most of the global significant exception is the depth of the slab below the frontal arc range (1-7 wt.%). Correlations between H2O/Ce and H2O/K2O and volcanoes, which transitions from a near global minimum in the non-volatile trace element ratios (e.g., Nb/Ce, Ba/La) give strong west (~65 km below sea level (BSL) to a more typical depth in the evidence that maximum water content melt inclusions preserve east (~100 km BSL). This makes our corridor an ideal locality to undegassed concentrations. Variation in water contents is reflected isolate the role of slab depth in driving magmatic processes. After in calc-alkaline vs. tholeiitic differentiation, consistent with earlier a one-year-long seismic deployment, forty five-gallon buckets of work. We find that all of these geochemical ratios, including those new rock samples, two stints aboard the R/V Maritime Maid, and involving H2O, are negatively correlated with slab depth. one PhD dissertation, we arrived at some answers. 55°N Rock samples Shishaldin Fisher 54°N Westdahl Akutan Bogoslof Makushin 53°N Okmok Vsevidof Cleveland Seguam 52°N Alaska 51°N 100 km 172°W 170°W 168°W 166°W 164°W 162°W Figure 1. The central-eastern Aleutian arc. Map of the field area with historically active volcanic centers labeled. 134 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

joint GeoPRISMS, Alaska Volcano Observatory, and Deep Carbon Observatory field party in 2016 onboard the R/V Maritime Maid. 60 300 ΔT (K) 50 60 150 100 70 a Cleveland b 70 Seguam Slab depth (km) Okmok 80 80 Akutan 90 Makushin 90 100 Westdahl 4 100 110 Fisher Shishaldin 110 0 5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 1 23 H2O/Ce * 10-3 Dy/Yb Figure 2. Variation in volatile and trace element compositions of magmas with slab depth (Rasmussen et al., 2019). (a) H2O/Ce is a proxy for slab temperature. Temperature relative to the wet sediment solidus (ΔT) is calculated using the thermometer of Plank et al. (2009). (b) Increased Dy/ Yb may indicate an increased role of garnet with depth in the slab. Moreover, the geochemical variations are consistent with higher These results indicate that slab depth controls the composition, slab temperature and mantle melting pressure for those volcanoes differentiation and volatile content of arc magmas. We speculate that overlie greater depths to the slab, consistent with nominally that the range in slab depths in the central-eastern Aleutians relates vertical transport paths from slab to the surface. Slab-mantle coupling depths must be shallower (~ 50 km) in the western part ■to the thickness of trenchfill sediments, which decreases from west of this corridor than is typically inferred worldwide (~80 km). (Seguam) to east (Shishaldin). References Plank, T., L.B. Cooper, C.E. Manning (2009). Emerging geothermometers for estimating slab surface temperatures. Nat Geosci, 2, 9, 611 Rasmussen D.J., T.A. Plank, D.C. Roman (2019). Magmatic water content controls magma storage depth. AGU Fall Meeting 2019, abstract V21A-06 Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 135 

Science Nuggets Water and oxygen fugacity controls on continental signatures in western Aleutian arc magmas Janine Andrys, Katherine A. Kelley, Laura Waters, Elizabeth Cottrell, Michelle Coombs, and Matthew Jackson Lavas of varying calc-alkaline affinity, from strongly calc- new constraints on the AfOle2utainand aHrc2O, ancdontteesntststhoef variably calc- alkaline to mildly tholeiitic, erupt along the western Aleutian alkaline magmas in the links between arc, making it an ideal natural laboratory for constraining ftOhe2, magmatic Hpl2aOte, ,maangdmtahteicsdlaibffearnendtiwateiodnge, contributions from the petrogenesis of these magma types. Our team collected tephra subducted thermal structure, and lava samples from Buldir (184.1°W), Segula (181.8°W), with the goal of resolving the key factors that trigger calc-alkaline Semisopochnoi (180.7°W), Gareloi (178.8°W) and Tanaga magmatic trends and the production of continental crust at (178°W) Islands during the 2015 field season on leg 3 of NSF subduction zones. Our findings indicate a gradient of increasing GeoPRISMS shared platform for Aleutians research. We measured amrca,gmwhatiicchHw2eOlicnokntteonctshatonwgeasrdins the western end of the Aleutian dissolved volatiles, Fe3+/ΣFe ratios, and major and trace elements of the slab thermal structure, and melt inclusions from these volcanoes in tandem with petrological ■further resolve a key role for fO2 in controlling the differentiation experiments at controlled H2O and fO2. Our work provides critical trends of arc magmas. 8 Plot of maximum dissolved H2O content measured in olivine- hosted melt inclusions from tephras collected at five western Max H2O (wt%) 7 Segula Tanaga Aleutian volcanoes, as a function of distance along strike. Results 6 Buldir Semisopochnoi Gareloi show a westward increase in magmatic H2O content. 5 4 Photo below. Liz Cottrell and Katherine Sheppard collecting 3 exposed tephra on Buldir Island during the 2015 NSF GeoPRISMS 2 Aleutians field campaign. Photo credit: M. Coombs 1 0 184 183 182 181 180 179 178 Longitude oW 136 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Science Nuggets Carbon in the mantle lithosphere Meghan R. Guild Carbon is one of the most important elements on Earth, In this study, this significant gap in knowledge will be addressed as it is the basis of all life, it fuels human civilization, by determining the amount of carbon stored in rocks from and modulates climate. On Earth's surface, carbon is Earth's interior, sampled by volcanoes, namely xenoliths. Through exchanged between the atmosphere, terrestrial biosphere, oceans, the analysis of carbonates and fluid inclusions hosted in those and sediments. The timescales and amount of carbon exchanged xenoliths, the abundance, sources, and forms of carbon stored on Earth's surface are well characterized. However, the carbon in the mantle lithosphere using carbon and oxygen isotopes, on Earth's surface makes up only ~1% of Earth's total carbon determining the entrapment temperature and composition of budget. The vast majority of Earth's carbon (~99%) is found in fluid inclusions, and investigating aqueous carbon speciation at its interior (mantle and core), making it much more challenging entrapment temperatures by theoretical thermodynamic modeling to investigate. To better understand the long-term contribution of of fluid inclusion chemistry at lithospheric mantle conditions. This Earth's interior carbon to climate and life through geologic time, study will provide much needed constraints on the amount and our understanding of Earth's deep carbon must be improved. ■sources of carbon in the mantle, specifically targeting samples from the southwestern United States. a. A’ b. inactive Colorado Plateau Colorado Mojave arc New Mexico N Nevada Utah continental crust lithospheric Arizona lithaoltseprehderoiccemaannmctrleuasnttle California Navajo Volcanic Field A A A’ Figure 1. Geologic setting for the proposed study. (a) Map view of the southwestern U.S. with state-lines. Blue field represents the extent of the Navajo Volcanic Field. Dotted line A-A' shows the location of the cross section represented in b. (b) Cross section of Farallon flat-slab subduction zone during Eocene based on Humphreys et al. (2003). Gradient represents dehydration of serpentine and percolation of fluids upwards to hydrate and metasomatize the Colorado Plateau. After Marshall et al. (2017). References Humphreys, E., E.Hessler, K. Dueker, G.L. Farmer, E. Erslev, T. Atwater (2003). How Laramide-age hydration of North American lithosphere by the Farallon Slab controlled subsequent activity in the western United States, Int Geol Rev, 45, 575-595, doi.org/10.2747/0020-6814.45.7.575 Marshall, E.D., J.D. Barnes, J.C. Lassiter (2017). The role of serpentinite-derived fluids in metasomatism of the Colorado Plateau (USA) lithospheric mantle. Geology, 45, 12, 1103-1106, doi.org/10.1130/G39444.1 Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 137 

Science Nuggets Exploring the Alaska Peninsula subduction zone with AACSE: The Alaska Amphibious Community Seismic Experiment Grace Barcheck, Geoffrey A. Abers, Aubreya N. Adams, Anne Bécel, Peter J. Haeussler, Emily Roland, Natalia Ruppert, Susan Y. Schwartz, Anne F. Sheehan, Donna J. Shillington, Spahr Webb, Douglas A. Wiens, and Lindsay L. Worthington North America’s largest earthquakes and most powerful Broadband seismometers were complemented by a suite of volcanic eruptions occur along the Alaska Peninsula additional geophysical instrumentation, including strong motion subduction zone. Despite the region’s high hazard, it has sensors, absolute and differential seafloor pressure gauges, received relatively little attention because of its remoteness. That hydrophones, and temperature and salinity sensors. OBS were changed with the GeoPRISMS focus on the Aleutian-Alaska deployed by the R/V Sikuliaq during two cruises in May and July system, and a focused call for major community data collection 2018 and retrieved by the R/V Sikuliaq and R/V Langseth during highlighted by AACSE. The Alaska Amphibious Community two cruses in August-September 2019. Seismic Experiment (AACSE) is a large shoreline-crossing AACSE instruments were deployed concurrently with the passive- and active-source seismic experiment that lasted from EarthScope Transportable Array (TA) across Alaska, extending the May 2018 through August 2019 (Abers et al., 2019; Barcheck et footprint of the TA offshore near the Alaska Peninsula. The AACSE al., 2020). The AACSE experimental footprint spanned the Alaska footprint also includes part of the aftershock region of the January Peninsula subduction zone from Kodiak Island in the northeast 2018 M7.9 offshore Kodiak earthquake (Ruppert et al., 2018), as to the Shumagin Islands in the southwest, crossing along-trench well as the Shumagin Islands region that later ruptured in the July variation in coupling, seismicity rates, incoming plate structure 2020 M7.8 and October 2020 M7.6 earthquakes. A complementary and hydration, and arc volcanic chemistry (von Huene et al., 2012; array of 398 nodal instruments was deployed for four weeks in May- Buurman et al., 2014; Shillington et al., 2015; Li and Freymueller, June 2019 on Kodiak Island, and an offshore active source survey 2018; see Fig.  1). The experiment consisted of 105 broadband was conducted concurrently by the R/V Langseth in June 2019 to seismometers: 75 offshore in ocean-bottom seismometer (OBS) shoot into the AACSE broadband network and the nodes. AACSE packages deployed from the outer rise to the trench to the shelf, is a \"community\" experiment, with all data becoming available and thirty seismometers onshore on Kodiak Island, the Alaska openly as soon as possible – the full data set was released by early Peninsula, and the Shumagin Islands. 2020. This approach enables rapid engagement with a range of science questions by all researchers interested in this fascinating The R/V Sikuliaq offshore the Kenai Peninsula. Photo credit: R. Martin-Short ■subduction zone. Data availability for AACSE is described in detail in Barcheck et al., 2020. 138 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

-164° -162° -160° -158° -156° -154° -152° -150° -148° 60° 58° Aerial view heading towards a station near King Salmon, AK. 56° Photo credit: A. Adams. 54° 52° Figure 1. The Alaska Amphibious Community Seismic Experiment (AACSE) broadband array. Symbols corresponding to different instrument types are shown in the legend. Gray dashed lines outline rupture areas of historical earthquakes (Davies et al., 1981). Dark blue lines are the 1000 and 5000 m oceanic depth contours. Empty circles indicate sites that were either not recoverable or that failed soon after deployment and recorded little to no data of any kind. AEC, Alaska Earthquake Center; APG, absolute pressure gauge; AVO, Alaska Volcano Observatory; OBS, ocean-bottom seismometer; TA, Transportable Array; TRM, trawl- resistant mounted. Modified from Barcheck et al. (2020). Figure 2. Median noise spectra of broadband seismometers in the AACSE network. (A,B) Median noise spectra on (A) vertical and (B) horizontal channels of all AACSE sites, colored by region of the network. Colors match Figure 1, except for Alaska Peninsula sites, which are white here. (C,D) Median noise spectra on (C) vertical and (D) horizontal channels of OBS sites, colored by water depth. Data plotted in all panels are median noise spectra at each site for the duration of the experiment. Sites with excessive instrumental noise or data quality issues are removed. Thick black lines are the high- and low-noise reference models of Peterson (1993). Modified from Barcheck et al. (2020). References can be found in Abers, G.A., A.N. Adams, P.J. Haeussler, E. Roland, P.J. Shore, D.A. Wiens, S.Y. Schwartz, A.F. Sheehan, D.J. Shillington, S. Webb, L.L. Worthington (2019). Examining Alaska’s earthquakes on land and sea. EOS Trans. AGU, doi.org/10.1029/2019EO117621 Barcheck, B., G.A. Abers, A.N. Adams, A. Bécel, J. Collins, J.B. Gaherty, P.J. Haeussler, Z. Li, G. Moore, E. Onyango, E. Roland, D.E. Sampson, S.Y. Schwartz, A.F. Sheehan, D.J. Shillington, P.J. Shore, S. Webb, D.A. Wiens, L.L. Worthington (2020). The Alaska Amphibious Community Seismic Experiment. Seismol Res Lett, doi:10.1785/0220200189 Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 139 

Science Nuggets After field work leaves the field – The AACSE active source supplement Emma Myers It has been over a year since the Alaska Amphibious Community 2019). Her work showed that despite the low fold coverage for Seismic Experiment (AACSE) Active Source Supplement cruise this MCS data, we clearly image along-strike variations in the with co-Principal Investigtors Anne Bécel and Anne Sheehan. incoming plate sediments and bathymetric features, as well as With many cruises this year having been cancelled, delayed, or changing accretionary prism deformation structures. On-going certainly forced to become much more complex in the response additional work include shallow fault mapping incorporating to the global pandemic, reflecting on this cruise provides a greater a subset of imaging from this cruise (Peter Haeussler) and 3D recognition of how important these experiences are. tomography using the refraction data from this active source cruise This cruise was part of the multiple AACSE research legs in 2018 (Anne Bécel) which will later be used in a joint inversion with the and 2019 focusing on the Alaska Subduction Zone, a GeoPRISMS passive local earthquake dataset (Juan-Pablo Canales). An SRL primary site and EarthScope target. The Active Source Supplement paper has also been published this year summarizing the breadth was a fortunate addition between the primary ocean-bottom of data acquisition and research covered by the entire AACSE seismometer (OBS) deployment and retrieval operations, spanning (Barcheck et al., 2020). three weeks in June 2019 aboard the R/V Marcus G. Langseth. Following up with students that I instructed at sea, while not Using the active source airgun array, this cruise was designed to working on the collected data, they shared the on-going effect the provide 400-m shot spacing arrivals for 3D refraction imaging cruise had on their own research. One student, Gökçe Astekin using the AACSE ocean bottom seismometer array as well as the (Masters student at Oklahoma State University), said the best part onshore seismometers and nodal array. Additionally, coincident was “getting together and working with students with different deployment of a 4 km-long streamer, allowed us to acquire 1751 km backgrounds, and listening to their own way of interpretation of multi-channel seismic (MCS) data. This imaging promotes the and learning from them.” Another student, Carlos Gomez (PhD investigation of the subducting sediments and the topography Candidate at Southern Illinois University), agreed, saying, “Good of the down-going plate over the SW Kodiak asperity and the research should be equal parts discovery and theory-refinement, Semidi segment. It was also used as part of an educational field modeling and field work, wherever that may take you.” These experience for several undergraduate and graduate Apply-to-Sail students help to highlight the success of this cruise not only in its students, many with no prior experience collecting, processing, data acquisition for on-going research, but in demonstrating the and interpreting seismic reflection data. Since the cruise, Anne Bécel presented the acquired data and ■importance of field experiences such as these for the future when reflection processing at AGU in December last year (Bécel et al., circumstances improve. References Bécel, A., A.F. Sheehan, E.K. Myers, G.A. Abers, D.S. Foster, L.L. Worthington, A.N. Adams, P.J. Haeussler, E.C. Roland, S.Y. Schwartz, D.J. Shillington, D. Wiens, S.C. Webb (2019). Along-strike variations in sediment input at the Alaska Peninsula subduction zone from new open access multichannel seismic reflection data of the Alaska Amphibious Community Seismic Experiment, Abstract [T51F-0330] 2019 Fall Meeting, San Francisco, CA Barcheck, B., G.A. Abers, A.N. Adams, A. Bécel, J. Collins, J.B. Gaherty, P.J. Haeussler, Z. Li, G. Moore, E. Onyango, E. Roland, D.E. Sampson, S.Y. Schwartz, A.F. Sheehan, D.J. Shillington, P.J. Shore, S. Webb, D.A. Wiens, L.L. Worthington (2020). The Alaska Amphibious Community Seismic Experiment. Seismol Res Lett, doi:10.1785/0220200189 Right page. The AACSE science party from top to bottom, left to right, Hongda Wang, Lucia Gonzalez, Gökçe Astekin, Emma Myers, Anne Bécel, Carlos Gomez, Mitchell Spangler, Ellyn Huggins, Brandon VanderBeek, Anne Sheehan, and William Frazer. Photo credit: A. Sheehan 140 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Example of an MCS profile crossing the accretionary prism, trench, and sediment-covered incoming plate (Bécel personal communication). Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 141 

Science Nuggets Plumes, plate thinning, and magmatism in the East African Rift Tyrone O. Rooney 15.9 HIMU The East African Rift System is among the most magmatically 15.8 E M2 active continental rifts, exhibiting a wide diversity of PA products that includes flood basalts, highly alkaline lavas, 207P b/204P b explosive silicic eruptions, stratiform rift basalts, and aligned 15.7 chains of cinder cones. The generation of these products derives from the melting of thermo-chemically anomalous material in the 15.6 East African upper mantle, and decompression of the upper mantle caused by lithospheric thinning during rift evolution. The central 15.5 Afar P lume focus of this project is the interaction of these two processes during (\"C\" res ervoir) advanced rifting. 15.4 DM 0.5130 DM Mobile B elt Mixing Geochemical and geophysical studies of East Africa have converged C raton Mixing on the necessity for the presence of hot and chemically distinct material in the East African upper mantle likely derived from 0.5128 Afar P lume one of two antipodal thermo-chemical anomalies located in the (\"C\" res ervoir) deep mantle – the African Large Low Shear Velocity Province. Derivation of material from this common source helps resolve the HIMU considerable debate over the number, composition, and location of these anomalies in East Africa. As part of this project, a synthesis 14 3 N d/14 4 N d 0.5126 CLM of existing isotopic data has been undertaken throughout East E M2 Africa that has revealed a common endmember in lava suites from Rungwe in the very south of the East African Rift System to Afar 0.5124 in the north. This common geochemical endmember resembles the existing hypothesized composition for the Afar plume and implies 0.5122 that material rising from the African Large Low Shear Velocity Province has this composition. These results confirm that the upper 87S r/86S r 0.5120 PA Yemen mantle beneath East Africa is contaminated with material derived 0 . 7 10 Afar from the deep mantle. 0.709 E M2 ME R 0.708 PA Turkana Figure 1. Quaternary Data: Isotopic variation of samples 0.707 NKR erupted 0.5 Ma to present in East Africa showing a 0.706 CLM S K R /NT D convergence in mixing arrays on the Afar Plume isotopic 0.705 R ungwe composition. Samples erupted in regions influenced by the K ivu-V irunga Craton are shown in cool colors, while warm colors are for C arbonatites those erupted within the Mobile belt. Mixing arrays (drawn by hand) are also shown as grey and orange arrows. The Afar P lume Pan African Lithosphere (PA), Afar Plume, CLM, HIMU, and (\"C\" res ervoir) Enriched Mantle 2 (EM2) are hypothetical endmembers derived from the existing literature. The Depleted Mantle 0.704 (DM) endmember is not shown in these plots due to the compressive impact on the plotting of the dataset. These 0.703 DM HIMU endmembers are shown as discrete points but in reality these 21 22 are regions of isotopic space and are shown as the center 17 18 19 20 points for clarity. Small symbols represent samples that are 206P b/204P b either 5 wt. % MgO. Figure is modified from Rooney (2020) 142 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Continued work on this project examines the link between Pliocene stratiform basalts from the Turkana and Afar Depressions extensional events in the East African Rift System and the location and the transition to more focused magmatism at both locales. and timing of volcanism. Research undertaken as part of this The aim of this ongoing work is to develop a conceptual model project has already identified distinct pulses of basaltic magmatism that correlate with known episodes of extension (Early Miocene, ■that uses magma generation processes as a probe of plate thinning Mid Miocene, and Pliocene). Our ongoing work is focused upon during rifting. Rooney, T. (2020). The Cenozoic magmatism of East Africa: Part II – Rifting of the mobile belt . Lithos, 360-361, 105291, doi: 10.1016/j.lithos.2019.105291 Figure 2. Cartoon depicting of melt generation crust processes within the Eastern Branch of the East lithospheric mantle African Rift System. The lithospheric mantle has been enriched with pyroxenite-bearing metasomes amphibole metasome (deeper) and amphibole-bearing metasomes pyroxenite metasome (shallower) generated by chromatographic metasomatism as fluids/melts passed through astenosphere the lithospheric mantle. The asthenosphere in this area has been hybridized and homogenized by mixing initially with lithospheric materials (depicted as down going arrows). Figure is modified from Rooney (2020) Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 143 

Science Nuggets Paleogene flood basalt stratigraphy in East Africa Tyrone O. Rooney, R. Alex Steiner, and John W. Kappelman One of the most pronounced Cenozoic features of 18°N East Africa are the Oligocene flood basalts of the 15°N northwestern Ethiopian plateau. However, there is a 12°N growing awareness that flood basalt magmatism in East Africa is more spatially and temporally extensive than had been initially 9°N considered, with implications for the interaction of deep mantle 6°N thermo-chemical anomalies with the continental lithosphere 3°N and rift development. Recent geophysical surveys and some limited high precision geochronology within the Turkana Depression (located between the Ethiopian and Kenyan domes) identified thick magmatic sequences that may represent flood basalt events. The primary goal of field work for this project were to undertake stratigraphically constrained flow-by-flow sampling of the newly identified Eocene-Oligocene flood basalts sequences and of well-constrained Miocene sections important for paleontological studies. The project field work successfully identified continuous sections of flood basalts in Turkana that bear a striking field resemblance to the more well-known flood basalts of the NW Ethiopian Plateau, from which a combined petrographic and paleomagnetic section is being assembled. 32°E 36°E 40°E 44°E 144 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Our study locales in West Turkana hold great significance for Using the well-constrained section we have developed in Turkana, studies of mammalian and especially primate evolution. Some of and equivalent sections of flood basalts in Southern Ethiopia and the sections that we are studying contain the earliest evidence for the northwest Ethiopian Plateau, we seek to link these disparate the origin and evolution of Old World monkeys and apes, but the magmatic systems by characterizing a spatially distributed suite ages are not tightly constrained. This fossil record also documents of Eocene-Oligocene lavas that extend from the Kenyan border to the major Oligo-Miocene dispersal event that introduced Eurasian the boundary of the flood basalts of the northwestern Ethiopian taxa to Africa, thus modernizing the faunal composition of the Plateau. Collectively, we hope that our observations will provide continent, along with the extinction of many taxa from the archaic new insights into the distribution of thermo-chemically anomalous fauna. Our work will better constrain the dates of these important material in the East African upper mantle derived from the African events and in doing so, facilitate testing of the various hypotheses thought to be responsible for these important changes. ■Large Low Shear Velocity Province during the initial stages of lithospheric destabilization and subsequent Cenozoic rifting. We are currently combining petrographic, geochemical, geochronologic, and paleomagnetic data on Eocene-Oligocene and Miocene basalt sections to explore the temporal evolution of magmatism in this important region. Ongoing work examines the evolution of the magmatic systems that produce flood basalt lavas by novel interpretation of crystal compositions within the context of geochemical modelling of open magmatic systems. Tentative results suggest these techniques can establish timescales of magma residence within the crust and contribute to models of flood basalt magma generation, storage, homogenization, and eruption. Figure 1 (left page). Map showing the distribution of the Eocene and Oligocene flood basalt provinces in East Africa. LT and HT represent Low Ti and High Ti basalts respectively. Figure 2 (right). Schematic figure showing a hypothetical trans- crustal magmatic plumbing system of a continental flood basalt. Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 145 

Science Nuggets Deformation of the East African Rift Kyle D. Murray and Bridget Smith-Konter The objective of this project is to produce precise maps of temporally and spatially dense crustal deformation time series across the central EARS, documenting tectonic, volcanic, and anthropogenic related deformation signals. This allows us to characterize and compare distributions of strain rates between the proposed plume-controlled eastern segment, and fault-controlled western segment, which branch around the Tanzania Craton ■- placing these observations in the context of the geologic history and pre-existing structures. The East Africa Rift. GPS velocity field (black vectors) show general eastward velocity increase as the Nubia plate separates from the Somalia plate. Historical earthquakes (red circles), active volcanoes (green triangles), and faults (orange lines) are also shown. The East African Rift System primary site exhibits a wide variety of rift processes and characteristics, making it an ideal target for GeoPRISMS goals. GeoPRISMS research around this site focused around the Eastern Rift - from the Tanzanian divergence in the south to Lake Turkana and southern Ethiopia to the north - the Afar and Main Ethiopian Rift, and the Western Rift and southwest branch. 146 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

Science Nuggets EARThD: A compilation of explosive volcanism in East Africa Erin DiMaggio and Sara Mana Tephra deposits are prolific and widespread across Students gain valuable skills in literature research, and data mining much of the East African Rift. They serve as important and archiving - skills that have supported multiple senior thesis chronostratigraphic marker beds in regional sedimentary projects and a masters degree. archives. Typically, tephra are studied in isolation by researchers As of November 2020, the EARThD project has reviewed over 850 from different disciplines, in most cases with little cross- academic papers and books. Of those, 405 had tephra data that disciplinary communication. This limits the use of East African could be entered into the EARThD database. Currently, data from tephra records for constraining the timing, rates, and evolution of 54 of those publications are publicly available through IEDA, and rifting, volcanism, climate events, and human origins. will ultimately be searchable using EarthChem. The EARThD East African Rift Tephra Database (EARThD) is a collaborative data reference list is available to the public through our website and compilation project that aims to maximize the scientific potential on Mendeley. The EARThD website (https://sites.psu.edu/earthd/) of regional tephra records and support and foster new collaborative documents project progress, offers a venue for community input, a research avenues. EARThD is digitizing, standardizing, and map for visualizing tephra locations, and provides instructions and integrating published geochemical, geochronological, and physical direct links for searching, accessing, and downloading datasets. tephra datasets from publications primarily within the East African With this effort, we aim to fulfill a crucial data integration role for Rift and making them widely available. We utilize an existing NSF- researchers working in East Africa and the increasingly complex supported community-based data facility, the Interdisciplinary Earth Data Alliance (IEDA), to store, curate, and provide access to ■and multidisciplinary research questions being studied in this the compiled data. region. Over the past two years, thirteen undergraduate students and one Google Earth map showing the approximate locations of each of the graduate student from Salem State University and Pennsylvania 400 publications entered into EARThD. Markers are color coded by State University have led the extensive data mining and archiving efforts. Our team is majority female with international students publication year; red (1932-1962), orange (1963-1973), yellow (1974- and members of underrepresented minorities in geoscience. 1985), green (1986-1996), blue (1997-2008, violet (2009-2020). Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 147 

Science Nuggets Displaced cratonic mantle concentrates deep carbon during continental rifting James D. Muirhead, Tobias P. Fischer, Sarah J. Oliva, Amani Laizer, Jolante van Wijk, Claire A. Currie, Hyunwoo Lee, Emily .J. Judd, Emmanuel Kazimoto, Yuji Sano, Naoto Takahata, Christel Tiberi, Stephen F. Foley, Josef Dufek, Miriam C. Reiss, and Cynthia J. Ebinger Continental rifts are important sources of mantle carbon dioxide 25o s(CtoOre2d) emission into Earth’s atmosphere. Because deep carbon is 0o 0o EB A for long periods in the lithospheric mantle, rift trCaOns2pfolurtx. WB study site N 1oS depends on lithospheric processes that control melt and volatile -25o In particular, the influence of compositional and thickness differences between Archaean (cratonic) and Proterozoic (orogenic) lithosphere on deep-carbon fluxes remains untested. Magadi basin In this project (Muirhead et al., 2020), we collected water samples from springs in the Natron, Manyara and Balangida regions of Tanzania that cross the boundary between orogenic and cratonic B 2oS 3oS lithosphere (Fig. 1). We also measured Natron basin odniffutsheeCOhe2 lifulumx in these regions. Based and carbon isotopes 29.8 g m d and CdOisp2 laflcuexmeesntobtoafinecdar, bwone-epnrroicphoesde magmatic CO2 that Tanzanian cratonic mantle concentrates C deep carbon below parts of the East African Rift System, which is released during rifting and also facilitates the rainftdinhgepliruomceossv.eSrotuhricse3s5a0n-dkiflloumxeestroef-lCoOng2 transect crossing the boundary between Manyara basin orogenic (Natron and Magadi basins) and cratonic (Balangida and Manyara basins) lithosphere from south to north show 4oS striking and systematic differences. Areas of diffuse tChOe 2 rdifetgatrsasinnsgitieoxnhsibfirtoimncrAearscihnagemanan(tclreaCtoOni2cf)lutxo and 3He/4He 25 kmBalangida basin ratios as Proterozoic (orogenic) lithosphere. Active carbonatite magmatism also occurs near the 36oW craton edge supplying significant amounts odfamtaawntilteh-dgeeroivpehdysCicOal2 to the Figure 1. Example of gas data (modified from Muirhead et atmosphere. We combine our geochemical results that illuminate the rift-craton transition in the lithosphere and mantle al. 2020) collected and analysed by the MODeSt project, as well as with existing numerical models of rift evolution. This multi- with these figures and data presented in full in Muirhead et disciplinary analyses of all data indicates that advection of the root of thick Archaean lithosphere laterally to the base of the much thinner adjacent al. (2020). A) Locations of examined rift basins in the East Proterozoic lithosphere creates a zone of highly concentrated deep carbon African Rift System. Cross-section line from X to X’ shows (Fig. 2). This mode of deep-carbon extraction pmroayduinctciroenasaenCdOlo2 cfalutixoens in the southern and north extents of the seismic velocity data some continental rifts, helping to control the of profile presented in C. B) Diffuse CO2 vs latitude for data ■carbonate-rich magmas as well as facilitate the rifting process. collected in the study region presented in A. C) Cross-section through the lithosphere density model of Tiberi et al. (2019). 148 • GeoPRISMS Newsletter Issue No. 43 Fall 2020

adi Photo above. Tobias Fischer sampling an actively degassing vent on the western border fault to the Natron Basin in Tanzania, with the Oldoinyo Lengai in the background. Photo credit: J. Muirhead. Figure 2. Schematic diagram from Muirhead et al., (2020) showing the process of lateral advection of the root of thick Archean lithosphere to the base of much thinner adjacent Proterozoic lithosphere. This process creates a zone of highly concentrated deep carbon that can facilitate melting and rifting. References Muirhead, J.D., T.P. Fischer, S.J. Oliva, A. Laizer, J. van Wijk, C. Currie, N. Takahata, C. Tiberi, S.F. Foley, J. Dufek, M.C. Reiss, C.J. Ebinger (2020). Displaced cratonic mantle concentrates deep carbon during continental rifting. Nature, 582,7810, 67-72 Tiberi, C., S. Gautier, C. Ebinger, S. Roecker, M. Plasman, J. Albaric, J. Deverchere, S. Peyrat, J. Perrot, R.F. Wambura, M. Msabi, A Muzuka, G. Mulibo, G. Kianji (2019). Lithospheric modification by extension and magmatism at the craton-orogenic boundary: North Tanzania Divergence, East Africa. Geophys J Int, 216, 3, 1693-1710 Fall 2020 Issue No. 43 GeoPRISMS Newsletter • 149 

Science Nuggets Revealing asthenospheric rheology beneath continental regions Tahiry A. Rajaonarison, D. Sarah Stamps, Stewart Fishwick, Sascha Brune, Anne Glerum, and Jinping Hu Seismic anisotropy is often used as a proxy for investigations Using the geodynamics code ASPECT, we simulated the two of past and present-day deformation of the lithosphere and modes of mantle convection to a depth of 660 km and used them to understand flow patterns in the sub-lithospheric mantle. to calculate the lattice-preferred orientation (LPO) that develops Beneath continents, poor depth constraints on seismic anisotropy along mantle flow pathlines. The predicted LPO was then used to measurements make it challenging to distinguish between shallow, calculated synthetic splitting parameters, which were compared lithospheric sources and deeper, asthenospheric contributions. with the azimuths of available seismic anisotropy across the island. Madagascar, the easternmost segment of the East African Rift Through a series of comparisons, we found that asthenospheric System, provides a study region where the source(s) of seismic flow resulting from undulations in lithospheric topography is the anisotropy can be investigated with geodynamic modeling. In dominant source of the seismic anisotropy, but fossilized structures Rajaonarison et al. (2020), we tested the hypothesis that observed from an ancient shear zone may play a role in southern Madagascar. seismic anisotropy across the continental island arises from either Our results suggest that the rheological conditions needed for the asthenospheric flow driven by thermal perturbations that arise from variations in lithospheric topography or mantle flow derived ■formation of seismic anisotropy, dislocation creep, may dominate from mantle wind interactions with lithospheric topography. the upper asthenosphere beneath other continental regions. Figure 1. A. Initial temperature condition for numerical model set-up. B. Initial viscosity model for numerical model set-up. C. Comparison of seismic anisotropy for preferred mantle convection model derived from lithospheric topography variations. 150 • GeoPRISMS Newsletter Issue No. 43 Fall 2020