TY - JOUR
AU1 - Craig, Magee,
AU2 - E, Stevenson, Carl T
AU3 - K, Ebmeier, Susanna
AU4 - Derek, Keir,
AU5 - S, Hammond, James O
AU6 - H, Gottsmann, Joachim
AU7 - A, Whaler, Kathryn
AU8 - Nick, Schofield,
AU9 - A-L, Jackson, Christopher
AU1 - S, Petronis, Michael
AU1 - Brian, O’Driscoll,
AU1 - Joanna, Morgan,
AU1 - Alexander, Cruden,
AU1 - A, Vollgger, Stefan
AU1 - Greg, Dering,
AU1 - Steven, Micklethwaite,
AU1 - D, Jackson, Matthew
AB - Abstract Over the last few decades, significant advances in using geophysical techniques to image the structure of magma plumbing systems have enabled the identification of zones of melt accumulation, crystal mush development, and magma migration. Combining advanced geophysical observations with petrological and geochemical data has arguably revolutionised our understanding of, and afforded exciting new insights into, the development of entire magma plumbing systems. However, divisions between the scales and physical settings over which these geophysical, petrological, and geochemical methods are applied still remain. To characterise some of these differences and promote the benefits of further integration between these methodologies, we provide a review of geophysical techniques and discuss how they can be utilised to provide a structural context for and place physical limits on the chemical evolution of magma plumbing systems. For example, we examine how Interferometric Synthetic Aperture Radar (InSAR), coupled with Global Positioning System (GPS) and Global Navigation Satellite System (GNSS) data, and seismicity may be used to track magma migration in near real-time. We also discuss how seismic imaging, gravimetry and electromagnetic data can identify contemporary melt zones, magma reservoirs and/or crystal mushes. These techniques complement seismic reflection data and rock magnetic analyses that delimit the structure and emplacement of ancient magma plumbing systems. For each of these techniques, with the addition of full-waveform inversion (FWI), the use of Unmanned Aerial Vehicles (UAVs) and the integration of geophysics with numerical modelling, we discuss potential future directions. We show that approaching problems concerning magma plumbing systems from an integrated petrological, geochemical, and geophysical perspective will undoubtedly yield important scientific advances, providing exciting future opportunities for the volcanological community. INTRODUCTION Igneous petrology and geochemistry are concerned with the chemical and physical mechanisms governing melt genesis, mobilisation, and segregation, as well as the transport/ascent, storage, evolution, and eruption of magma. The reasons for studying these fundamental processes include understanding volcanic eruptions, modelling the mechanical development of magma conduits and reservoirs, finding magma-related economic ore deposits, exploring for active geothermal energy sources, and determining the impact of magmatism in different plate tectonic settings on the evolution of the lithosphere and crustal growth. However, whilst petrological and geochemical studies over the last century have shaped our understanding of the physical and chemical evolution of magma plumbing systems, assessing the distribution, movement, and accumulation of magma in the Earth’s crust from these data remains challenging. A key frontier in igneous petrological and geochemical research thus involves deciphering how and where magma forms, the routes it takes toward the Earth’s surface, and where exactly it is stored. This contribution will demonstrate how geophysical data can be used to determine the architecture of magma plumbing systems, providing a structural framework for the interpretation of petrological and geochemical data. To aid the alignment of petrological, geochemical, and geophysical disciplines it is first important to delineate what we mean by ‘magma’. We follow Glazner et al. (2016) and define magma as, ‘naturally occurring, fully or partially molten rock material generated within a planetary body, consisting of melt with or without crystals and gas bubbles and containing a high enough proportion of melt to be capable of intrusion and extrusion’. Importantly, this definition specifically considers that magma: (i) forms through the migration and accumulation of partial melt that is initially distributed throughout pore spaces in a rock volume; and (ii) is a suspension of particles (i.e. crystals, xenoliths, and/or bubbles) within melt (see Cashman et al., 2017). As magma starts to solidify, the proportion of suspended crystals and thus the relative viscosity of the magma increases until a relatively immobile, continuous network of crystals and interstitial melt develops; we term this a ‘crystal mush’ (e.g. Hildreth, 2004; Glazner et al., 2016; Cashman et al., 2017). The rheological transition from a magma to a crystal mush is partly dependent on its chemistry, but typically occurs abruptly when the particle volume increases above the 50–65% range (Cashman et al., 2017). Crystal mushes thus exist at or above the solidus and generally cannot be erupted, although they may be partly entrained in eruptible magma as glomerocrysts, cumulate nodules, or restite (Cashman et al., 2017). Migration of interstitial melt within a crystal mush can lead to its accumulation and, thus, formation of a magma. A magma plumbing system, therefore, consists of interconnected magma conduits and reservoirs, which store magma as it evolves into a crystal mush, ultimately fed from a zone of partial melting (e.g. Fig. 1). These definitions are supported by geophysical imaging and analyses of contemporary magma reservoirs, which show that melt volumes in the mid- to upper crust are typically low (<10%) and likely exist within a crystal mush (e.g. Paulatto et al., 2010; Koulakov et al., 2013; Hammond, 2014; Ward et al., 2014; Comeau et al., 2015,, 2016; Delph et al., 2017). These definitions and geophysical data question the traditional view that magma resides in long-lived, liquid-rich, and volumetrically significant magma chambers. Following this, the emerging paradigm for igneous systems is thus that liquid-rich magma chambers are short-lived, transient phenomena with: (i) melt typically residing in mushes that develop through the incremental injection of small, distinct magma batches; and (ii) magma accumulating in thin lenses (e.g. Hildreth, 2004; Annen et al., 2006, 2015; Annen, 2011; Miller et al., 2011; Solano et al., 2012; Cashman & Sparks, 2013; Cashman et al., 2017). We are now starting to view magmatic systems as vertically extensive, transcrustal, interconnected networks of magma conduits and magma/mush reservoirs (Fig. 1) (e.g. Cashman et al., 2017). Fig. 1. View largeDownload slide Schematic of a vertically extensive, transcrustal magma plumbing system involving transient, interconnected, relatively low-volume tabular magma intrusions (e.g. dykes, sills, and laccoliths) within a crystal mush (based on Cashman et al., 2017; Cruden & Weinberg, 2018). Fig. 1. View largeDownload slide Schematic of a vertically extensive, transcrustal magma plumbing system involving transient, interconnected, relatively low-volume tabular magma intrusions (e.g. dykes, sills, and laccoliths) within a crystal mush (based on Cashman et al., 2017; Cruden & Weinberg, 2018). The current use of geophysical techniques within the igneous community can be separated into two distinct areas focused on either characterising active volcanic domains or investigating the structure and emplacement of ancient magma plumbing systems. For example, in areas of active volcanism, our understanding of magma plumbing system structure principally comes from the application of geophysical techniques that detect sites of magma movement or accumulation (e.g. Sparks et al., 2012; Cashman & Sparks, 2013). Such geophysical techniques include Interferometric Synthetic Aperture Radar (InSAR; e.g. Biggs et al., 2014), seismicity (e.g. recording of earthquakes associated with magma movement; e.g. White & McCausland, 2016), various seismic imaging methods (e.g. Paulatto et al., 2010; Hammond, 2014), gravimetry (e.g. Battaglia et al., 1999; Rymer et al., 2005), and electromagnetic techniques (e.g. Desissa et al., 2013; Comeau et al., 2015). These techniques allow examination of: (i) the temporal development of magma plumbing systems (e.g. Pritchard & Simons, 2004; Sigmundsson et al., 2010); (ii) vertical and lateral movements of magma (e.g. Keir et al., 2009; Jay et al., 2014); (iii) the relationship between eruption dynamics, volcano deformation, and intrusion (e.g. Sigmundsson et al., 2010, 2015); and (iv) estimates of melt sources and melt fractions (e.g. Desissa et al., 2013; Johnson et al., 2016). However, inversion of these geophysical data typically results in non-unique, relatively low-resolution models of subsurface structures. Furthermore, some methods only capture active processes, which may be short-lived or even instantaneous, potentially providing information on only a small fraction of the magma plumbing system. In contrast to the study of active volcanic domains, the analysis of ancient plumbing systems through field observations, geophysical imaging techniques (e.g. reflection seismology, gravity, and magnetic data) and/or rock magnetic experiments can provide critical insights into magma emplacement, mush evolution, and allow the geometry of entire plumbing systems to be reconstructed (e.g. Cartwright & Hansen, 2006; Stevenson et al., 2007a; Petronis et al., 2013; Muirhead et al., 2014; O’Driscoll et al., 2015; Magee et al., 2016). Whilst such studies of ancient plumbing systems provide a framework for interpreting the structure of active intrusion networks, capturing a snapshot of how magma moved and melt was distributed through the system at any one time is difficult because magmatism has long since ceased. All the techniques employed to define active and ancient plumbing systems, including petrological and chemical analyses, provide information at different spatial and/or temporal resolutions. Answering the major outstanding questions in studies of magma plumbing systems, therefore, requires the integration of complementary petrological, geochemical, geophysical, geochronological, and structural techniques. Here, we examine active plumbing systems using InSAR, seismicity, seismic imaging, gravimetry, and electromagnetic techniques. To provide a context for the interpretation of data pertaining to the active systems, we also discuss how seismic reflection data and rock magnetic techniques can be used to derive the structure and evolution of ancient intrusion networks. The potential of emerging techniques involving seismic full-waveform inversion (FWI) and unmanned aerial vehicles (UAVs) are also considered, as is the role of numerical modelling in bringing together outputs from different datasets. For each technique described, we briefly discuss the methodology and limitations and provide a summary of the key findings and potential uses, with a focus on integration with petrological and geochemical data. The aim of this review is to facilitate and promote integration between petrologists, geochemists, geochronologists, structural geologists, and geophysicists interested in addressing outstanding problems in studies of magma plumbing systems. UNDERSTANDING MAGMA PLUMBING SYSTEM STRUCTURE Here, we discuss a range of techniques that can be utilised to define different aspects of magma plumbing system structure and evolution. In particular, we describe how InSAR, seismicity, seismic imaging (e.g. seismic tomography), gravity, and electromagnetic data are used to determine melt fractions and distribution, track movement of magma in near real-time and/or locate sites and examine the evolution of magma or mush storage. Overall, these geophysical techniques allow the structure of active plumbing systems and their transient evolution to be assessed. We also discuss how seismic reflection data can provide unprecedented images of ancient plumbing systems and associated host rock deformation in three-dimensions at resolutions of tens of metres. Finally, we examine the application of rock magnetic techniques to assess magma flow and crystallisation processes at a range of scales. Although beyond the scope of this review, it is critical to highlight that interpreting the geophysical response of a rock, magma, or mush relies on understanding its physical and chemical properties (e.g. density, temperature and melt fraction). Laboratory experiments testing how rock or magma properties influence geophysically measured parameters (e.g. seismic velocities and resistivity) thus provide context for interpreting magma plumbing system structure and evolution from geophysical data (e.g. Gaillard, 2004; Pommier et al., 2010; Pommier, 2014). Insights into magma plumbing systems from ground deformation data Techniques Changes in volume within magma plumbing systems can deform the host rock, potentially resulting in displacement of the Earth’s surface. Such displacements are a unique source of information for volcanologists and can be modelled to estimate geodetic source depth and, to varying extents, the source geometry and volume change (e.g. Segall, 2010). Measuring the deformation of the Earth’s surface can thus provide information about the characteristics and timing of magma movement and accumulation, as well as variations in internal reservoir conditions. Traditionally, deformation measurements are made using levelling, electronic distance meters, tiltmeters, and Global Positioning System (GPS), all of which have proven to be reliable methods and thus are widely used in volcano monitoring (e.g. Dzurisin, 2006). For example, GPS measurements retrieve the relative positions of receivers on the Earth’s surface from dual frequency carrier phase signals transmitted from GPS or Global Navigation Satellite System (GNSS) satellites with precisely known orbits. Distances between satellites and receivers are assessed from the travel-time, i.e. the measured difference between the transmitted and received times of a unique ranging code, allowing movement of the Earth’s surface over time to be monitored (see review by Dixon, 1991). Permanently installed receivers record position data continuously, but receivers can also be deployed for a limited time during GPS campaigns to provide additional measurements, normally made relative to a standard benchmark location (e.g. Dvorak & Dzurisin, 1997). Whilst tiltmeters and GPS can provide continuous measurements, their spatial resolution is limited by logistical constraints such as cost and accessibility, which may be restricted at active volcanoes. The geographic reach of volcano geodesy has been greatly expanded over the past two decades by the application of Interferometric Synthetic Aperture Radar (InSAR), an active remote sensing technique that uses microwave electromagnetic radiation to image the Earth’s surface (e.g. Simons & Rosen, 2007; Pinel et al., 2014). Surface displacements can be measured by constructing interferograms, where the difference in phase between radar echoes from time-separated images appear as ‘fringes’ of variation in the line of sight distance to the satellite (Fig. 2). The patterns of fringes in individual interferograms are distinctive for different deformation source geometries, such as for horizontal (sill-like) or vertical (dyke-like) opening of intrusions, or the pressurisation of a spheroidal reservoir (i.e. a Mogi source) (e.g. Fig. 2b). However, magma intrusion processes can rarely be uniquely identified from geodetic source geometry alone and distinguishing between magmatic, hydrothermal, structural (e.g. faulting and compaction), and combinations of elastic and inelastic sources is particularly challenging (e.g. Galland, 2012; Holohan et al., 2017). Fig. 2. View largeDownload slide (a) Interferograms showing fringes caused by the pressurisation of a point source directly beneath a stratovolcano from both ascending and descending satellite lines of sight. Note that the centre of the fringes are slightly offset from the summit of the volcano (marked by a black triangle). (b) Typical fringe patterns for analytical deformation sources in an elastic half space from ascending satellite geometry: (i) Mogi source at 5 km depth; (ii) dyke extending between depths of 3 and 9 km; (iii) rectangular sill; and (iv) a penny-shaped horizontal crack both at 5 km depth. Fig. 2. View largeDownload slide (a) Interferograms showing fringes caused by the pressurisation of a point source directly beneath a stratovolcano from both ascending and descending satellite lines of sight. Note that the centre of the fringes are slightly offset from the summit of the volcano (marked by a black triangle). (b) Typical fringe patterns for analytical deformation sources in an elastic half space from ascending satellite geometry: (i) Mogi source at 5 km depth; (ii) dyke extending between depths of 3 and 9 km; (iii) rectangular sill; and (iv) a penny-shaped horizontal crack both at 5 km depth. Whilst a single interferogram only provides displacements in satellite line-of-sight, a pseudo-3D displacement field can be estimated by combining multiple images from polar orbits that are ascending (i.e. satellite moves roughly northward, looking east) and descending (i.e. satellite moves roughly southward, looking west) (Fig. 2a), especially where GNSS measurements can also be incorporated. The lateral spatial resolution of most InSAR data is on the order of metres to tens of metres, whilst vertical movements can be resolved on the order of centimetres and sometimes millimetres. Temporal resolution depends on the satellite revisit time and ranges between days to months depending upon the sensor type and satellite orbit. This means that InSAR can be used to regularly assess ground deformation at virtually any volcano worldwide situated above sea level, with a higher spatial density of measurements than achieved using from ground-based instrumentation. Magmatic processes are only observable by InSAR when either magma movement or internal reservoir processes (e.g. cooling and contraction, phase changes) cause changes in pressure and thereby instigate deformation of the host rock and free surface. The best-fit parameters of a deformation source (e.g. an intruding magma body) are most often assessed by inverting measured displacements using analytical elastic-half space models of simple source geometries, although there are often trade-offs between parameters such as source depth and volume change (e.g. Pritchard & Simons, 2004). Complex and more realistic deformation source geometries may be retrieved using finite element-based linear inversion of displacement fields (e.g. Ronchin et al., 2017). A proportion of any pressure change may be accommodated by magma compressibility, leading to underestimation of volume changes (e.g. Rivalta & Segall, 2008; McCormick Kilbride et al., 2016). Assessing both volume changes and especially the total volume of a magma reservoir from geodetic data, therefore, remains challenging. Furthermore, host rocks in areas of repeated intrusion that have been heated above the brittle–ductile transition are better described by a viscoelastic rheology (e.g. Newman et al., 2006; Yamasaki et al., 2018), while ductile accommodation of volume changes may occur at greater depth. Where some constraints are available for the structure and rheology of Earth’s crust, finite or boundary element models may achieve a more realistic model of the deformation source (e.g. Masterlark, 2007; Gottsmann et al., 2017; Hickey et al., 2017). Observations Measurements of volcano deformation preceding and, or, accompanying eruption have provided insights into the extent and structure of magma plumbing systems and, in some instances, the dynamics of magma movement. For example, InSAR-based observations at Eyjafjallajökull, Iceland, have recognised the intrusion of multiple, distinct sills over a decade and their subsequent extraction when tapped during an explosive eruption (e.g. Pedersen & Sigmundsson, 2006; Sigmundsson et al., 2010). Extensive lateral connections via dykes and sills between reservoirs and/or volcanoes have been illuminated by eruptions or unrest accompanied by ground deformation tens of kilometres away and by the existence of multiple deformation sources (e.g. Alu-Dalafilla shown in Fig. 3a and b, Pagli et al., 2012; Korovin, Lu & Dzurisin, 2014; Cordon-Caulle, Jay et al., 2014; Kenyan volcanoes, Biggs et al., 2014; global synthesis, Ebmeier et al., 2018). Inter-eruptive deformation at calderas is especially complex and seems to be particularly frequent and high magnitude (e.g. Laguna del Maule: Fournier et al., 2010; Singer et al., 2014; Le Mével et al., 2015), with the location of the deformation sources inferred to vary over time (e.g. Campi Flegrei, Trasatti et al., 2008; Yellowstone, Wicks et al., 2006). The geometries of dykes and sills inferred from InSAR data inform our understanding of changing subsurface stress fields (e.g. Afar, Hamling et al., 2010; Fernandina, Bagnardi et al., 2013), as do measurements of displacements caused by moderate earthquakes in close proximity to magma plumbing systems (e.g. Kilauea, Wauthier et al., 2013; Chiles-Cerro Negro, Ebmeier et al., 2016). Fig. 3. View largeDownload slide (a) Ascending line of sight (LOS) co-eruptive interferogram from the 2008 basalt lava extrusion between the Alu and Alu South domes and the Dalafilla stratovolcano (modified from Pagli et al., 2012). (b) Inversion of uplift and subsidence patterns, recorded by InSAR during the 2008 basalt lava eruption at the Alu dome in the Danakil Depression, suggested ground deformation could be attributed to a combination of: (i) deflation of a reservoir, modelled as a Mogi source, at ∼4 km depth; (ii) inflation and deflation of a tabular sill at ∼1 km depth; and (iii) opening of a dyke beneath the eruptive fissure (Fig. 3a and b) (Pagli et al., 2012). See Fig. 3a for location. (c) Geological map showing that lava flows radiate out from Alu and originate from the periphery of the dome, which is cross-cut by an array of randomly oriented faults (modified from Magee et al., 2017a). (d) Magee et al., (2017a) inferred Alu is underlain by a saucer-shaped sill plumbing system, based on field observations and comparison to seismic reflection data, not a tabular sill (b). Fig. 3. View largeDownload slide (a) Ascending line of sight (LOS) co-eruptive interferogram from the 2008 basalt lava extrusion between the Alu and Alu South domes and the Dalafilla stratovolcano (modified from Pagli et al., 2012). (b) Inversion of uplift and subsidence patterns, recorded by InSAR during the 2008 basalt lava eruption at the Alu dome in the Danakil Depression, suggested ground deformation could be attributed to a combination of: (i) deflation of a reservoir, modelled as a Mogi source, at ∼4 km depth; (ii) inflation and deflation of a tabular sill at ∼1 km depth; and (iii) opening of a dyke beneath the eruptive fissure (Fig. 3a and b) (Pagli et al., 2012). See Fig. 3a for location. (c) Geological map showing that lava flows radiate out from Alu and originate from the periphery of the dome, which is cross-cut by an array of randomly oriented faults (modified from Magee et al., 2017a). (d) Magee et al., (2017a) inferred Alu is underlain by a saucer-shaped sill plumbing system, based on field observations and comparison to seismic reflection data, not a tabular sill (b). At a transcrustal scale, deformation measurements have contributed to evidence for temporal variations in magma supply rates (e.g. in Hawaii, Poland et al., 2012). Volume increases in the mid- to lower-crust, notably in the Central Andes, have provided the first observations of deep pluton growth (Pritchard & Simons, 2004). Furthermore, uplift during episodes of unrest that have not (yet) resulted in eruption have been detected at a broad range of volcanoes (e.g. Westdahl, Mount Peulik, Lu & Dzurisin, 2014; Alutu & Corbetti, Biggs et al., 2011) and, in some cases, have been interpreted as evidence for the ‘pulsed’ accumulation of potentially eruptible magma (e.g. Santorini, Parks et al., 2012). In addition to magma movement, volume changes associated with internal reservoir processes can also cause deformation of the host rock and free surface. For example, InSAR measurements have recorded subsidence linked to cooling and crystallisation of sills (Medicine Lake, Parker, 2016; Taupo Volcanic Zone, Hamling et al., 2015). Transient periods of subsidence during inter-eruptive uplift have been attributed to phase transitions in response to the addition of more juvenile magma (e.g. Okmok, Caricchi et al., 2014). Implications and integration InSAR has increased the number of volcanoes at which measurements of ground deformation have been made, from less than 50 in the late 1990s to over 200 today (Biggs & Pritchard, 2017; Ebmeier et al., 2018). This increase in coverage has been particularly influential in the developing world where monitoring infrastructure is typically poor (Chaussard & Amelung, 2013; Ebmeier et al., 2013), with InSAR often providing the first evidence of magmatic activity at many volcanoes previously considered to be inactive (e.g. Pritchard & Simons, 2004; Biggs et al., 2009,, 2011; Lu & Dzurisin, 2014). A continued increase in the number and range of satellite- and large-scale UAV-based SAR instruments, as well as enhancements to their spatial and temporal resolution, will allow the detection of a greater range of volcanic ground deformation (e.g. Salzer et al., 2014; Schaefer et al., 2015; Stephens et al., 2017). Overall, improved InSAR coverage will also increase the number of volcanoes where deformation measurements have been made across multiple cycles of eruption and deformation, increasing its usefulness for both hazard assessment and for characterising the extent, geometry, and changes in magma plumbing systems. Geodetic measurements provide information only about the parts of a plumbing system that are currently active, and do not necessarily reflect the full extent and character of the intrusion network (e.g. Sigmundsson, 2016). However, geodetic analyses of ground deformation provide critical insight into the spatial and temporal development of active plumbing systems. Comparing observations of ancient plumbing systems (e.g. Magee et al., 2013a; Schofield et al., 2014), integration of ground deformation measurements with petrological observations (e.g. Caricchi et al., 2014; Jay et al., 2014) or thermal models (Parker et al., 2016), as well as tomographic geophysical imaging, will increase the sophistication of models of magmatic systems. Integrating InSAR with gravity or electromagnetic measurements is particularly powerful, as it can allow discrimination between melt, volatiles, and hydrothermal fluids for which deformation signals are similar (see below) (e.g. Tizzani et al., 2009). Seismicity and magma plumbing systems Technique Seismicity (i.e. earthquakes) at volcanoes is primarily caused by the dynamic interaction of magma and hydrothermal fluids with the solid host rock (e.g. Chouet & Matoza, 2013), as well as by fracturing and fragmentation of silicic magma (e.g. Tuffen et al., 2008). There are a number of primary physical mechanisms for causing volcano seismicity (e.g. faulting), each of which typically produces seismic signals of specific frequency content (Chouet & Matoza, 2013). Recording and isolating different volcano seismicity signals, therefore, allows a variety of plumbing system processes to be assessed. The majority of volcano monitoring agencies have now deployed or aim to use a network of distributed seismic sensors, including broadband seismometers, to monitor volcano activity (Neuberg et al., 1998; Sparks et al., 2012). Furthermore, an increase in computing power and reduction in cost of seismic sensors means that researchers are now developing fast, fully automated detection and real-time location techniques that can locate seismicity to sub-decimetre precision (e.g. Drew et al., 2013; Sigmundsson et al., 2015). Observations Volcano-tectonic (VT) seismicity generally produces relatively high frequency (1–20 Hz), short period signals, involving clear primary (P), secondary (S) and surface waves, which are caused by displacement on new or existing faults in the host rock in response to fluid-induced stress changes (e.g. Rubin & Gillard, 1998; Roman & Cashman, 2006; Tolstoy et al., 2008). These earthquakes commonly occur near the propagating edge of intrusions, meaning the space-time evolution of VT earthquake locations can be used to track the horizontal and vertical growth of sills and dykes (e.g. Keir et al., 2009; Sigmundsson et al., 2010, 2015). Inflation of a magma or mush body can also induce VT seismicity on any preferentially oriented faults surrounding the intrusion, thereby recording the delivery time and locus of new magma injected into a reservoir (e.g. Roman & Cashman, 2006; Vargas-Bracamontes & Neuberg, 2012). Earthquakes with longer period seismic signals and low-frequencies (0·5–2 Hz) are thought to be generated near the interface between magma and solid rock (Chouet & Matoza, 2013). The earthquake source proximity to the magma causes the seismic signal to resonate in parts of the plumbing system (e.g. conduits, dykes, and cracks), leading to a reduction in its frequency content (Chouet & Matoza, 2013). These earthquakes can potentially be caused by stick-slip motion between the magma and wall-rock or fracturing of cooling magma near the conduit wall (Neuberg et al., 2006; Tuffen et al., 2008). Such earthquakes typically occur at restricted portions of conduits where the magma flow and shear strain rate are highest (Neuberg et al., 2006; Tuffen et al., 2008). Very long period seismicity (VLP) of 10 s of seconds to several minutes period are typically attributed to inertial forces associated with perturbations in the flow of magma and gases through conduits (Chouet & Matoza, 2013). These signals can record the response of the host rock to reservoir inflation and deflation and may be used to model conduit shape and size (Chouet et al., 2008). To do this requires a better understanding of the links between flow processes and resultant pressure/momentum changes using laboratory experiments and numerical models that include the elastic response to magma flow across multiple signal frequency bands (e.g. Thomas & Neuberg, 2012). Implications and integration Studies of evolving reservoirs now aim to link episodes of seismicity related to new magma injection to petrological evidence for timing of reservoir recharge events, thereby providing independent constraints on day to year-long time-scales of magma residence and input prior to eruptions. For example, Fe–Mg diffusion chronometry modelling of orthopyroxene crystals from the 1980–1986 eruption of Mount St. Helens indicates that compositionally distinct rims grew within 12 months prior to eruption (Fig. 4) (Saunders et al., 2012). Peaks in crystal growth correlated extremely well with increased seismicity and SO2 flux (Fig. 4), confirming the relationship between seismicity and magma movement, as well as demonstrating how a combination of seismicity and petrological information can be used to detect magma injections (Saunders et al., 2012). Fig. 4. View largeDownload slide Example of integrating seismology and petrology to constrain time-scales of magma storage and recharge (from Saunders et al., 2012). Calculated Fe–Mg diffusion time scales of orthopyroxene crystals compared to monitoring data for the same eruptive period for Mount St. Helens. (a) The seismic record of depth against time of the 1980–1986 eruption sequence. (b) Measured flux of SO2 gas. (c) Calculated age of orthopyroxene rim growth binned by month for the entire population. The age recorded is the month in which the orthopyroxene rim growth was triggered by magmatic perturbation. The black line displays the running average (over five points, equivalent to the average calculated uncertainty in calculated time scales) of all the data. The peaks in the diffusion time series correspond to episodes of deep seismicity in 1980 and 1982 and to elevated SO2 flux in 1980 and possibly 1982. (d) Running average of the orthopyroxene rim time scales, displaying reverse zonation (Mg-rich rims) in blue and normal zonation (Fe-rich rims) in green. There are reverse zonation peaks in early 1980, probably due to rejuvenation of the magma system by hotter pulses, whereas Fe-rich rims are more dominant from 1982 on. Vertical dashed grey lines represent the volcanic eruptions. Fig. 4. View largeDownload slide Example of integrating seismology and petrology to constrain time-scales of magma storage and recharge (from Saunders et al., 2012). Calculated Fe–Mg diffusion time scales of orthopyroxene crystals compared to monitoring data for the same eruptive period for Mount St. Helens. (a) The seismic record of depth against time of the 1980–1986 eruption sequence. (b) Measured flux of SO2 gas. (c) Calculated age of orthopyroxene rim growth binned by month for the entire population. The age recorded is the month in which the orthopyroxene rim growth was triggered by magmatic perturbation. The black line displays the running average (over five points, equivalent to the average calculated uncertainty in calculated time scales) of all the data. The peaks in the diffusion time series correspond to episodes of deep seismicity in 1980 and 1982 and to elevated SO2 flux in 1980 and possibly 1982. (d) Running average of the orthopyroxene rim time scales, displaying reverse zonation (Mg-rich rims) in blue and normal zonation (Fe-rich rims) in green. There are reverse zonation peaks in early 1980, probably due to rejuvenation of the magma system by hotter pulses, whereas Fe-rich rims are more dominant from 1982 on. Vertical dashed grey lines represent the volcanic eruptions. Petrology and seismicity can also be integrated with other methods, such as GPS and InSAR. Field et al. (2012) analysed volatiles in melt inclusions trapped in phenocrysts within peralkaline lavas from historic eruptions at the Dabbahu Volcano in Afar, Ethiopia. Volatile saturation pressures at typical magmatic temperatures were constrained to be in the range 43–207 MPa, consistent with the phenocryst assemblage being stable at 100–150 MPa. The interpreted magma/mush storage depths for these historic eruptions are ∼1–5 km, consistent with the depths of earthquakes associated with reservoir inflation following dyke intrusion in 2005–2006 (Fig. 5) (Ebinger et al., 2008; Field et al., 2012). Additionally, the best-fit result for modelling of uplift patterns recorded by InSAR data, which were collected over the same time period as seismicity measurements, suggests the magma/mush reservoir comprises a series of stacked sills over a ∼1–5 km depth range (Fig. 5) (Ebinger et al., 2008). The consistency of depth estimates based on petrological study of ancient eruptions, along with the seismicity and inflation of the Dabbahu Volcano following axial dyke intrusion in 2005–2006, implies a vertically extensive and potentially long-lived magma/mush storage region. Such multidisciplinary studies demonstrate that joint observations and modelling of seismic signals, petrological data, and other techniques (e.g. geodesy and gas emissions) significantly strengthen interpretation of the physical structure, emplacement, and evolution of magma plumbing systems. Fig. 5. View largeDownload slide Plot of melt inclusion saturation and earthquake hypocentre depths, which suggest magma storage occurred at 1–5 km depth beneath the Dabbahu volcanic system in Afar, Ethiopia (modified from Field et al., 2012). Melt inclusion data obtained from analyses of alkali feldspar, clinopyroxene, and olivine phenocrysts within Dabbahu lavas <8 Kyr (Field et al., 2012). Earthquake data recorded during the 2005 dyke event (Ebinger et al., 2008). Fig. 5. View largeDownload slide Plot of melt inclusion saturation and earthquake hypocentre depths, which suggest magma storage occurred at 1–5 km depth beneath the Dabbahu volcanic system in Afar, Ethiopia (modified from Field et al., 2012). Melt inclusion data obtained from analyses of alkali feldspar, clinopyroxene, and olivine phenocrysts within Dabbahu lavas <8 Kyr (Field et al., 2012). Earthquake data recorded during the 2005 dyke event (Ebinger et al., 2008). Identifying melt in plumbing systems using seismic imaging Techniques Both active and passive source seismological techniques, which utilise man-made seismic events and natural earthquakes respectively, can be used to identify areas where the presence of partial melt or magma causes a local reduction in seismic wavespeed, an increase in anisotropy, or an increase in attenuation (e.g. Berryman, 1980; Hammond & Humphreys, 2000a, 2000b). With the recent availability of dense seismic networks, resolution of the crust and mantle seismic velocity structure has improved to the degree that active source seismic experiments can: (i) use tomographic techniques to image likely storage regions in the upper crust beneath ocean island volcanoes (e.g. Soufrière Hills Volcano, Montserrat; Fig. 6) (Paulatto et al., 2010; Shalev et al., 2010) and, occasionally, onshore volcanoes (e.g. Mt Erebus, Antarctica, Zandomeneghi et al., 2013; Mt. St. Helens, Kiser et al., 2014); and (ii) utilise reflected data to image individual sills beneath mid-ocean ridges (e.g. Kent et al., 2000; Marjanovic et al., 2014). A further example from Katla volcano Iceland, demonstrates how active source seismic experiments can be used to identify S-wave shadow zones (i.e. S-waves cannot travel through fluids) and delays in P-waves, which may be used to infer the location and geometry of shallow-level magma reservoirs (Gudmundsson et al., 1994). However, recent modelling approaches suggest that the upper crust likely represents only a small portion of magma plumbing systems and long-term storage is dominated by mushy zones throughout the lower crust (e.g. Annen et al., 2006). Active source seismic experiments, particularly on land where the crust is thick and coverage less uniform, cannot penetrate to these depths efficiently. Furthermore, whilst seismic tomographic methods using local earthquakes offer 3 D images of crustal velocity beneath many volcanoes (e.g. Mt. St. Helens, Waite & Moran, 2009; Askja, Iceland, Mitchell et al., 2013), they can only resolve areas directly above the deepest earthquakes. Non-uniform coverage thus makes interpreting tomographic images difficult as resolution varies across the model (see review by Lees, 2007). Fig. 6. View largeDownload slide P-wave (Vp) tomography beneath Montserrat (black outline), highlighting the location of fast and slow seismic velocity anomalies (i.e. >6% faster or slower than average) relative to the location of the Silver Hills (SH), Central Hills (CH), and Soufrière Hills (SHV) volcanoes (modified from Shalev et al., 2010). The fast velocity anomalies, interpreted to represent solidified andesitic intrusions underlie the volcanoes (Shalev et al., 2010). Fig. 6. View largeDownload slide P-wave (Vp) tomography beneath Montserrat (black outline), highlighting the location of fast and slow seismic velocity anomalies (i.e. >6% faster or slower than average) relative to the location of the Silver Hills (SH), Central Hills (CH), and Soufrière Hills (SHV) volcanoes (modified from Shalev et al., 2010). The fast velocity anomalies, interpreted to represent solidified andesitic intrusions underlie the volcanoes (Shalev et al., 2010). To illuminate lower crustal regions, seismologists rely on passive seismology. Extending seismic tomographic images of magma plumbing systems to lower crustal depths requires the use of teleseismic body-wave and surface wave data, which emanate far (>1000 km) from the measurement site. However, these data are dominated by longer period signals, meaning their resolution is relatively low. For example, the Fresnel zone (i.e. the region within ¼ seismic wavelength and an estimate of the minimum resolvable structure) for active source data at 10 Hz is on the order of 3 km in the upper crust compared to 10–15 km for 1 Hz teleseismic data used in receiver function or tomography studies. Observations Active and passive seismological techniques provide crucial insight into transcrustal melt and magma distribution. For example, P-wave seismic travel-time tomography across Monserrat and the Soufrière Hills Volcano images a series of relatively fast seismic velocity zones, which are interpreted as solidified andesitic intrusions, surrounded by regions of slow seismic velocities likely related to either areas of hydrothermal alteration or buried volcaniclastic deposits (Fig. 6) (Paulatto et al., 2010; Shalev et al., 2010). Within the lower crust, inversions using surface wave data generated by ambient seismic noise and receiver function data, which isolates P-wave to S-wave conversions at major discontinuities in the Earth, have identified low shear-wave velocities probably related to melt presence beneath several volcanic settings (e.g. New Zealand, Bannister et al., 2007; Toba, Sumatra, Stankiewicz et al., 2010; Ethiopia, Hammond et al., 2011; Jaxybulatov et al., 2014; Costa Rica, Harmon & Rychert, 2015). When trying to determine how much melt or magma is present, numerous studies have shown that seismic velocities are more sensitive to the shapes of melt/magma-filled spaces on a range of scales compared to the melt fraction (e.g. Hammond & Humphreys, 2000a, 2000b; Miller & Savage, 2001; Johnson & Poland, 2013; Hammond & Kendall, 2016). On the grain-scale, melt commonly wets grain boundaries, forming planar pockets (e.g. Takei, 2002; Garapic et al., 2013; Miller et al., 2014), whereas on the larger scale magma may form planar intrusions of either mush (e.g. Annen et al., 2006), or liquid-rich dykes or sills. If these features are preferentially aligned, they will appear as a distributed region of melt to seismic waves and the analyses described will not be able to discriminate between a melt-poor region dominated by aligned melt-pockets on grain boundaries and an elongate melt-rich body such as an intrusion (e.g. Hammond & Kendall, 2016). A further problem is that seismic velocities are affected by variations in temperature (Jackson et al., 2002), composition (Karato & Jung, 1998), and attenuation (Goes et al., 2012). Therefore, relating seismic velocity anomalies to melt fraction is difficult without some prior knowledge of melt distribution (Hammond & Kendall, 2016). One possible approach to investigate melt distributions further is through measuring seismic anisotropy. If melt has some preferential distribution on a length-scale smaller than the seismic wavelength, such as a stacked network of sills or an anisotropic permeability on the grain scale, then the seismic wavespeed will vary with direction of propagation, i.e. be anisotropic. As a result, measuring the effects of seismic anisotropy allows inferences about sub-seismic wavelength structures, leading to estimates of the preferential orientation of melt distribution. It is common to observe strong anisotropy beneath volcanoes and this has been used to place constraints on melt distribution. For example, high degrees of shear-wave splitting from volcanic earthquakes can either directly map out regions of significant quantities of melt aligned in pockets (Keir et al., 2011), or map out stress changes related to overpressure from injections of magma into the upper crust (Gerst & Savage, 2004; Roman et al., 2011). To image the deeper crustal magmatic system, azimuthal variations in the ratio of P-wave to S-wave speeds (i.e. Vp/VS) from receiver functions led to the interpretation that a stacked network of sills is present in the lower crust beneath the Afar Depression, Ethiopia (Hammond, 2014). Differences in the velocity of Rayleigh Waves and Love Waves, which are vertically polarised shear waves and horizontally polarised shear waves, respectively, suggest a similar anisotropic melt distribution is present beneath the Toba Caldera, Sumatra (Jaxybulatov et al., 2014) and Costa Rica (Harmon & Rychert, 2015). Implications and integration Due to the interference of signals denoting the geometry of melt-filled pockets and the volumetric proportion of that melt, estimating melt fraction remains difficult using seismology alone. Some attempt has been made to directly infer magma/mush reservoir properties from seismic velocities. For example, Paulatto et al., (2012) used thermal modelling to test the range of melt fractions that could account for the low velocity zones imaged in the upper crust beneath Soufrière Hills Volcano (Fig. 6), Montserrat, and concluded the melt fraction is between 3 and 10%. However, accounting for resolution of the tomography, together with uncertainties in the distribution and geometry of melt, means >30% melt may be present more locally in the low velocity zones defined beneath Soufrière Hills Volcano (Paulatto et al., 2012). Possible ways forward involve integrating seismological data with: (i) petrological data that can place limits on likely melt fractions and/or emplacement depths (e.g. McKenzie & O’Nions, 1991; Comeau et al., 2016); (ii) geochemical techniques that can help determine timescales of melt and magma evolution (e.g. Hawkesworth et al., 2000); and (iii) geodetic or other monitoring data, which helps determine magma movement (Sturkell et al., 2006). Recent efforts applying industry software, such as full waveform inversions (FWI; Warner et al., 2013), which is discussed below, are also pushing the potential application of seismological data further and mean that it may be possible to resolve features to sub-kilometre levels, particularly in the upper crust. Together, these techniques may allow us to directly relate seismic velocity anomalies to melt fractions and distributions in the whole crust. Studying magma plumbing systems using gravimetry Techniques Gravimetry measures the gravitational field and its changes over space and time, which can be related to variations in the subsurface distribution and redistribution of mass (e.g. magma). A variety of gravimeter instruments (e.g. free-fall, superconducting, and spring-based) and techniques (e.g. ground-based, sea-floor, ship-borne, and air-borne instrumentations) are available. Spring gravimeters, where a test mass is suspended on a spring, are mostly used to study magmatic and volcanic processes in ground-based surveys (e.g. Carbone et al., 2017; Van Camp et al., 2017). Changes in the gravitational acceleration across a survey area shorten or lengthen the spring, which is recorded electronically and converted to gravity units. These changes are evaluated across a survey network in relation to a reference and are hence termed ‘relative measurements’. Absolute gravimetry can also be measured, i.e. the value of gravitational acceleration, and serves primarily to create a reference frame into which other geodetic methods (e.g. InSAR, GNSS, levelling, relative gravimetry) can be integrated for joint data evaluation. Recent reviews by Carbone et al., (2017) and Van Camp et al., (2017) provide a broad account of gravimetric instruments, measurement protocols, and data processing relevant for the study of magmatic systems. Static gravimetric techniques obtain a single snap-shot of the subsurface mass distribution. For example, Bouguer anomaly maps are perhaps the best-known products of static gravity surveys and capture spatial variations in gravity over an area of interest, providing insight into anomalous mass distribution in the subsurface. Within magmatic studies, computational modelling and inversion of Bouger anomaly data allows identification of shallow intrusions (e.g. dykes and sills; Rocchi et al., 2007), magma-related ore bodies (Hammer, 1945; Bersi et al., 2016), and plutons (e.g. Fig. 7a and b) (e.g. Vigneresse, 1995; Vigneresse et al., 1999; Petford et al., 2000) exhibiting a density contrast with their host rocks. Fig. 7. View largeDownload slide Static and dynamic gravimetric investigations of two active silicic magmatic systems in the Andes: Uturuncu volcano (Bolivia; a, c, e) and the Laguna del Maule volcanic field (Chile; b, d.f). (a) 3D view of the isosurface corresponding to the -120 kg m3 density contrast beneath Uturuncu volcano, derived from Bouguer gravity data, interpreted to reflect a large (∼750 km3) plumbing system composed of a lower (<-10 km) partially molten reservoir and upper, fractured and fluid-bearing solidified intrusions above sea level (after del Potro et al., 2013). (b) 3D view of the -600 kg m3 density contrast isosurface beneath the Laguna del Maule, which is interpreted to define a magma reservoir (>50% melt) within a larger region of a crystal mush system; the 2D planes show slices through the dataset (Miller et al., 2017). Elevation above sea level (a.s.l.) shown. See (d) for area of data coverage. (c) Map of the 55 km long, dynamic gravity network (white circles) installed to track changes in gravity over time and space at Uturuncu volcano between 2010 and 2013 (modified from Gottsmann et al., 2017). (d) Spatio-temporal residual gravity changes at Laguna del Maule recorded from 2013–2014, after correcting for deformation effects (modified from Miller et al., 2017). (e) Gravity and deformation data, recorded from Uturuncu from 2010–2013, plotted against the measured free-air gravity gradient (solid red line) and associated errors (broken red lines) (modified from Gottsmann et al., 2017). The data follow the gradient and are indicative of a subsurface density change as a cause of the uplift, possibly reflecting the release of fluids from a large deep-seated magma reservoir (i.e. the Altiplano-Puna Magmatic Body; Chmielowski et al., 1999) through the vertically extensive crystal mush system shown in (A) (Gottsmann et al., 2017). (f) Plot of gravity against horizontal distance for the source centre at Laguna del Maule (modified from Miller et al., 2017). The increase in gravity of up to 120 µGal is explained by a hydrothermal fluid injection focused along a fault system, shown in (d), at 1·5–2 km depth as a result of a deeper seated magma injection, and is best modelled by a vertical rectangular prism source. Fig. 7. View largeDownload slide Static and dynamic gravimetric investigations of two active silicic magmatic systems in the Andes: Uturuncu volcano (Bolivia; a, c, e) and the Laguna del Maule volcanic field (Chile; b, d.f). (a) 3D view of the isosurface corresponding to the -120 kg m3 density contrast beneath Uturuncu volcano, derived from Bouguer gravity data, interpreted to reflect a large (∼750 km3) plumbing system composed of a lower (<-10 km) partially molten reservoir and upper, fractured and fluid-bearing solidified intrusions above sea level (after del Potro et al., 2013). (b) 3D view of the -600 kg m3 density contrast isosurface beneath the Laguna del Maule, which is interpreted to define a magma reservoir (>50% melt) within a larger region of a crystal mush system; the 2D planes show slices through the dataset (Miller et al., 2017). Elevation above sea level (a.s.l.) shown. See (d) for area of data coverage. (c) Map of the 55 km long, dynamic gravity network (white circles) installed to track changes in gravity over time and space at Uturuncu volcano between 2010 and 2013 (modified from Gottsmann et al., 2017). (d) Spatio-temporal residual gravity changes at Laguna del Maule recorded from 2013–2014, after correcting for deformation effects (modified from Miller et al., 2017). (e) Gravity and deformation data, recorded from Uturuncu from 2010–2013, plotted against the measured free-air gravity gradient (solid red line) and associated errors (broken red lines) (modified from Gottsmann et al., 2017). The data follow the gradient and are indicative of a subsurface density change as a cause of the uplift, possibly reflecting the release of fluids from a large deep-seated magma reservoir (i.e. the Altiplano-Puna Magmatic Body; Chmielowski et al., 1999) through the vertically extensive crystal mush system shown in (A) (Gottsmann et al., 2017). (f) Plot of gravity against horizontal distance for the source centre at Laguna del Maule (modified from Miller et al., 2017). The increase in gravity of up to 120 µGal is explained by a hydrothermal fluid injection focused along a fault system, shown in (d), at 1·5–2 km depth as a result of a deeper seated magma injection, and is best modelled by a vertical rectangular prism source. In contrast to static surveys, dynamic gravimetric observations allow spatio-temporal mass changes to be tracked. Dynamic gravimetric studies investigate how the subsurface architecture changes over time and is usually performed by measuring variations in gravity across a network of survey points (e.g. Fig. 7c) or, in a few exceptional cases, by installing a network of continuously operating gravimeters. Dynamic observations demand one-to-two orders of magnitude higher data precision (i.e. to a few µGal where 1 µGal = 10-8 m/s2) compared to static surveys, making them an elaborate and time-consuming exercise. However, dynamic gravity data yields important insights into the source processes behind non-tectonic volcano and crustal deformation, particularly if combined with surface deformation data (e.g. InSAR and GNSS) as subsurface mass and volume changes can be employed to characterise the density of the material behind the stress changes (Figs 7c–f and 8) (e.g. Jachens & Roberts, 1985; Battaglia & Segall, 2004; Poland & Carbone, 2016). There are also cases where volcano unrest, due either to magma intrusion into a ductile host rock or to volatile migration at shallow depths, does not result in resolvable surface deformation; in these scenarios, gravity data have provided vital clues about subsurface processes otherwise hidden from conventional monitoring techniques (e.g. Gottsmann et al., 2006,, 2007; Miller et al., 2017). Fig. 8. View largeDownload slide Gravity changes and deformation at the restless Long Valley caldera. (a) Map of the Long Valley caldera, California, USA, which hosts a resurgent dome (black outline), to highlight changes in residual gravity between 1982 and 1999 (modified from Tizzani et al., 2009). (b) Plot of ground uplift and residual gravity changes with radial distance from the centre of the resurgent dome in (a) (modified from Tizzani et al., 2009). The correlation between uplift and positive gravity residuals across the resurgent dome indicates ground deformation was instigated by intrusion of magma (Tizzani et al., 2009). Fig. 8. View largeDownload slide Gravity changes and deformation at the restless Long Valley caldera. (a) Map of the Long Valley caldera, California, USA, which hosts a resurgent dome (black outline), to highlight changes in residual gravity between 1982 and 1999 (modified from Tizzani et al., 2009). (b) Plot of ground uplift and residual gravity changes with radial distance from the centre of the resurgent dome in (a) (modified from Tizzani et al., 2009). The correlation between uplift and positive gravity residuals across the resurgent dome indicates ground deformation was instigated by intrusion of magma (Tizzani et al., 2009). Whilst static and dynamic gravimetric observations offer considerable insight into the structure and dynamics of magma plumbing systems, care must be exercised when collecting and interpreting gravity data from active magmatic areas where seasonal variations in hydrothermal systems, aquifers, or the vadose zone can influence subsurface mass distribution (e.g. Hemmings et al., 2016). These seasonal changes can, in some cases, result in data aliasing artefacts and inhibit the quantification of deeper-seated magmatic processes (e.g. Gottsmann et al., 2005,, 2007). Observations Gravimetric investigations have been at the heart of studies into the subsurface structure of active and ancient magma plumbing systems for more than 80 years (e.g. Carbone et al., 2017; Van Camp et al., 2017). Using techniques initially designed for imaging salt domes, silicic plutons were the first components of magma plumbing systems to be examined using gravimetry because their low density relative to surrounding rocks produces clear, negative gravity anomalies of ∼10 to ∼40 mGal amplitude (e.g. Reich, 1932; Bucher, 1944; Bott, 1953). Gravity data have been instrumental in the investigation of upper-crustal, silicic magma plumbing systems, helping to reveal: (i) the 3 D geometry of plutons by allowing floor morphologies (e.g. flat-floored or wedge-shaped) to be determined (e.g. Vigneresse et al., 1999; Petford et al., 2000); and (ii) how plutons are constructed, for example, by the amalgamation of multiple intrusions fed from depth by dykes (e.g. Vigneresse, 1995). Furthermore, recent high-precision static surveys over active silicic volcanoes have enabled detailed modelling of the sub-volcanic magma plumbing system, commonly demonstrating the occurrence of vertically extensive, transcrustal magma bodies (Fig. 7a and b) (e.g. Gottsmann et al., 2008; del Potro et al., 2013; Saxby et al., 2016; Miller et al., 2017). In addition to examining silicic magma plumbing systems, negative gravity anomalies with typical amplitudes of up to 60 mGal and up to 100 km wavelength can be associated with, and provide insight into, the geometry and size of silicic ash-flow calderas (e.g. Eaton et al., 1975; Masturyono et al., 2001). Positive gravity anomalies with amplitudes of up to 30 mGal and wavelengths of up to 20 km are commonly identified at mafic volcanoes and likely result from dense intrusive complexes (e.g. Rymer & Brown, 1986). Dynamic gravity observations have provided unprecedented insight into the evolution of magma plumbing systems over timescales of seconds to decades, including: (i) the characterisation of multi-year lava lake dynamics (e.g. Poland & Carbone, 2016); (ii) mass budgets of magma intrusions (e.g. Fig. 8) (e.g. Battaglia et al., 1999; Jousset et al., 2000; Rymer et al., 2005; Bonforte et al., 2007; Tizzani et al., 2009); (iii) shallow hydrothermal fluid flow processes induced by deeper magmatic unrest (e.g. Battaglia et al., 2006; Gottsmann et al., 2007; Miller et al., 2017); and (iv) parameters of magmatic geothermal reservoirs (e.g. Hunt & Bowyer, 2007; Sofyan et al., 2011). For example, using data from a network of continuously recording gravimeters, Carbone et al., (2013) calculated the density of the Kilauea lava lake as 950 ± 300 kg m3, i.e. similar to and potentially less than that of water, suggesting that the magma column within the upper portions of the volcanic edifice is gas-rich. Because density and volatile content are critical controls on magma rheology, identification of a gas-rich magma column and lava lake at Kilauea is crucial to modelling and understanding convection and eruption dynamics (Carbone et al., 2013). Implications and integration The advent of data-rich geodetic observations from satellite-remote sensing (e.g. InSAR), in conjunction with spatio-temporal gravity studies, provides unprecedented opportunities to characterise magma plumbing system dynamics and the driving mechanisms behind volcano deformation. At Long Valley caldera, for example, a residual gravity increase of more than 60 μGal between 1982 and 1999 indicates a mass addition at depth (Battaglia et al., 1999). Joint inversion of InSAR and gravity data from Long Valley derives a best-fit source density of 2509 kg m3 and is indicative of a magmatic intrusion (Fig. 8) (Tizzani et al., 2009). At the deforming Laguna del Maule volcanic centre, Chile, multi-year InSAR and dynamic gravity records demonstrate that uplift and extension above an inflating sill-like reservoir at ∼5 km depth promoted migration of hydrothermal fluids along a fault to shallow (1–2 km) depths (Miller et al., 2017). Alternatively, although no ground deformation is observed at Tenerife, Canary Islands, Spain, deconvolution of dynamic gravity into a shallow and deep gravity field provides evidence of unrest (Prutkin et al., 2014). The gravity data suggest hybrid processes have generated the unrest, whereby fluids were released and migrated upward along deep-rooted faults from an intrusion at ∼9 km beneath the summit of Teide Volcano (Prutkin et al., 2014). Overall, combining ground deformation and gravimetric observations has highlighted complex processes both within magma reservoirs (e.g. mass addition by magma input, density decrease by volatile exsolution, or density increase by crystallisation; Fig. 7c–f) and in the surrounding host rock (e.g. migration of magmatic fluids, phase changes in hydrothermal systems). Key to a better understanding of the processes governing these magma plumbing system and volcano deformation dynamics is the integration of gravimetric and geodetic data with other geophysical data (e.g. seismicity or magnetotellurics) and petrological data. Coupled with advanced numerical modelling, such multi-parameter studies promise exciting new insights into the inner workings of sub-volcanic magma plumbing systems (e.g. Currenti et al., 2007; Hickey et al., 2016; Currenti et al., 2017; Gottsmann et al., 2017; Miller et al., 2017). Resolving magma plumbing system structure with electromagnetic methods Techniques Electromagnetic (EM) methods probe subsurface electrical resistivity or its inverse, i.e. electrical conductivity. Spatial variations in resistivity control the position, strength, and geometry of local electrical eddy currents and the magnetic fields they produce. These electrical eddy currents are induced by time-varying, naturally occurring magnetic fields external to Earth, which forms the basis of the magnetotelluric (MT) technique, or by controlled sources. Monitoring these decaying electrical and magnetic fields, therefore, allows the subsurface resistivity distribution to be inferred. Controlled source methods generally probe only the shallow subsurface, but MT has a greater depth range as it uses longer-period signals to penetrate deeper. The signals propagate diffusively, which means EM methods typically have a lower resolution than seismic techniques. However, melt, magma, and magmatic hydrothermal fluids are generally considerably less resistive than solid rock and can thus easily be detected by EM methods, which are sensitive to conductive materials (e.g. Whaler & Hautot, 2006; Wannamaker et al., 2008; Desissa et al., 2013; Comeau et al., 2015). Therefore, EM methods, particularly MT, have been used extensively to study magmatic systems in various tectonic settings. MT equipment, data acquisition and processing are described by Simpson & Bahr (2005) and Ferguson (2012). Measured field variations have very low amplitudes, meaning equipment needs to be positioned and installed carefully to reduce vibrational (e.g. from wind, vegetation, or vehicles) and electrical (e.g. from power lines) noise. If data are recorded synchronously at a second, less noisy site, remote reference methods can be used to improve the data quality (e.g. Gamble et al., 1979). One further problem is that small-scale resistivity anomalies in the shallow subsurface generate galvanic (non-inductive) effects that distort MT data. The distortion is identified and corrected for, which may involve using controlled source transient electromagnetic data to ensure complete removal (e.g. Sternberg et al., 1988), at the same time as assessing whether the data can be modelled with a one-, two- or three-dimensional resistivity structure (e.g. Jones, 2012). Failure to remove galvanic distortion can result in models having resistivity features at the wrong depth. For example, there has been controversy as to whether a conductor beneath Vesuvius Volcano, Italy is caused by a deep (∼8–10 km depth) magma reservoir (Di Maio et al., 1998) or a shallow brine layer (Manzella et al., 2004). These and other factors can be a significant problem when using MT to study magmatic systems, especially on volcanic islands. The relationship between MT data and subsurface resistivity is strongly non-linear meaning that inversion is fundamentally non-unique and computationally expensive (e.g. Bailey, 1970; Parker, 1980; Weaver, 1994). Most practical algorithms for inverting MT data obtain a unique result by minimising a combination of misfit to the data and a measure of model roughness (e.g. Constable et al., 1987). This approach poorly delimits how magma is distributed in the subsurface, whether it is in sills, dykes, or larger reservoirs (Johnson et al., 2016). Whilst MT data are sensitive to the top surface of a conductor, its base may not be detected because conductive material reduces the penetration depth of the signal. Sensitivity analysis is used to ascertain the model features required to fit the MT data, which allows a conductor to be confined to a certain depth range and thereby constrains its base (e.g. Desissa et al., 2013). Furthermore, if the resistivity of a conductor can be inferred, its conductance (i.e. a product of thickness and conductivity) can be used to determine its thickness (e.g. Comeau et al., 2016). Observations EM induction surveys have been conducted on most major sub-aerial volcanoes and magmatic systems; only a few will be mentioned here to illustrate the type of information on magma plumbing systems that has been obtained. MT data have been used to image several low resistivity features in the central Andes, particularly beneath the uplifting (10–15 mm/yr) Volcán Uturuncu, Bolivia (Fig. 9a) (Comeau et al., 2015,, 2016). The deepest of these bodies has resistivities of <3 Ω m, has a top contact at ∼15–20 km depth (beneath Uturuncu), likely has a thickness of >6 km, and extends E–W for ∼170 km (Fig. 9) (Comeau et al., 2015,, 2016). This large-scale structure is interpreted to be the Altiplano-Puna magma body (APMB), which has been identified in other geophysical datasets (e.g. Fig. 7a) (e.g. gravimetry, del Potro et al., 2013), with its low resistivity attributed to the presence of at least 20% andesitic melt and/or magma. Extending from the top of the APMB towards the surface are several vertical, narrow (<10 km wide), low resistivity (<10 Ω m) zones that coincide with areas of seismicity and negative gravity anomalies (Fig. 9). These zones likely reflect a network of dykes and upper crustal magma reservoirs (Jay et al., 2012; del Potro et al., 2013; Comeau et al., 2015,, 2016). Fig. 9. View largeDownload slide (a) Map showing MT stations deployed around Volcán Uturuncu (U) and Volcán Quetena (Q), relative to areas of uplift and subsidence (modified from Comeau et al., 2015). The white box shows area of modelled 3D MT data (Comeau et al., 2015). (b) Regional 2D magnetotelluric line through the Altiplano-Puna magma body (APMB) highlighting the position of Volcán Uturuncu (modified from Comeau et al., 2015). The APMB corresponds to a large, conducive (i.e. low-resistivity) body (Comeau et al., 2015, 2016). Above the APMB are other areas of low-resistivity (e.g. C4) that are likely upper crustal magma reservoirs and dykes (Comeau et al., 2016). C1–C7 and R1–R2 identify discrete zones of marked conductivity or resistivity, respectively (see Comeau et al., 2015, 2016 for details). The white box shows area of modelled 3D MT data (Comeau et al., 2015). See (a) for location. Fig. 9. View largeDownload slide (a) Map showing MT stations deployed around Volcán Uturuncu (U) and Volcán Quetena (Q), relative to areas of uplift and subsidence (modified from Comeau et al., 2015). The white box shows area of modelled 3D MT data (Comeau et al., 2015). (b) Regional 2D magnetotelluric line through the Altiplano-Puna magma body (APMB) highlighting the position of Volcán Uturuncu (modified from Comeau et al., 2015). The APMB corresponds to a large, conducive (i.e. low-resistivity) body (Comeau et al., 2015, 2016). Above the APMB are other areas of low-resistivity (e.g. C4) that are likely upper crustal magma reservoirs and dykes (Comeau et al., 2016). C1–C7 and R1–R2 identify discrete zones of marked conductivity or resistivity, respectively (see Comeau et al., 2015, 2016 for details). The white box shows area of modelled 3D MT data (Comeau et al., 2015). See (a) for location. Monitoring of magmatic systems can also be undertaken by both time-lapse and continuous EM measurement. For example, MT data collected immediately after the 1977–1978 eruption at Usu volcano, Japan revealed a conductive zone (<100 Ω m) beneath the summit that probably corresponded to intruded magma. By 2000, MT data revealed that this conductive body had become resistive (500–1000 Ω m) as the intrusion cooled, from 800°C to 50°C, and crystallised (Matsushima et al., 2001). Continuous MT monitoring of Sakurajima volcano, Japan between May 2008 and July 2009 revealed temporal changes in resistivity in the data, some of which were correlated to periods of surface deformation and were inferred to reflect mixing between groundwater and volatiles exsolved from an underlying magma body (Aizawa et al., 2011). Continuous MT monitoring at La Fournaise, Réunion Island recorded apparent resistivity decreases associated with the large 1998 eruption, which were attributed to the injection of a N–S striking dyke (Wawrzyniak et al., 2017). Several EM studies have focussed on magma plumbing systems at divergent margins, including mid-ocean ridges and continental rifts. For example, at the fast-spreading East Pacific Rise, a ∼10 km wide, sub-vertical conductor, slightly displaced from the ridge axis and connected to a deep, broad conductive zone was interpreted as a channel efficiently transporting melt to the base of the crust (Baba et al., 2006; Key et al., 2013). Imaging of a crustal conductor for the first time beneath a slow-spreading ridge, i.e. the Reykjanes Ridge in the Atlantic Ocean, suggests that magma injection into crustal reservoirs is intermittent but rapid (MacGregor et al., 1998; Heinson et al., 2000). Conversely, slow-spreading continental rifting in the Dabbahu magma segment, Afar, Ethiopia, appears to be underlain by a large conductor, either at the top of the mantle or straddling the Moho, containing more melt (>300 km3) than is intruded during a typical rifting episode (Desissa et al., 2013). The volume of this large conductor implies it is a long-lived feature that could source magmatic activity for tens of thousands of years (Desissa et al., 2013). Implications and integration It is clear from MT studies of the APMB that other geophysical techniques aid and/or corroborate data interpretation (Fig. 9) (e.g. Comeau et al., 2015,, 2016). Over the last two decades, numerous geophysical studies have been applied to examine magma and melt distribution beneath various portions of the East African Rift, providing an excellent opportunity to test how different techniques and data can be integrated. For example, extensive zones of melt beneath the Afar region in Ethiopia inferred from MT data by Desissa et al., (2013) are supported by: (i) the occurrence of coincident, low P-wave velocity (down to 7·2 km s-1) zones identified from analysis of seismic Pn waves that propagate along the Moho (Stork et al., 2013); (ii) surface wave studies that reveal lower crustal areas in magmatic domains with low S-wave velocities (∼3·2 km s-1) (Guidarelli et al., 2011); and (iii) high Vp/VS ratios and low amplitude receiver functions, which are indicative of the presence of melt (Hammond et al., 2011; Hammond, 2014). Similarly, crustal conductors along the northern flanks of the Main Ethiopian Rift, interpreted to represent melt or magma (Whaler & Hautot, 2006; Samrock et al., 2015), coincide with locations where receiver functions either have amplitudes too low to interpret or indicate high Vp/VS values (Dugda et al., 2005; Stuart et al., 2006). Electrical anisotropy can be inferred directly from MT data consistent with a two-dimensional subsurface resistivity distribution (Padilha et al., 2006; Hamilton et al., 2006). Large amounts of electrical anisotropy were found in the lower crust beneath Quaternary magmatic segments in Afar, Ethiopia, where there is also significant crustal seismic anisotropy (see figure 11 of Ebinger et al., 2017); oriented melt-filled pockets are the probable cause of both. Although EM methods can image subsurface conductors that are interpreted to represent magma bodies or zones of partial melt (i.e. crystal mushes), additional information is required to determine their composition, volume, and/or melt fraction. However, there are several challenges in inverting measured bulk resistivities to recover this information. Two-phase mixing laws predict bulk resistivity is primarily a function of melt resistivity and geometry in the rock matrix when the fluid phase has low resistivity, as in the case of partial melt. Well-connected melt gives a lower bulk resistivity than isolated melt pockets, for the same melt fraction and resistivity (e.g. Hashin & Shtrikman, 1962; Roberts & Tyburczy, 1999; Schmeling, 1986). Whilst resistivities of basaltic and rhyolitic melts have been measured in laboratory experiments (e.g. Laumonier et al., 2015; Guo et al., 2016), they are strongly dependent on temperature, pressure, silica, sodium and water content, making extrapolation uncertain. The web-based SIGMELTS tool can, however, be used to predict melt and bulk resistivities for a wide range of compositions and conditions (Pommier & Le Trong, 2011). Importantly, petrological and geochemical characterisation of eruptive products can help inform interpretations of associated, subsurface conductors, but it is difficult to ascertain either whether their composition reflects the current magma/melt present in the plumbing system or whether melt pockets are interconnected. These large uncertainties in the resistivity of the melt and the requirement to make assumptions about its geometry mean direct inference of melt fraction is difficult. Nonetheless, information from laboratory studies, petrology and geochemistry aids interpreting resistivity anomalies in magmatic regions (see review by Pommier, 2014). Imaging ancient magma plumbing systems in seismic reflection data Techniques Over the last two decades, major advances have been made in imaging deep crustal melt beneath active volcanic terrains using P- and S-wave tomographic data (e.g. Yellowstone, Husen et al., 2004; Mt. St. Helens, Lees, 2007; Hawaii, Okubo et al., 1997). These data image deep (>7 km), often laterally extensive (up to 20 km), sill-like magma reservoirs (e.g. Paulatto et al., 2012). However, like many geophysical and geodetic techniques applied to study active magma plumbing systems, these data typically lack the spatial resolution to resolve the detailed geometry of pathways transporting magma to the Earth’s surface. Active source seismic reflection data, which have a spatial resolution of metres-to-decametres down to depths of ∼5 km, can provide unprecedented images of and insights into the geometry and dynamics of shallow-level, crystallised, magma plumbing systems (e.g. Fig. 10) (e.g. Planke et al., 2000; Smallwood & Maresh, 2002; Thomson & Hutton, 2004; Cartwright & Hansen, 2006; Jackson et al., 2013; Magee et al., 2016; Schofield et al., 2017). Whilst seismic reflection data are traditionally used to find and assist in the production of hydrocarbons in sedimentary basins (Cartwright & Huuse, 2005), here we discuss and support its application to volcanological problems. Fig. 10. View largeDownload slide (a) Interpreted seismic section and geological map showing the distribution of, and connectivity between, sills within the Faroe-Shetland Basin (modified from Schofield et al., 2017). Mapping of magma flow patterns within individual sills reveals that the sill-complex facilitates extensive vertical and lateral magma transport. Magma was fed into the sedimentary basin via basement-involved faults. TWT, two-way travel time. (b) Interpreted seismic section and geological map describing the spatial relationship between volcanoes/vents and sills, inferred to represent the magma plumbing system, emplaced at ∼42 Ma (modified from Jackson et al., 2013; Magee et al., 2013b). Sills are laterally offset from the volcanoes/vents summits. No ‘magma chambers’ are observed in the seismic data, which images down to ∼8 s TWT (i.e. ∼10 km) (Magee et al., 2013b). Fig. 10. View largeDownload slide (a) Interpreted seismic section and geological map showing the distribution of, and connectivity between, sills within the Faroe-Shetland Basin (modified from Schofield et al., 2017). Mapping of magma flow patterns within individual sills reveals that the sill-complex facilitates extensive vertical and lateral magma transport. Magma was fed into the sedimentary basin via basement-involved faults. TWT, two-way travel time. (b) Interpreted seismic section and geological map describing the spatial relationship between volcanoes/vents and sills, inferred to represent the magma plumbing system, emplaced at ∼42 Ma (modified from Jackson et al., 2013; Magee et al., 2013b). Sills are laterally offset from the volcanoes/vents summits. No ‘magma chambers’ are observed in the seismic data, which images down to ∼8 s TWT (i.e. ∼10 km) (Magee et al., 2013b). Acquiring active source seismic reflection data involves firing acoustic energy (i.e. seismic waves) into the subsurface and measuring the surface arrival times (i.e. the travel-time) of reflected energy. Processing of these arrival time data allows reconstruction of the location and geometry of the geological interfaces from which acoustic energy was reflected. Mafic intrusive igneous rocks are generally well-imaged in seismic reflection data because they typically have greater densities (>2·5 g/cm3) and acoustic velocities (i.e. >4000 m s) than encasing sedimentary strata; these differences result in a high acoustic impedance contrast, causing more seismic energy to be reflected back to the surface compared to low acoustic impedance boundaries (Smallwood & Maresh, 2002; Brown, 2004). In contrast, silicic igneous rocks have similar acoustic properties to encasing sedimentary strata, meaning that felsic intrusions are rarely imaged in seismic reflection data (Mark et al., 2018; Rabbel et al., 2018)). Furthermore, because reflection seismology relies on the return of acoustic energy to the surface, seismic reflection data favourably image mafic, sub-horizontal-to-moderately inclined intrusions (e.g. sills, inclined sheets, and laccoliths; Smallwood & Maresh, 2002; Jackson et al., 2013; Magee et al., 2016). Sub-vertical dykes reflect only a limited amount of acoustic energy back to the surface and are thus typically poorly imaged in seismic reflection data (e.g. Smallwood & Maresh, 2002; Planke et al., 2005; Thomson, 2007; Wall et al., 2010; Eide et al., 2017b; Phillips et al., 2017). Observations Sills and inclined sheets are commonly observed in seismic reflection data as laterally discontinuous, high-amplitude reflections, which may cross-cut the host rock strata (Fig. 10) (e.g. Symonds et al., 1998, Smallwood & Maresh, 2002; Planke et al., 2005, Magee et al., 2015). Many of the sills and inclined sheets imaged in seismic reflection data are, however, expressed as tuned reflection packages, whereby discrete reflections from the top and base contacts interfere on their return to the surface and cannot be distinguished (e.g. Figs 10 and 11a) (e.g. Smallwood & Maresh, 2002; Peron-Pinvidic et al., 2010; Magee et al., 2015; Eide et al., 2017b; Rabbel et al., 2018). Therefore, it is difficult to assess either intrusion thicknesses, or to detect whether imaged sills are composite bodies made of numerous, stacked, thin sheets. Either way, subtle vertical offsets and corresponding amplitude variations of sill reflections can often be mapped, defining linear structures that radiate out from either the central, deepest portions of sills or areas where underlying intrusions intersect the sill (e.g. Schofield et al., 2012b; Magee et al., 2014,, 2016). These structures are interpreted to relate to magma flow indicators such as intrusive steps, broken bridges, and magma fingers (e.g. Schofield et al., 2010, 2012a; Magee et al., 2017b). Fig. 11. View largeDownload slide (a) Interpreted seismic section from the Exmouth Sub-basin offshore NW Australia, which images a saucer-shaped sill that is overlain by a forced fold and feeds a small vent from its inclined limb (modified from Magee et al., 2013a). See Fig. 11b for line location). (b) Time-structure map of the folded horizon (thick black line) in (a), highlighting fault traces and vent locations and thicknesses (modified from Magee et al., 2013a). (c) Seismic section from the Farsund Basin, offshore southern Norway, which images part of a dyke-swarm that has been rotated by basin flexure post-emplacement (modified from Phillips et al., 2017). Fig. 11. View largeDownload slide (a) Interpreted seismic section from the Exmouth Sub-basin offshore NW Australia, which images a saucer-shaped sill that is overlain by a forced fold and feeds a small vent from its inclined limb (modified from Magee et al., 2013a). See Fig. 11b for line location). (b) Time-structure map of the folded horizon (thick black line) in (a), highlighting fault traces and vent locations and thicknesses (modified from Magee et al., 2013a). (c) Seismic section from the Farsund Basin, offshore southern Norway, which images part of a dyke-swarm that has been rotated by basin flexure post-emplacement (modified from Phillips et al., 2017). A recurring observation from seismic reflection-based studies of extinct and buried intrusive systems is that complexes of interconnected sills and inclined sheets, which may cover >3 × 106 km2, can dominate magma plumbing systems (e.g. Fig. 10b) (e.g. Svensen et al., 2012; Magee et al., 2016). Importantly, where buried volcanic edifices are imaged in seismic reflection data, they rarely appear to be underlain by ‘magma chambers’ (i.e. a spheroidal or ellipsoidal body of now-crystallised magma). Instead, these imaged volcanoes commonly appear laterally offset from genetically related sills and/or laccoliths that are inferred to represent their feeder reservoirs (e.g. Fig. 10b) (Magee et al., 2013b; McLean et al., 2017). The geometry, location and connectivity of these intrusions, which can represent magma storage sites and conduits to the surface, are often heavily influenced by both the host rock structure and lithology (see review by Magee et al., 2016). For example, magma may flow along pronounced discontinuities (e.g. bedding) or within specific stratigraphic units (e.g. coal) for considerable distances, occasionally climbing to higher stratigraphic levels by instigating deformation of the host rock or by exploiting pre-existing faults (e.g. Jackson et al., 2013; Magee et al., 2016; Schofield et al., 2017; Eide et al., 2017a). It is clear from seismic reflection data that shallow-level tabular intrusions are commonly accommodated by roof uplift to form a flat-topped or dome-shaped forced fold (e.g. Fig. 11a and b) (e.g. Trude et al., 2003; Hansen & Cartwright, 2006; Jackson et al., 2013; Magee et al., 2013a). Moreover, if the age of reflections onlapping onto these intrusion-induced forced folds can be ascertained, the timing and to some extent the duration of magmatic activity can be determined (e.g. Trude et al., 2003; Hansen & Cartwright, 2006; Magee et al., 2014; Reeves et al., 2018). Although most seismic-based studies examine intrusions within sedimentary basins, saucer-shaped sills and laterally extensive sill-complexes emplaced into crystalline basement rock are also imaged (e.g. Ivanic et al., 2013; McBride et al., 2018). Lastly, seismic reflection data can also be used to image the internal structure of layered ultramafic–mafic intrusions (e.g. the Bushveld Layered Intrusion, Malehmir et al., 2012) and, in some instances, identify dykes (e.g. Fig. 11c) (e.g. Wall et al., 2010; Abdelmalak et al., 2015; Bosworth et al., 2015; Phillips et al., 2017). Implications and integration Despite being limited in terms of their spatial resolution (typically a few tens of metres) and ability to image steeply dipping features (i.e. dykes), they provide unprecedented snapshots into the final 3 D structure of magma plumbing systems. Beyond quantifying the structure and connectivity of magma plumbing systems, seismic-based studies have shown that: (i) magma flow patterns mapped across entire sill-complexes indicate they can transport melt from source to surface over great lateral (>100 s km) and vertical distances (10 s km), potentially without significant input from dykes (Fig. 10a) (e.g. Thomson & Hutton, 2004; Cartwright & Hansen, 2006; Magee et al., 2014, 2016; Schofield et al., 2017); and (ii) a variety of elastic and inelastic mechanisms can accommodate host rock deformation during magma emplacement, meaning that the location and size of ground deformation does not necessarily equal that of the forcing intrusion (e.g. Jackson et al., 2013, Magee et al., 2013a). Importantly, observations from seismic reflection data highlight that the lateral dimension should be considered when modelling the transit of magma in the crust, posing problems for the widely held and simple assumption that magma simply travels vertically from melt source to eruption site. Seismic-based studies have also shown that direct comparison to active deformation structures can be informative. For example, through comparing mapped lava flows and structures associated with the Alu dome to similar features observed in seismic reflection data, Magee et al., (2017a, 2017b) concluded that the shallow-level sill likely has a saucer-shaped, as opposed to the sill-like tabular morphology inferred from an episode of deformation measured using InSAR (Fig. 3c and d). Despite its benefits, it is important to remember that seismic reflection data typically reveal only the final geometry of the magma plumbing system. Thus there remains a challenge in using these data to understand areas where deformation captures potentially transient, active processes, rather than structures resulting from (multiple) periods of intrusion and cooling (Reeves et al., 2018). One potential and exciting way forward is the development of Virtual Reflection Seismic Profiling, by which microseismicity at active volcanoes may be used to image magma reservoirs and subsurface structure in 4 D (Kim et al., 2017). Although challenges exist in dataset integration, the imaging power afforded by modern seismic reflection data thus presents a unique opportunity to further unite field-, petrological-, geochemical- and other geophysical-based analyses within more realistic structural frameworks (e.g. Figs 3, 11a and b). In our view, however, seismic reflection data are under-utilised in igneous research, remaining an unfamiliar technique to many Earth Scientists in the volcanic and magmatic community. Rock magnetism Technique Whilst seismic reflection data provide unique 3 D images of ancient magma plumbing systems, which can be used to infer magma flow patterns across entire intrusion networks, we commonly lack sufficient data (e.g. boreholes) to test seismic-based hypotheses. Therefore, it is critical to compare seismic interpretations to field analogues where magma flow patterns, emplacement mechanics, and intrusion evolution can be investigated via other techniques. In this section, we examine how rock magnetic analyses can be used to systematically study magnetic mineralogy and petrofabrics, thereby illuminating the structure and history of igneous intrusions. There are two principal types of rock magnetic study: magnetic remanence and magnetic susceptibility, where the total magnetisation (M) of a rock is the sum of the magnetic remanence (Mrem) and the induced magnetisation (Mind), which is a product of the susceptibility (K) and applied field strength (H) (Dunlop & Özdemir, 2001). Remanence carries a geological record of the various magnetisations acquired over time and is central to palaeomagnetic studies. However, we focus on magnetic fabric analysis, which relies on measurements of the anisotropy of magnetic susceptibility (AMS). The AMS signal of a rock carries information from all constituent grains. Although mineral phases that have a paramagnetic behaviour (i.e. they are weakly attracted to externally applied magnetic fields) volumetrically dominate most igneous rocks (e.g. olivine, clinopyroxene, biotite), ferromagnetic mineral phases (e.g. titanomagnetite) are highly susceptible to magnetization and, therefore, tend to dominate K (e.g. Dunlop & Özdemir, 2001; Biedermann et al., 2014). Magnetic fabrics, therefore, typically reflect the preferential orientation of crystallographic axes (i.e. crystalline anisotropy), the shape-preferred orientation of individual crystals (i.e. shape anisotropy) and/or the alignment of closely spaced crystals (i.e. distribution anisotropy) of Fe-bearing silicate and oxide phases (e.g. Voigt & Kinoshita, 1907; Graham, 1954; Hrouda, 1982; Tarling & Hrouda, 1993; Dunlop & Özdemir, 2001). The principal axes of the magnetic fabrics measured by AMS can thus be related to the orientation, shape and distribution of individual grains (i.e. the petrofabric) (e.g. Fig. 12a). Fig. 12. View largeDownload slide (a) At the sample scale, all magnetic grains create a magnetic fabric. (i) Dominantly prolate fabric, where K2 and K3 are least certain and form a girdle. Only the magnetic lineation (K1) can be confidently determined. (ii) When K1>K2>K3, both a foliation (K1–K2) and a lineation (K1) may be discerned, defining a triaxial fabric. (iii) When K1 and K2 are equally uncertain and form a girdle, K3 is perpendicular to a foliation. (b) Schematic representation of how magma flow within a planar sheet intrusion can produce imbricated magnetic fabrics at its margins, the closure of which define the magma flow direction (after Féménias et al., 2004). (c) AMS data and interpretations from part of the Trawenagh Bay Granite, NW Ireland (adapted from Stevenson et al., 2007a). (i) AMS foliation traces are shown in blue and lineation traces in red. Lobes were defined in this intrusion based on foliations curving around a lineation axis. In some lobes, the magnetic lineation trend was parallel to this axis, whilst in others they tended to splay or converge down flow. (ii) 3D sketch showing the geometry of three of the lobes (numbered in part i). Fig. 12. View largeDownload slide (a) At the sample scale, all magnetic grains create a magnetic fabric. (i) Dominantly prolate fabric, where K2 and K3 are least certain and form a girdle. Only the magnetic lineation (K1) can be confidently determined. (ii) When K1>K2>K3, both a foliation (K1–K2) and a lineation (K1) may be discerned, defining a triaxial fabric. (iii) When K1 and K2 are equally uncertain and form a girdle, K3 is perpendicular to a foliation. (b) Schematic representation of how magma flow within a planar sheet intrusion can produce imbricated magnetic fabrics at its margins, the closure of which define the magma flow direction (after Féménias et al., 2004). (c) AMS data and interpretations from part of the Trawenagh Bay Granite, NW Ireland (adapted from Stevenson et al., 2007a). (i) AMS foliation traces are shown in blue and lineation traces in red. Lobes were defined in this intrusion based on foliations curving around a lineation axis. In some lobes, the magnetic lineation trend was parallel to this axis, whilst in others they tended to splay or converge down flow. (ii) 3D sketch showing the geometry of three of the lobes (numbered in part i). Regardless of whether mineral phases crystallise early or late, whereby their orientation and distribution typically mimics the earlier silicate framework, it is expected that the initial petrofabric developed in intrusive rocks will likely be sensitive to alignment of crystals during primary magma flow. However, it is also critical to recognise that later magmatic processes (e.g. convection and melt extraction) and syn- or post-emplacement tectonic deformation can modify or overprint primary magma flow fabrics during intrusion, solidification (i.e. mush development), or sub-solidus conditions (e.g. Borradaile & Henry, 1997; Bouchez, 1997; O’Driscoll et al., 2015; Kavanagh et al., 2018). Whilst anisotropy of magnetic susceptibility (AMS) can thus rapidly and accurately detect weak or subtle mineral alignments within igneous intrusions, which may be attributable to magmatic and/or tectonic processes, evaluating the origin and evolution of petrofabric development requires additional information (e.g. Borradaile & Henry, 1997; Bouchez, 1997). For example, shape-preferred orientation analyses and comparison to visible flow indicators (e.g. intrusive steps and bridge structures) allow magma flow axes and directions that have been inferred from magnetic fabrics to be verified (e.g. Launeau & Cruden, 1998; Callot et al., 2001; Magee et al., 2012b). For a useful précis of AMS-related magnetic theory in igneous rocks, the reader is referred to early works by Balsley & Buddington (1960) and Khan (1962), and more recent summaries provided by Martín-Hernández et al., (2004), O’Driscoll et al., (2008), and O’Driscoll et al., (2015). The principle behind AMS relies on the measurement of the bulk susceptibility (Km) of a single sample in different orientations to determine the susceptibility anisotropy tensor, which relates the induced magnetisation (Mind) to the applied field (H) in three dimensions (Tarling & Hrouda, 1993). The orientation and magnitude of the eigenvectors and eigenvalues of this tensor define an ellipsoid with three principal axes; the long axis of the ellipsoid, K1, defines the magnetic lineation and the short axis, K3, defines the normal (i.e. the pole) to the magnetic foliation plane (K1–K2; Fig. 12a) (Stacey et al., 1960; Khan, 1962; Tarling & Hrouda, 1993). In order to interpret magnetic fabrics, it is important to determine the mineralogy of the phases carrying the magnetic signal because the composition, grainsize, and distribution of magnetically dominant minerals (e.g. titanomagnetite) can control fabric orientation (e.g. Hargraves et al., 1991; Stephenson, 1994; Dunlop & Özdemir, 2001). In addition to primary crystallographic and textural controls on magnetic fabrics, subsequent oxidation of remaining melt and secondary hydrothermal alteration can affect the magnetic mineralogy and, thereby, the AMS signal (e.g. Trindade et al., 2001; Stevenson et al., 2007a). A variety of rock magnetic experiments are thus required to determine the magnetic mineralogy. The most widely used method involves measuring susceptibility, and thereby behaviour of magnetic materials, at varying temperatures ranging from -200°C to 700°C (i.e. thermomagnetic analysis sensuOrlický, 1990; Hrouda et al., 1997). For example, paramagnetic materials (e.g. biotite) follow the Curie-Weiss law, whereby their susceptibility drops hyperbolically with increasing temperature. In contrast, the thermomagnetic curve of ferromagnetic materials (e.g. titanomagnetite) displays little change in susceptibility with temperature, apart from when characteristic crystallographic transitions occur (e.g. the Curie point for pure magnetite at ∼580°C, Petrovský & Kapička, 2006). To determine the grain size of the ferromagnetic fraction in the magnetic susceptibility signal, the hysteretic property of the magnetisation is important (Dunlop, 2002). Other rock magnetic experiments (e.g. anisotropy of anhysteretic remanent magnetism (AARM) can be conducted to further isolate the relative importance of different paramagnetic and ferromagnetic phases (e.g. McCabe et al., 1985; Richter & van der Pluijm, 1994; Kelso et al., 2002). Observations Having established the magnetic mineralogy, AMS fabrics can be interpreted. Even in weakly anisotropic igneous rocks (i.e. visually isotropic), particularly sheet intrusions, it is now accepted that the magnetic lineation and foliation can provide information on magma migration (e.g. flow direction) or regional and local strain (e.g. Hrouda, 1982; Knight & Walker, 1988; Rochette et al., 1992; Bouchez, 1997; Tauxe, 1998; Callot et al., 2001; Féménias et al., 2004; Magee et al., 2012b). For example, comparisons to other indicators of magma flow (e.g. intrusive steps and visible mineral alignments) in sheet intrusions have shown that magnetic lineations commonly parallel the magma flow (e.g. Knight & Walker, 1988; Cruden & Launeau, 1994; Callot et al., 2001; Magee et al., 2012b), whilst imbrication of elongate crystals induced by simple shear at intrusion margins define the sense of magma flow (Fig. 12b) (e.g. Knight & Walker, 1988; Hargraves et al., 1991; Stephenson, 1994; Geoffroy et al., 2002; Féménias et al., 2004). Alternatively, contact-parallel magnetic fabrics generated during the formation and inflation of magma lobes can be used to determine flow and emplacement dynamics, even if other evidence for the presence of magma lobes is lacking (e.g. Fig. 12c) (Cruden et al., 1999; Stevenson et al., 2007a; Magee et al., 2012a). Identifying changes in fabric orientation within or between individual sheet intrusions is also important because these variations suggest that deformation, imparted by either the emplacement of adjacent magma bodies or tectonic processes, did not significantly modify magma emplacement fabrics (e.g. Clemente et al., 2007). Post solidification textural modification and the possibility of overlap in tectonic and magmatic strain fields during protracted emplacement is a particular complication when studying granitoid and gabbroic plutons (e.g. Mamtani et al., 2013; O’Driscoll et al., 2015; Cheadle & Gee et al., 2017). In fact, most early studies of granitoid emplacement using AMS, in conjunction with many other structural analysis tools, concluded that tectonic strain was the main source of subtle fabrics (e.g. Brun et al., 1990; Bouchez, 1997; de Saint Blanquat & Tikoff, 1997; Neves et al., 2003; Mamtani & Greiling, 2005). Although primary magma flow fabrics in granitic and gabbroic plutons may thus be overprinted, the magnetic fabrics characterised by AMS can still provide fundamental insights into emplacement mechanics (e.g. Stevenson et al., 2007a; Petronis et al., 2012) and magma/mush evolution (e.g. formation of layering; O’Driscoll et al., 2015). Implications and integration Overall, AMS has provided vital magma flow and evolution information that has helped to understand mafic and silicic magma plumbing systems (e.g. Knight & Walker, 1988; Ernst & Baragar, 1992; Glen et al., 1997; Aubourg et al., 2008; Petronis et al., 2013,, 2015). Critical insights emanating from these AMS studies have revealed that: (i) flow trajectories predicted by classic emplacement models (e.g. for ring dykes and cone sheets) are not always consistent with measured AMS fabrics and supporting data, which thereby call into question the application of such models (e.g. Stevenson et al., 2007b; Magee et al., 2012b); (ii) lateral magma flow is recorded in many shallow, planar intrusions associated with volcanic magma plumbing systems (e.g. Ernst & Baragar, 1992; Cruden & Launeau, 1994; Cruden et al., 1999; Herrero-Bervera et al., 2001; Magee et al., 2012b; Petronis et al., 2013,, 2015); and (iii) plutons, particularly those with a granitic composition, commonly consist of incrementally emplaced magma pulses that often develop lobate geometries (e.g. Fig. 12c) (e.g. Stevenson et al., 2007a). Analysing AMS fabrics from layered mafic–ultramafic intrusions can also provide evidence for magma reservoir processes, including crystal settling, or post-cumulus modification of crystal mushes (O’Driscoll et al., 2008,, 2015). Importantly, AMS and related analyses provide robust, testable, and repeatable methods to constrain subtle shape and crystallographic orientations of crystals in igneous rocks. Rock magnetic instrumentation technology continues to advance with better automation of measurement protocols, sensitivity of measurements, and a greater ability to unravel contributors to the AMS signal. The direction and scope of these developments are improving the holistic integration of AMS with other structural, microstructural, geophysical, petrological and geochemical techniques, promising to advance our understanding of magmatism and crustal evolution. Future advances Our understanding of magma plumbing system structure and evolution has been significantly enhanced by the geophysical techniques described above. We have demonstrated that there is scope for advancement within individual methodologies and through the integration of different techniques, particularly involving the synthesis of geophysical, petrological, and geochemical data. In this section, we discuss two new techniques that will potentially revolutionize our understanding of magma plumbing systems. We also briefly discuss how integration of geophysical data with numerical modelling can enhance our knowledge of reservoir construction and evolution. Full-waveform inversion Technique We have demonstrated that seismic reflection data can provide unique insight into the 3D structure of magma plumbing systems (e.g. see review by Magee et al., 2016). In addition to using seismic reflection data to image the subsurface, we can also invert the measured travel-times of reflected acoustic energy to model subsurface P-wave velocities. Full-waveform inversion (FWI) is a rapidly developing technology using active source seismic data to generate models that reproduce both the travel-times and full waveform of the arriving wavefield, thereby matching observed seismic data (Tarantola, 1984). Because FWI considers the full wavefield, as opposed to conventional techniques that only model travel-times, it is a technique capable of recovering high-resolution models of subsurface P-wave velocities and other physical properties (Warner et al., 2013; Routh et al., 2017). The FWI technique begins with a best-guess starting velocity model for the subsurface geology, which is then iteratively updated using a local linearised inversion until the observed seismic data are matched (Virieux & Operto, 2009). FWI is much more computationally expensive than travel-time tomography, as a full-physics implementation of the wave equation is required to generate the predicted seismic data at all energy source and receiver locations for each iteration (Routh et al., 2017). FWI, however, has the advantage of being able to resolve much finer-scale structure than conventional techniques. Observations To date, 3D FWI has principally been applied within the petroleum sector to obtain high-resolution velocity models that can be used to improve depth-migrated (i.e. travel-time is converted to depth in metres) reflection images of petroleum reservoirs and their overburden (Sirgue et al., 2010; Vigh et al., 2010; Kapoor et al., 2013; Warner et al., 2013; Routh et al., 2017). FWI can also produce interpretable, quantitative models of the physical properties of rocks in the subsurface that can be related directly to compaction, permeabilit, and overpressure as measured in subsurface boreholes (Lazaratos et al., 2011; Mancini et al., 2015). Of relevance here, is that mafic intrusions, which appear as high-amplitude reflections in seismic reflection data (Figs 10 and 11a), are recovered as high-velocity features in FWI velocity models (Fig. 13) (Mancini et al., 2015; Kalinicheva et al., 2017). For example, successful application of 3D FWI to a marine ocean bottom seismometer dataset acquired across the Endeavour segment of the Juan de Fuca Ridge led to generation of a velocity model that had a resolution up to four times greater than travel-time tomography (Morgan et al., 2016). Within this new, high-resolution velocity model, several velocity anomalies were identified and interpreted to indicate localised magma recharge of the axial reservoir, induced seismogenic cracking, and increased permeability (Arnoux et al., 2017). Fig. 13. View largeDownload slide (a) Starting model derived from smoothed, pre-stack, time-migrated (PSTM) stacking velocities. (b) Final 2D FWI-derived velocity model obtained using 10 km streamer data and inversion frequencies of between 2·5 and 24 Hz. (c) FWI velocity model overlain by the 2D pre-stack, depth-migrated (PSDM) section. Strong irregular reflections in the lower half of the section are from basaltic intrusions, which appear as high-velocity anomalies in the FWI velocity model. Both the FWI velocity model and the PSDM pick out a major unconformity and show shallow channels in the upper parts of the section (redrawn from Kalinicheva et al., 2017). Fig. 13. View largeDownload slide (a) Starting model derived from smoothed, pre-stack, time-migrated (PSTM) stacking velocities. (b) Final 2D FWI-derived velocity model obtained using 10 km streamer data and inversion frequencies of between 2·5 and 24 Hz. (c) FWI velocity model overlain by the 2D pre-stack, depth-migrated (PSDM) section. Strong irregular reflections in the lower half of the section are from basaltic intrusions, which appear as high-velocity anomalies in the FWI velocity model. Both the FWI velocity model and the PSDM pick out a major unconformity and show shallow channels in the upper parts of the section (redrawn from Kalinicheva et al., 2017). Implications and integration Active magma plumbing systems comprise a complex network of interconnected conduits and reservoirs with variable geometries and sizes, which likely contain magmatic vapour-rich, liquid-rich and mush zones (Christopher et al., 2015). These intrusions will all be associated with reduced P-wave velocities, which could be resolved in high-resolution, 3D FWI datasets as supported by successes in the fine-scale imaging of: (i) low-velocity gas clouds (Warner et al., 2013); (ii) axial reservoirs at an oceanic spreading centre (Arnoux et al., 2017); (iii) relatively narrow, low-velocity fault zones within an antiform (Morgan et al., 2013); and (iv) a subduction zone using 2D FWI (Kamei et al., 2017). A suite of synthetic tests has been performed to investigate whether 3D FWI could be applied to better understand magma plumbing systems (Morgan et al., 2013). These tests indicate that it is possible to recover high-resolution models of P-wave velocity beneath volcanoes, which can then be used to better determine where magma/mush is stored beneath the surface. In particular, these synthetic tests suggest that FWI could be used to: (i) distinguish between continuous zones of mush and individual magma reservoirs; (ii) image sills and conduits of magma and/or fluids that are a few 10s metres across (Fig. 13); and (iii) image the deeper (lower crustal) part of the magma system. Therefore, we consider that 3D FWI affords an unprecedented opportunity to obtain high-resolution images of actual magma plumbing systems beneath active volcanoes. To this end, the ongoing PROTEUS (Plumbing Reservoirs of The Earth Under Santorini) experiment was specifically designed to use 3D FWI to investigate the Santorini magma plumbing system (Hooft et al., 2016). Unmanned aerial vehicle photogrammetry Technique Despite major advances in satellite-based remote sensing systems and aeromagnetic surveys, very high-resolution (i.e. mm–cm scale ground sampling distance) imagery of dykes and other igneous intrusions has been limited to low altitude aerial photography. This in turn has created a critical scale gap in intrusion studies, which range from <1 mm at thin section scale to the metres to 100s of metres scale provided by outcrop analysis, conventional remote sensing, and geophysical data. Fortunately, the emerging capability of unmanned aerial vehicle (UAV) photogrammetry fills this gap (e.g. Eisenbeiss, 2009; Westoby et al., 2012; Bemis et al., 2014; Eide et al., 2017a). It is also noteworthy that several studies have demonstrated that digital photogrammetry can deliver high quality datasets with accuracies similar to more established laser scanning techniques (e.g. Leberl et al., 2010; Hodgetts, 2013; Thiele et al., 2015). The basic setup required to carry out UAV (or drone) photogrammetry is commercially available and relatively inexpensive, comprising a fixed wing or rotary wing UAV, a digital camera, and access to a suitable digital photogrammetry software package (e.g. Agisoft Photoscan Pro, Pix4Dmapper Pro, VisualSFM). UAV photogrammetry combines a simple and cost-effective method to acquire geospatially referenced, overlapping digital aerial images, from which structure-from-motion algorithms can generate spatial 3D datasets (Bemis et al., 2014; Vollgger & Cruden, 2016). Such an approach can be used for high spatial resolution mapping of all types of well-exposed igneous intrusions. The resulting data greatly enhance the effectiveness of traditional field mapping, particularly the characterisation of contact relationships and internal and external structure (e.g. fractures, fabrics, and phase distributions) of intrusive rocks, complementing AMS and petrological analyses. Observations A photogrammetric workflow was applied to examine a swarm of 5 cm to 1 m wide Palaeogene dolerite and dacite dykes exposed on coastal outcrops at Bingie Bingie Point, SE Australia (Fig. 14). The orthophotograph of the entire wave-cut platform shows the distribution of the Palaeogene dolerite and dacite dykes and their Devonian host rock lithologies, including a prominent moderately NE-dipping aplite dyke (Fig. 14a). Linear ENE–WSW terrain features pick out the traces of dyke-parallel joints (Fig. 14a). The Palaeogene dykes trend 063° parallel to a major set of joints in the country rock that likely formed contemporaneously with syn-dyking extension (Fig. 14b). Subsidiary joint sets trend NNW–SSE, sub-perpendicular to the Palaeogene dykes, N–S and E–W (Fig. 14b). The Palaeogene dykes display considerable structural complexity such as bridge structures, intrusive steps and apophyses (Fig. 14c). Where present, the steps mostly occur where dykes cross country rock contacts (e.g. the aplite–tonalite contact in the NE; Fig. 14c). Fig. 14. View largeDownload slide (a) UAV orthophotograph of the wave cut platform at Bingie Point, NSW, Australia showing the distribution of Palaeogene dolerite (Dol) and dacite (Dac) dykes within Devonian tonalite (Ton), diorite (Di), and aplite (Ap) host rocks. (b) Circular histogram of joint sets measured in the Devonian rocks from the orthophotograph; the dominant (purple) set is parallel to and likely contemporaneous with the Palaeogene dykes. (c) Annotated close-up image highlighting dykes and structural features. The northern dacite dyke shows two broken bridge (BB) structures, whilst the central dolerite dyke displays prominent step structures (S). Narrow apophyses are also associated with the broken bridges and steps. Fig. 14. View largeDownload slide (a) UAV orthophotograph of the wave cut platform at Bingie Point, NSW, Australia showing the distribution of Palaeogene dolerite (Dol) and dacite (Dac) dykes within Devonian tonalite (Ton), diorite (Di), and aplite (Ap) host rocks. (b) Circular histogram of joint sets measured in the Devonian rocks from the orthophotograph; the dominant (purple) set is parallel to and likely contemporaneous with the Palaeogene dykes. (c) Annotated close-up image highlighting dykes and structural features. The northern dacite dyke shows two broken bridge (BB) structures, whilst the central dolerite dyke displays prominent step structures (S). Narrow apophyses are also associated with the broken bridges and steps. Implications and integration Data such as the orthophotograph collected at Bingie Bingie Point indicate that high-resolution structural and lithological mapping and measurement can be carried out much more rapidly than by traditional survey methods (e.g. plane table or grid mapping). However, the use of conventional RGB cameras restricts the resulting image data to reflected visible light. Future applications will include the deployment of multispectral and hyperspectral sensors (infrared to short wave infrared to thermal infrared) as well as potential field geophysical or geodetic instruments (e.g. Sparks et al., 2012). A further challenge for UAV applications in many countries concerns the regulatory framework around the use of drones for research. The global trend is moving to require non-recreational UAV operators to have remotely piloted aircraft licences and for the associated organisation to be certified for UAV operations. Innovations in sensor types and design, attachment of geophysical instruments, machine learning, and integration with complementary techniques such as AMS will open up new avenues for UAV applications in the study of magma plumbing systems. Numerical modelling of magma reservoir processes constrained by geophysical data Geophysical imaging of both active and ancient magma plumbing systems is delivering new insights into the 3D geometry of magma reservoirs, the timing and rates of melt and magma transport, the pathways followed by magmas as they ascend through the crust, and typical stored melt fractions in mushes. These data can be used to constrain and calibrate numerical models of reservoir processes. Numerical models are used ubiquitously to understand and predict the behaviour of other subsurface crustal reservoirs, such as hydrocarbon reservoirs, groundwater resources, and targets for geological CO2 storage (e.g. Chen et al., 2003; Class et al., 2009; Dean & Chen, 2011). However, there has been relatively little focus to date on developing numerical models for magma/mush reservoirs. Yet such models can integrate across different data sources and types, provide quantitative estimates of rates, volumes and timescales, and provide a framework for data interpretation. For example, numerical modelling of heat transfer within the plumbing system at Okmok Volcano in Alaska, which was informed by analytical models of geodetic data and estimated magma compositions of erupted material, allowed estimation of the role magma injection, crystallisation and degassing processes had on volume changes over time (Caricchi et al., 2014). Numerical thermal modelling has also helped interpret seismic data from the Soufrière Hills Volcano, Montserrat, suggesting a higher melt fraction in the underlying magma reservoir than was inferred from seismic data alone (Paulatto et al., 2012). More recent numerical models focus on crystal mushes, evaluating melt transport and reaction at low melt fractions, and these show that temperature and melt fraction in mushes can be decoupled; i.e. maximum temperature occurs close to the centre of the reservoir, but maximum melt fraction occurs close to the top (Solano et al., 2012). This decoupling impacts how seismic velocities and electrical conductivities will be modified within the mush (Solano et al., 2012). Other numerical models show the important role played by exsolution, crystallisation, and the viscoelastic response of the crust in driving magma mobilisation in and eruption from shallow reservoirs (e.g. Degruyter & Huber, 2014; Parmigiani et al., 2016), as well as providing insights into the mixing mechanisms of melt and crystals in mushes (Bergantz et al., 2015). However, most models to date have a lower dimensionality (zero dimension box models, or one/two dimensions) and capture only a small subset of the key physical and chemical processes that are likely to occur in crustal magma reservoirs or crystal mushes. Moreover, few studies have integrated modelling with geophysical data (cf. Gutierrez et al., 2013). This is in marked contrast to the 3D modelling routinely undertaken of other crustal reservoirs (e.g. hydrocarbon reservoirs), which is commonly integrated with and delimited by geophysical data. There is thus significant scope for improved, and integrated, numerical modelling of crustal magma reservoirs. CONCLUSIONS Determining the structure of magma plumbing systems is critical to understanding where melt and magma are stored in the crust, which can influence the location of volcanic eruptions and economic ore deposits, providing an important framework for interpreting the physical and chemical evolution of magma from petrological and geochemical datasets. Geophysical techniques have revealed unique insights into the architecture of active and ancient magma plumbing systems, which when integrated with traditional structural, petrological and geochemical results has yielded exciting advances in our understanding of magmatic processes. However, divisions between communities applying these methodologies still exist, contributing to diverging views on the nature of magma plumbing systems. To help promote collaboration, we have reviewed a range of geophysical techniques and discussed how they could be integrated with structural, petrological and geochemical datasets to answer outstanding questions in the volcanological community. In particular, we demonstrate how a range of geophysical techniques can be applied to track melt migration in near real-time, map entire intrusion networks in 3D, examine magma emplacement mechanics, and understand the evolution of crystal mushes. For example, Interferometric Synthetic Aperture Radar (InSAR) allows measurement of the development of active magmatic systems by successive intrusion, the vertical and lateral movements of magma, and the relationship between magma plumbing system dynamics and eruption. Seismicity beneath volcanoes can, when the magma interacts dynamically with the host rock, illuminate in high-resolution the time and spatial scales of the motion of magma and hydrothermal fluids. Seismic imaging of magma plumbing systems allows the spatial distribution of melt and magma to be determined, whilst the inclusion of anisotropy within seismic techniques even allows sub-seismic wavelength features to be identified. Gravimetry can characterise the distribution and redistribution of mass (e.g. magma) in the subsurface over high spatial and temporal resolutions, helping to reveal the structure and composition of magma plumbing systems and the source(s) of volcano deformation. Electromagnetic methods, particularly magnetotellurics, can identify fluids within magmatic systems (e.g. melt, magma, and hydrothermal fluids). Seismic reflection data provide unprecedented 3D images of ancient magma plumbing systems and have revealed that laterally extensive, interconnected networks of sills and inclined sheets can play a pivotal role in transporting magma through the crust to eruption sites potentially located >100 km away from the melt source. Rock magnetics can provide fabric data pertaining to magma flow, deformation or crystallisation. All these methodologies discussed have provided unique insights into the structure of igneous intrusions and, through integration with petrological and geochemical datasets, are beginning to help unravel the entire evolution of magma plumbing systems. In addition to the ongoing application and advancement of these geophysical techniques, emerging methodologies look set to radically improve our understanding of magma plumbing systems. For example, full-waveform inversion can image and characterise physical properties across plumbing systems at an unprecedented resolution, whereas unmanned aerial vehicle photogrammetry provides a tool for high spatial resolution of outcrop scale intrusions that bridges the scale gap between seismic reflection data and traditional mapping of magma plumbing systems. The geophysical techniques discussed also provide critical constraints on input parameters for numerical modelling. Overall, we consider that the future of magma plumbing system studies will benefit greatly from the synthesis of geophysics and more traditional petrological and geochemical approaches. ACKNOWLEDGEMENTS We would like to thank Marian Holness for inviting us to put together this review article and for editorial handling. We are very grateful to Juliet Biggs, Martyn Unsworth, John Bartley, and Magnús Gudmundsson for their extensive and constructive reviews. FUNDING CM is funded by an Imperial College Research Fellowship at Imperial College London. SKE is funded by an Early Career Fellowship from the Leverhulme Trust. KAW is funded by Natural Environment Research Council grant NE/L013932/1. REFERENCES Abdelmalak M. M. , Andersen T. B. , Planke S. , Faleide J. I. , Corfu F. , Tegner C. , Shephard G. E. , Zastrozhnov D. , Myklebust R. ( 2015 ). The ocean-continent transition in the mid-Norwegian margin: insight from seismic data and an onshore Caledonian field analogue . Geology 43 , 1011 – 1014 . Google Scholar Crossref Search ADS Aizawa K. , Kanda W. , Ogawa Y. , Iguchi M. , Yokoo A. , Yakiwara H. , Sugano T. ( 2011 ). Temporal changes in electrical resistivity at Sakurajima volcano from continuous magnetotelluric observations . Journal of Volcanology and Geothermal Research 199 , 165 – 175 . Google Scholar Crossref Search ADS Annen C. ( 2011 ). Implications of incremental emplacement of magma bodies for magma differentiation, thermal aureole dimensions and plutonism–volcanism relationships . Tectonophysics 500 , 3 – 10 . Google Scholar Crossref Search ADS Annen C. , Blundy J. D. , Sparks R. S. J. ( 2006 ). The genesis of intermediate and silicic magmas in deep crustal hot zones . Journal of Petrology 47 , 505 – 539 . Google Scholar Crossref Search ADS Annen C. , Blundy J. D. , Leuthold J. , Sparks R. S. J. ( 2015 ). Construction and evolution of igneous bodies: towards an integrated perspective of crustal magmatism . Lithos 230 , 206 – 221 . Google Scholar Crossref Search ADS Arnoux G. M. , Toomey D. , Hooft E. , Wilcock W. , Morgan J. , Warner M. , VanderBeek B. ( 2017 ). Seismic evidence that black smoker heat flux is rate-limited by crustal permeability . Geophysical Research Letters 44 , 1687 – 1695 . Google Scholar Crossref Search ADS Aubourg C. , Tshoso G. , le Gall B. , Bertrand H. , Tiercelin J.-J. , Kampunzu A. B. , Dyment J. , Modisi M. ( 2008 ). Magma flow revealed by magnetic fabric in the Okavango giant dyke swarm, Karoo igneous province, northern Botswana . Journal of Volcanology and Geothermal Research 170 , 247 – 261 . Google Scholar Crossref Search ADS Baba K. , Chave A. D. , Evans R. L. , Hirth G. , Mackie R. L. ( 2006 ). Mantle dynamics beneath the East Pacific Rise at 17°S: insights from the Mantle Electromagnetic and Tomography (MELT) experiment . Journal of Geophysical Research: Solid Earth 111 , B02101. Bagnardi M. , Amelung F. , Poland M. P. ( 2013 ). A new model for the growth of basaltic shields based on deformation of Fernandina volcano, Galápagos Islands . Earth and Planetary Science Letters 377–378 , 358 – 366 . Google Scholar Crossref Search ADS Bailey R. C. ( 1970 ). Inversion of the geomagnetic induction problem . Proceedings of the Royal Society 315 , 185 – 194 . Google Scholar Crossref Search ADS Balsley J. R. , Buddington A. F. ( 1960 ). Magnetic susceptibility anisotropy and fabric of some Adirondack granites and orthogneisses . American Journal of Science A258 , 6 – 20 . Bannister S. , Reyners M. , Stuart G. , Savage M. ( 2007 ). Imaging the Hikurangi subduction zone, New Zealand, using teleseismic receiver functions: crustal fluids above the forearc mantle wedge . Geophysical Journal International 169 , 602 – 616 . Google Scholar Crossref Search ADS Battaglia M. , Segall P. ( 2004 ). The interpretation of gravity changes and crustal deformation in active volcanic areas . Pure and Applied Geophysics 161 , 1453 – 1467 . Google Scholar Crossref Search ADS Battaglia M. , Roberts C. , Segall P. ( 1999 ). Magma intrusion beneath Long Valley caldera confirmed by temporal changes in gravity . Science 285 , 2119 – 2122 . Google Scholar Crossref Search ADS PubMed Battaglia M. , Troise C. , Obrizzo F. , Pingue F. , De Natale G. ( 2006 ). Evidence for fluid migration as the source of deformation at Campi Flegrei caldera (Italy) . Geophysical Research Letters 33 , L01307. Bemis S. P. , Micklethwaite S. , Turner D. , James M. R. , Akciz S. , Thiele S. , Bangash H. A. ( 2014 ). Ground-based and UAV-Based photogrammetry: a multi-scale, high-resolution mapping tool for structural geology and paleoseismology . Journal of Structural Geology 69 , 163 – 178 . Google Scholar Crossref Search ADS Bergantz G. W. , Schleicher J. M. , Burgisser A. ( 2015 ). Open-system dynamics and mixing in magma mushes . Nature Geoscience 8 , 793 – 796 . Google Scholar Crossref Search ADS Berryman J. G. ( 1980 ). Long‐wavelength propagation in composite elastic media II. Ellipsoidal inclusions . The Journal of the Acoustical Society of America 68 , 1820 – 1831 . Google Scholar Crossref Search ADS Bersi M. , Saibi H. , Chabou M. C. ( 2016 ). Aerogravity and remote sensing observations of an iron deposit in Gara Djebilet, southwestern Algeria . Journal of African Earth Sciences 116 , 134 – 150 . Google Scholar Crossref Search ADS Biedermann A. R. , Pettke T. , Reusser E. , Hirt A. M. ( 2014 ). Anisotropy of magnetic susceptibility in natural olivine single crystals . Geochemistry, Geophysics, Geosystems 15 , 3051 – 3065 . Google Scholar Crossref Search ADS Biggs J. , Pritchard M. E. ( 2017 ). Global volcano monitoring: what does it mean when volcanoes deform? Elements 13 , 17 – 22 . Google Scholar Crossref Search ADS Biggs J. , Anthony E. Y. , Ebinger C. J. ( 2009 ). Multiple inflation and deflation events at Kenyan volcanoes, East African Rift . Geology 37 , 979 – 982 . Google Scholar Crossref Search ADS Biggs J. , Bastow I. D. , Keir D. , Lewi E. ( 2011 ). Pulses of deformation reveal frequently recurring shallow magmatic activity beneath the Main Ethiopian Rift . Geochemistry, Geophysics, Geosystems 12 , Q0AB10. Biggs J. , Ebmeier S. K. , Aspinall W. P. , Lu Z. , Pritchard M. E. , Sparks R. S. J. , Mather T. A. ( 2014 ). Global link between deformation and volcanic eruption quantified by satellite imagery . Nature Communications 5 , 3471 . Google Scholar Crossref Search ADS PubMed Bonforte A. , Carbone D. , Greco F. , Palano M. ( 2007 ). Intrusive mechanism of the 2002 NE-rift eruption at Mt Etna (Italy) modelled using GPS and gravity data . Geophysical Journal International 169 , 339 – 347 . Google Scholar Crossref Search ADS Borradaile G. J. , Henry B. ( 1997 ). Tectonic applications of the magnetic susceptibility and its anisotropy . Earth Science Reviews 42 , 49 – 93 . Google Scholar Crossref Search ADS Bosworth W. , Stockli D. F. , Helgeson D. E. ( 2015 ). Integrated outcrop, 3D seismic, and geochronologic interpretation of Red Sea dike-related deformation in the Western Desert, Egypt–the role of the 23Ma Cairo “mini-plume” . Journal of African Earth Sciences 109 , 107 – 119 . Google Scholar Crossref Search ADS Bott M. H. P. ( 1953 ). Negative gravity anomalies over acid “intrusions” and their relation to the structure of the Earth’s crust . Geological Magazine 90 , 257 – 267 . Google Scholar Crossref Search ADS Bouchez J. L. ( 1997 ). Granite is never isotropic: an introduction to AMS studies of granitic rocks. In: Bouchez J. L. , Hutton D. H. W. , Stephens W. E. (eds) Granite: From Segregation of Melt to Emplacement Fabrics 8 . Dordrecht : Kluwer Academic , pp. 95 – 112 . Brown A. R. ( 2004 ). Interpretation of Three-Dimensional Seismic Data . Society of Exploration Geophysicists and American Association of Petroleum Geologists, Tulsa. Brun J. P. , Gapais D. , Cogne J. P. , Ledru P. , Vigneresse J. L. ( 1990 ). The Flamanville granite (northwest France): an unequivocal example of a syntectonically expanding pluton . Geological Journal 25 , 271 – 286 . Google Scholar Crossref Search ADS Bucher W. H. ( 1944 ). Discussion in Romberg, F., and Barnes, V. E. Correlation of gravity observations with the geology of the Smoothingiron granite mass, Llano County, Texas . Geophysics 9 , 79 – 93 . Google Scholar Crossref Search ADS Callot J.-P. , Geoffroy L. , Aubourg C. , Pozzi J. , Mege D. ( 2001 ). Magma flow directions of shallow dykes from the East Greenland volcanic margin inferred from magnetic fabric studies . Tectonophysics 335 , 313 – 329 . Google Scholar Crossref Search ADS Carbone D. , Poland M. P. , Diament M. , Greco F. ( 2017 ). The added value of time-variable microgravimetry to the understanding of how volcanoes work . Earth-Science Reviews 169 , 146 – 179 . Google Scholar Crossref Search ADS Carbone D. , Poland M. P. , Patrick M. R. , Orr T. R. ( 2013 ). Continuous gravity measurements reveal a low-density lava lake at Kīlauea Volcano, Hawai’i . Earth and Planetary Science Letters 376 , 178 – 185 . Google Scholar Crossref Search ADS Caricchi L. , Biggs J. , Annen C. , Ebmeier S. ( 2014 ). The influence of cooling, crystallisation and re-melting on the interpretation of geodetic signals in volcanic systems . Earth and Planetary Science Letters 388 , 166 – 174 . Google Scholar Crossref Search ADS Cartwright J. , Hansen D. M. ( 2006 ). Magma transport through the crust via interconnected sill complexes . Geology 34 , 929 – 932 . Google Scholar Crossref Search ADS Cartwright J. , Huuse M. ( 2005 ). 3D seismic technology: the geological ‘Hubble’ . Basin Research 17 , 1 – 20 . Google Scholar Crossref Search ADS Cashman K. V. , Sparks R. S. J. ( 2013 ). How volcanoes work: a 25 year perspective . Geological Society of America Bulletin 125 , 664 – 690 . Google Scholar Crossref Search ADS Cashman K. V. , Sparks R. S. J. , Blundy J. D. ( 2017 ). Vertically extensive and unstable magmatic systems: a unified view of igneous processes . Science 355 , eaag3055 . Google Scholar Crossref Search ADS PubMed Chaussard E. , Amelung F. ( 2013 ). Depth of magma storage in volcanic arcs: testing the influence of regional parameters using a global data compilation. In EGU General Assembly Conference Abstracts (Vol. 15), Copernicus, Gottingen. Cheadle M. J. , Gee J. S. ( 2017 ). Quantitative textural insights into the formation of gabbro in mafic intrusions . Elements 13 , 409 – 414 . Google Scholar Crossref Search ADS Chen Y. , Durlofsky L. J. , Gerritsen M. , Wen X. H. ( 2003 ). A coupled local–global upscaling approach for simulating flow in highly heterogeneous formations . Advances in Water Resources 26 , 1041 – 1060 . Google Scholar Crossref Search ADS Chmielowski J. , Zandt G. , Haberland C. ( 1999 ). The central Andean Altiplano‐Puna magma body . Geophysical Research Letters 26 , 783 – 786 . Google Scholar Crossref Search ADS Chouet B. , Matoza R. S. ( 2013 ). A Multi-decadal view of seismic methods for detecting precursors of magma movement and eruption . Journal of Volcanology and Geothermal Research 252 , 108 – 175 . Google Scholar Crossref Search ADS Chouet B. , Dawson P. , Martini M. ( 2008 ). Shallow-conduit dynamics at Stromboli Volcano, Italy, imaged from waveform inversions. In: Lane S. J. , Gilbert J. S. (eds) Fluid Motions in Volcanic Conduits: A Source of Seismic and Acoustic Signals . Geological Society of London Special Publication 307, 57 – 84 . Christopher T. E. , Blundy J. D. , Cashman K. , Cole P. , Edmonds M. , Smith P. J. , Sparks R. S. J. , Stinton A. ( 2015 ). Crustal‐scale degassing due to magma system destabilization and magma‐gas decoupling at Soufrière Hills Volcano, Montserrat . Geochemistry, Geophysics, Geosystems 16 , 2797 – 2811 . Google Scholar Crossref Search ADS Class H. , Ebigbo A. , Helmig R. , Dahle H. K. , Nordbotten J. M. , Celia M. A. , Audigane P. , Darcis M. , Ennis-King J. , Fan Y. , Flemisch B. , Gasda S. E. , Jin M. , Krug S. , Labregere D. , Naderi Beni A. , Pawar R. J. , Sbai A. , Thomas S. G. , Trenty L. , Wei L. ( 2009 ). A benchmark study on problems related to CO2 storage in geologic formations . Computational Geosciences 13 , 409 – 434 . Google Scholar Crossref Search ADS Clemente C. S. , Amorós E. B. , Crespo M. G. ( 2007 ). Dike intrusion under shear stress: effects on magnetic and vesicle fabrics in dikes from rift zones of Tenerife (Canary Islands) . Journal of Structural Geology 29 , 1931 – 1942 . Google Scholar Crossref Search ADS Comeau M. J. , Unsworth M. J. , Cordell D. ( 2016 ). New constraints on the magma distribution and composition beneath Volcán Uturuncu and the southern Bolivian Altiplano from magnetotelluric data . Geosphere 12 , 1391 – 1421 . Google Scholar Crossref Search ADS Comeau M. J. , Unsworth M. J. , Ticona F. , Sunagua M. ( 2015 ). Magnetotelluric images of magma distribution beneath Volcán Uturuncu, Bolivia: implications for magma dynamics . Geology 43 , 243 – 246 . Google Scholar Crossref Search ADS Constable S. C. , Parker R. L. , Constable C. G. ( 1987 ). Occam’s inversion: a practical algorithm for generating smooth models from electromagnetic sounding data . Geophysics 52 , 289 – 300 . Google Scholar Crossref Search ADS Cruden A. R. , Launeau P. ( 1994 ). Structure, magnetic fabric and emplacement of the Archean Lebel Stock, SW Abitibi greenstone belt . Journal of Structural Geology 16 , 677 – 691 . Google Scholar Crossref Search ADS Cruden A. R. , Weinberg R. F. ( 2018 ). Mechanisms of magma transport and storage in the lower and Middle crust—magma segregation, ascent and emplacement. In: Burkhardt S. (ed.) Volcanic and Igneous Plumbing Systems . Amsterdam : Elsevier . Cruden A. R. , Tobisch O. T. , Launeau P. ( 1999 ). Magnetic fabric evidence for conduit fed emplacement of a tabular granite: Dinkey Creek Pluton, central Sierra Nevada Batholith, California . Journal of Geophysical Research: Solid Earth 104 , 10511 – 10531 . Google Scholar Crossref Search ADS Currenti G. , Del Negro C. , Ganci G. ( 2007 ). Modelling of ground deformation and gravity fields using finite element method: an application to Etna volcano . Geophysical Journal International 169 , 775 – 786 . Google Scholar Crossref Search ADS Currenti G. , Napoli R. , Coco A. , Privitera E. ( 2017 ). Effects of hydrothermal unrest on stress and deformation: insights from numerical modeling and application to Vulcano Island (Italy) . Bulletin of Volcanology 79 , 28 . Google Scholar Crossref Search ADS Dean O. S. , Chen Y. ( 2011 ). Recent progress on reservoir history matching: a review . Computational Geosciences 15 , 185 – 221 . Google Scholar Crossref Search ADS DeGruyter W. , Huber C. ( 2014 ). A model for eruption frequency of upper crustal silicic magma chambers . Earth and Planetary Science Letters 403 , 117 – 130 . Google Scholar Crossref Search ADS Delph J. R. , Ward K. M. , Zandt G. , Ducea M. N. , Beck S. L. ( 2017 ). Imaging a magma plumbing system from MASH zone to magma reservoir . Earth and Planetary Science Letters 457 , 313 – 324 . Google Scholar Crossref Search ADS del Potro R. , Díez M. , Blundy J. , Gottsmann J. , Camacho A. ( 2013 ). Diapiric ascent of silicic magma beneath the Bolivian Altiplano . Geophysical Research Letters 40 , 2044 – 2048 . Google Scholar Crossref Search ADS de Saint Blanquat M. , Tikoff B. ( 1997 ). Development of magmatic to solid-state fabrics during syntectonic emplacement of the Mono Creek Granite, Sierra Nevada Batholith. In: Bouchez, J. L., Hutton, D. & Stephens, W. E. (eds). Granite: From Segregation of Melt to Emplacement Fabrics . Netherlands : Springer , pp. 231 – 252 . Desissa M. , Johnson N. E. , Whaler K. A. , Hautot S. , Fisseha S. , Dawes G. ( 2013 ). A mantle magma reservoir beneath an incipient mid-ocean ridge in Afar, Ethiopia . Nature Geoscience 6 , 861 – 865 . Google Scholar Crossref Search ADS Di Maio R. , Mauriello P. , Patella D. , Petrillo Z. , Piscitelli S. , Siniscalchi A. ( 1998 ). Electric and electromagnetic outline of the Mount Somma–Vesuvius structural setting . Journal of Volcanology and Geothermal Research 82 , 219 – 238 . Google Scholar Crossref Search ADS Dixon T. H. ( 1991 ). An introduction to the Global Positioning System and some geological applications . Reviews of Geophysics 29 , 249 – 276 . Google Scholar Crossref Search ADS Drew J. , White R. S. , Tilmann F. , Tarasewicz J. ( 2013 ). Coalescence microseismic mapping . Geophysical Journal International 195 , 1773 – 1785 . Google Scholar Crossref Search ADS Dugda M. T. , Nyblade A. A. , Julia J. , Langston C. A. , Ammon C. J. , Simiyu S. ( 2005 ). Crustal structure in Ethiopia and Kenya from receiver function analysis: implications for rift development in eastern Africa . Journal of Geophysical Research 110 , B01303 . Google Scholar Crossref Search ADS Dunlop D. J. ( 2002 ). Theory and application of the Day plot (Mrs/Ms versus Hcr/Hc) 1. Theoretical curves and tests using titanomagnetite data . Journal of Geophysical Research 107 , B3, 2056. Dunlop D. J. , Özdemir Ö. ( 2001 ). Rock Magnetism: Fundamentals and Frontiers , Vol. 3. Cambridge University Press, Cambridge . Dvorak J. J. , Dzurisin D. ( 1997 ). Volcano geodesy: the search for magma reservoirs and the formation of eruptive vents . Reviews of Geophysics 35 , 343 – 384 . Google Scholar Crossref Search ADS Dzurisin D. ( 2006 ). Volcano Deformation: New Geodetic Monitoring Techniques . Springer Science & Business Media , Berlin. Eaton G. P. , Christiansen R. L. , Iyer H. M. , Pitt A. D. , Mabey D. R. , Blank H. R. , Zietz I. , Gettings M. E. ( 1975 ). Magma beneath yellowstone national park . Science 188 , 787 . Google Scholar Crossref Search ADS PubMed Ebinger C. , Keir D. , Ayele A. , Calais E. , Wright T. J. , Belachew M. , Hammond J. O. S. , Campbell E. , Buck W. R. ( 2008 ). Capturing magma intrusion and faulting processes during continental rupture: seismicity of the Dabbahu (Afar) rift . Geophysical Journal International 174 , 1138 – 1152 . Google Scholar Crossref Search ADS Ebinger C. J. , Keir D. , Bastow I. D. , Whaler K. , Hammond J. O. , Ayele A. , Miller M. S. , Tiberi C. , Hautot S. ( 2017 ). Crustal structure of active deformation zones in Africa: Implications for global crustal processes . Tectonics 36 , 3298 – 3332 . Google Scholar Crossref Search ADS Ebmeier S. K. , Biggs J. , Mather T. A. , Amelung F. ( 2013 ). On the lack of InSAR observations of magmatic deformation at Central American volcanoes . Journal of Geophysical Research: Solid Earth 118 , 2571 – 2585 . Google Scholar Crossref Search ADS Ebmeier S. K. , Elliott J. R. , Nocquet J. M. , Biggs J. , Mothes P. , Jarrín P. , Yépez M. , Aguaiza S. , Lundgren P. , Samsonov S. V. ( 2016 ). Shallow earthquake inhibits unrest near Chiles–Cerro Negro volcanoes, Ecuador–Colombian border . Earth and Planetary Science Letters 450 , 283 – 291 . Google Scholar Crossref Search ADS Ebmeier S. K. , Andrews B. J. , Araya M. C. , Arnold D. W. D. , Biggs J. , Cooper C. , Cottrell E. , Furtney M. , Hickey J. , Jay J. , Lloyd R. , Parker A. L. , Pritchard M. E. , Robertson E. , Venzke E. , Williamson J. L. ( 2018 ). Synthesis of global satellite observations of magmatic and volcanic deformation: implications for volcano monitoring & the lateral extent of magmatic domains . Journal of Applied Volcanology 7 , 2 . Google Scholar Crossref Search ADS Eide C. H. , Schofield N. , Jerram D. A. , Howell J. A. ( 2017a ). Basin-scale architecture of deeply emplaced sill complexes: Jameson Land, East Greenland . Journal of the Geological Society of London 174 , 23 – 40 . Google Scholar Crossref Search ADS Eide C. H. , Schofield N. , Lecomte I. , Buckley S. J. , Howell J. A. ( 2017b ). Seismic interpretation of sill complexes in sedimentary basins: implications for the sub-sill imaging problem . Journal of the Geological Society of London . Eisenbeiss H. ( 2009 ). UAV photogrammetry. Ph.D. thesis, ETH Zurich. Ernst R. E. , Baragar W. R. A. ( 1992 ). Evidence from magnetic fabric for the flow pattern of magma in the Mackenzie giant radiating dyke swarm . Nature 356 , 511 . Google Scholar Crossref Search ADS Féménias O. , Diot H. , Berza T. , Gauffriau A. , Demaiffe D. ( 2004 ). Asymmetrical to symmetrical magnetic fabric of dikes: paleo-flow orientations and Paleo-stresses recorded on feeder-bodies from the Motru Dike Swarm (Romania) . Journal of Structural Geology 26 , 1401 – 1418 . Google Scholar Crossref Search ADS Ferguson I. J. ( 2012 ). Instrumentation and field procedures. In: Chave A. D. , Jones A. G. (ed.) The Magnetotelluric Method: Theory and Practice . Cambridge University Press , Cambridge. Field L. , Blundy J. D. , Brooker R. A. , Wright T. , Yirgu G. ( 2012 ). Magma storage conditions beneath Dabbahu volcano (Ethiopia) constrained by petrology, seismicity and satellite geodesy . Bulletin of Volcanology 74 , 981 – 1004 . Google Scholar Crossref Search ADS Fournier T. J. , Pritchard M. E. , Riddick S. N. ( 2010 ). Duration, magnitude, and frequency of subaerial volcano deformation events: New results from Latin America using InSAR and a global synthesis . Geochemistry, Geophysics, Geosystems 11 , Q01003. Gaillard F. ( 2004 ). Laboratory measurements of electrical conductivity of hydrous and dry silicic melts under pressure . Earth and Planetary Science Letters 218 , 215 – 228 . Google Scholar Crossref Search ADS Galland O. ( 2012 ). Experimental modelling of ground deformation associated with shallow magma intrusions . Earth and Planetary Science Letters 317–318 , 145 – 156 . Google Scholar Crossref Search ADS Gamble T. D. , Goubau W. M. , Clarke J. ( 1979 ). Magnetotellurics with a remote magnetic reference . Geophysics 44 , 53 – 68 . Google Scholar Crossref Search ADS Garapic G. , Faul U. H. , Brisson E. ( 2013 ). High-resolution imaging of the melt distribution in partially molten upper mantle rocks: evidence for wetted two-grain boundaries . Geochemistry, Geophysics, Geosystems 14 , 556 – 566 . Google Scholar Crossref Search ADS Geoffroy L. , Callot J. P. , Aubourg C. , Moreira M. ( 2002 ). Magnetic and plagioclase linear fabric discrepancy in dykes: a new way to define the flow vector using magnetic foliation . Terra Nova 14 , 183 – 190 . Google Scholar Crossref Search ADS Gerst A. , Savage M. K. ( 2004 ). Seismic anisotropy beneath Ruapehu volcano: a possible eruption forecasting tool . Science 306 , 1543 – 1547 . Google Scholar Crossref Search ADS PubMed Glazner A. F. , Bartley J. M. , Coleman D. S. ( 2016 ). We need a new definition for “magma” . Eos 97 . Glen J. M. , Renne P. R. , Milner S. C. , Coe R. S. ( 1997 ). Magma flow inferred from anisotropy of magnetic susceptibility in the coastal Parana-Etendeka igneous province: evidence for rifting before flood volcanism . Geology 25 , 1131 – 1134 . Google Scholar Crossref Search ADS Goes S. , Armitage J. , Harmon N. , Smith H. , Huismans R. ( 2012 ). Low seismic velocities below mid-ocean ridges: attenuation versus melt retention . Journal of Geophysical Research: Solid Earth 117 , B12403. Gottsmann J. , Blundy J. D. , Henderson S. , Pritchard M. E. , Sparks R. S. J. ( 2017 ). Thermomechanical modeling of the Altiplano-Puna deformation anomaly: multiparameter insights into magma mush reorganization . Geosphere 13 , 1042 – 1065 . Gottsmann J. , Rymer H. , Wooller L. K. ( 2005 ). On the interpretation of gravity variations in the presence of active hydrothermal systems: insights from the Nisyros Caldera, Greece . Geophysical Research Letters 32 , L23310 . Google Scholar Crossref Search ADS Gottsmann J. , Camacho A. G. , Tiampo K. F. , Fernández J. ( 2006 ). Spatiotemporal variations in vertical gravity gradients at the Campi Flegrei caldera (Italy): a case for source multiplicity during unrest? Geophysical Journal International 167 , 1089 – 1096 . Google Scholar Crossref Search ADS Gottsmann J. , Carniel R. , Coppo N. , Wooller L. , Hautmann S. , Rymer H. ( 2007 ). Oscillations in hydrothermal systems as a source of periodic unrest at caldera volcanoes: multiparameter insights from Nisyros, Greece . Geophysical Research Letters 34 , L07307 . Google Scholar Crossref Search ADS Gottsmann J. , Camacho A. G. , Marti J. , Wooller L. , Fernández J. , Garcia A. , Rymer H. ( 2008 ). Shallow structure beneath the Central Volcanic Complex of Tenerife from new gravity data: implications for its evolution and recent reactivation . Physics of the Earth and Planetary Interior 168 , 212 – 230 . Google Scholar Crossref Search ADS Graham J. W. ( 1954 ). Magnetic anisotropy, an unexploited petrofabric element . Geological Society of America Bulletin 65 , 1257 – 1258 . Gudmundsson O. , Brandsdottir B. , Menke W. , Sigvaldason G. E. ( 1994 ). The crustal magma chamber of the Katla volcano in south Iceland revealed by 2-D seismic undershooting . Geophysical Journal International 119 , 277 – 296 . Google Scholar Crossref Search ADS Guidarelli M. , Stuart G. , Hammond J. O. , Kendall J. M. , Ayele A. , Belachew M. ( 2011 ). Surface wave tomography across Afar, Ethiopia: crustal structure at a rift triple-junction zone . Geophysical Research Letters 38 , L24313. Guo X. , Zhang L. , Behrens H. , Ni H. ( 2016 ). Probing the status of felsic magma reservoirs: Constraints from the P–T–H2O dependences of electrical conductivity of rhyolitic melt . Earth and Planetary Science Letters 433 , 54 – 62 . Google Scholar Crossref Search ADS Gutierrez F. , Payacan I. , Gelman S. E. , Bachmann O. , Parada M. A. ( 2013 ). Late-stage magma flow in a shallow felsic reservoir: merging the anisotropy of magnetic susceptibility record with numerical simulations in La Gloria Pluton, central Chile . Journal of Geophysical Research: Solid Earth 118 , 1984 – 1998 . Google Scholar Crossref Search ADS Hamilton M. P. , Jones A. G. , Evans R. L. , Evans S. , Fourie C. J. S. , Garcia X. , Mountford A. , Spratt J. E. & SAMTEX MT Team. ( 2006 ). Electrical anisotropy of South African lithosphere compared with seismic anisotropy from shear-wave splitting analyses . Physics of the Earth and Planetary Interiors 158 , 226 – 239 . Google Scholar Crossref Search ADS Hamling I. J. , Hreinsdottir S. , Fournier N. ( 2015 ). The ups and downs of the TVZ: geodetic observations of deformation around the Taupo Volcanic Zone, New Zealand . Journal of Geophysical Research: Solid Earth 120 , 4667 – 4679 . Google Scholar Crossref Search ADS Hamling I. J. , Wright T. J. , Calais E. , Bennati L. , Lewi E. ( 2010 ). Stress transfer between thirteen successive dyke intrusions in Ethiopia . Nature Geoscience 3 , 713 – 717 . Google Scholar Crossref Search ADS Hammer S. ( 1945 ). Estimating ore masses in gravity prospecting . Geophsyics 10 , 50 – 62 . Google Scholar Crossref Search ADS Hammond J. O. S. ( 2014 ). Constraining melt storage geometries beneath the Afar Depression, Ethiopia from teleseismic receiver functions: the anisotropic H-κ stacking technique . Geochemistry, Geophysics, Geosystems 15 , 1316 – 1332 . Google Scholar Crossref Search ADS Hammond J. O. S. , Kendall J. ( 2016 ). Constraints on melt distribution from seismology: a case study in Ethiopia. In: Wright T. J. , Ayele A. , Ferguson D. J. , Kidane T. , Vye-Brown C. (eds) Magmatic Rifting and Active Volcanism . Geological Society, Special Publication, London 420 , 127 – 147 . Hammond J. O. S. , Kendall J. , Stuart G. W. , Keir D. , Ebinger C. J. , Ayele A. , Belachew M. ( 2011 ). The nature of the crust beneath the Afar triple junction: evidence from receiver functions . Geochemistry, Geophysics, Geosystems 12 , Q12004. Hammond W. , Humphreys E. ( 2000a ). Upper mantle seismic wave velocity: effects of realistic partial melt geometries . Journal of Geophysical Research: Solid Earth 105 , 10975 – 10986 . Google Scholar Crossref Search ADS Hammond W. , Humphreys E. ( 2000b ). Upper mantle seismic wave attenuation: effects of realistic partial melt distribution . Journal of Geophysical Research: Solid Earth 105 , 10987 – 10999 . Google Scholar Crossref Search ADS Hansen D. M. , Cartwright J. ( 2006 ). The three-dimensional geometry and growth of forced folds above saucer-shaped igneous sills . Journal of Structural Geology 28 , 1520 – 1535 . Google Scholar Crossref Search ADS Hargraves R. B. , Chan C. Y. , Johnson D. ( 1991 ). Distribution anisotropy: the cause of AMS in igneous rocks? Geophysical Research Letters 18 , 2193 – 2196 . Google Scholar Crossref Search ADS Harmon N. , Rychert C. A. ( 2015 ). Seismic imaging of deep crustal melt sills beneath Costa Rica suggests a method for the formation of the Archean continental crust . Earth and Planetary Science Letters 430 , 140 – 148 . Google Scholar Crossref Search ADS Hashin Z. , Shtrikman S. ( 1962 ). A variational approach to the theory of the effective magnetic permeability of multiphase materials . Journal of Applied Physics 33 , 3125 – 3131 . Google Scholar Crossref Search ADS Hawkesworth C. J. , Blake S. , Evans P. , Hughes R. , Macdonald R. , Thomas L. E. , Turner S. P. , Zellmer G. ( 2000 ). Time scales of crystal fractionation in magma chambers—integrating physical, isotopic and geochemical perspectives . Journal of Petrology 41 , 991 – 1006 . Google Scholar Crossref Search ADS Heinson G. , Constable S. , White A. ( 2000 ). Episodic melt transport at mid-ocean ridges inferred from magnetotelluric sounding . Geophysical Research Letters 27 , 2317 – 2320 . Google Scholar Crossref Search ADS Hemmings B. , Coco A. , Gottsmann J. , Whitaker F. ( 2016 ). Investigating hydrological contributions to volcano monitoring signals . Geophysical Journal International 207 , 259 – 273 . Google Scholar Crossref Search ADS Herrero-Bervera E. , Walker G. P. L. , Cañon-Tapia E. , Garcia M. O. ( 2001 ). Magnetic fabric and inferred flow direction of dikes, conesheets and sill swarms, Isle of Skye, Scotland . Journal of Volcanology and Geothermal Research 106 , 195 – 210 . Google Scholar Crossref Search ADS Hickey J. , Gottsmann J. , Nakamichi H. , Iguchi M. ( 2016 ). Thermomechanical controls on magma supply and volcanic deformation: application to Aira caldera, Japan . Scientific Reports 6 , 32691 . Google Scholar Crossref Search ADS PubMed Hickey J. , Gottsmann J. , Mothes P. , Odbert H. , Prutkin I. , Vajda P. ( 2017 ). The Ups and Downs of Volcanic Unrest: Insights from Integrated Geodesy and Numerical Modelling. In: Advances in Volcanology. Springer, Berlin, pp. 1–17. Hildreth W. ( 2004 ). Volcanological perspectives on Long Valley, Mammoth Mountain, and Mono Craters: several contiguous but discrete systems . Journal of Volcanology and Geothermal Research 136 , 169 – 198 . Google Scholar Crossref Search ADS Hodgetts D. ( 2013 ). Laser scanning and digital outcrop geology in the petroleum industry: a review . Marine and Petroleum Geology 46 , 335 – 354 . Google Scholar Crossref Search ADS Hooft E. E. E. , Morgan J. V. , Nomikou P. , Toomey D. R. , Papazachos C. V. , Warner M. , Heath B. , Christopoulou M.-E. , Lampridou D. , Kementzetzidou D. ( 2016 ). The PROTEUS experiment: active source seismic imaging of the crustal magma plumbing structure of the Santorini Arc volcano . AGU Fall Meeting Abstract# DI23B-2619. Holohan E. P. , Sudhaus H. , Walter T. R. , Schöpfer M. P. , Walsh J. J. ( 2017 ). Effects of host-rock fracturing on elastic-deformation source models of volcano deflation . Scientific Reports 7 , 10970 . Google Scholar Crossref Search ADS PubMed Hrouda F. ( 1982 ). Magnetic anisotropy of rocks and its application in geology and geophysics . Geophysical Surveys 5 , 37 – 82 . Google Scholar Crossref Search ADS Hrouda F. , Schulmann K. , Suppes M. , Ullemayer K. , de Wall H. , Weber K. ( 1997 ). Quantitative relationship between low-field AMS and phyllosilicate fabric: a review . Physics and Chemistry of the Earth 22 , 153 – 156 . Google Scholar Crossref Search ADS Hübert J. , Whaler K. A. , Fisseha S. ( 2018 ). Preliminary interpretation of the 2010 Youanmi deep seismic reflection lines and magnetotelluric data for the Windimurra Igneous Complex . Journal of Geophysical Research . Hunt T. , Bowyer D. ( 2007 ). Reinjection and gravity changes at Rotokawa geothermal field, New Zealand . Geothermics 36 , 421 – 435 . Google Scholar Crossref Search ADS Husen S. , Taylor R. , Smith R. B. , Healser H. ( 2004 ). Changes in geyser eruption behavior and remotely triggered seismicity in Yellowstone National Park produced by the 2002 M 7.9 Denali fault earthquake, Alaska . Geology 32 , 537 – 540 . Google Scholar Crossref Search ADS Ivanic T. , Zibra I. , Doublier M. , Wyche S. ( 2013 ). Preliminary interpretation of the 2010 Youanmi deep seismic reflection lines and magnetotelluric data for the Windimurra Igneous Complex. In: Wyche S. , Ivanic T. , Zibra I. (eds) Youanmi and Southern Carnarvon Seismic and Magnetotelluric (MT) Workshop 2013 . Geological Survey of Western Australia , Record 6, 93–107. Jachens R. C. , Roberts C. W. ( 1985 ). Temporal and areal gravity investigations at Long Valley Caldera, California . Journal of Geophysical Research 90 , 11210 – 11218 . Google Scholar Crossref Search ADS Jackson C. A.-L. , Schofield N. , Golenkov B. ( 2013 ). Geometry and controls on the development of igneous sill–related forced folds: a 2-D seismic reflection case study from offshore southern Australia . Geological Society of America Bulletin 125 , 1874 – 1890 . Google Scholar Crossref Search ADS Jackson I. , Fitz Gerald J. D. , Faul U. H. , Tan B. H. ( 2002 ). Grain-size-sensitive seismic wave attenuation in polycrystalline olivine . Journal of Geophysical Research: Solid Earth 107 , 2156 – 2202 . Google Scholar Crossref Search ADS Jaxybulatov K. , Shapiro N. M. , Koulakov I. , Mordret A. , Landès M. , Sens-Schönfelder C. ( 2014 ). A large magmatic sill complex beneath the Toba caldera . Science 346 , 617 – 619 . Google Scholar Crossref Search ADS PubMed Jay J. , Costa F. , Pritchard M. , Lara L. , Singer B. , Herrin J. ( 2014 ). Locating magma reservoirs using InSAR and petrology before and during the 2011–2012 Cordón Caulle silicic eruption . Earth and Planetary Science Letters 395 , 254 – 266 . Google Scholar Crossref Search ADS Jay J. A. , Pritchard M. E. , West M. E. , Christensen D. H. , Haney M. , Minaya E. , Sunagua M. , McNutt S. R. , Zabala M. ( 2012 ). Shallow seismicity, triggered seismicity, and ambient noise tomography at the long-dormant Uturuncu volcano, Bolivia . Bulletin of Volcanology 74 , 817 – 837 . Google Scholar Crossref Search ADS Johnson J. H. , Poland M. P. ( 2013 ). Seismic detection of increased degassing before Kīlauea’s 2008 summit explosion . Nature Communications 4 , 1668 . Google Scholar Crossref Search ADS PubMed Johnson N. E. , Whaler K. A. , Hautot S. , Fisseha S. , Desissa M. , Dawes G. J. K. ( 2016 ). Magma imaged magnetotellurically beneath an active and an inactive magmatic segment in Afar, Ethiopia . Geological Society, London, Special Publications 420 , 105 – 125 . Google Scholar Crossref Search ADS Jones A. G. ( 2012 ). Distortion of magnetotelluric data: its identification and removal. In: Chave A. D. , Jones A. G. (ed.) The Magnetotelluric Method: Theory and Practice . Cambridge University Press . Jousset P. , Mori H. , Okada H. ( 2000 ). Possible magma intrusion revealed by temporal gravity, ground deformation and ground temperature observations at Mount Komagatake (Hokkaido) during the 1996-1998 crisis . Geophysical Journal International 143 , 557 – 574 . Google Scholar Crossref Search ADS Kalinicheva T. , Warner M. , Ashley J. , Mancini F. ( 2017 ). Two vs three-dimensional FWI in a 3D world. SEG Technical Program Expanded Abstracts 2017, Society of Exploration Geophysicists, 1383–1387. Kamei R. , Jang U. G. , Lumley D. , Mouri T. , Nakatsukasa M. , Kato A. , Takanashi M. ( 2017 ). Time-lapse full waveform inversion for monitoring near-surface microbubble injection. 79th Conference and Exhibition, EAGE, Extended Abstracts. Kapoor S. , Vigh D. , Wiarda E. , Alwon S. ( 2013 ). Full waveform inversion around the world. 75th Conference and Exhibition, EAGE, Extended Abstracts. Karato S. , Jung H. ( 1998 ). Water, partial melting and the origin of the seismic low velocity and high attenuation zone in the upper mantle . Earth and Planetary Science Letters 157 , 193 – 207 . Google Scholar Crossref Search ADS Kavanagh J. L. , Burns A. J. , Hazim S. H. , Wood E. , Martin S. A. , Hignett S. , Dennis D. J. ( 2018 ). Challenging dyke ascent models using novel laboratory experiments: implications for reinterpreting evidence of magma ascent and volcanism . Journal of Volcanology and Geothermal Research 354 , 87–101. Keir D. , Belachew M. , Ebinger C. J. , Kendall J.-M. , Hammond J. O. S. , Stuart G. W. , Ayele A. , Rowland J. R. ( 2011 ). Mapping the evolving strain field during continental breakup from crustal anisotropy in the Afar Depression . Nature Communications 2 , 285 – 287 . Google Scholar Crossref Search ADS PubMed Keir D. , Hamling I. J. , Ayele A. , Calais E. , Ebinger C. J. , Wright T. J. , Jacques E. , Mohamed K. , Hammond J. O. S. , Belachew M. , Baker E. , Rowland J. V. , Lewi E. , Bennati L. ( 2009 ). Evidence for focused magmatic accretion at segment centers from lateral dike injection captured beneath the Red Sea rift of Afar . Geology 37 , 59 – 62 . Google Scholar Crossref Search ADS Kelso P. R. , Tikoff B. , Jackson M. , Sun W. ( 2002 ). A new method for the separation of paramagnetic and ferromagnetic susceptibility anisotropy using low field and high field methods . Geophysical Journal International 151 , 345 – 359 . Google Scholar Crossref Search ADS Kent G. M. , Singh S. C. , Harding A. J. , Sinha M. C. , Orcutt J. A. , Barton P. J. , White R. S. , Bazin S. , Hobbs R. W. , Tong C. H. , Pye J. W. ( 2000 ). Evidence from three-dimensional seismic reflectivity images for enhanced melt supply beneath mid-ocean-ridge discontinuities . Nature 406 , 614 – 618 . Google Scholar Crossref Search ADS PubMed Key K. , Constable S. , Liu L. , Pommier A. ( 2013 ). Electrical image of passive upwelling beneath the northern East Pacific Rise . Nature 495 , 499 – 502 . Google Scholar Crossref Search ADS PubMed Khan M. A. ( 1962 ). The anisotropy of magnetic susceptibility of some igneous and metamorphic rocks . Journal of Geophysical Research 67 , 2873 – 2885 . Google Scholar Crossref Search ADS Kim D. , Brown L. D. , Árnason K. , Águstsson K. , Blanck H. ( 2017 ). Magma reflection imaging in Krafla, Iceland, using microearthquake sources . Journal of Geophysical Research: Solid Earth 122 , 5228 – 5242 . Google Scholar Crossref Search ADS Kiser E. , Levander A. , Schmandt B. , Palomeras I. , Harder S. H. , Creager K. C. , Vidale J. E. , Malone S. D. ( 2014 ). Field report on the iMUSH active source seismic experiment . AGU Fall Meeting Abstracts 1 , 2 . Knight M. D. , Walker G. P. ( 1988 ). Magma flow directions in dikes of the Koolau Complex, Oahu, determined from magnetic fabric studies . Journal of Geophysical Research: Solid Earth 93 , 4301 – 4319 . Google Scholar Crossref Search ADS Koulakov I. , Gordeev E. I. , Dobretsov N. L. , Vernikovsky V. A. , Senyukov S. , Jakovlev A. , Jaxybulatov K. ( 2013 ). Rapid changes in magma storage beneath the Klyuchevskoy group of volcanoes inferred from time-dependent seismic tomography . Journal of Volcanology and Geothermal Research 263 , 75 – 91 . Google Scholar Crossref Search ADS Laumonier M. , Gaillard F. , Sifre D. ( 2015 ). The effect of pressure and water concentration on the electrical conductivity of dacitic melts: Implication for magnetotelluric imaging in subduction areas . Chemical Geology 418 , 66 – 76 . Google Scholar Crossref Search ADS Launeau P. , Cruden A. R. ( 1998 ). Magmatic fabric acquisition mechanisms in a syenite: results of a combined anisotropy of magnetic susceptibility and image analysis study . Journal of Geophysical Research: Solid Earth 103 , 5067 – 5089 . Google Scholar Crossref Search ADS Lazaratos S. , Chikichev I. , Wang K. ( 2011 ). Improving the convergence rate of full wavefield inversion using spectral shaping. SEG Technical Program Expanded Abstracts 2011, Society of Exploration Geophysicists, 2428–2432. Leberl F. , Irschara A. , Pock T. , Meixner P. , Gruber M. , Scholz S. , Wiechert A. ( 2010 ). Point Clouds: Lidar versus 3D Vision . Photogrammetric Engineering & Remote Sensing 76 , 1123 – 1134 . Google Scholar Crossref Search ADS Lees J. M. ( 2007 ). Seismic tomography of magmatic systems . Journal of Volcanology and Geothermal Research 167 , 37 – 56 . Google Scholar Crossref Search ADS Le Mével H. , Feigl K. L. , Córdova L. , DeMets C. , Lundgren P. ( 2015 ). Evolution of unrest at Laguna del Maule volcanic field (Chile) from InSAR and GPS measurements, 2003 to 2014 . Geophysical Research Letters 42 , 6590 – 6598 . Google Scholar Crossref Search ADS Lu Z. , Dzurisin D. ( 2014 ). InSAR Imaging of Aleutian Volcanoes . Berlin, Heidelberg : Springer , pp. 87 – 345 . MacGregor L. M. , Constable S. , Sinha M. C. ( 1998 ). The RAMESSES experiment—III. Controlled-source electromagnetic sounding of the Reykjanes Ridge at 57° 45′ N . Geophysical Journal International 135 , 773 – 789 . Google Scholar Crossref Search ADS Magee C. , Briggs F. , Jackson C. A.-L. ( 2013a ). Lithological controls on igneous intrusion-induced ground deformation . Journal of the Geological Society 170 , 853 – 856 . Google Scholar Crossref Search ADS Magee C. , Hunt-Stewart E. , Jackson C. A. L. ( 2013b ). Volcano growth mechanisms and the role of sub-volcanic intrusions: Insights from 2D seismic reflection data . Earth and Planetary Science Letters 373 , 41 – 53 . Google Scholar Crossref Search ADS Magee C. , Jackson C. L. , Schofield N. ( 2014 ). Diachronous sub‐volcanic intrusion along deep‐water margins: insights from the Irish Rockall Basin . Basin Research 26 , 85 – 105 . Google Scholar Crossref Search ADS Magee C. , Maharaj S. M. , Wrona T. , Jackson C. A.-L. ( 2015 ). Controls on the expression of igneous intrusions in seismic reflection data . Geosphere 11 , 1024 . GES01150.01151. Google Scholar Crossref Search ADS Magee C. , Stevenson C. T. E. , O’Driscoll B. , Petronis M. S. ( 2012a ). Local and regional controls on the lateral emplacement of the Ben Hiant Dolerite intrusion, Ardnamurchan (NW Scotland) . Journal of Structural Geology 39 , 66 – 82 . Google Scholar Crossref Search ADS Magee C. , Stevenson C. T. E. , O’Driscoll B. , Schofield N. , McDermott K. ( 2012b ). An alternative emplacement model for the classic Ardnamurchan cone sheet swarm, NW Scotland, involving lateral magma supply via regional dykes . Journal of Structural Geology 43 , 73 – 91 . Google Scholar Crossref Search ADS Magee C. , Bastow I. D. , van Wyk de Vries B. , Jackson C. A. L. , Hetherington R. , Hagos M. , Hoggett M. ( 2017a ). Structure and dynamics of surface uplift induced by incremental sill emplacement . Geology 45 , 431 – 434 . Google Scholar Crossref Search ADS Magee C. , Muirhead J. , Schofield N. , Walker R. , Galland O. , Holford S. , Spacapan J. , Jackson C., A.-L. , McCarthy W. ( 2017b ). Structural signatures of igneous sheet intrusion propagation . EarthArXiv doi:10.17605/OSF.IO/PDN42. Magee C. , Muirhead J. D. , Karvelas A. , Holford S. P. , Jackson C. A. L. , Bastow I. D. , Schofield N. , Stevenson C. T. E. , McLean C. , McCarthy W. , Shtukert O. ( 2016 ). Lateral magma flow in mafic sill complexes . Geosphere 12 , 809 . Google Scholar Crossref Search ADS Malehmir A. , Durrheim R. , Bellefleur G. , Urosevic M. , Juhlin C. , White D. J. , Milkereit B. , Campbell G. ( 2012 ). Seismic methods in mineral exploration and mine planning: a general overview of past and present case histories and a look into the future . Geophysics 77 , WC173 – WC190 . Google Scholar Crossref Search ADS Mamtani M. A. , Greiling R. O. ( 2005 ). Granite emplacement and its relation with regional deformation in the Aravalli Mountain Belt (India) --inferences from magnetic fabric . Journal of Structural Geology 27 , 2008 – 2029 . Google Scholar Crossref Search ADS Mamtani M. A. , Pal T. , Greiling R. O. ( 2013 ). Kinematic analysis using AMS data from a deformed granitoid . Journal of Structural Geology 50 , 119 – 132 . Google Scholar Crossref Search ADS Mancini F. , Moss J. , Prindle K. , Ridsdill-Smith T. ( 2015 ). Where can full waveform inversion have the biggest impact in the exploration and production cycle? 77th International Conference and Exhibition, EAGE, Extended Abstracts. Manzella A. , Volpi G. , Zaja A. , Meju M. ( 2004 ). Combined TEM-MT investigation of shallow-depth resistivity structure of Mt Somma-Vesuvius . Journal of Volcanology and Geothermal Research 131 , 19 – 32 . Google Scholar Crossref Search ADS Marjanovic M. , Carbotte S. M. , Carton H. , Nedimovic M. R. , Mutter J. C. , Canales J. P. ( 2014 ). A multi-sill magma plumbing system beneath the axis of the East Pacific Rise . Nature Geoscience 7 , 825 – 829 . Google Scholar Crossref Search ADS Mark N. J. , Schofield N. , Pugliese S. , Watson D. , Holford S. , Muirhead D. , Brown R. , Healy D. ( 2017 ). Igneous intrusions in the Faroe Shetland basin and their implications for hydrocarbon exploration; new insights from well and seismic data . Marine and Petroleum Geology 9 2 , 733–753. Martín-Hernández F. , Lüneburg C. M. , Aubourg C. , Jackson M. ( 2004 ). Magnetic fabric: methods and applications–an introduction. In: Martín-Hernández F. , Lüneburg C. M. , Aubourg C. & Jackson M. (eds) Magnetic Fabric: Methods and Applications. Geological Society , London , Special Publications 238 , 1 - 7 . Masterlark T. ( 2007 ). Magma intrusion and deformation predictions: sensitivities to the Mogi assumptions . Journal of Geophysical Research 112 , B06419. Masturyono , McCaffrey R. , Wark D. A. , Roecker S. W. , Fauzi Ibrahim G. , Sukhyar . ( 2001 ). Distribution of magma beneath Toba caldera complex, north Sumatra, Indonesia, constrained by three-dimensional P wave velocities, seismicity, and gravity . Geochemitry Geophysics Geosystems 2 , 1014 . Matsushima N. , Oshima H. , Ogawa Y. , Takakura S. , Satoh H. , Utsugi M. , Nishida Y. ( 2001 ). Magma prospecting in Usu volcano, Hokkaido, Japan, using magnetotelluric soundings . Journal of Volcanology and Geothermal Research 109 , 263 – 277 . Google Scholar Crossref Search ADS McBride J. H. , William Keach R. , Leetaru H. E. , Smith K. M. ( 2018 ). Visualizing Precambrian basement tectonics beneath a carbon capture and storage site, Illinois Basin . Interpretation 6 , T257 – T270 . Google Scholar Crossref Search ADS McCabe C. , Jackson M. , Ellwood B. B. ( 1985 ). Magnetic anisotropy in the Trenton limestone: results of a new technique, anisotropy of anhysteretic susceptibility . Geophysical Research Letters 12 , 333 – 336 . Google Scholar Crossref Search ADS McCormick Kilbride B. , Edmonds M. , Biggs J. ( 2016 ). Observing eruptions of gas-rich compressible magmas from space . Nature Communications 7 , 13744 . Google Scholar Crossref Search ADS PubMed McKenzie D. A. N. , O’Nions R. K. ( 1991 ). Partial melt distributions from inversion of rare earth element concentrations . Journal of Petrology 32 , 1021 – 1091 . Google Scholar Crossref Search ADS McLean C. E. , Schofield N. , Brown D. J. , Jolley D. W. , Reid A. ( 2017 ). 3D seismic imaging of the shallow plumbing system beneath the Ben Nevis Monogenetic Volcanic Field: Faroe-Shetland Basin . Journal of the Geological Society 174 , 468 – 485 . Google Scholar Crossref Search ADS Miller C. A. , Le Mével H. , Currenti G. , Williams-Jones G. , Tikoff B. ( 2017 ). Microgravity changes at the Laguna del Maule volcanic field: Magma-induced stress changes facilitate mass addition . Journal of Geophysical Research: Solid Earth 122 , 3179 – 3196 . Google Scholar Crossref Search ADS Miller C. F. , Furbish D. J. , Walker B. A. , Claiborne L. L. , Koteas G. C. , Bleick H. A. , Miller J. S. ( 2011 ). Growth of plutons by incremental emplacement of sheets in crystal-rich host: evidence from Miocene intrusions of the Colorado River region, Nevada, USA . Tectonophysics 500 , 65 – 77 . Google Scholar Crossref Search ADS Miller K. J. , Zhu W. , Montesi L. G. J. , Gaetani G. A. ( 2014 ). Experimental quantification of permeability of partially molten mantle rock . Earth and Planetary Science Letters 388 , 273 – 282 . Google Scholar Crossref Search ADS Miller V. , Savage M. ( 2001 ). Changes in seismic anisotropy after volcanic eruptions: evidence from Mount Ruapehu . Science 293 , 2231 – 2233 . Google Scholar Crossref Search ADS PubMed Mitchell M. A. , White R. S. , Roecker S. , Greenfield T. ( 2013 ). Tomographic image of melt storage beneath Askja Volcano, Iceland using local microseismicity . Geophysical Research Letters 40 , 5040 – 5046 . Google Scholar Crossref Search ADS Morgan J. V. , Warner M. , Arnoux G. , Hooft E. , Toomey D. , VanderBeek B. , Wilcock W. ( 2016 ). Next-generation seismic experiments—II: wide-angle, multi-azimuth, 3-D, full-waveform inversion of sparse field data . Geophysical Journal International 204 , 1342 – 1363 . Google Scholar Crossref Search ADS Morgan J. V. , Warner M. , Bell R. E. , Ashley J. , Barnes D. , Little R. , Roele K. , Jones C. ( 2013 ). Next-generation seismic experiments: wide-angle, multi-azimuth, three-dimensional, full-waveform inversion . Geophysical Journal International 195 , 1657 – 1678 . Google Scholar Crossref Search ADS Muirhead J. D. , Airoldi G. , White J. D. , Rowland J. V. ( 2014 ). Cracking the lid: sill-fed dikes are the likely feeders of flood basalt eruptions . Earth and Planetary Science Letters 406 , 187 – 197 . Google Scholar Crossref Search ADS Neuberg J. , Baptie B. , Luckett R. , Stewart R. ( 1998 ). Results from the broadband seismic network on Montserrat . Geophysical Research Letters 25 , 3661 – 3664 . Google Scholar Crossref Search ADS Neuberg J. , Tuffen H. , Collier L. , Green D. , Powell T. , Dingwell D. ( 2006 ). The trigger mechanism of low-frequency earthquakes on Montserrat . Journal of Volcanology and Geothermal Research 153 , 37 – 50 . Google Scholar Crossref Search ADS Neves S. P. , Araújo A. M. , Correia P. B. , Mariano G. ( 2003 ). Magnetic fabrics in the Cabanas Granite (NE Brazil): interplay between emplacement and regional fabrics in a dextral transpressive regime . Journal of Structural Geology 25 , 441 – 453 . Google Scholar Crossref Search ADS Newman A. V. , Dixon T. H. , Gourmelen N. ( 2006 ). A four-dimensional viscoelastic deformation model for Long Valley Caldera, California, between 1995 and 2000 . Journal of Volcanology and Geothermal Research 150 , 244 – 269 . Google Scholar Crossref Search ADS O’Driscoll B. , Stevenson C. T. E. , Troll V. R. ( 2008 ). Mineral lamination development in layered gabbros of the British Palaeogene Igneous Province: A combined anisotropy of magnetic susceptibility, textural and mineral chemistry study . Journal of Petrology 49 , 1187 – 1221 . Google Scholar Crossref Search ADS O’Driscoll B. , Ferré E. C. , Stevenson C. T. E. , Magee C. ( 2015 ). The significance of magnetic fabrics in layered mafic-ultramafic intrusions. In: Charlier B. , Latypov R. , Namur O. , Tegner C. (eds) Layered Intrusions . Springer , Berlin, pp. 295 – 329 . Okubo P. G. , Benz H. M. , Chouet B. A. ( 1997 ). Imaging the crustal magma sources beneath Mauna Loa and Kilauea volcanoes, Hawaii . Geology 25 , 867 – 870 . Google Scholar Crossref Search ADS Orlický O. ( 1990 ). Detection of magnetic carriers in rocks: results of susceptibility changes in powdered rock samples induced by temperature . Physics of the Earth and Planetary Interiors 63 , 66 – 70 . Google Scholar Crossref Search ADS Padilha A. L. , Vitorello Í. , Pádua M. B. , Bologna M. S. ( 2006 ). Lithospheric and sublithospheric anisotropy beneath central-southeastern Brazil constrained by long period magnetotelluric data . Physics of the Earth and Planetary Interiors 158 , 190 – 209 . Google Scholar Crossref Search ADS Pagli C. , Wright T. J. , Ebinger C. J. , Yun S. H. , Cann J. R. , Barnie T. , Ayele A. ( 2012 ). Shallow axial magma chamber at the slow-spreading Erta Ale Ridge . Nature Geoscience 5 , 284 – 288 . Google Scholar Crossref Search ADS Parker A. L. ( 2016 ). InSAR Observations of Ground Deformation: Application to the Cascades Volcanic Arc . Springer Theses , Cham, Switzerland. Parker R. L. ( 1980 ). The inverse problem of electromagnetic induction: Existence and construction of solutions based on incomplete data . Journal of Geophysical Research: Solid Earth 85 , 4421 – 4428 . Google Scholar Crossref Search ADS Parker A. L. , Biggs J. , Lu Z. ( 2016 ). Time-scale and mechanism of subsidence at Lassen Volcanic Center, CA, from InSAR . Journal of Volcanology and Geothermal Research 320 , 117 – 127 . Google Scholar Crossref Search ADS Parks M. M. , Biggs J. , England P. , Mather T. A. , Nomikou P. , Palamartchouk K. , Papanikolaou X. , Paradissis D. , Parsons B. , Pyle D. M. , Raptakis C. , Zacharis V. ( 2012 ). Evolution of Santorini Volcano dominated by episodic and rapid fluxes of melt from depth . Nature Geoscience 5 , 749 – 754 . Google Scholar Crossref Search ADS Parmigiani A. , Faroughi S. A. , Huber C. , Bachmann O. , Su Y. ( 2016 ). Bubble accumulation and its role on the evolution of upper crustal magma reservoirs . Nature 532 , 492 – 495 . Google Scholar Crossref Search ADS PubMed Paulatto M. , Annen C. , Henstock T. J. , Kiddle E. , Minshull T. A. , Sparks R. S. J. , Voight B. ( 2012 ). Magma chamber properties from integrated seismic tomography and thermal modeling at Montserrat . Geochemistry, Geophysics, Geosystems 13 , Q01014. Paulatto M. , Minshull T. A. , Baptie B. , Dean S. , Hammond J. O. S. , Henstock T. , Kenedi C. L. , Kiddle E. J. , Malin P. , Peirce C. , Ryan G. , Shalev E. , Sparks R. S. J. , Voight B. ( 2010 ). Upper crustal structure of an active volcano from refraction/reflection tomography, Montserrat, Lesser Antilles . Geophysical Journal International 180 , 685 – 696 . Google Scholar Crossref Search ADS Pedersen R. , Sigmundsson F. ( 2006 ). Temporal development of the 1999 intrusive episode in the Eyjafjallajökull volcano, Iceland, derived from InSAR images . Bulletin of Volcanology 68 , 377 – 393 . Google Scholar Crossref Search ADS Peron-Pinvidic G. , Shillington D. J. , Tucholke B. E. ( 2010 ). Characterization of sills associated with the U reflection on the Newfoundland margin: evidence for widespread early post-rift magmatism on a magma-poor rifted margin . Geophysical Journal International 182 , 113 – 136 . Petford N. , Cruden A. R. , McCaffrey K. J. W. , Vigneresse J. L. ( 2000 ). Granite magma formation, transport and emplacement in the Earth’s crust . Nature 408 , 669 – 673 . Google Scholar Crossref Search ADS PubMed Petronis M. S. , O'Driscoll B. , Stevenson C. T. E. , Reavy R. J. ( 2012 ). Controls on emplacement of the Caledonian Ross of Mull Granite, NW Scotland: Anisotropy of magnetic susceptibility and magmatic and regional structures . Geological Society of America Bulletin 124 , 906 – 927 . Google Scholar Crossref Search ADS Petronis M. S. , Delcamp A. , de Vries B. V. W. ( 2013 ). Magma emplacement into the Lemptégy scoria cone (Chaîne Des Puys, France) explored with structural, anisotropy of magnetic susceptibility, and Paleomagnetic data . Bulletin of Volcanology 75 , 753 . Google Scholar Crossref Search ADS Petronis M. S. , Brister A. R. , Rapprich V. , de Vries B. V. W. , Lindline J. , Misurec J. ( 2015 ). Emplacement history of the Trosky basanitic volcano (Czech Republic): paleomagnetic, rock magnetic, petrologic, and anisotropy of magnetic susceptibility evidence for lingering growth of a monogenetic volcano . Journal of Geosciences 60 , 129 – 147 . Google Scholar Crossref Search ADS Petrovský E. , Kapička A. ( 2006 ). On determination of the Curie point from thermomagnetic curves . Journal of Geophysical Research: Solid Earth 111 , B12S27. Phillips T. B. , Magee C. , Jackson C. A. L. , Bell R. E. ( 2017 ). Determining the three-dimensional geometry of a dike swarm and its impact on later rift geometry using seismic reflection data . Geology 46 , 119 – 122 . Google Scholar Crossref Search ADS Pinel V. , Poland M. P. , Hooper A. ( 2014 ). Volcanology: lessons learned from synthetic aperture radar imagery . Journal of Volcanology and Geothermal Research 289 , 81 – 113 . Google Scholar Crossref Search ADS Planke S. , Rasmussen T. , Rey S. S. , Myklebust R. ( 2005 ). Seismic characteristics and distribution of volcanic intrusions and hydrothermal vent complexes in the Vøring and Møre basins. In: Doré A. G. (ed.) Petroleum Geology: North-West Europe and Global Perspectives—Proceddings of the 6th Petroleum Geology Conference: Geological Society 6, London 833 – 844 . Planke S. , Symonds P. A. , Alvestad E. , Skogseid J. ( 2000 ). Seismic volcanostratigraphy of large‐volume basaltic extrusive complexes on rifted margins . Journal of Geophysical Research: Solid Earth 105 , 19335 – 19351 . Google Scholar Crossref Search ADS Poland M. P. , Carbone D. ( 2016 ). Insights into shallow magmatic processes at Kīlauea Volcano, Hawai‘i, from a multiyear continuous gravity time series . Journal of Geophysical Research: Solid Earth 121 , 5477 – 5492 . Google Scholar Crossref Search ADS Poland M. P. , Miklius A. , Sutton A. J. , Thornber C. R. ( 2012 ). A mantle-driven surge in magma supply to Kilauea Volcano during 2003-2007 . Nature Geoscience 5 , 295 – 300 . Google Scholar Crossref Search ADS Pommier A. ( 2014 ). Interpretation of magnetotelluric results using laboratory measurements . Surveys in Geophysics 35 , 41 – 84 . Google Scholar Crossref Search ADS Pommier A. , Le Trong E. ( 2011 ). SIGMELTS: a web portal for electrical conductivity calculations in geosciences . Computational Geoscience 37 , 1450 – 1459 . Google Scholar Crossref Search ADS Pommier A. , Tarits P. , Hautot S. , Pichavant M. , Scaillet B. , Gaillard F. ( 2010 ). A new petrological and geophysical investigation of the present‐day plumbing system of Mount Vesuvius . Geochemistry, Geophysics, Geosystems 11 , Q07013. Pritchard M. E. , Simons M. ( 2004 ). An InSAR‐based survey of volcanic deformation in the central Andes . Geochemistry, Geophysics, Geosystems 5 , Q02002. Prutkin I. , Vajda P. , Gottsmann J. ( 2014 ). The gravimetric picture of magmatic and hydrothermal sources driving hybrid unrest on Tenerife in 2004/5 . Journal of Volcanology and Geothermal Research 282 , 9 – 18 . Google Scholar Crossref Search ADS Rabbel O. , Galland O. , Mair K. , Lecomte I. , Senger K. , Spacapan J. B. , Manceda R. ( 2018 ). From field analogues to realistic seismic modelling: a case study of an oil-producing andesitic sill complex in the Neuquén Basin, Argentina . Journal of the Geological Society 175 , 580 – 593 . Google Scholar Crossref Search ADS Reeves J. , Magee C. , Jackson C. A. L. ( 2018 ). Unravelling intrusion-induced forced fold kinematics and ground deformation using 3D seismic reflection data . Volcanica 1 , 1 – 17 . Google Scholar Crossref Search ADS Reich H. ( 1932 ). Die Bedeutung der finnischen Schweremessungen fur die angewandte Geophysik . Ergänz.-Hefte für die angewandte Geophysik 2 , 1 – 13 . Richter C. , van der Pluijm B. A. ( 1994 ). Separation of paramagnetic and ferrimagnetic susceptibilities using low temperature magnetic susceptibilities and comparison with high field methods . Physics of the Earth and Planetary Interiors 82 , 113 – 123 . Google Scholar Crossref Search ADS Rivalta E. , Segall P. ( 2008 ). Magma compressibility and the missing source for some dike intrusions . Geophysical Research Letters 35 , L04306. Roberts J. J. , Tyburczy J. A. ( 1999 ). Partial-melt electrical conductivity: Influence of melt composition . Journal of Geophysical Research: Solid Earth 104 , 7055 – 7065 . Google Scholar Crossref Search ADS Rocchi S. , Mazzotti A. , Marroni M. , Pandolfi L. , Costantini P. , Giuseppe B. , Biase D. D. , Federici F. , Lo P. G. ( 2007 ). Detection of Miocene saucer‐shaped sills (offshore Senegal) via integrated interpretation of seismic, magnetic and gravity data . Terra Nova 19 , 232 – 239 . Google Scholar Crossref Search ADS Rochette P. , Jackson M. , Aubourg C. ( 1992 ). Rock magnetism and the interpretation of anisotropy of magnetic susceptibility . Reviews of Geophysics 30 , 209 – 226 . Google Scholar Crossref Search ADS Roman D. C. , Cashman K. V. ( 2006 ). The origin of volcano-tectonic earthquake swarms . Geology 34 , 457 – 460 . Google Scholar Crossref Search ADS Roman D. C. , Savage M. K. , Arnold R. , Latchman J. L. , De Angelis S. ( 2011 ). Analysis and forward modeling of seismic anisotropy during the ongoing eruption of the Soufrière Hills Volcano, Montserrat, 1996–2007 . Journal of Geophysical Research 116 , B03201 . Google Scholar Crossref Search ADS Ronchin E. , Masterlark T. , Dawson J. , Saunders S. , Martì Molist J. ( 2017 ). Imaging the complex geometry of a magma reservoir using FEM-based linear inverse modeling of InSAR data: application to Rabaul Caldera, Papua New Guinea . Geophysical Journal International 209 , 1746 – 1760 . Google Scholar Crossref Search ADS Routh P. , Neelamani R. , Lu R. , Lazaratos S. , Braaksma H. , Hughes S. , Saltzer R. , Stewart J. , Naidu K. , Averill H. , Gottumukkula V. , Homonko P. , Reilly J. , Leslie D. ( 2017 ). Impact of high-resolution FWI in the western black sea: revealing overburden and reservoir complexity . The Leading Edge 36 , 60 – 66 . Google Scholar Crossref Search ADS Rubin A. , Gillard D. ( 1998 ). Dike-induced earthquakes: theoretical considerations . Journal of Geophysical Research: Solid Earth 103 , 10017 – 10030 . Google Scholar Crossref Search ADS Rymer H. , Brown G. C. ( 1986 ). Gravity fields and the interpretation of volcanic structures: Geological discrimination and temporal evolution . Journal of Volcanology and Geothermal Research 27 , 229 – 254 . Google Scholar Crossref Search ADS Rymer H. , Locke C. A. , Brenes J. , Williams-Jones G. ( 2005 ). Magma plumbing processes for persistent activity at Poas volcano, Costa Rica . Geophysical Research Letters 32 , L08307 . Google Scholar Crossref Search ADS Salzer J. T. , Nikkhoo M. , Walter T. R. , Sudhaus H. , Reyes-Dávila G. , Bretón M. , Arámbula R. ( 2014 ). Satellite radar data reveal short-term pre-explosive displacements and a complex conduit system at Volcán de Colima, Mexico . Frontiers in Earth Science 2 , 12 . Google Scholar Crossref Search ADS Samrock F. , Kuvshinov A. , Bakker J. , Jackson A. , Fisseha S. ( 2015 ). 3D analysis and interpretation of Magnetotelluric data from the Aluto-Langano geothermal field, Ethiopia . Geophysical Journal International 202 , 1923 – 1948 . Google Scholar Crossref Search ADS Saunders K. , Blundy J. D. , Dohmen R. , Cashman K. ( 2012 ). Linking Petrology and Seismology at an active volcano . Science 336 , 1023 – 1027 . Google Scholar Crossref Search ADS PubMed Saxby J. , Gottsmann J. , Cashman K. , Gutiérrez E. ( 2016 ). Magma storage in a strike-slip caldera . Nature Communications 7 , 12295 . Google Scholar Crossref Search ADS PubMed Schaefer L. N. , Lu Z. , Oommen T. ( 2015 ). Dramatic volcanic instability revealed by InSAR . Geology 43 , 743 – 746 . Google Scholar Crossref Search ADS Schmeling H. ( 1986 ). Numerical models on the influence of partial melt on elastic, anelastic and electrical properties of rocks. Part II: electrical conductivity . Physics of Earth and Planetary Interiors 43 , 123 – 136 . Google Scholar Crossref Search ADS Schofield N. J. , Brown D. J. , Magee C. , Stevenson C. T. E. ( 2012a ). Sill morphology and comparison of brittle and non-brittle emplacement mechanisms . Journal of the Geological Society of London 169 , 127 – 141 . Google Scholar Crossref Search ADS Schofield N. , Heaton L. , Holford S. P. , Archer S. G. , Jackson C. A. L. , Jolley D. W. ( 2012b ). Seismic imaging of ‘broken bridges’: linking seismic to outcrop-scale investigations of intrusive magma lobes . Journal of the Geological Society of London 169 , 421 – 426 . Google Scholar Crossref Search ADS Schofield N. J. , Alsop I. , Warren J. , Underhill J. R. , Lehné R. , Beer W. , Lukas V. ( 2014 ). Mobilizing salt: Magma-salt interactions . Geology 42 , 599 – 602 . Google Scholar Crossref Search ADS Schofield N. , Holford S. , Millett J. , Brown D. , Jolley D. , Passey S. , Muirhead D. , Grove C. , Magee C. , Murray J. , Hole M. , Jackson C. , Stevenson C. ( 2017 ). Regional Magma Plumbing and emplacement mechanisms of the Faroe-Shetland Sill Complex: implications for magma transport and petroleum systems within sedimentary basins . Basin Research 29 , 41 – 63 . Google Scholar Crossref Search ADS Schofield N. , Stevenson C. T. E. , Reston T. ( 2010 ). Magma fingers and host rock fluidization in the emplacement of sills . Geology 38 , 63 – 66 . Google Scholar Crossref Search ADS Segall P. ( 2010 ). Earthquake and Volcano Deformation . Princeton University Press, Princeton . Shalev E. , Kenedi C. L. , Malin P. , Voight V. , Miller V. , Hidayat D. , Sparks R. S. J. , Minshull T. , Paulatto M. , Brown L. , Mattioli G. ( 2010 ). Three‐dimensional seismic velocity tomography of Montserrat from the SEA‐CALIPSO offshore/onshore experiment . Geophysical Research Letters 37 , L00E17. Sigmundsson F. ( 2016 ). New insights into magma plumbing along rift systems from detailed observations of eruptive behavior at Axial volcano . Geophysical Research Letters 43 , 12423 – 12427 . Google Scholar Crossref Search ADS Sigmundsson F. , Hooper A. , Hreinsdóttir S. , Vogfjörd K. S. , Ófeigsson B. G. , Heimisson E. R. , Dumont S. , Parks M. , Spaans K. , Gudmundsson G. B. , Drouin V. , Árnadóttir T. , Jónsdóttir K. , Gudmundsson M. T. , Högnadóttir T. , Fridriksdóttir H. M. , Hensch M. , Einarsson P. , Magnússon E. , Samsonov S. , Brandsdóttir B. , White R. S. , Ágústsdóttir T. , Greenfield T. , Green R. G. , Hjartardóttir Á. R. , Pedersen R. , Bennett R. A. , Geirsson H. , La Femina P. C. , Björnsson H. , Pálsson F. , Sturkell E. , Bean C. J. , Möllhoff M. , Braiden A. K. , Eibl E. P. S. ( 2015 ). Segmented lateral dyke growth in a rifting event at Bardabunga volcanic system, Iceland . Nature 517 , 191 – 195 . Google Scholar Crossref Search ADS PubMed Sigmundsson F. , Hreinsdóttir S. , Hooper A. , Arnadóttir T. , Pedersen R. , Roberts M. J. , Oskarsson N. , Auriac A. , Decriem J. , Einarsson P. , Geirsson H. , Hensch M. , Ofeigsson B. G. , Sturkell E. , Sveinbjörnsson H. , Feigl K. L. ( 2010 ). Intrusion triggering of the 2010 Eyjafjallajokull explosive eruption . Nature 468 , 426 – 430 . Google Scholar Crossref Search ADS PubMed Simons M. , Rosen P. A. ( 2007 ). Interferometric synthetic aperture radar Geodesy. In: Schubert, G. (ed.) Treatise on Geophysics—Geodesy , Vol. 3, Amsterdam : Elsevier , pp. 391 – 446 . Simpson F. , Bahr K. ( 2005 ). Practical Magnetotellurics . Cambridge : Cambridge University Press . Singer B. S. , Andersen N. L. , Le Mével H. , Feigl K. L. , DeMets C. , Tikoff B. , Thurber C. H. , Jicha B. R. , Cardona C. , Córdova L. , Gil F. , Unsworth M. J. , Williams-Jones G. , Miller C. , Fierstein J. , Hildreth W. , Vazquez J. ( 2014 ). Dynamics of a large, restless, rhyolitic magma system at Laguna del Maule, southern Andes, Chile . GSA Today 24, 4 – 10 . Sirgue L. , Barkved O. I. , Dellinger J. , Etgen J. , Albertin U. , Kommedal J. H. ( 2010 ). Full waveform inversion: the next leap forward in imaging at Valhall . First Break 28 , 65 – 70 . Google Scholar Crossref Search ADS Smallwood J. R. , Maresh J. ( 2002 ). The properties, morphology and distribution of igneous sills: modelling, borehole data and 3D seismic from the Faroe-Shetland area. In: Jolley D. W. , Bell B. R. (eds) The North Atlantic Igneous Province: Stratigraphy, Tectonic, Volcanic and Magmatic Processes . Geological Society, London , Special Publications 197, 271 – 306 . Sofyan Y. , Kamah Y. , Nishijima J. , Fujimitsu Y. , Ehara S. , Fukuda Y. , Taniguchi M. ( 2011 ). Mass variation in outcome to high production activity in Kamojang Geothermal Field, Indonesia: a reservoir monitoring with relative and absolute gravimetry . Earth, Planets and Space 63 , 1157 – 1167 . Google Scholar Crossref Search ADS Solano J. M. S. , Jackson M. D. , Sparks R. S. J. , Blundy J. D. , Annen C. ( 2012 ). Melt segregation in deep crustal hot zones: a mechanism for chemical differentiation, crustal assimilation and the formation of evolved magmas . Journal of Petrology 53 , 1999 – 2026 . Google Scholar Crossref Search ADS Sparks R. , Biggs J. , Neuberg J. ( 2012 ). Monitoring volcanoes . Science 335 , 1310 – 1311 . Google Scholar Crossref Search ADS PubMed Stacey F. D. , Joplin G. , Lindsay J. ( 1960 ). Magnetic anisotropy and fabric of some foliated rocks from S.E. Australia . Pure Applied Geophysics 47 , 30 – 40 . Google Scholar Crossref Search ADS Stankiewicz J. , Ryberg T. , Haberland C. , Natawidjaja D. ( 2010 ). Lake Toba volcano magma chamber imaged by ambient seismic noise tomography . Geophysical Research Letters 37 , L17306 . Google Scholar Crossref Search ADS Stephens T. L. , Walker R. J. , Healy D. , Bubeck A. , England R. W. , McCaffrey K. J. W. ( 2017 ). Igneous sills record far-field and near-field stress interaction during volcano construction: Isle of Mull, Scotland . Earth and Planetary Science Letters 478 , 159 – 174 . Google Scholar Crossref
TI - Magma Plumbing Systems: A Geophysical Perspective
JF - Journal of Petrology
DO - 10.1093/petrology/egy064
DA - 2018-06-01
UR - https://www.deepdyve.com/lp/oxford-university-press/magma-plumbing-systems-a-geophysical-perspective-twCqEGGxAE
SP - 1217
VL - 59
IS - 6
DP - DeepDyve
ER -