Bull Volcanol (2017) 79:59 DOI 10.1007/s00445-017-1140-x

REVIEW ARTICLE

Volcano geodesy in the Cascade arc, USA Michael P. Poland 1 & Michael Lisowski 1 & Daniel Dzurisin 1 & Rebecca Kramer 1 & Megan McLay 1 & Ben Pauk 1

Received: 7 March 2017 / Accepted: 18 June 2017 # US Government (outside the USA) 2017

Abstract Experience during historical time throughout the Cascade arc and the lack of deep-seated deformation prior to the two most recent eruptions of Mount St. Helens might lead one to infer that Cascade volcanoes are generally quiescent and, specifically, show no signs of geodetic change until they are about to erupt. Several decades of geodetic data, however, tell a different story. Ground- and space-based deformation studies have identified surface displacements at five of the 13 major Cascade arc volcanoes that lie in the USA (Mount Baker, Mount St. Helens, South Sister, Medicine Lake, and Lassen volcanic center). No deformation has been detected at five volcanoes (Mount Rainier, Mount Hood, Newberry Volcano, Crater Lake, and Mount Shasta), and there are not sufficient data at the remaining three (Glacier Peak, Mount Adams, and Mount Jefferson) for a rigorous assessment. In addition, gravity change has been measured at two of the three locations where surveys have been repeated (Mount St. Helens and Mount Baker show changes, while South Sister does not). Broad deformation patterns associated with heavily forested and ice-clad Cascade volcanoes are generally characterized by low displacement rates, in the range of millimeters to a few centimeters per year, and are overprinted by larger tectonic motions of several centimeters per year. Continuous GPS is therefore the best means of tracking temporal changes in deformation of Cascade volcanoes and also for characterizing tectonic signals so that they may be distinguished from Editorial responsibility: A. Harris * Michael P. Poland [email protected]

1

U.S. Geological Survey, David A. Johnston Cascades Volcano Observatory, 1300 SE Cardinal Ct., Suite 100, Vancouver, WA 98683, USA

volcanic sources. Better spatial resolution of volcano deformation can be obtained through the use of campaign GPS, semipermanent GPS, and interferometric synthetic aperture radar observations, which leverage the accumulation of displacements over time to improve signal to noise. Deformation source mechanisms in the Cascades are diverse and include magma accumulation and withdrawal, post-emplacement cooling of recent volcanic deposits, magmatic-tectonic interactions, and loss of volatiles plus densification of magma. The Cascade Range thus offers an outstanding opportunity for investigating a wide range of volcanic processes. Indeed, there may be areas of geodetic change that have yet to be discovered, and there is good potential for addressing a number of important questions about how arc volcanoes work before, during, and after eruptions by continuing geodetic research in the Cascade Range. Keywords Volcano geodesy . Cascade arc . Cascade Range . Deformation . Gravity change

Introduction Since the establishment of the US Geological Survey’s Cascades Volcano Observatory following the catastrophic eruption of Mount St. Helens in 1980, terrestrial and/or remote sensing geodetic studies have been conducted at most of the volcanic centers in the Cascade arc of the western USA (Fig. 1). This work has evolved from temporally sporadic and spatially sparse terrestrial reconnaissance surveys to include networks of continuously recording Global Positioning System (CGPS) stations, semipermanent GPS stations (SPGPS, which are deployed for weeks to months at a time; see Dzurisin et al. 2017), and space-based interferometric synthetic aperture radar (InSAR) observations with excellent

59

Bull Volcanol (2017) 79:59

Page 2 of 33

N

BRITISH COLUMBIA Mount Baker Glacier Peak

WASHINGTON Mount Rainier Mount St. Helens Mount Adams

Mount Hood Mount Jefferson South Sister Newberry Volcano Crater Lake

OREGON

Mount Shasta

NEVADA

Medicine Lake

Lassen volcanic center

CALIFORNIA

50 km Fig. 1 Map of the northwest coast of the USA and southern British Columbia showing volcanoes in the US portion of the Cascade volcanic arc

spatiotemporal resolution. After more than 35 years of effort, a reasonable question to ask is Bwhat has been learned?^ Despite their overall high productivity during the Quaternary compared to other volcanic arcs around the world, Cascade volcanoes erupt infrequently, averaging only one to two eruptions per century in postglacial times (Hildreth 2007). The only two eruptions in the arc that have been monitored geophysically, at Mount St. Helens in 1980–1986 and 2004– 2008, were associated with spectacular localized deformation but not preceded by deep-seated inflation indicative of magma accumulation several kilometers below the surface (Dzurisin et al. 2008). Given these facts, it is tempting to infer that Cascade volcanoes do not deform until just before they erupt, and that geodetic surveys might therefore be of limited use for monitoring and research within the arc. Indeed, no unambiguous instances of volcano deformation were detected by electronic distance measurement (EDM) surveys throughout the arc during the 1980s (Chadwick et al. 1985; Iwatsubo et al. 1988; Iwatsubo and Swanson 1992a). Geodetic monitoring that focused on the Mount St. Helens lava dome during 1980–1986 demonstrated the critical importance of nearfield deformation monitoring to eruption prediction in the case of repeated and self-similar dome-building episodes (Swanson et al. 1983), but prior to 2000 broad ground motion at quiescent volcanoes had only been detected at Medicine Lake volcano in northern California, which is subsiding (Dzurisin et al. 1991, 2002; Poland et al. 2006; Parker et al. 2014). Starting in the 2000s, however, InSAR and GPS began providing data that resulted in the recognition of previously unknown deformation at several volcanoes in the arc. Of the 13 major Cascade volcanic centers in the USA, deformation has now been detected at five: Mount Baker, Mount St. Helens, South Sister, Medicine Lake volcano, and Lassen volcanic center. The style of deformation varies from inflation caused by magma accumulation (e.g., Wicks et al. 2002; Dzurisin et al. 2006, 2009) to subsidence thought to be a function of tectonic setting and crustal loading (Dzurisin et al. 1991, 2002; Poland et al. 2006). In many cases, deformation may result from a complex interplay of processes that include subduction zone strain, forearc rotation, episodic tremor and slip, and magmatism (Poland et al. 2006; Lisowski et al. 2008). Even the confirmed lack of deformation at some volcanoes is fodder for additional research. For instance, with Medicine Lake volcano and Newberry Volcano sharing similar tectonic settings, morphologies, and compositions, why is Medicine Lake subsiding while Newberry is not? Microgravity measurements also have a role to play in any understanding of Cascade volcano geodesy. Two of the three places where repeated microgravity measurements have been completed thus far—Mount Baker and Mount St. Helens—show evidence of gravity change over time, indicating subsurface mass accumulation (Crider et al. 2008; Battaglia et al. 2015).

Bull Volcanol (2017) 79:59

Compared to other volcanic arcs around the world that have been examined systematically for evidence of surface deformation, the Cascade arc is intermediate in terms of its activity. For example, Lu and Dzurisin (2014) used InSAR to survey the Aleutian volcanic arc and found that, of 52 historically active volcanoes studied, only 12 showed no evidence of deformation of any kind from the early 1990s to 2010. Two of those 12 erupt repeatedly without deforming, and results for eight were inconclusive owing to decorrelation or poor spatial resolution of the InSAR images. Lu and Dzurisin (2014) concluded that deformation mechanisms were varied, including at least one episode of magma intrusion beneath 21 of the 44 volcanoes with adequate results (48%), deflation of crustal magma reservoirs beneath three volcanoes (7%), and shallow-seated subsidence caused by contraction of recent volcanic deposits or by hydrothermal processes at 11 volcanoes (25%). During the ~20-year period of observation, 17 of the 52 volcanoes examined erupted at least once (33%). Volcano deformation in the Central American volcanic arc seems to fall near the opposite end of the spectrum—relatively few cases have been detected there. Ebmeier et al. (2013) made L-band InSAR observations over a 3-year period of 20 of the 26 historically active volcanoes in the arc and found evidence for surface deformation at only three of them (15%). In all cases, the deformation was shallow-seated and likely associated with contraction of recent deposits or edifice loading; none was attributed directly to magmatic processes. The paucity of deformation may reflect rapid rise of basaltic magma from depth, the geometry of magma storage, or high bubble contents that result in very compressible magma. As exemplified by the cases above, most deformation surveys of entire volcanic arcs rely upon InSAR, since groundbased measurements are expensive in terms of equipment, implementation, personnel, and time. In this sense, however, the Cascades represent an outlier. Of the 13 major volcanoes in the arc, 10 have been surveyed repeatedly by ground-based methods. The results of these surveys have been critical in identifying deformation that might otherwise be missed—for example, at Medicine Lake (Dzurisin et al. 1991) and Mount Baker (Hodge and Crider 2010). InSAR has also played a role in recognizing previously unknown volcano deformation in the Cascade Range, specifically at South Sister (Wicks et al. 2002) and Lassen volcanic center (Poland et al. 2004). The combination of ground- and space-based geodetic observations thus is critical for fully characterizing the geodetic activity in a volcanic arc. In a global context, the Cascade arc should by no means be considered geodetically dormant, and it provides a good example of the importance of a strategy that combines groundand space-based observations. Surface motion has occurred or is occurring at several Cascade volcanoes, providing abundant opportunity for exploration of processes associated with magma ascent and accumulation, volcanic unrest, and volcano-

Page 3 of 33 59

tectonic interaction. Here, we review the results of over 40 years of geodetic studies in the Cascade Range of the USA, focusing on the major stratovolcanoes that define the volcanic arc and the large shields of the backarc. Insights from this work provide the rationale for a geodetic monitoring and research blueprint for the region, while also identifying problems and places worthy of additional attention.

Geodetic measurements in the Cascades The first geodetic measurements in the Cascade arc were not intended to quantify surface deformation. Instead, they were parts of land surveys, especially to determine vertical elevations, conducted in the early and mid-twentieth century. Nevertheless, the quality of these measurements ensured that they might be used, in some cases, as baselines against which subsequent surveys could be compared. It is thanks to these measurements that, for example, subsidence at Medicine Lake volcano was discovered (Dzurisin et al. 1991). Deformation surveys specifically targeting Cascade volcanoes began in 1972, with electronic distance measurement (EDM; also called slope-distance measurement or trilateration) surveys at a few sites on Mount St. Helens and Mount Hood, but these were not repeated until the onset of unrest at Mount St. Helens in early 1980 (Iwatsubo and Swanson 1992a). Tilt measurements using borehole, surface, and spirit-level methods were undertaken in 1975–1976 in response to fumarolic unrest at Mount Baker, but they showed no significant changes (Frank et al. 1977; Crider et al. 2011). Microgravity surveys were also conducted at Mount Baker in the late 1970s, and gravity decrease over time suggested mass loss that Malone (1979) ascribed to volatile loss through fumarolic degassing. Geodetic work on Cascade volcanoes expanded considerably following the 1980 eruption of Mount St. Helens. EDM, tiltmeter, crack measurements, leveling, and other methods were used to forecast and track dome-building eruptions at that volcano during 1980–1986 (Dzurisin et al. 1983a; Swanson et al. 1983; Chadwick et al. 1988; Iwatsubo and Swanson 1992b; Dzurisin 1992a; Iwatsubo et al. 1992a, 1992b). In addition, EDM networks (Chadwick et al. 1985; Iwatsubo et al. 1988; Iwatsubo and Swanson 1992a) and tilt and leveling arrays (Dzurisin et al. 1982, 1983b; Dzurisin 1992b; Yamashita 1992) were established at many volcanoes throughout the arc. Although these networks provided important baselines for future work, EDM and tilt arrays have uncertainties of a up to a few centimeters and tens of microradians, respectively (in comparison, modern geodetic methods, like continuous GPS and borehole tilt, have respective uncertainties that are at the level of a mm and a microradian). EDM line lengths generally have errors on the order of 2–4 ppm (depending on whether or not the lines were

59

Page 4 of 33

flown to measure along-path air temperature and pressure, which might bring errors down to 1 ppm), and most line lengths in the Cascades are on the order of 3–5 km (Chadwick et al. 1985; Iwatsubo and Swanson 1992a), suggesting that line length changes less than a few centimeters may not be significant. The resolution of tilt leveling is on the order of 10 μrad for single-setup leveling, and 6.5 μrad for leveling lines shorter than 500 m (for 1-km leveling lines, the resolution increases to 2 μrad) (Yamashita 1992). More details about geodetic techniques used in the Cascades during the 1980s are given in Ewert and Swanson (1992). GPS (used here to generically refer to all Global Navigation Satellite Systems) measurements in the Cascades began in 1990 (e.g., Yamashita and Wieprecht 1995), and repeated campaigns have been conducted at most volcanoes in the arc since that time. In many cases, GPS data have been collected from EDM benchmarks, allowing for evaluation of decadal-scale changes in distances between pairs of benchmarks (e.g., Dzurisin et al. 2006; Lisowski et al. 2008; Hodge and Crider 2010). The first CGPS sites dedicated to monitoring Cascade volcanoes were installed in 1997 at Mount St. Helens (Lisowski et al. 2008) and in the early 2000s at South Sister (Dzurisin et al. 2006). GPS monitoring grew rapidly with the onset of renewed eruptive activity at Mount St. Helens in 2004 (Lisowski et al. 2008), which coincided with the establishment of the Plate Boundary Observatory and expanded use of CGPS throughout the region, as well as installation of borehole tilt and strain meters at Mount St. Helens. Starting in 2009, annual deployments of an SPGPS network were made at the Three Sisters volcanic center (Dzurisin et al. 2017). GPS displacements due to volcanic activity are particularly challenging to resolve given the numerous active deformation processes in the region, which include subduction zone strain, episodic tremor and slip events, and forearc rotation (Poland et al. 2006; Lisowski et al. 2008). To isolate volcanic deformation from regional processes in GPS datasets, we developed models to minimize the motion at far-field GPS sites surrounding volcanic areas and subtracted those displacements from near-field continuous and campaign data, following the method of Savage et al. (2001). We solved for GPS station velocities using time series of daily positions, accounting for trends, offsets, seasonal changes, and temporally correlated noise in the GPS time series. To approximate scatter in GPS time series, we followed the model of Langbein (2008), which includes white, flicker, and random-walk noise. For continuous GPS station displacements, we used Langbein’s (2008) noise model from his Table 8. For campaign GPS displacements, we increased the flicker and random walk components by a factor of 3 based on comparisons of time series from nearby CGPS and campaign GPS stations. The exploitation of InSAR data accelerated beginning in the early 2000s, when the method was used to discover

Bull Volcanol (2017) 79:59

inflation centered near South Sister (Wicks et al. 2002) and track deformation at Mount St. Helens (Poland and Lu 2008). Use of InSAR in the Cascades has been relatively sporadic owing to low rates of deformation and poor coherence, but the technique has provided critical data to help constrain the spatio-temporal patterns of deformation at the Lassen volcanic center (Poland et al. 2004; Parker et al. 2016), Medicine Lake volcano (Poland et al. 2006; Parker et al. 2014), South Sister (Wicks et al. 2002; Dzurisin et al. 2006, 2009; Riddick and Schmidt 2011), and Mount St. Helens (Poland and Lu 2008). Atmospheric artifacts are a significant impediment to InSAR studies of Cascade volcanoes, however, and mitigation requires application of weather models and/or the use of large datasets to average random noise (Parker et al. 2015). Imagery from the C-band Sentinel-1 satellite system will eventually provide a much denser time series of InSAR data over the Cascade arc, so our discussion below mostly focuses on the potential of C-band interferometry. Although the L-band has greater ability to penetrate vegetative cover, we find that yearto-year L-band interferograms offer little overall improvement compared to C-band data, unless used in multi-temporal analyses (e.g., Parker et al. 2014, 2016). As of 2017, the primary tool used to study the geodetic character of the Cascades is GPS, in campaign, semipermanent, and continuous modes. InSAR provides excellent regional coverage, but incoherence due to vegetation, snow, and ice limits applicability of the method. Borehole instrumentation is expensive and difficult to maintain, and so is only used at the most active volcano in the arc—Mount St. Helens. Even CGPS stations suffer from considerable challenges— buildup of snow and ice on antennas installed high on volcano flanks can result in positioning errors that make data acquired during winter months unreliable, and generally harsh conditions can cause interruptions in telemetered data streams. Nevertheless, GPS campaigns or SPGPS surveys conducted every few years provide sufficient resolution to detect horizontal deformation on the order of 1 cm, and CGPS stations, despite winter outages, can track changes in surface displacements over time periods ranging from weeks to years with precision on the order of 1 mm/year.

Geodesy of Cascade volcanoes Below, we discuss the geodetic behavior, to the extent it is known, of that portion of the Cascade Range that lies in the USA, categorizing the volcanoes based on (1) confirmed geodetic change, (2) confirmed lack of geodetic change, and (3) unknown geodetic change due to lack of data. We focus on the 13 major volcanic centers of the arc that are located in the USA, many of which have been the sites of repeated geodetic studies (Table 1), but acknowledge that there are a number of areas of distributed Holocene volcanism, especially south of

1989 1981, 1982, 1984

2002, 2009, 2016 2009 1990, 1996, 1999, 2000, 2001, 2004, 2011 1990, 1996, 1999, 2004, 2012 2004, 2006, 2012

2001–2006, 2008–2013

2005– 2007–

2011– 2009– 2006–

2001–

1997, 1998, 2002, 2008, 2010, 2006– 2014, 2015

2006– 1997– 2005–

b

Includes both short tilt arrays and long transects

Does not include crack/fault measurements and tiltmeter deployments in the crater used to monitor dome growth during 1980–1986

SSL single-setup leveling, B. tilt borehole tilt, B. strain borehole strain

a

1976 1976, 1980, 2010, 2012, 2014, 2016

1975–1978, 1981, 2005, 2006

Microgravity

1954, 1988, 1989, 1990, 1999 1932, 1991, 2004, 2006, 2012

1981

1985, 1986, 2001, 2002, 2003, 2002, 2004, 2005, 2008, 2004 2009, 2016 1932, 1985, 1986, 1994, 2002 2016 1985, 1986 1981

CGPS B. strain Levelingb

InSAR applications are not given, but data for the entire arc are available starting in 1991 and continuously thereafter

Medicine Lake volcano Lassen volcanic center 1981, 1982, 1984, 2004, 2006, 2012

1981, 1982, 1984

1985, 1986 1981–1984, 1988 1981, 1982, 1984

Newberry Volcano Crater Lake Mount Shasta

1972, 1983, 1984

1985, 1986

1983, 1984

Mount Jefferson South Sister

Mount Adams Mount Hood

2008– 1982, 1983, 1988, 1989 1994, 2008, 2009, 2016 2005– 1972, 1980–1991 2000, 2003, 2004, 2005, 2014

2004, 2006, 2007, 2009, 2015

1982, 1983, 1988

1981, 1983

1975

Camp. GPS

1975, 1977, 1981

EDM

Mount Baker Glacier Peak Mount Rainier Mount St. Helensa

B. tilt

SSL

List of ground-based geodetic techniques and the years they were used at major Cascade volcanic centers

Volcano

Table 1

Bull Volcanol (2017) 79:59 Page 5 of 33 59

59

Page 6 of 33

Mount Rainier, such as Indian Heaven, that deserve more study. In his treatise on Quaternary magmatism in the Cascades, Hildreth (2007, p. 78) noted that B…the Quaternary Cascade arc has 100 times more small volcanoes than big ones. Half or more of the postglacial eruptions…have been from monogenetic vents—distributed or peripheral cones, fissure-fed chains, or short-lived shields…^ There has been little ground-based geodetic work in these areas, but, given those statistics, more is clearly warranted. Volcanic centers of the Canadian portion of the Cascade arc also merit additional work, given their evidence of Holocene eruptive activity (Hildreth 2007). Volcanoes with known geodetic change Geodetic change (deformation and/or gravity variation) has been documented at five volcanoes in the Cascade Range: Mount Baker, Mount St. Helens, South Sister, Medicine Lake volcano, and Lassen volcanic center. The rates, styles, and mechanisms of the activity are diverse and provide opportunities for examining a variety of volcanic processes. Mount Baker The northernmost of the US Cascade volcanoes, Mount Baker (3286 m), is an ice-clad stratovolcano located in northern Washington. Eruptive activity in the region has been nearly continuous over the past 1.3 Ma, but the present cone was built during 40–12 ka (Hildreth et al. 2003; Hildreth 2007). Two Holocene magmatic eruptions are known: 9.8 ka (a scoria cone and lava flow on the volcano’s south flank) and 6.5 ka (a subplinan ash eruption from the volcano’s summit (Hildreth 2007)). A phreatic eruption was reported from Sherman Crater, just south of the volcano’s summit, in 1843 (Tucker et al. 2007). In March 1975, an order-of-magnitude increase in thermal emission was noted at Mount Baker, causing substantial melting of snow and ice in the area of Sherman Crater (Eichelberger et al. 1976; Frank et al. 1977). In response, a variety of tilt measurements (including spirit-level, tilt-bar, and borehole stations) were collected in 1975–1976, but results were inconclusive, and no consistent pattern of deformation was noted (Frank et al. 1977; Crider et al. 2011). Microgravity measurements were also conducted and indicated a large gravity decrease (hundreds of microgals, μgal) at Sherman Crater during mid-1975, possibly due to mass loss from volatile release through fumarolic activity. No significant gravity change was detected from late 1975 into 1976, and a minor gravity increase, similar to the levels of uncertainty, was noted during 1976–1978 (Malone 1979). The cause of the increase in thermal emissions in 1975 could not be identified at that time (Frank et al. 1977), although there was a consensus that unrest was driven by a change in hydrothermal circulation

Bull Volcanol (2017) 79:59

patterns beneath Sherman Crater (Crider et al. 2011). Signs of thermal activity waned after 1976, and EDM surveys at 14 benchmarks surrounding the volcano in 1981 and 1983 showed no significant line-length changes (Chadwick et al. 1985). Likewise, tilt-leveling surveys at three stations detected no significant changes from 1977 to 1981 or 1983. After 1983, geodetic work at Baker stalled for over 20 years. In 2005 and 2006, the gravity stations established by Malone (1979) were reoccupied by Crider et al. (2008), who found a massive (thousands of μgal) gravity increase that they attributed to densification of a magma body emplaced in 1975 and/or shallow mineral precipitation due to hydrothermal activity (glacier ice had actually diminished over the time spanned, so it could not have contributed to the mass increase). In addition, following a small-scale pilot study in 2004, GPS was used in 2006 and 2007 to reoccupy the EDM network and compare to slope distances measured in 1981 and 1983. Results indicated shortening of most lines, possibly indicating deep deflation of the edifice over that time period. Hodge and Crider (2010) modeled the deflation as due to pressure loss in a source located about 6 km beneath the volcano’s northeast flank. These new geodetic data, coupled with analysis of gas emissions and chemistry, led to a retrospective interpretation of the 1975 thermal unrest as due to magmatic activity—either intrusion of magma at midcrustal levels or establishment of a connection between a deep magma body and the surface to allow volatile release (Werner et al. 2009; Crider et al. 2011). Campaign GPS reoccupation of the Baker EDM network is not a viable long-term means of monitoring deformation of the volcano. EDM sites are high on the edifice and, without helicopter support (which is expensive and not generally used in wilderness areas), reaching them by foot is time-consuming and hazardous given the extensive mountaineering and glacier travel that is required. As a result, campaign GPS sites were established at more accessible locations, mostly on the lower flanks of the volcano, in 2009. A reoccupation of many of these sites, in addition to a few of the original EDM sites, was completed in 2015. The velocity field calculated from reoccupations of sites surrounding the volcano indicates continued contraction of the edifice at low rates (velocities of 1– 2 mm/year) (Fig. 2). The overall pattern and rate of displacements are consistent with the results of Hodge and Crider (2010) for a source of volume loss beneath the northeast flank of the volcano. Unfortunately, there are no CGPS stations on Mount Baker (the closest PBO sites are over 25 km from the summit), so the temporal character of Baker deflation is unknown. While InSAR provides a potential means of assessing spatial and temporal variability in deformation, coherence in the region is poor owing to snow and ice on the edifice and vegetation on the lower flanks. Reasonable coherence on the middle flanks of the edifice can be obtained in 1year C-band interferograms spanning late summer to late

Bull Volcanol (2017) 79:59 Fig. 2 Horizontal displacement rates from campaign GPS surveys in 2009 and 2015 at Mount Baker, Washington. White triangle gives location of Baker’s summit. Confidence ellipses, here and in other velocity plots, represent uncertainty at the 2σ (95%) level. Although most rates are within uncertainty, the consistent inward pattern at proximal stations suggests deflation. White circle is approximate location of modeled deflation source from Hodge and Crider (2010)

Page 7 of 33 59

48° 50’

48° 40’ 1 mm/yr 5 km -121° 50’

summer (Fig. 3), but the spatially restricted nature of the coherence makes it a considerable challenge to identify any volcano-related deformation, especially displacements occurring at a rate of ~1 mm/year. Advanced multi-temporal InSAR methods (e.g., Hooper 2008) would be critical for any analysis of InSAR data at Baker. Mount St. Helens The youngest and most active Cascade volcano is Mount St. Helens (2950 m). The current volcanic cone was built mostly within the past 2.5 ka, but its summit and north flank were

-121° 40’

destroyed by the May 18, 1980, sector collapse and eruption (Christansen and Peterson 1981; Mullineaux and Crandell 1981). Following the collapse of the edifice, a series of dome-forming eruptions took place within the new crater, ultimately building a lava dome complex that reached a volume of ~90 × 106 m3 over the course of 1980–1986 (Thompson and Schilling 2007, p. 218). The volcano remained quiescent for the following 18 years, with the exception of a series of small steam and gas explosions during 1989–1991 (Mastin 1994) and seismicity interpreted to indicate magma chamber recharge (Moran 1994; Musumeci et al. 2002). Eruptive activity resumed in 2004 with the extrusion of a new lava dome

59

Bull Volcanol (2017) 79:59

Page 8 of 33

Mount Baker

Mount Hood

Mount Jefferson

2015/09/07 - 2016/08/08

2015/08/21 - 2016/09/08

2015/08/21 - 2016/09/08

Glacier Peak

Three Sisters

Crater Lake

2015/09/14 - 2016/06/28

2015/08/11 - 2016/08/05

2015/09/14 - 2016/09/08

Mount Rainier

Mount Shasta

Lassen volcanic center

2015/08/21 - 2016/08/15

2015/09/14 - 2016/07/22

2015/08/21 - 2016/09/08

Mount St. Helens

Newberry Volcano

Medicine Lake volcano

2015/08/11 - 2016/08/05

2015/09/14 - 2016/07/22

2015/08/21 - 2016/08/15

Mount Adams

LOS displacement 2015/08/21 - 2016/08/15

0

5 km

0

2.83 cm

Bull Volcanol (2017) 79:59

RƒFig. 3

Representative Sentinel-1a interferograms spanning summer 2015 to summer 2016 for the 13 major volcanic centers in the Cascade arc of the USA. No deformation is suspected (or resolvable) in any of the images, so most apparent phase changes are probably a result of atmospheric delay anomalies. The interferograms provide a sense of the 1-year summer-to-summer coherence at C-band that can be expected at major Cascade volcanic centers. Interferograms spanning shorter time periods within a summer show better coherence

in the crater. The dome ultimately reached a volume of 93 × 106 m3 by the time the eruption ended in early 2008 (Scott et al. 2008). Few deformation data are available from the years preceding the May 18, 1980, eruption. A single EDM line on the east side of the volcano (Smith Creek Butte–East Dome, 7.6 km line length) was measured in 1972 and again in April 1980, a few weeks after the start of pre-eruption seismic unrest and phreatic explosions. No significant change was detected, and the Occam’s razor principle argues that it is most likely that none actually occurred, although there is a possibility that the magma reservoir geometry and depth were such that the two end points of the EDM line were displaced by roughly the same amount, resulting in no net change in line length (Dzurisin 2000). During the weeks before the May 18, 1980, eruption, geologists also monitored the shoreline of Spirit Lake, treating the lake as a giant carpenter’s level that might tilt due to deep-seated changes beneath the volcano; however, the shore of Spirit Lake was not observed to be tilting, arguing against deep accumulation or withdrawal of magma immediately prior to the eruption (Swanson 1992). Despite the lack of broad, deep-seated, precursory deformation, massive bulging of the volcano’s north flank due to the intrusion of a cryptodome accompanied the unrest, with displacement rates reaching 2.5 m/day during late April–mid-May (Lipman et al. 1981). No other deformation was noted prior to the catastrophic eruption, and gravity changes measured between March 29 and May 9 at sites on the north and east flanks of the volcano were small, suggesting that the source of the magma feeding the cryptodome was at least 7 km below the surface (Jachens et al. 1981). Deformation monitoring was rapidly reestablished after the May 18, 1980, eruption (Swanson et al. 1981). The Smith Creek Butte–East Dome EDM line was measured again in June 1980; April–June shortening by 18 mm was close to or within measurement error (Swanson et al. 1981, p. 161). Repeat EDM measurements to targets elsewhere on the cone revealed inflation focused within the new crater of the eviscerated volcano prior to explosive eruptions in July, August, and October 1980. These observations motivated intensive geodetic work in the crater. Subsequent dome-building eruptions were predicted with remarkable success (Swanson et al. 1983) owing, in large part, to pre-eruptive accelerations in surface displacement rates as measured by EDM (Swanson et al. 1983), ground tilt (Dzurisin et al. 1983a), and the

Page 9 of 33 59

formation and displacement of ground cracks and thrust faults on the crater floor adjacent to the growing lava dome (Chadwick et al. 1988). Following the cessation of dome building in 1986, few geodetic measurements were attempted. A regional trilateration survey centered on the volcano was completed in 1991, duplicating an earlier survey from 1982. Linelength changes over this period all showed extension above uncertainty, suggesting areal dilatation that might be a result of recharge of a deep magma reservoir (Lisowski et al. 2008). This result is consistent with seismicity indicative of magma recharge over the same time period (Moran 1994). A few sporadic GPS measurements were collected in the 1990s, and the EDM network was completely replaced by a campaign GPS network in 2000; however, no significant changes were detected from a comparison of EDM- and GPS-derived line lengths (Lisowski et al. 2008) despite the continuation of seismicity indicative of magma accumulation at depth (Musumeci et al. 2002). The GPS campaign in 2000 included the establishment of several dozen new stations, and the entire network was reoccupied in 2003, but no significant volcanorelated displacements were detected, with one exception. A campaign station located on the 1980–86 lava dome moved toward the dome center by almost 3 cm/year and subsided by 9 cm/year, probably due to post-emplacement cooling and contraction of the dome (Lisowksi et al. 2008). InSAR likewise revealed no volcano-wide deformation prior to 2004, although localized patches of subsidence were identified in the 1980 debris avalanche deposit (Poland and Lu 2008). CGPS monitoring began at Mount St. Helens in 1997 with the installation of station JRO1 at the Johnston Ridge Observatory about 9 km north of the volcano’s crater (Fig. 4) (Lisowski et al. 2008). Johnston Ridge Observatory is a US Forest Service visitor center named to commemorate USGS geologist David A. Johnston, who died near the site as a result of the May 18, 1980, eruption. On September 23, 2004, Mount St. Helens reawakened with a seismic swarm that gradually intensified until a phreatic eruption occurred on October 1, followed by the emergence of a new lava dome on October 11 (Scott et al. 2008). The eruption was not preceded by any measurable ground deformation at JRO1, but the station started moving south-southeastward (toward the crater) and downward on the day the seismic swarm started (Fig. 4) (Lisowski et al. 2008). A singlefrequency CGPS station on the 1980–1986 lava dome, which had been moving downward and toward the dome center prior to losing power in February 2004, was reactivated on September 27, by which time it had moved north, away from the dome center, by over 20 cm (LaHusen et al. 2008; Lisowski et al. 2008). On about the same day, a welt appeared on the surface of Crater Glacier just south of the 1980–1986 dome. This welt grew rapidly in the ensuing days and was the site of the subsequent phreatic eruptions and dome extrusion

Bull Volcanol (2017) 79:59

Page 10 of 33

48° 20’

JRO1

TSEP

48° 10’

5 km -121° 40’

-121° 50’

40

Displacement (mm)

(Schilling et al. 2008; Dzurisin et al. 2008). In retrospect, JRO1 southward motion reflected an ultimate volume loss of 16–24 × 106 m3 from a geodetically identified subvolcanic magma reservoir centered at 7–8 km depth (5–6 km below sea level) (Lisowski et al. 2008). This volume is several times smaller than the volume of the new lava dome, reflecting some combination of magma compressibility effects and possible recharging of the reservoir (Lisowski et al. 2008; Mastin et al. 2009). The northward motion of the GPS station on the dome was due to bulging of the eventual vent area as magma ascended toward the surface, but the magnitude of the displacement was less than expected from the size of the welt, perhaps due to decoupling of the conduit from the rest of the volcano (Dzurisin et al. 2008). The southward motion of JRO1 shortened the distance between that site and the volcano by several centimeters; in 1980, if such a decrease in line length were observed by EDM, which has no absolute reference frame, it might have been interpreted as volcano inflation (since inflation might drive proximal areas of the volcano outward more than distal areas, producing a net contraction). A CGPS network was rapidly established on and around Mount St. Helens in September–October 2004 and expanded in 2005, with stations installed by both the USGS Cascades Volcano Observatory and the Plate Boundary Observatory. Data from these sites indicated deflation throughout the eruption, with the deflation rate decaying exponentially over time. Cumulative displacements at most sites were less than a few centimeters over the course of the eruption (Lisowski et al. 2008). Although complicated by incoherence and low deformation rates, InSAR also indicated deflation of the same reservoir as that modeled from GPS (Poland and Lu 2008). Over the course of the eruption, a variety of additional deformation measurements were made, including borehole tilt, terrestrial and airborne photogrammetry, and GPS at a number of single-frequency stations deployed on and around the growing lava dome and glacier (Dzurisin et al. 2008; LaHusen et al. 2008; Schilling et al. 2008; Anderson et al. 2010; Diefenbach et al. 2012). As the 2004–2008 eruption ended, the CGPS network around the volcano recorded a transition from deflation to subtle, time-decaying inflation, apparently due to repressurization of the same magma reservoir that deflated during the eruption (Dzurisin et al. 2015). Inflationary deformation appeared to decay to background levels by early 2013 (Fig. 4). Campaign gravity measurements from stations located on and around the volcano were collected in 2010, 2012, 2014, and 2016. Changes over time indicate mass accumulation beneath the crater; the mass increase is much more than expected given the low rates of deformation during the same time period (Battaglia et al. 2015). As of 2017, deformation at Mount St. Helens is monitored by nearly two dozen CGPS stations and four borehole strainmeters. Displacements at CGPS stations around Mount St. Helens are dominated by tectonic sources, including

-122° 00’

JRO1

east

20 north 0 -20 -40 1998

Displacement (mm)

59

100 80 60 40 20 0 -20 -40 -60 80 -100 2012

up 2002

2006

2010

2014

TSEP east north

up 2013

2014

2015

2016

2017

Fig. 4 CGPS stations (red dots) around Mount St. Helens (top). Time series are shown for JRO1 (middle), 9 km north of Mount St. Helens crater, and TSEP (bottom), on the 1980–1986 lava dome. Gray shaded area in JRO1 time series indicates period of 2004–2008 eruption, during which time southeast and downward motions of the GPS station reflect depressurization of the magma reservoir beneath the volcano (subtle northeast and upward motions after the eruption are consistent with a small amount of reservoir repressurization). East-northeast and downward motions at TSEP suggest post-emplacement cooling and contraction of the 1980–1986 lava dome. In this and other GPS time series data, scatter indicates uncertainty in the measurement

subduction zone strain, forearc rotation, and episodic tremor and slip events. Recognition of small-scale changes due to the magmatic system is therefore a considerable challenge, but

Bull Volcanol (2017) 79:59

one that can be overcome by careful analysis of the various contributing processes (Lisowski et al. 2008; Dzurisin et al. 2015). The only persistent source of deformation on the volcano is the complex of lava domes within the crater. A CGPS station on the 1980–1986 dome complex indicates several centimeters per year of subsidence and horizontal motion toward the center of the lava dome (station TSEP in Fig. 4), presumably due to cooling and contraction of the dome. The rate of station movement toward the center of the lava dome is similar to that measured before the 2004–2008 eruption, but the rate of subsidence has slowed by about half, probably because the rate of dome cooling has decreased over time. South Sister Of the Three Sisters edifices, only South Sister (3157 m) has experienced Holocene eruptive activity, despite the fact that Middle Sister is the youngest cone of the three (Hildreth 2007). Two closely spaced episodes between 2.2 and 2.0 ka produced rhyolite tephra, pyroclastic flows, lava flows, and lava domes from vents on the volcano’s south, southeast, east, and north flanks, suggesting the existence of a silicic magma reservoir beneath the edifice (Scott 1987). A number of postglacial basaltic and andesitic eruptions also occurred from monogenetic events around the Three Sisters, emphasizing the magmatic vigor of the region (Hildreth 2007). An EDM network was established at South Sister in 1985 and remeasured in 1986; no significant changes were noted (Iwatsubo et al. 1988). Also in 1985, four short (~200–300 m) tilt-leveling lines were established on the flanks of the edifice; these, too, showed no changes when resurveyed in 1986 (Yamashita and Doukas 1987; Dzurisin 1992b). No additional geodetic work was attempted on South Sister through the remainder of the 1980s or 1990s. South Sister became the focus of intensive geodetic work starting in 2001, when InSAR data revealed inflation centered about 6 km west of the summit of the volcano (Wicks et al. 2002). Analysis of a suite of interferograms suggests that the inflation started between 1996 and 1998 at a maximum rate of 3–5 cm/year that decayed over time (Wicks et al. 2002; Dzurisin et al. 2006, 2009; Riddick and Schmidt 2011). Measurement of the EDM network with GPS and reoccupation of the tilt-leveling arrays at South Sister revealed changes consistent with the inflation, especially on the volcano’s west flank (Dzurisin et al. 2006). In addition to resurveying existing sites, an extensive geodetic network, including campaign GPS and leveling stations, was installed on the west side of the volcano centered on the region of uplift. Together with InSAR results, these data suggest magma accumulation at a depth of 5–7 km as the mechanism for the uplift (Wicks et al. 2002; Dzurisin et al. 2006, 2009; Riddick and Schmidt 2011). Campaign microgravity measurements were collected from the deforming region in 2002, but no significant changes were

Page 11 of 33 59

noted by a resurvey of the network in 2009, suggesting the possibility that the deformation was viscoelastic and no mass change occurred after 2002 (Zurek et al. 2012). A subsequent resurvey and expansion of the microgravity network in 2016 also found no changes greater than the uncertainty in the measurements. As of 2017, the only seismicity associated with the inflation was a swarm of small earthquakes in March 2004 (Dzurisin et al. 2006, 2009). Three CGPS stations have been installed near South Sister to monitor the inflation (Fig. 5a). One, PMAR, is on the south side of the deforming area. HUSB and WIFC/WIFR are close to the center of uplift, and the long time series from HUSB, which starts in 2001, is particularly revealing in terms of the temporal evolution of the deformation. Data from HUSB reveal a decay in inflation over time that, when projected backward, suggests the inflation began in September 1997 (Dzurisin et al. 2009). Inflation at South Sister continues as of 2017, and rates are low enough (~5 mm/year maximum uplift rate) that CGPS data from HUSB are the best means of tracking the temporal evolution of the activity (Fig. 5b). The 16-year-long time series from HUSB suggests an Omori logarithmic decay with a time constant of 2.2 years. Applying this relation to PMAR and campaign GPS sites in the vicinity yields the clearest picture yet of inflation west of South Sister (Fig. 5a). In addition to CGPS, 12–14 SPGPS stations have been installed in and around the deforming area for a few months each summer since 2009. These data provide further constraints on the evolution of the uplift over time without the need for committing infrastructure to continuous stations that are difficult to maintain during harsh winter months (Dzurisin et al. 2017). Indeed, the antenna mount for WIFC was destroyed by snow loading, and the replacement, WIFR, has been compromised as well. Medicine Lake Medicine Lake (2412 m) is a shield volcano located in the back arc of the Cascade Range, and it is subject to the eastwest tectonic extension that characterizes the Basin and Range province, as indicated by numerous north-south normal faults on and around the volcano (Donnelly-Nolan et al. 2008). Only one ignimbrite is known to have erupted from Medicine Lake, about 180 ka. That eruption cannot account fully for formation of the 7 × 12 km summit caldera; draining of a magma reservoir during voluminous effusive eruptions probably also played a role (Donnelly-Nolan and Nolan 1986; DonnellyNolan 1988; Donnelly-Nolan et al. 2008). The volcano’s eruptive activity spans 500,000 years, and 17 eruptions have occurred since the end of glaciation, about 13 ka. Although the volcano is dominated by mafic vents, especially on the flanks, and mafic eruptions occurred shortly after deglaciation, late Holocene eruptions have included silicic lava domes and tephra from the periphery of the caldera or, in one instance, the

59

Bull Volcanol (2017) 79:59

Page 12 of 33

Fig. 5 GPS data from South Sister, Oregon. a Horizontal displacements from campaign (black vectors) and continuous (red vectors, with station names given) GPS stations, as well as vertical displacements (indicated by color of GPS station symbol). Displacements are cumulative between 2001 and 2017 based on a logarithmic fit to the time series from CGPS site HUSB and applied to all other campaign and CGPS stations. Summit of South Sister is indicated by white triangle, and white circle marks location of modeled volume increase (Dzurisin et al. 2006, 2009). b East, north, and up time series for CGPS station HUSB

a

vertical displacement (cm) HUSB

-2

44° 10’

4

10

16

22

5 km

PMAR

44° 00’ 5 cm -122° 00’

-121° 50’

-121° 40’

80 b

Displacement (mm)

HUSB up

40 0 HUSB east

-40 HUSB north

-80

-120 2000

2004

caldera floor (Donnelly-Nolan et al. 1990; Donnelly-Nolan et al. 2016). The most recent eruption, 950 years ago, was that of Glass Mountain, which included rhyolite and dacite tephra and lava erupted from a 5-km-long fissure that ultimately formed a series of small lava domes and one large compositionally graded flow; the total erupted volume was 1 km3 (Donnelly-Nolan et al. 2016). Because chemically similar rhyolite eruptions occurred in close succession but on opposite sides of the caldera in the late Holocene, it had been hypothesized that a large silicic

2008

2012

2016

reservoir might be present beneath the volcano (e.g., Eichelberger 1981). Geophysical investigations, however, have mostly suggested that the volcano is underlain by solidified intrusive bodies at shallow levels, and the only possible melt is a low-velocity, high-attenuation structure 3–5 km beneath the caldera (Evans and Zucca 1988; Chiarabba et al. 1995). Donnelly-Nolan (1988) suggested that tectonic extension prevents the formation of a large magma body, and instead the magmatic system of Medicine Lake is made up of a network of dikes, sills, and small magma bodies.

Bull Volcanol (2017) 79:59

Medicine Lake was the first Cascade volcano other than Mount St. Helens at which active crustal deformation was identified. In 1989, Dzurisin et al. (1991) reoccupied a leveling line, first surveyed in 1954, and found that the volcano was subsiding at a maximum rate of about 11 mm/year, centered on the summit caldera. Additional surveys and a reassessment of the 1954–1989 data led to a refined subsidence rate of 8.6 mm/year. A resurvey of part of the leveling line across the caldera in 1999 confirmed that the 1954–1989 subsidence shape and rate remained the same during 1989–1999. In addition to the subsidence, leveling results highlighted two faulting events, only one of which was accompanied by recorded seismic activity (Dzurisin et al. 2002). Although the subsidence could be approximated by a source of volume loss at ~10 km beneath the caldera (Dzurisin et al. 2002), a lack of geophysical evidence for magma storage at that level led Dzurisin et al. (1991, 2002) to favor a subsidence mechanism of extension of thermally weakened crust beneath the massive load of the edifice. A single EDM line was measured between the rim and center of the caldera in 1989, but it was never resurveyed because a campaign GPS network was established in 1990. An incomplete GPS satellite constellation at the time resulted in relatively large uncertainty in the initial measurement, but subsequent surveys and expansion of the network demonstrated unequivocal radially inward displacements across the volcano, perhaps due to some combination of crustal loading, tectonic extension, and cooling and crystallization of magma bodies at depth (Poland et al. 2006). Geologic constraints suggest that the current rate of subsidence at Medicine Lake is relatively recent. Drilling indicates that the 180-ka ignimbrite is depressed beneath the caldera, but the rate of subsidence needed to lower the level of the tuff by the observed amount is a small fraction of the current subsidence rate (Donnelly-Nolan et al. 2008). The subsidence may be associated with the postglacial pulse of volcanism that began 13,000 years ago (Poland et al. 2006). In addition to leveling and GPS, subsidence at Medicine Lake has also been measured using InSAR. A stack of interferograms, which emphasized deformation occurring at a steady rate (as was known to be the case from leveling), supported the radially symmetric nature of the subsidence (Poland et al. 2006). A more rigorous analysis of InSAR data from 2004 to 2011 demonstrates that subsidence was continuing at the same rate measured by the original leveling surveys (Parker et al. 2014). The data could be fit by a source of contraction at ~10 km depth, leading Parker et al. (2014) to conclude that the tectonic extension inferred from other studies was accompanied by subsurface volume loss. Three CGPS stations are in operation at Medicine Lake volcano, installed as part of the Plate Boundary Observatory in 2005-2006. Only one station is located within the caldera, with the other two on the lower flanks. The caldera GPS

Page 13 of 33 59

station is subsiding at a rate of 4–5 mm/year and is located a few kilometers from the source of subsidence modeled from leveling data (Fig. 6). In summary, all geodetic data collected at Medicine Lake—leveling, campaign GPS, InSAR, and CGPS—suggest steady, symmetrical subsidence of the volcano at a maximum rate of 8–9 mm/year since at least 1954. Lassen volcanic center Although the Lassen volcanic center is east of the main axis of Cascade volcanism and is subject to extensional Basin and Range tectonism (like Medicine Lake volcano), as indicated by numerous normal faults in the region, it is generally considered to be part of the arc, as opposed to the back arc (Hildreth 2007). Magmatism in the region has been ongoing since about 1 Ma and has progressed from a silicic caldera sequence, through the growth and degradation of an andesitedacite stratovolcano, to the emplacement of the current field of dacite and rhyodacite lava domes, the latter stage of which includes the Lassen Peak (3187 m) rhyodacite dome. In addition, volcanism peripheral to, and contemporaneous with, the dome field has resulted in the formation of numerous cones and lava flows (Hildreth 2007). Three eruptions are known to have occurred in the Holocene (Muffler and Clynne 2015). The six domes of the Chaos Crags complex erupted 1100 years ago and may have been active for centuries; one of the domes collapsed partially in a series of cold rockfall events about 275 years ago (Clynne et al. 2008). The Cinder Cone eruption northwest of Lassen Peak produced basaltic andesite and andesite scoria and lava flows in about 1666 CE. The most recent eruption occurred during 1914–1917 from Lassen Peak and consisted of phreatic explosions and dacite lava and tephra, in addition to pyroclastic and debris flows (Muffler and Clynne 2015). As was the case with most Cascade volcanic centers, the first geodetic monitoring to be completed at Lassen was EDM, with a network of 14 stations established on and around Lassen Peak in 1981 and remeasured in 1982 and 1984. Minor contraction of the network was suggested by changes measured between 1982 and 1984, but the 1984 survey may be compromised by adverse weather, and no deformation is suspected (Chadwick et al. 1985). Single-setup leveling stations were also installed in 1981 (Dzurisin et al. 1982); reoccupations in 1982 and 1984 did not identify significant changes. Leveling lines several kilometers northwest and south of Lassen Peak were established in 1932–1934 as part of regional land surveys and resurveyed in 1991; again, no significant changes were detected (Dzurisin 1999). Renewed focus on potential deformation of Lassen was prompted by the 2004 discovery of deflation centered 5 km southeast of Lassen Peak and occurring at a maximum rate of ~10 mm/year based on InSAR data spanning 1996–2000 (Poland et al. 2004). A simple model of the deformation

59

Bull Volcanol (2017) 79:59

Page 14 of 33

a 80 P672

b

P673 east

60

P673 north

40

P674 P673

Poland et al., 2006 Parker et al., 2014

41° 30’

Displacement (mm)

41° 40’

20 P673 up

0 P674 up

-20 -40 P674 east

1 mm/yr 5 km -121° 40’

-60 P674 north

-121° 30’

-80 2006

2008

2010

2012

2014

2016

Fig. 6 a Station map and displacement rates from CGPS stations installed at Medicine Lake volcano in 2005–2006. Dashed ellipse shows rough outline of caldera. White circles denote spherical

subsidence sources modeled by Poland et al. (2006) from GPS data and Parker et al. (2014) from InSAR data. b Time series of north, east, and up displacements from CGPS stations P673 and P674

assuming a point source of contraction (Mogi 1958) suggests a depth of 11.6 km and volume loss rate of 7 × 106 m3/year (Fig. 7). A subsequent, more thorough, analysis by Parker et al. (2016) expanded the InSAR dataset to multiple satellites covering 1992–2000 and 2004–2010. They, too, found subsidence, which they modeled as due to a point source at the same location as that of Poland et al. (2004) but at 8.3 km depth with a volume loss rate of 2.86 × 106 m3/year, which was constant over the entire time period to within uncertainty (although there is a suggestion that the 2004–2010 rate of volume loss may have been slightly less than that of 1992– 2000). The cause of the subsidence is probably some combination of tectonic extension and magmatic/hydrothermal processes (Poland et al. 2004; Parker et al. 2016). Additional leveling and GPS surveys were spurred by the subsidence discovery. A subset of the EDM network was reoccupied in 2004, resulting in the recovery of five line lengths, all of which shortened by several to over 10 cm (Poland et al. 2004). Almost the entire EDM network was occupied by GPS in 2006. Comparison to the average EDM line lengths in 1981–1984 indicates contraction of all lines, some by well over 10 cm (Fig. 8). Several new, more accessible campaign GPS stations were established in 2004 and 2006, and the network was revisited in 2012. Displacements for the 2006–2012 period are several millimeters per year at many stations, and nearly all displacement vectors are downward and inward toward a point about 2 km southeast of Lassen Peak,

which is slightly northwest of the source modeled from InSAR (Fig. 9). These data suggest that subsidence at Lassen Peak has been ongoing since at least the 1980s (the subsidence may not have been detected by EDM surveys in 1981, 1982, and 1984 because the amount of line-length change over 3 years was too small to be unambiguously measurable). Before that time, little can be inferred, since the leveling transects occupied by Dzurisin (1999) and spanning the 1930s to 1991 are too far from, and not radially oriented with respect to, the locus of deformation. In addition to GPS campaign work, three short (few-hundred-meter) leveling arrays were established, one in 2004 and two in 2006 (Fig. 8). Resurveys in 2012 indicate vertical displacements of a few millimeters that, at the King’s Creek and North Summit Lake arrays, are consistent with contraction of a source southeast of Lassen Peak. Given the deformation, Lassen was chosen as a site of special emphasis for PBO CGPS deployments, with eight stations installed in the volcanic field in 2007 and 2008. Displacements at these sites are consistent with those from campaign GPS (Fig. 9a) and indicate relatively steady subsidence over time (Fig. 9b). Additional evidence for continued deformation at Lassen comes from a stack of TerraSAR-X interferograms spanning 2011–2015 that shows line-of-sight deflation of at least 6 mm/year; the core area of subsidence, where rates should be at a maximum, is not coherent (Fig. 10). Broad deflation of the Lassen volcanic center thus appears to be a long-term phenomenon.

Bull Volcanol (2017) 79:59

Page 15 of 33 59

Observed

Mount Rainier

Lassen Peak

LOS Displacement (mm) -60 -45

-30 -15

0

15

Modeled

5 km Residual

Fig. 7 Observed (top), modeled (middle), and residual (bottom) line-ofsight displacements from an ERS-2 interferogram spanning August 17, 1996–September 30, 2000. Source location given by white circle. Dashed line is boundary of Lassen Volcanic National Park

Volcanoes with known lack of geodetic change At five Cascade volcanoes—Mount Rainier, Mount Hood, Newberry Volcano, Crater Lake caldera, and Mount Shasta—there is a high degree of certainty that no volcanic deformation has occurred since the 1980s. Knowledge of the deformation states of these volcanoes establishes an important baseline for future work and will be especially valuable in interpreting the sources of any future unrest.

The highest peak in the Cascade arc, Mount Rainier (4392 m), has been a site of eruptions for ~500 ka, although significant explosive events have been rare. The volcano has erupted numerous small-volume tephras and lava flows throughout the Holocene, however, and the edifice collapsed around 5.6 ka, generating the large Osceola mudflow; the collapse scar was subsequently refilled by eruptions of lava (Sisson and Lanphere 2000; Vallance and Donaghue 2000; Hildreth 2007). The most recent magmatic eruption of Mount Rainier occurred about 1000 years ago and produced a thin finegrained tephra deposit; far-traveled lahars with similar ages might be associated with the same eruption (Sisson and Vallance 2009). The only geodetic work on Mount Rainier prior to 1980 involved the installation of a few microgravity sites in the late 1970s, inspired by gravity change associated with thermal unrest at Mount Baker in 1975 (Malone 1979). An EDM network was established on the volcano in 1982 and reoccupied in 1983, 1988, and 1989, but no significant changes were detected (Chadwick et al. 1985; Iwatsubo and Swanson 1992a). In addition, seven single-setup leveling stations were established around the volcano in 1982 (Dzurisin et al. 1983b); resurveys in 1983 and 1988 showed no changes above uncertainty. Sporadic campaign GPS measurements were collected from a few sites in the 1990s, and the EDM network was reoccupied with GPS in 2008 and 2009, at which time new, more accessible, GPS sites were also established on the lower flanks of the volcano. Comparing the 1982/1983 line lengths to those recovered by GPS in 2008/2009 shows no changes above a few centimeters, which is within the uncertainty of the measurements. Several of the campaign GPS sites were reoccupied in 2016, and velocities over the 7–8 years between campaigns show no significant pattern that suggests volcano-wide deformation (Fig. 11). Ice-free ridges and the lower flanks of the volcano show reasonable coherence in 1-year, C-band, summerto-summer interferograms (Fig. 3), raising hopes that multi-temporal InSAR methods can be used to monitor any future deformation. Given the extreme hazard that debris flows from Mount Rainier pose to the Puget Sound area, the USGS Cascades Volcano Observatory installed seven CGPS sites and two borehole tiltmeters on the flanks of the volcano in 2006– 2008. These instruments complement a number of far-field stations that were installed in the mid-2000s as part of the Plate Boundary Observatory. Like the campaign results, continuous GPS data show no significant volcano-related displacements (Fig. 11). Maintaining the stations has been a considerable challenge, owing to snow and ice buildup at high

59

Bull Volcanol (2017) 79:59

Page 16 of 33

40° 35' Road EDM site (occupied with GPS in 2006) EDM site (not occupied with GPS in 2006) -3

1981/84-2006 line length change (cm) Leveling array

-1

-11 -4

-9 40° 30'

-21

-13 NSL

-7 -5 C -10 -9

-12

-15 -12 -14 -3

KC

modeled deflation source

LOS displacement (cm) 0 40° 25' 121° 35'

1.4

Lassen Volcanic National Park boundary

2.8 121° 30'

1 km 121° 25'

Fig. 8 Changes in line lengths (in cm) measured by EDM (average from 1981, 1982, and 1984) and GPS (in 2006). Background is ERS-2 interferogram spanning August 17, 1996–September 30, 2000. Leveling arrays (blue circles) are as follows: C Tilt array c (Dzurisin et al. 1982); KC

King’s Creek; NSL North Summit Lake. Approximate location of deflation source modeled by Poland et al. (2004) and Parker et al. (2016) is shown by white circle. White triangle gives location of summit

elevations, and data collected at several of the CGPS sites are often not reliable during winter months.

known from the Holocene: 1500 years ago and in the late eighteenth century. Both periods involved lava dome growth and small pyroclastic flows from nearsummit vents (Scott et al. 1997; Pierson et al. 2010). In addition, a peripheral vent 12 km north of the summit erupted a basaltic-andesite lava flow 7700 years ago (Hildreth 2007).

Mount Hood Mount Hood (3425 m) has been active for about 500 ka (Hildreth 2007), although only two eruptive periods are

Bull Volcanol (2017) 79:59 48° 35’

Page 17 of 33 59

a

5 mm/yr vertical velocity (mm/yr) -10 -8 -6 -4 -2 0

48° 30’

P664 InSAR-modeled deflation source 48° 25’

5 km -122° 30’

-122° 35’

-122° 25’

60

Displacement (mm)

b

P664 up

40

P664 east

20 0 -20

P664 north

-40 -60 2006

2008

2010

2012

2014

2016

Fig. 9 a Horizontal displacement rates from campaign GPS surveys in 2004, 2006, and 2012 (black vectors) and CGPS sites installed in 2007 and 2008 (red vectors), as well as vertical displacement rates (indicated by color of GPS station symbol), at Lassen volcanic center, California. White triangle gives location of Lassen Peak’s summit. White circle is

location of deflation source modeled from InSAR (Poland et al. 2004; Parker et al. 2016). b Time series of displacements from CGPS station P664. Gaps in the time series are due to snow and ice accumulation on the GPS antenna, which cause the data to be unreliable and so are omitted here

An EDM network was established on Mount Hood in 1980, expanded in 1983, and remeasured in 1984; no changes above the uncertainties were detected (Chadwick et al. 1985).

Five tilt-leveling stations were established in 1983 and remeasured in 1984, also with null results. Campaign GPS measurements were first conducted in 1997 at a benchmark

59

Bull Volcanol (2017) 79:59

Page 18 of 33

Fig. 10 Stack of 201 descendingmode TerraSAR-X interferograms derived from 34 individual SAR images acquired during summer months between 2011 and 2015. All interferograms span 1 to 4 years. Although coherence is poor (probably due to the X-band wavelength combined with heavy vegetative and snow and ice cover), line-of-sight subsidence of at least several millimeters per year located southeast of Lassen Peak (white triangle) is clear

LOS displacement

N 0

near Timberline Lodge on the volcano’s south flank, and the site was reoccupied several times through the late 2000s. A larger GPS campaign of some of the existing EDM sites and a number of new flank stations was completed in 2010. Only four of the original EDM line lengths could be recovered, and none of these showed changes greater than 3 cm, which is approximately the uncertainty in the measurements. The GPS network was resurveyed in 2015; comparison to 2010 results shows no consistent pattern of deformation that can be associated with activity at the volcano (Fig. 12). In addition to campaign GPS, starting in 2006, two CGPS stations were installed on the flanks of the volcano and a third was placed off the south flank. Displacements at these stations also show no changes indicative of volcano deformation (Fig. 12). Mount Hood appears to have been geodetically quiescent over the past few decades. C-band coherence around Mount Hood is poor owing to extensive vegetation up to the flanks of the volcano (Fig. 3), but this could probably be improved by the application of multi-temporal methods. Newberry Volcano Lying 60 km east of the main axis of Cascade volcanism, Newberry Volcano (2434 m) is the largest edifice by volume in the Cascade Range (MacLeod and Sherrod 1988; DonnellyNolan et al. 2011). Growth of Newberry began about 400 ka and has resulted in the formation of a broad shield capped by a 6 × 8-km summit caldera that formed during an explosive eruption at 75 ka (MacLeod and Sherrod 1988; Hildreth 2007; Donnelly-Nolan et al. 2011; Mandler et al. 2014). Newberry is at the intersection of the Brothers, Sisters, and Walker Lane fault zones (MacLeod and Sherrod 1988), and

5 km

0

8 mm/yr

the volcano defines the young end of the High Lava Plains trend of age-progressive rhyolitic volcanism (Walker 1974). Eruptions at Newberry are bimodal, with rhyolitic vents in the summit area and mafic vents on the flanks, similar to Medicine Lake volcano in this regard. Both compositions have been produced during the Holocene, when at least six eruptions occurred, the most recent of which (~1300 ka) produced the rhyolitic Big Obsidian Flow within the caldera (MacLeod and Sherrod 1988; Hildreth 2007). This eruption was accompanied by a Plinian column that reached a height of about 20 km, testifying to the explosive potential of the volcano (Gardner et al. 1998). The long history of rhyolitic eruptions, plus the high geothermal gradient within the caldera, suggests the existence of a long-lived silicic magma reservoir beneath the summit caldera (MacLeod and Sherrod 1988)—a hypothesis confirmed by seismic velocity studies of the subsurface. In the 1980s, active-source seismic tomography experiments identified a low-velocity anomaly about 3 km beneath the caldera, suggesting the presence of a small magma reservoir (Achauer et al. 1988), but the anomaly’s average attenuation led Zucca and Evans (1992) to propose that it was solidified. Subsequent active-source and teleseismic tomography studies from a 2008 experiment, however, favor the presence of a zone of partial melt 3–5 km beneath the caldera (Beachley et al. 2012; Heath et al. 2015). Geologic evidence indicates Holocene deformation of the caldera floor. Wave-cut terraces around one of the two lakes that occupy Newberry Caldera were differentially uplifted sometime within the past ~5 ka, with maximum uplift toward the caldera center (Jensen and Chitwood 2000). Leveling surveys were completed on an east-west transect across much of

Bull Volcanol (2017) 79:59

Page 19 of 33 59

47° 00’

46° 50’

1 mm/yr 5 km -121° 50’

-121° 40’

Fig. 11 Horizontal displacement rates from campaign GPS surveys in 2008 and 2016 (black vectors) and CGPS stations installed in 2006–2008 (red vectors) at Mount Rainier, Washington. White triangle gives location

of summit. Motion at some CGPS stations can be attributed to local site effects—for example, freeze-thaw cycles, snow loading, and unstable antenna mounts

the volcano in 1931 and 1994, and comparison of the two surveys suggests volcano-wide uplift reaching a maximum of ~10 cm (~1.6 mm/year) in the caldera. However, at least some of this difference may be due to uncorrected rod-scale and refraction error in the 1931 survey, so the actual amount of deformation, if any, is unclear. Shorter leveling transects

across the caldera were completed in 1985 and 1986 and, when compared to the same segment of the 1994 survey, indicate no deformation across the caldera within uncertainty (Dzurisin 1999). A 12-station EDM network was established on the volcano in 1985 and remeasured in 1986; no changes larger than uncertainty were noted (Iwatsubo et al. 1988).

59

Bull Volcanol (2017) 79:59

Page 20 of 33

45° 30’

45° 20’

1 mm/yr 5 km -121° 50’

-121° 40’

-121° 30’

Fig. 12 Horizontal displacement rates from campaign GPS surveys in 2010 and 2015 (black vectors) and CGPS stations installed in 2006 and 2008 (red vectors) at Mount Hood, Oregon. White triangle gives location of summit

In 2002, the full leveling line was resurveyed, and a campaign GPS network that included the EDM stations was established. Comparison of the 1994 and 2002 leveling data indicates no change outside of uncertainty (Fig. 13). Likewise, comparison of 1985–1986 EDM line lengths to those recovered by GPS in 2002 shows no changes above a few centimeters—within the measurement uncertainty. Subsequent GPS campaigns were conducted in 2009 and 2016; velocities determined from those results, in addition to the 2002 survey, show only small displacements on the volcano, and no recognizable pattern that could be associated with magmatic activity (Fig. 14). An eight-station CGPS network was established

on the volcano in 2011, and these stations also indicate no anomalous surface displacements (Fig. 14). In 2016, a microgravity survey was conducted across the volcano’s west flank and in the caldera, following the leveling line. Future occupations of this network will reveal any changes in subsurface mass that may be associated with magmatic activity. Newberry’s location behind the main axis of Cascade volcanism, where vegetative cover is sparse, makes it more conducive to InSAR studies than most of the edifices in the range. Cband interferograms spanning 1 year have reasonable coherence on the volcano, especially on recent lava flows (Fig. 3), but no deformation has been noted. In addition to standard

3

1931-1994

1985-1986

1994-2002

2 1 0 -1 2000

-2 caldera

-3

0

5

10

15

20

25

30

1000 35

Elevation (m)

Fig. 13 Leveling results from an east-west transect across Newberry Volcano spanning 1931–1994 (blue diamonds), 1985–1986 (red squares), and 1994–2002 (yellow triangles). Topography is given by white diamonds. Uplift is suggested by 1931–1994 results, but is difficult to confirm. No vertical deformation occurred during 1985–1986 or 1994–2002. Location of leveling line is shown in Fig. 14

Page 21 of 33 59

Vertical displacement rate (mm/yr)

Bull Volcanol (2017) 79:59

Distance along leveling traverse (km) satellite interferometric techniques, work at Newberry has included the use of corner reflectors and ground-based InSAR to track changes that might be associated with geothermal development (Vincent et al. 2013). Crater Lake The edifice of Mount Mazama, later to become the Crater Lake caldera, began growing about 400 ka, reaching proportions similar to those of other large Cascade volcanoes (Bacon 1983; Bacon and Lanphere 2006). Starting about 27 ka, silicic magma began to accumulate beneath the volcano, building a large reservoir over the ensuing 20,000 years. In the hundreds of years before the volcano’s 7.7-ka collapse, numerous rhyodacite eruptions occurred from this leaking reservoir. The collapse itself took place upon the expulsion of about 50 km3 of material, which blanketed the region with tephra (Mazama Ash). Since the climactic eruption, a number of andesitic eruptions have taken place within the caldera (such as that which formed Wizard Island), mostly in the hundreds of years following collapse and as Crater Lake began to fill the newly formed caldron. The most recent eruption emplaced a rhyodacite dome on the northeast flank of Wizard Island about 4.8 ka (Bacon and Lanphere 2006). In 1981, a five-station EDM network was established, with lines spanning Crater Lake. The network was reoccupied in 1982, 1983, and 1984, but no changes greater than the uncertainty in the measurements were detected (Chadwick et al. 1985). Three of the EDM sites were recovered with GPS in 2009, but only one EDM line could be repeated. The change on this line was less than 2 cm, within the uncertainty in the EDM measurement. Short leveling lines were also established in 1985 and remeasured in 1986, with no significant changes noted (Yamashita and Doukas 1987). To provide a continuous

record of surface displacement, four CGPS stations were installed around the rim of Crater Lake and on Wizard Island in 2009–2010. No anomalous displacements on these stations have been observed (Fig. 15). C-band interferometric coherence is fair in the vicinity of the volcano, especially to the east (Fig. 3), raising hopes than any anomalous future deformation will be detectable by InSAR. Mount Shasta Mount Shasta (4316 m) is the most voluminous Cascade stratovolcano (Hildreth 2007). Cone building has been ongoing since 600 ka. An ancestral version of the volcano was destroyed by a massive sector collapse between 300 and 400 ka, resulting in the largest known terrestrial debris avalanche deposit on Earth (Crandell et al. 1984; Crandell 1989; Hildreth 2007). The current edifice is composed of four overlapping cones, two of which (Shastina and Hotlum) amount to a cumulative volume of about 60 km3 and were constructed entirely within the Holocene. In addition to frequent eruptions from the stratocone, a number of peripheral vents have been active in the past 10,000 years, including the prominent Black Butte dacite complex 13 km southwest of Shasta’s summit (Miller 1980; Hildreth 2007). The most recent eruption occurred 200–300 years ago, possibly in 1786 CE according to an unconfirmed report by explorer La Perouse (Miller 1980). The first geodetic monitoring of Shasta was by EDM surveys in 1981, 1982, and 1984. Comparison of the 1982 and 1984 surveys suggest slight contraction of the edifice, but this might have been due to error induced by windy conditions during the 1984 survey, and so evidence of deformation is equivocal (Chadwick et al. 1985). Five single-setup leveling sites were also established in 1981 (Dzurisin et al. 1982); reoccupations later in the 1980s revealed no significant

59

Bull Volcanol (2017) 79:59

Page 22 of 33

-121° 30’

44° 00’

-121° 00’

-120° 30’

43° 30’

1 mm/yr 10 km

1 mm/yr Fig. 14 Horizontal displacement rates from campaign GPS surveys in 2002, 2009, and 2016 (black vectors) and CGPS stations installed in 2011 (red vectors) at Newberry Volcano, Oregon. Dashed box outlines area of

summit inset at lower right. Black line in inset is leveling transect, and dotted line gives approximate outline of caldera

changes. A regional network of campaign GPS stations, mostly intended to focus on Medicine Lake volcano (55 km to the east-northeast) but also surrounding the outermost flanks of Shasta, was established in 1990 (Yamashita and Wieprecht 1995) and subsequently reoccupied several times, but no displacements related to the volcano were identified (Poland et al. 2006). A reoccupation of some EDM stations with GPS occurred in 2000 and was repeated in 2004 and 2011, but

only a few sites were surveyed owing to the extreme challenge and hazard in reaching high-elevation sites on foot. The three EDM lines that were recovered show no changes greater than 2 cm, which is less than the uncertainty of the original line-length measurements. GPSmeasured displacements during 2000–2011 on and around the volcanic edifice are on the order of ~1 mm and show no pattern that can be related to the volcano (Fig. 16). As part of the Plate Boundary Observatory, six CGPS stations

Bull Volcanol (2017) 79:59

Page 23 of 33 59

Fig. 15 Horizontal displacement rates of CGPS stations at Crater Lake, Oregon. Although the pattern suggests potential inflation of the edifice, the velocities are very low, and significant snow and ice accumulation on the antennas during winter months causes much of the data to be unreliable

43° 00’

42° 50’

5 km

1 mm/yr -121° 50’

were established on the flanks of Mount Shasta in 2006 and 2007, but none of the sites display any evidence of volcano-related deformation (Fig. 16). C-band interferograms have good coherence on the edifice, but coherence

Fig. 16 Horizontal displacement rates from campaign GPS surveys in 1996, 1999, 2000, 2004, and 2011 (black vectors) and CGPS stations installed in 2006 and 2007 (red vectors) at Mount Shasta, California. White triangle gives location of summit

-122° 00’

is lost rapidly on the lower flanks due to extensive vegetative cover (Fig. 3). Multi-temporal InSAR techniques would likely increase the area over which deformation could be assessed.

41° 40’

1 mm/yr 10 km

41° 30’

41° 20’

41° 10’

-122° 30’

-122° 20’

-122° 10’

-122° 00’

-121° 50’

59

Bull Volcanol (2017) 79:59

Page 24 of 33

Volcanoes where geodetic change can be neither confirmed nor denied There are three major volcanic centers in the Cascade Range where no ground-based geodetic studies have been attempted: Glacier Peak, Mount Adams, and Mount Jefferson. All three are, at least to some extent, located in remote wilderness areas, making field work a logistical challenge. InSAR has not detected deformation at any of the three centers based on a limited analysis of summer-to-summer interferograms constructed from data collected starting in the 1990s, but the low displacement rates that characterize most deforming Cascade volcanoes may not be detectable given atmospheric delay anomalies and poor coherence. A more rigorous analysis using multi-temporal methods and modeling of atmospheric delays would, however, yield a more definitive assessment (e.g., Chaussard and Amelung 2012; Ebmeier et al. 2013). Glacier Peak The geologic history of Glacier Peak (3213 m) is poorly known owing to extensive glaciation. Recent volcanism, however, has included several explosive eruptions ranging from early postglacial (~13 ka) to the seventeenth or eighteenth century (Beget 1982; Gardner et al. 1998; Hildreth 2007). These eruptions seem to recur on millennial timescales, and many are associated with significant pyroclastic and debris flows (Hildreth 2007). Although remote, hazards from Glacier Peak explosive eruptions are nonetheless significant, with ash fall deposits suggesting that plumes can reach heights that would impact aviation (Gardner et al. 1998). Because of its remote location and designation as a wilderness, no ground-based geodetic surveys have been completed at Glacier Peak. Summer-to-summer C-band interferograms spanning 1 year show some coherence on the flanks of the volcano, but the summit is ice-covered, and surroundings are heavily vegetated (Fig. 3). Identifying any magmatic signal at Glacier Peak, especially one occurring at low rates, using InSAR will be difficult and certainly require a large dataset analyzed with multi-temporal methods. Mount Adams Mount Adams (3742 m) is in the middle of an extensive distributed volcanic field and has been active since about 520 ka (Hildreth and Lanphere 1994; Hildreth 2007). In the past ~15 ka, since deglaciation, 10 eruptions of lava accompanied by minor tephra emissions have occurred. Although several of the lava flows extend more than 10 km from their vents, debris avalanches may constitute a more serious hazard, given their common occurrence during the Holocene and the extensive hydrothermal alteration of the cone. The most recent eruption

at Mount Adams occurred about 1000 years ago (Hildreth and Fierstein 1997). Like Glacier Peak, there has been no ground-based geodetic work on Mount Adams (much of the volcano has been designated a wilderness area). InSAR data, however, have good coherence on the flanks of the volcano during summer months owing to light vegetative cover behind the main axis of the Cascades (Fig. 3). Stacks of interferograms from different satellites and look angles spanning the 1990s and early 2000s also show good coherence but no obvious signs of deformation (Poland and Lu 2008).

Mount Jefferson No Holocene eruptive activity has occurred at Mount Jefferson (3199 m), although post-glacial monogenetic eruptions of basaltic andesite have taken place up to ~12 km south of the summit, suggesting that volcanism in the region is not extinct (Scott 1977; Hildreth 2007). The volcano’s remote location and lack of recent eruptive activity have made it a low priority for ground-based geodetic surveys. InSAR data, however, cover the region, and C-band interferograms are coherent on the flanks of the edifice, especially on the lessvegetated east side of the volcano (Fig. 3). No indications of deformation have been recognized in any InSAR data from Jefferson, from the 1990s into the 2000s.

Discussion Of the 13 major volcanic centers in the US portion of the Cascade arc, deformation has been identified at five (Baker, St. Helens, South Sister, Medicine Lake, and Lassen), while at three volcanoes (Glacier Peak, Adams, and Jefferson) no ground-based investigations have been attempted. In other words, half of the volcanoes in the Cascade Range studied by ground-based methods are deforming. Despite long periods of quiescence and relatively infrequent eruptive activity, the Cascades display a rich diversity of volcano deformation processes, although the low signal-to-noise ratio requires careful analysis of geodetic data. The situation is similar to that of Cascade volcano seismicity. Some volcanoes were thought to be seismically quiescent, but that may be because seismometer deployments were sparse in certain segments of the arc. For example, local earthquakes began to be detected at Newberry Volcano only after the installation of a seismic network there in 2011 (Thelen 2016). Indeed, a thorough review of seismic catalogs shows that deep long-period earthquakes, which may indicate the presence of magma, have occurred at six of the 10 Cascade volcanoes in Washington and Oregon (Nichols et al. 2011).

Bull Volcanol (2017) 79:59

Source processes Volcano deformation in the Cascades appears to be controlled by a variety of source processes. Subsidence is occurring or has occurred at four centers: Mount Baker, Mount St. Helens, Medicine Lake volcano, and the Lassen volcanic center. At Mount Baker, the ongoing subsidence may be associated with the 1975 unrest that is speculated to have been caused by a magmatic intrusion or the opening of a pathway between the surface and a deep source of magmatic volatiles (Crider et al. 2011). That the subsidence has been accompanied by a massive gravity increase seems contradictory, but the two measurements are compatible with loss of volatiles and densification of the magmatic system over time. There are at least three distinct forms of subsidence at Mount St. Helens. Subsidence of patches of the 1980 debris avalanche deposit is likely caused by post-emplacement processes, such as dewatering or cooling (Poland and Lu 2008), while contraction of the lava dome complex is almost certainly related to cooling and compaction of hot rock (Lisowski et al. 2008). Co-eruptive subsidence of the edifice during 2004–2008, however, reflects removal of magma from the subsurface reservoir (Lisowski et al. 2008; Poland and Lu 2008). The modeled volume loss at depth was much less than the volume of extruded lava, leading to questions of whether the reservoir was being recharged during the eruption, and if recharge might have triggered the eruption (Dzurisin et al. 2008). On the other hand, analysis of deformation and extrusion rate data through physics-based models suggests that no recharge is necessary to explain the discrepancy between modeled volume loss and extruded volume; an increase in the volume of the fluid magma reservoir core and compressibility of the reservoir and magma can account for the difference (Mastin et al. 2009; Anderson and Segall 2011; Segall 2013). Deflation patterns measured at Medicine Lake and Lassen share a number of similarities, including maximum subsidence rate (~1 cm/year), radial symmetry, and modeled source depth (~10 km), although the ratios of vertical to horizontal displacements may differ, suggesting a horizontally elongated source at Medicine Lake and a more equant source at Lassen (Parker et al. 2016). The volcanoes occupy similar tectonic settings along the western edge of the Basin and Range extensional province. Crustal thinning due to tectonic extension is therefore favored as a driving mechanism in both places, possibly augmented by surface loading, especially at Medicine Lake. Could magmatic activity also be a cause? Lassen erupted in the twentieth century and hosts a vigorous hydrothermal system, and Medicine Lake has experienced numerous post-glacial eruptions. The subsidence rates at both locations are also probably recent developments in a geologic context, owing to the longevity of volcanism in both locations and the fact that such high subsidence rates would produce

Page 25 of 33 59

prominent geomorphic bowls in a few thousand years (e.g., Poland et al. 2006). Additional data and modeling are therefore needed to assess the role that magmatic activity may play in the deformation. Inflation at Mount St. Helens and South Sister is probably related to magma accumulation. At South Sister, the inflation apparently began abruptly in the late 1990s and has waned with time, due to a decrease in magma accumulation rate or a viscoelastic response to sudden emplacement of magma or magmatic volatiles (Dzurisin et al. 2009; Zurek et al. 2012). Two episodes of post-eruptive inflation at Mount St. Helens, after the 1980–1986 and 2004–2008 eruptions, likely also indicate some recharge to the 7–8-km-depth magma reservoir that is known from deformation and seismic data (Lisowski et al. 2008; Dzurisin et al. 2015), although post-eruptive viscoelastic relaxation or expansion of compressible magma is also possible. Questions A tangible benefit to whole-arc geophysical or geologic surveys is that observed behaviors allow researchers to ask penetrating questions based on a comparative analysis of volcanic centers. Here, we explore a few unanswered but tractable questions related to deformation of Cascade volcanoes. Why is Newberry Volcano not subsiding? Subsidence is occurring at both Lassen and Medicine Lake, presumably due to some combination of tectonic and magmatic processes. Newberry, however, is not currently deforming (although evidence of Holocene uplift is preserved in the geologic record). This is somewhat surprising, given the similarities between Newberry and especially Medicine Lake. Both volcanoes are subject to Basin and Range extension (although Newberry is at the northern extent of the Basin and Range), have shield morphologies, erupted rhyolite lava domes and tephras 950–1300 years ago, are underlain by magma reservoirs at ~3 km depth, and are the two largest volcanoes by volume in the range. Any volcano-tectonic or gravitational loading process acting at Medicine Lake would seem likely to also affect Newberry. That Newberry is stable suggests that Medicine Lake subsidence may be driven by a transient process, perhaps related to recent magmatism, or that the properties of the crust beneath the two volcanoes are very different. A similar situation exists in Iceland, where the Askja central volcano is subsiding at a high rate while other nearby volcanoes with similar characteristics and tectonic settings are not (or are subsiding at much lower rates). de Zeeuw-van Dalfsen et al. (2013) suggested that properties of Askja’s magmatic system or the rheological properties of the surrounding crust may provide an explanation for the different behaviors.

59

Page 26 of 33

What causes volcanic subsidence? Related to the question of subsidence at Medicine Lake and not at Newberry, why do volcanoes subside in the first place? In many cases, subsidence is co-eruptive or co-intrusive, caused by volume loss at depth due to magma ascent and possibly extrusion at the surface, as was the case at Mount St. Helens during 2004–2008. But what about volcanoes that have not erupted recently or shown other signs of unrest (e.g., seismicity or other deformation)? Several mechanisms are possible, including (1) magma withdrawal to deeper levels; (2) cooling and crystallization of a magma body; (3) gravitational loading; (4) crustal thinning; and (5) reduction of porefluid pressure due to cooling of a magmatic system (Dzurisin et al. 2002; Lu et al. 2002; Poland et al. 2006; Kwoun et al. 2006). These mechanisms cannot be distinguished using deformation alone, but additional datasets can offer much needed constraints. For example, even though subsidence at Askja volcano suggests volume loss at depth, mass loss determined from repeat microgravity surveys cannot explain the entirety of the subsidence, arguing that the deformation is caused by a combination of magma withdrawal (facilitated by crustal extension) and contraction of a shallow magma body due to cooling and crystallization (de Zeeuw-van Dalfsen et al. 2005). The same approach of combining deformation and gravity measurements was applied to Mount Baker, where deflation is accompanied by, paradoxically, a mass increase, possibly indicating densification of the edifice due to mineral precipitation (Crider et al. 2008). Do Cascade volcanoes inflate before they erupt? The question of deep-seated precursory inflation is perhaps the most important from a hazards point of view, but unfortunately few data exist to support any conclusion—inflation has been measured only at South Sister and Mount St. Helens, and only the latter has erupted since the start of geodetic monitoring (no eruptive activity has occurred in the vicinity of South Sister for ~2000 years). What is more, broad inflation at Mount St. Helens has been observed only for several years after the 1980–1986 and 2004–2008 eruptions, not before, and is thought to indicate magma reservoir recharge following withdrawal during those eruptions. At the very least, deep-seated inflation at both volcanoes indicates the presence of subsurface magma that may be eruptible. The occurrence of such inflation should be enough to warrant the intensification of monitoring efforts and the development of communication and mitigation plans in advance of possible future eruptive activity. Although no deep-seated deformation was detected at Mount St. Helens in the years prior to either the 1980–1986 or 2004–2008 eruptions, the apparent absence of broad precursory deformation is subject to qualification in both cases. Prior to the onset of seismic unrest and phreatic explosions in

Bull Volcanol (2017) 79:59

1980, the geodetic network was sparse, and only one preexisting EDM line was remeasured before the May 18, 1980, eruption. No significant length change had occurred in the line, but Dzurisin (2000) noted that its end points could have been displaced by roughly the same amount due to inflation of the mid-crustal magma reservoir that fed the May 18 eruption. Various types of geodetic measurements were made during the buildup to the paroxysmal eruption, but by that time magma had already risen into the edifice to form a cryptodome, as evidenced by the north flank bulge, and deformation was confined to the edifice itself. When and from what depth did the cryptodome magma rise into the cone? The first anomalous seismicity was noted on March 20, 1980— 2 months before the climactic eruption (Endo et al. 1981). Very soon thereafter, the bulge became noticeable, indicating that magma had already risen into the edifice. Did the cryptodome magma rise from the 7–8-km deep reservoir that fed the eruption (Scandone and Malone 1985) to within a few hundred meters of the surface in a matter of days, or was there subtler, broader-scale deformation indicative of a deeper source in the years or months prior to the onset of noticeable unrest? Unfortunately, we do not know. The question was also not answered definitively prior to the beginning of the next eruption in 2004, partly because only one CGPS station (JRO1) was operating at the time and partly because the magma involved in that eruption might have been remnant from eruptions in 1980–1986. We only know that JRO1 detected no volcano-centric deformation from the time it was installed in 1997 until September 23, 2004, when precursory seismicity began, and that it recorded deflation of a 7–8-km-deep source simultaneously with the seismic onset. In the years following the 1980 eruption, the broad trilateration network around the volcano indicated areal dilatation centered on Mount St. Helens (Lisowski et al. 2008). Inflation was also recorded by the augmented CGPS network for several years after the end of the 2004–2008 eruption (Dzurisin et al. 2015). Perhaps, at least in the case of Mount St. Helens, inflation that is Bprecursory^ to the next eruption happens early in the eruption cycle, i.e., during and immediately after the previous eruption. As the most active volcano in the Cascade Range, Mount St. Helens will doubtless erupt again, perhaps within years to a few decades. When it does, the geodetic network that is in place, provided it is maintained and upgraded as new technology becomes available, will provide critical information to test models of magma accumulation and ascent beneath the volcano. A blueprint for geodetic research and monitoring in the cascades Several decades of geodetic work in the Cascades have provided some important lessons regarding the nature of volcano deformation in the arc, as well as the best ways to detect and

Bull Volcanol (2017) 79:59

track volcano-related ground displacements. Where volcanowide deformation has been detected, displacement rates are low—a few centimeters per year or less, mostly a few millimeters per year (only localized, immediate pre-eruption bulging of the north flank of Mount St. Helens in 1980 and of the crater floor in 2004 exceed that threshold). This principle holds from the most northern to the most southern Cascade volcano in the USA and may be a rule of thumb for the entire arc, at least during non-eruptive periods. To detect such low rates of motion, strategies for tracking deformation of Cascade volcanoes should either employ techniques that specialize in detecting small changes or should leverage long time periods that allow sufficient accumulation of displacements to emphasize signal over noise. Campaign GPS occupations every few years are one example of the latter strategy. Even where volcanic displacements are on the order of a few millimeters per year, as at Mount Baker, careful campaign GPS surveys spanning 6 years are capable of resolving patterns of volcanic displacement. The signal-to-noise ratio of campaign datasets can be improved with SPGPS deployments, where GPS stations are left in place for weeks to months at a time, as at South Sister (Dzurisin et al. 2017). To detect small-scale deformation, and especially variations over time (e.g., changes associated with eruptive activity or variable rates of non-eruptive inflation/deflation), CGPS data from a network of stations are necessary. CGPS is also critical for characterizing non-volcanic deformation (like subduction zone strain, episodic tremor and slip events, and forearc rotation), which overprints magmatic displacements throughout the arc. Regional networks of CGPS stations, like the Plate Boundary Observatory, or broad, regularly occupied campaign or SPGPS networks are critical for modeling and removing tectonic motion and isolating signals caused by volcanic processes (e.g., Lisowski et al. 2008). There is a growing emphasis on InSAR as a tool for monitoring entire volcanic arcs (e.g., Pritchard and Simons 2002, 2004; Chaussard and Amelung 2012; Ebmeier et al. 2013). While appropriate in many situations, it is not clear that InSAR is the best method for characterizing volcano deformation in the Cascades. Coherence at most volcanoes is poor due to persistent snow and ice cover on most edifices and heavy vegetation in some cases. Multi-temporal methods and modeling of atmospheric delay anomalies are necessary to emphasize signal over the sometimes substantial noise (e.g., Parker et al. 2015). As a research tool, InSAR has great promise, having already provided important insights into known deformation sources, like those active at Medicine Lake volcano (Parker et al. 2014) and the Lassen volcanic center (Parker et al. 2016). The technique should not be abandoned as a means of detecting previously unknown deformation, given the examples of Three Sisters (Wicks et al. 2002) and Lassen (Poland et al. 2004), but the low rates of deformation that characterize Cascade volcanoes should temper expectations

Page 27 of 33 59

that InSAR is a Bmagic bullet^ for discovering and characterizing surface displacements throughout the range. Finally, microgravity surveys should be a part of any monitoring and/or research strategy. Experience from volcanic systems around the world indicates that gravity change can occur in the absence of volcanic deformation, providing yet another tool for identifying restless volcanoes (Carbone et al. 2017). As of 2017, gravity measurements have suggested the possibility of subsurface mass variations at two of the three Cascade volcanoes where repeat data have been collected. The Mount Baker case was discussed above. At Mount St. Helens, precise gravity surveys in 2010, 2012, 2014, and 2016 recorded changes indicative of a mass increase, especially during 2010–2012, that Battaglia et al. (2015) interpreted as evidence for magma recharge but that might also reflect reaccumulation of groundwater in porous crater-fill material surrounding the magma conduit (especially given that the mass increase was much larger than expected from the slight amount of inflation that occurred during 2008–2011). Seismicity indicative of reservoir repressurization has been recorded since the end of the 2004–2008 eruption, but only sporadically. Together, geodetic and seismic observations suggest ongoing but episodic magma accumulation in a complex storage system beneath the volcano. Future work Despite the long record of geodetic studies in the Cascade volcanoes of the USA (Table 1), many exceptional opportunities still exist to gain new insights into the range and into volcanic processes in general, and additional study is needed. Indeed, there still may be undiscovered volcano deformation occurring at low rates in areas of the arc that have received scant attention thus far. Based on the lessons learned from the previous several decades of work, we suggest that the following elements be included in future efforts to build upon and expand knowledge of the geodetic characteristics of the Cascades: 1. Establish ground-based monitoring of major volcanic centers that have yet to be investigated. Especially important are Glacier Peak and Mount Adams. Mount Jefferson should also receive attention, but it has no record of Holocene eruptions, and Jefferson is classified as a very low threat volcano by Ewert et al. (2005) and therefore should be a lower priority. Such efforts would not be without their challenges, given the remote nature and wilderness designations at these volcanoes. However, recognizing the limitations on InSAR studies described above, no definitive assessment of deformation throughout the arc will be possible without terrestrial datasets. In particular, a network of CGPS stations (at least three sites) is key to understanding the temporal evolution of any magmatic activity. These stations need not be located high on

59

Page 28 of 33

the edifice. Permitting for such installations is currently underway for Glacier Peak. 2. Perform a systematic application of multi-temporal InSAR to all volcanoes in the arc using Sentinel-1 data, once sufficient imagery has been acquired. Sentinel-1 is a notable satellite mission, since it has as a primary goal the detection of changes in the land surface using interferometry, and images are acquired from the same viewing geometry on a regular basis (other radar satellites, with different mission objectives, acquire data in a less systematic way in terms of viewing geometry and imaging mode). The collection of numerous Sentinel-1 images in the Cascades every summer ensures that many summer-tosummer interferograms, spanning periods when snow and ice cover are minimized, will be possible. Within a few years, the archive of images will be such that even small-scale deformation of a few millimeters per year should be detectable, especially if atmospheric delay anomalies can be modeled (Parker et al. 2015). Examination of archive InSAR data from past SAR missions may also prove fruitful. L-band data collected by the ALOS-1 satellite, in particular, have shown value in tracking deformation at heavily vegetated volcanoes (Chaussard and Amelung 2012; Ebmeier et al. 2013). As demonstrated by multi-temporal InSAR methods applied to Lassen and Medicine Lake (Parker et al. 2014, 2016), the increased spatial resolution provided by SAR interferometry is a key aid for modeling source geometries and interpreting source mechanisms. 3. Investigate distributed volcanic fields, both on the ground and using InSAR. There are numerous fields of mafic volcanism throughout the Cascade arc south of Mount Rainier—in fact, scoria cones, shields, and fissure-fed lavas are the dominant products of volcanism in Oregon (Hildreth 2007). Many of these volcanic fields have experienced Holocene eruptions. Indian Heaven, between Mount Adams and Mount St. Helens, hosts ~50 mafic vents, the youngest of which erupted 9000 years ago (Hildreth 2007). In Oregon, Holocene mafic eruptions have occurred in the vicinity of Mount Jefferson (Hildreth 2007). Indeed, it is unclear if surface uplift west of South Sister is associated with that volcano, or if it indicates intrusion of magma that may one day feed a monogenetic mafic eruption peripheral to South Sister. Given the relatively high rates of magmatic productivity in central Oregon and northern California compared to the rest of the arc (Sherrod and Smith 1990; Hildreth 2007), distributed volcanic fields should not be overlooked as potential sites of geodetic change due to magmatic processes. 4. Establish and reoccupy microgravity networks. A growing body of literature suggests that gravity change on volcanoes is relatively common, and that it sometimes

Bull Volcanol (2017) 79:59

occurs in the absence of deformation and other indicators of unrest (Carbone et al. 2017). Limited work in the Cascades suggests that gravity may be a useful indicator of volcanic processes in that arc as well. Gravity increases at Mount Baker and Mount St. Helens have occurred in the absence of significant uplift, leading to the development of models that account for increases in mass without significant corresponding increases in volume. Even at South Sister, where no gravity change was detected by repeated surveys despite many centimeters of surface uplift, the lack of subsurface mass change can inform models of source mechanisms, perhaps indicating the importance of viscoelastic processes (Zurek et al. 2012). Obvious targets for future gravity work in the Cascades are Medicine Lake and Lassen, which are known to be subsiding at relatively high rates (as much as ~1 cm/year) due to uncertain processes. Knowledge of any mass change accompanying the deformation could provide important insights into subsidence source mechanisms, similar to the combined deformation and microgravity approach applied at Askja, Iceland (de Zeeuw-van Dalfsen 2005). A small microgravity network was installed at Lassen in 1981 (Jachens et al. 1983) but has not been reoccupied. Subsidence of 1 cm/year since that time implies a maximum gravity change of over 100 μGal as of 2017—well above measurement uncertainty—from the free-air effect alone (e.g., Carbone et al. 2017). 5. Continue episodic GPS and SPGPS surveys at Cascade volcanoes, and maintain and expand CGPS networks. The results detailed above emphasize the importance of GPS to any volcano geodesy strategy in the Cascades. Continuous networks are critical for assessing changes over time, as demonstrated by variability of deformation at Mount St. Helens and decay in inflation rate at South Sister, as well as establishing a regional context so that tectonic and magmatic deformation can be distinguished. Campaign or semipermanent GPS networks provide additional flexibility in terms of spatial and temporal coverage, and over just a few years are capable of resolving millimeter-scale volcano-related displacements. The strategy of revisiting each volcano in the Cascades every ~5 years should be continued, as it has yielded good results thus far, and wherever possible campaign stations should be supplemented with or replaced by SPGPS deployments, which produce better signal-tonoise results (Dzurisin et al. 2017). Efforts are currently underway to broaden the campaign GPS network and install SPGPS stations at Crater Lake. In addition, CGPS stations are planned for Mount Baker and Glacier Peak, and additional CGPS sites are being permitted for Mount Hood. Expansion of ground-based geodetic monitoring in the Cascades

Bull Volcanol (2017) 79:59

will remain a priority of the USGS Cascades Volcano Observatory and partner organizations for years to come.

Page 29 of 33 59 participated in the collection, analysis, and interpretation of geodetic data over the years. In particular, we wish to acknowledge Jack Kleinman, Ken Yamashita, and Gene Iwatsubo for their commitment to establishing, maintaining, and expanding geodetic measurements throughout the Cascades. Joe Bard’s expertise was critical for creating the maps in this report. Several figures were created using the Generic Mapping Tools software (Wessel et al. 2013). This manuscript was greatly improved by reviews from Don Swanson, Bill Chadwick, and Matt Pritchard.

Conclusions Even in quiescence, Cascade volcanoes are dynamic systems that display a range of geodetic behaviors. Deformation has been detected at five of the 10 USA Cascade volcanic centers that have been investigated using terrestrial and space-based geodetic methods. This proportion places the range intermediate between arcs characterized by numerous deforming volcanoes, like the Aleutians, and those with very few, like Central America. Results from the Cascades demonstrate the importance of ground-based geodetic data, particularly from GPS and gravity, for detecting and tracking low-magnitude displacements at heavily forested and ice-clad volcanoes. InSAR also provides a valuable tool for focused study of the spatial patterns of volcanic deformation. Although poor coherence limits the utility of the technique, multi-temporal methods applied to growing archives of data may be able to overcome this challenge. Except for localized deformation associated with Mount St. Helens’ recent eruptions, annual displacement rates due to volcanic sources are on the order of millimeters per year to, at most, a few centimeters per year. Source mechanisms active in the Cascades include post-emplacement cooling of recent volcanic deposits, magma accumulation, magma withdrawal, volatile exsolution, and magmatic-tectonic interactions. Deformation associated with eruptive activity, at least based on the example of Mount St. Helens, can be subtle in the far field and may not immediately precede the onset of eruption. A carefully planned observation strategy is therefore required to fully characterize deformation associated with eruption cycles—before, during, and after eruptive activity. Near-field deformation, however, can be extraordinary, reflecting magma ascent to, and accumulation at, shallow depths. Continuation and expansion of geodetic monitoring and research in the Cascades is of fundamental importance for better understanding the diverse manifestations and mechanisms of volcano deformation and gravity change. Such work is also critical for establishing a context for recognition and interpretation of precursors to the next eruption in the arc, whether that occurs at Mount St. Helens, one of the other major volcanic centers of the range, or at a distributed volcanic field between major edifices. Acknowledgments This research was supported by the US Geological Survey’s Volcano Hazards Program and Volcano Science Center. We are grateful to the numerous staff members, interns, and volunteers at the USGS Cascades Volcano Observatory, from 1980 to the present, who

References Achauer U, Evans JR, Stauber DA (1988) High-resolution seismic tomography of compressional wave velocity structure at Newberry volcano, Oregon Cascade Range. J Geophys Res 93(B9):10,135– 10,147. doi:10.1029/JB093iB09p10135 Anderson K, Lisowski M, Segall P (2010) Cyclic ground tilt associated with the 2004-2008 eruption of Mount St. Helens. J Geophys Res 115(B11201). doi:10.1029/2009JB007102 Anderson K, Segall P (2011) Physics-based models of ground deformation and extrusion rate at effusively erupting volcanoes. J Geophys Res 116(B07204). doi:10.1029/2010JB007939 Bacon CR (1983) Eruptive history of Mount Mazama and Crater Lake Caldera, Cascade Range, USA. J Volcanol Geotherm Res 18(1–4): 57–115. doi:10.1016/0377-0273(83)90004-5 Bacon CR, Lanphere MA (2006) Eruptive history and geochronology of Mount Mazama and the Crater Lake region, Oregon. Geol Soc Am Bull 118(11/12):1331–1359. doi:10.1130/B25906.1 Battaglia M, Lisowski M, Dzurisin D, Poland MP, Schilling SP, Diefenbach AK, Wynn J (2015) Mass intrusion at Mount St. Helens (WA) from temporal gravity variations. Am. Geophys. Un. Fall Meet., Dec. 14–18, San Fran., Calif., Abstract G41A-1012 Beachly MW, Hooft EEE, Toomey DR, Waite GP (2012) Upper crustal structure of Newberry Volcano from P-wave tomography and finite difference waveform modeling. J Geophys Res 117(B10311). doi: 10.1029/2012JB009458 Begét JE (1982) Recent volcanic activity at Glacier Peak. Science 215: 1389–1390. doi:10.1126/science.215.4538.1389 Carbone D, Poland MP, Diament M, Greco F (2017) The added value of time-variable microgravimetry to the understanding of how volcanoes work. Earth-Sci Rev 168:146–179. doi:10.1016/j.earscirev. 2017.04.014 Chadwick WW, Iwatsubo EY, Swanson DA, Ewert JW (1985) Measurements of slope distances and vertical angles at Mount Baker and Mount Rainier, Washington, Mount Hood and Crater Lake, Oregon, and Mount Shasta and Lassen Peak, California, 1980–1984. US Geol Surv Open File Rep 85-205, pp 96 Chadwick WW, Archuleta RJ, Swanson DA (1988) The mechanics of ground deformation precursory to dome-building extrusions at Mount St. Helens 1981–1982. J Geophys Res 93(B5):4351–4366. doi:10.1002/9781118782064.ch16 Chaussard E, Amelung F (2012) Precursory inflation of shallow magma reservoirs at west Sunda volcanoes detected by InSAR. Geophys Res Lett 39(L21311). doi:10.1029/2012GL053817 Chiarabba C, Amato A, Evans JR (1995) Variations on the NeHT highresolution tomography method: a test of technique and results for Medicine Lake Volcano, Northern California. J Geophys Res 100(B3):4035–4052. doi:10.1029/94JB02771 Christiansen RL, Peterson DW (1981) Chronology of the 1980 eruptive activity. In: Lipman PW, Mullineaux DR (eds) The 1980 eruptions of Mount St. Helens, Washington. US Geol Surv Prof Pap 1250:17– 30 Clynne MA, Christiansen RL, Trimble DA, McGeehin JP (2008) Radiocarbon dates from volcanic deposits of the Chaos Crags and

59

Page 30 of 33

Cinder Cone eruptive sequences and other deposits, Lassen Volcanic National Park and vicinity, California. US Geol Surv Open File Rep 02–290 Crandell DR (1989) Gigantic debris avalanche of Pleistocene age from ancestral Mount Shasta volcano, California, and debris-avalanche hazard zonation. US Geol Surv Bull 1861:32 Crandell DR, Miller CD, Glicken HX, Christiansen RL, Newhall CG (1984) Catastrophic debris avalanche from ancestral Mount Shasta volcano, California. Geology 12:143–146. doi:10.1130/00917613(1984)12<143:CDAFAM>2.0.CO;2 Crider JG, Frank D, Malone SD, Poland MP, Werner C, Caplan-Auerbach J (2011) Magma at depth: a retrospective analysis of the 1975 unrest at Mount Baker, Washington, USA. Bull Volcanol 73(2):175–189. doi:10.1007/s00445-010-0441-0 Crider JG, Johnsen KH, Williams-Jones G (2008) Thirty-year gravity change at Mount Baker Volcano, Washington, USA: extracting the signal from under the ice. Geophys Res Lett 35(20). doi:10.1029/ 2008GL034921 de Zeeuw-van Dalfsen E, Rymer H, Sigmundsson F, Sturkell E (2005) Net gravity decrease at Askja volcano, Iceland: constraints on processes responsible for continuous caldera deflation, 1988–2003. J Volcanol Geotherm Res 139(3–4):227–239. doi:10.1016/j. jvolgeores.2004.08.008 de Zeeuw-van Dalfsen E, Rymer H, Sturkell E, Pedersen R, Hooper A, Sigmundsson F, Ófeigsson B (2013) Geodetic data shed light on ongoing caldera subsidence at Askja, Iceland. Bull Volcanol 75(5): 709. doi:10.1007/s00445-013-0709-2 Diefenbach AK, Crider JG, Schilling SP, Dzurisin D (2012) Rapid, lowcost photogrammetry to monitor volcanic eruptions: an example from Mount St. Helens, Washington, USA. Bull Volcanol 74(2): 579–587. doi:10.1007/s00445-011-0548-y Donnelly-Nolan JM (1988) A magmatic model of Medicine Lake Volcano, California. J Geophys Res 93(B5):4412–4420. doi:10. 1029/JB093iB05p04412 Donnelly-Nolan JM, Champion DE, Grove TL (2016) Late Holocene volcanism at Medicine Lake Volcano, northern California Cascades. US Geol Surv Prof Pap 1822, pp 59. doi:10.3133/pp1822 Donnelly-Nolan JM, Champion DE, Miller CD, Grove TL, Trimble DA (1990) Post-11,000-year volcanism at Medicine Lake Volcano, Cascade Range, Northern California. J Geophys Res 95(B12):19, 693–19,704. doi:10.1029/JB095iB12p19693 Donnelly-Nolan JM, Grove TL, Lanphere MA, Champion DE, Ramsey DW (2008) Eruptive history and tectonic setting of Medicine Lake Volcano, a large rear-arc volcano in the southern Cascades. J Volcanol Geotherm Res 177(2):313–328. doi:10.1016/j.jvolgeores. 2008.04.023 Donnelly-Nolan JM, Nolan KM (1986) Catastrophic flooding and eruption of ash-flow tuff at Medicine Lake Volcano, California. Geology 14(10):875–878. doi:10.1130/0091-7613(1986)14<875: CFAEOA>2.0.CO;2 Donnelly-Nolan JM, Stovall WK, Ramsey DW, Ewert JW, Jensen RA (2011) Newberry Volcano—Central Oregon’s Sleeping Giant. US Geol Surv Fact Sheet 2011–3145 Dzurisin D (1992a) Electronic tiltmeters for volcano monitoring: lessons from Mount St. Helens. In: Ewert JW, Swanson DA (eds) Monitoring volcanoes: techniques and strategies used by the staff of the Cascades Volcano Observatory 1980–90. US Geol Surv Bull 1966:69–83 Dzurisin D (1992b) Geodetic leveling as a tool for studying restless volcanoes. In: Ewert JW, Swanson DA (eds) Monitoring volcanoes: techniques and strategies used by the staff of the Cascades Volcano Observatory 1980–90. US Geol Surv Bull 1966:125–134 Dzurisin D (1999) Results of repeated leveling surveys at Newberry Volcano, Oregon, and near Lassen Peak Volcano. Calif Bull Volcanol 61(1–2):83–91. doi:10.1007/s004450050264

Bull Volcanol (2017) 79:59 Dzurisin D (2000) Volcano geodesy: challenges and opportunities for the 21st century. Philos Trans R Soc Lond Ser A 358(1770):1547–1566. doi:10.1098/rsta.2000.0603 Dzurisin D, Donnelly-Nolan JM, Evans JR, Walter SR (1991) Crustal subsidence, seismicity, and structure near Medicine Lake Volcano, California. J Geophys Res 96(B10):16,319–16,333. doi:10.1029/ 91JB01452 Dzurisin D, Johnson DJ, Murray TL, Myers B (1982) Tilt networks of Mount Shasta and Lassen Peak, California. US Geol Surv Open File Rep 82–670, pp 42 Dzurisin D, Johnson DJ, Symonds RB (1983b) Dry tilt network at Mount Rainier, Washington. US Geol Surv Open File Rep 83–277, pp 18 Dzurisin D, Lisowski M, Poland MP, Sherrod DR, LaHusen RG (2008) Constraints and conundrums resulting from ground-deformation measurements made during the 2004–2005 dome-building eruption of Mount St. Helens, Washington. In: Sherrod DR, Scott WE, Stauffer PH (eds) A volcano rekindled: the renewed eruption of Mount St. Helens, 2004–2006. US Geol Surv Prof Pap 1850:281– 300 Dzurisin D, Lisowski M, Wicks CW (2009) Continuing inflation at Three Sisters volcanic center, central Oregon Cascade Range, USA, from GPS, leveling, and InSAR observations. Bull Volcanol 71(10): 1091–1110. doi:10.1007/s00445-009-0296-4 Dzurisin D, Lisowski M, Wicks CW (2017) Semipermanent GPS (SPGPS) as a volcano monitoring tool: rationale, method, and applications. J Volcanol Geotherm Res. doi:10.1016/jvolgeores2017. 03.007 Dzurisin D, Lisowski M, Wicks CW Jr, Poland MP, Endo ET (2006) Geodetic observations and modeling of magmatic inflation at the Three Sisters volcanic center, central Oregon Cascade Range, USA. J Volcanol Geotherm Res 150:35–54. doi:10.1016/j. jvolgeores.2005.07.011 Dzurisin D, Moran SC, Lisowski M, Schilling SP, Anderson KR, Werner C (2015) The 2004–2008 dome-building eruption at Mount St. Helens, Washington: epilogue. Bull Volcanol 77(10):89. doi:10. 1007/s00445-015-0973-4 Dzurisin D, Poland MP, Bürgmann R (2002) Steady subsidence of Medicine Lake Volcano, Northern California, revealed by repeated leveling surveys. J Geophys Res 107(B12):2372. doi:10.1029/ 2001JB000893 Dzurisin D, Westphal JA, Johnson DJ (1983a) Eruption prediction aided by electronic tiltmeter data at Mount St. Helens Sci 221(4618): 1381–1383. doi:10.1126/science.221.4618.1381 Ebmeier SK, Biggs J, Mather TA, Amelung F (2013) On the lack of InSAR observations of magmatic deformation at Central American volcanoes. J Geophys Res 118(5):2571–2585. doi:10.1002/jgrb. 50195 Eichelberger JC (1981) Mechanism of magma mixing at Glass Mountain, Medicine Lake Highland volcano, California. In: Johnston DA, Donnelly-Nolan JM (eds) Guides to some volcanic terranes in Washington, Idaho, Oregon, and northern California. US Geol Surv Circ 838:183–189. New fumarolic activity on Mt. Baker: observations during April through July, 1975 Eichelberger JC, Heiken G, Widdicombe R, Wright D, Keady CJ, Cobb DD (1976) New fumarolic activity on Mt. Baker: observations during April through July, 1975. J Volcanol Geotherm Res 1(1):35–53. doi:10.1016/0377-0273(76)90017-2 Endo ET, Malone SD, Noson LL, Weaver CS (1981) Locations, magnitudes, and statistics of the March 20–May 18 earthquake sequence. In: Lipman PW, Mullineaux DR (eds) The 1980 eruptions of Mount St. Helens, Washington. US Geol Surv Prof Pap 1250:93–107 Evans JR, Zucca JJ (1988) Active high-resolution seismic tomography of compressional wave velocity and attenuation structure at Medicine Lake Volcano, Northern California Cascade Range. J Geophys Res 93(B12):15,016–15,036. doi:10.1029/JB093iB09p10135

Bull Volcanol (2017) 79:59 Ewert JW, Swanson DA (eds) (1992) Monitoring volcanoes; techniques and strategies used by the staff of the Cascades Volcano Observatory, 1980–90. US Geol Surv Bull 1966 Ewert JW, Guffanti M, Murray TL (2005) An assessment of volcanic threat monitoring cababilities in the United States: framework for a National Volcano Early Warning System. US Geol Surv Open File Rep 2005–1164, pp 62 Frank D, Meier MF, Swanson DA (1977) Assessment of increased thermal activity at Mount Baker, Washington, March 1975– March 1976. US Geol Surv Prof Pap 1022-A, pp 49 Gardner JE, Carey S, Sigurdsson H (1998) Plinian eruptions at Glacier Peak and Newberry volcanoes, United States: implications for volcanic hazards in the Cascade Range. Geol Soc Am Bull 110(2):173– 187. doi:10.1130/0016-7606(1998)110<0173:PEAGPA>2.3.CO;2 Heath BA, Hooft EEE, Toomey DR, Bezada MJ (2015) Imaging the magmatic system of Newberry Volcano using joint active source and teleseismic tomography. Geochem Geophys Geosys 16(12): 4433–4448. doi:10.1002/2015GC006129 Hildreth W (2007) Quaternary magmatism in the Cascades—geologic perspectives. US Geol Surv Prof Pap 1744 Hildreth W, Fierstein J (1997) Recent eruptions of Mount Adams, Washington Cascades, USA. Bull Volcanol 58(6):472–490. doi:10. 1007/s004450050156 Hildreth W, Fierstein J, Lanphere M (2003) Eruptive history and geochronology of the Mount Baker volcanic field, Washington. Geol Soc Am Bull 115(6):729–764. doi:10.1130/0016-7606(2003) 115<0729:EHAGOT>2.0.CO;2 Hildreth W, Lanphere MA (1994) Potassium-argon geochronology of a basalt-andesite-dacite arc system: the Mount Adams volcanic field, Cascade Range of southern Washington. Geol Soc Am Bull 106(11):1413–1429. doi:10.1130/0016-7606(1994)106<1413: PAGOAB>2.3.CO;2 Hodge BE, Crider JG (2010) Investigating mechanisms of edifice deflation, 1981–2007, at Mount Baker volcano, Washington, United States. J Geophys Res 115(B04401). doi:10.1029/2009JB006730 Hooper A (2008) A multi-temporal InSAR method incorporating both persistent scatterer and small baseline approaches. Geophys Res Lett 35(16). doi:10.1029/2008GL034654 Iwatsubo EY, Ewert JW, Murray TL (1992a) Monitoring radial crack deformation by displacement meters. In: Ewert JW, Swanson DA (eds) Monitoring volcanoes: techniques and strategies used by the staff of the Cascades Volcano Observatory 1980–90. US Geol Surv Bull 1966:95–101 Iwatsubo EY, Swanson DA (1992a) Trilateration and distance-measuring techniques used at Cascades and other volcanoes. In: Ewert JW, Swanson DA (eds) Monitoring volcanoes: techniques and strategies used by the staff of the Cascades Volcano Observatory 1980–90. US Geol Surv Bull 1966:103–114 Iwatsubo EY, Swanson DA (1992b) Methods used to monitor deformation of the crater floor and lava dome at Mount St. Helens, Washington. In: Ewert JW, Swanson DA (eds) Monitoring volcanoes: techniques and strategies used by the staff of the Cascades Volcano Observatory 1980–90. US Geol Surv Bull 1966:53–68 Iwatsubo EY, Topinka L, Swanson DA (1992b) Slope-distance measurements to the flanks of Mount St. Helens, late 1980 through 1989. In: Ewert JW, Swanson DA (eds) Monitoring volcanoes: techniques and strategies used by the staff of the Cascades Volcano Observatory 1980–90. US Geol Surv Bull 1966:85–94 Iwatsubo EY, Topinika L, Swanson DA (1988) Measurements of slope distances and zenith angles at Newberry and South Sister volcanoes, Oregon, 1985–1986. US Geol Surv Open File Rep 88–377 Jachens RC, Dzurisin D, Elder WP, Saltus RW (1983) Precision gravity networks at Lassen Peak and Mount Shasta, California. US Geol Surv Open File Rep 83–192 Jachens RC, Spydell DR, Pitts GS, Dzurisin D, Roberts CW (1981) Temporal gravity variations at Mount St. Helens, March-May

Page 31 of 33 59 1980. In: Lipman PW, Mullineaux DR (eds) The 1980 eruptions of Mount St. Helens, Washington. US Geol Surv Prof Pap 1250: 175–182 Jensen RA, Chitwood LA (2000) Late Holocene uplift of caldera floor, Newberry Volcano, central Oregon. What’s new at Newberry Volcano, Oregon: guidebook for the friends of the Pleistocene eighth annual Pacific northwest cell field trip 88–96. Kwoun O-I, Lu Z, Neal C, Wicks C Jr (2006) Quiescent deformation of the Aniakchak Caldera, Alaksa, mapped by InSAR. Geology 34(1): 5–8. doi:10.1130/G22015.1 LaHusen RG, Swinford KJ, Logan M, Lisowski M (2008) Instrumentation in remote and dangerous settings; examples using data from GPS Bspider^ deployments during the 2004–2005 eruption of Mount St. Helens, Washington. In: Sherrod DR, Scott WE, Stauffer PH (eds) A volcano rekindled: the renewed eruption of Mount St. Helens, 2004–2006. US Geol Surv Prof Pap 1850:335– 345 Langbein J (2008) Noise in GPS displacement measurements from Southern California and Southern Nevada. J Geophys Res 113(B5). doi:10.1029/2007JB005247 Lipman PW, Moore JG, Swanson DA (1981) Bulging of the north flank before the May 18 eruption—geodetic data. In: Lipman PW, Mullineaux DR (eds) The 1980 eruptions of Mount St. Helens, Washington. US Geol Surv Prof Pap 1250:143–155 Lisowski M, Dzurisin D, Denlinger RP, Iwatsubo EY (2008) Analysis of GPS-measured deformation associated with the 2004–2006 domebuilding eruption of Mount St. Helens, Washington. In: Sherrod DR, Scott WE, Stauffer PH (eds) A volcano rekindled: the renewed eruption of Mount St. Helens, 2004–2006. US Geol Surv Prof Pap 1850: 301–333 Lu Z and Dzurisin D (2014) InSAR imaging of Aleutian volcanoes— monitoring a volcanic arc from space: Springer-Praxis Books, Geophysical Sciences, 978-3-642-00347-9, 388 p Lu Z, Masterlark T, Power JA, Dzurisin D, Wicks C (2002) Subsidence at Kiska Volcano, Western Aleutians, detected by satellite radar interferometry. Geophys Res Let 29(18):1855. doi:10.1029/ 2002GL014948 MacLeod NS Jr, Sherrod DR (1988) Geologic evidence for a magma chamber beneath Newberry Volcano, Oregon. J Geophys Res 93(B9):10,067–10,079. doi:10.1029/JB093iB09p10067 Malone SD (1979) Gravity changes accompanying increased heat emission at Mount Baker, Washington. J. Volcanol Geotherm Res 6(3– 4):241–256. doi:10.1016/0377-0273(79)90004-0 Mandler BE, Donnelly-Nolan JM, Grove TL (2014) Straddling the tholeiitic/calc-alkaline transition: the effects of modest amounts of water on magmatic differentiation at Newberry Volcano, Oregon. Contrib Mineral Petrol 168(4):1066. doi:10.1007/s00410-0141066-7 Mastin LG (1994) Explosive tephra emissions at Mount St. Helens, 19891991: the violent escape of magmatic gas following storms? Geol Soc AmBull 106(2):175–185. doi:10.1130/0016-7606(1994) 106<0175:ETEAMS>2.3.CO;2 Mastin LG, Lisowski M, Roeloffs E, Beeler N (2009) Improved constraints on the estimated size and volatile content of the Mount St. Helens magma system from the 2004–2008 history of dome growth and deformation. Geophys Res Lett 36(L20304). doi:10.1029/ 2009GL039863 Miller CD (1980) Potential hazards from future eruptions in the vicinity of Mount Shasta Volcano, Northern California. US Geol Surv Bull 1503 Mogi K (1958) Relations between the eruptions of various volcanoes and the deformations of the ground surfaces around them. Bull Earthquake Res Inst 36(2):99–134 Moran SC (1994) Seismicity at Mount St. Helens, 1987-1992: evidence for repressurization of an active magmatic system. J Geophys Res 99(B3):4341–4354. doi:10.1029/93JB02993

59

Page 32 of 33

Muffler LP, Clynne MA (2015) Geologic field-trip guide to Lassen Volcanic National Park and vicinity, California. US Geol Surv Sci Investig Rep 2015–5067, pp 67. doi:10.3133/sir20155067 Mullineaux DR, Crandell DR (1981) The eruptive history of Mount St. Helens. In: Lipman PW, Mullineaux DR (eds) The 1980 eruptions of Mount St. Helens, Washington. US Geol Surv Prof Pap 1250:3–15 Musumeci C, Gresta S, Malone SD (2002) Magma system recharge of Mount St. Helens from precise relative hypocenter location of microearthquakes. J Geophys Res 107(B10). doi:10.1029/ 2001JB000629 Nichols ML, Malone SD, Moran SC, Thelen WA, Vidale JE (2011) Deep long-period earthquakes beneath Washington and Oregon volcanoes. J Volcanol Geotherm Res 200(3–4):116–128. doi:10.1016/j. jvolgeores.2010.12.005 Parker AL, Biggs J, Lu Z (2014) Investigating long-term subsidence at Medicine Lake Volcano, CA, using multitemporal InSAR. Geophys J Int 199(2):844–859. doi:10.1093/gji/ggu304 Parker AL, Biggs J, Lu Z (2016) Time-scale and mechanism of subsidence at Lassen Volcanic Center, CA, from InSAR. J Volcanol Geotherm Res 320:117–127. doi:10.1016/j.jvolgeores.2016.04.013 Parker AL, Biggs J, Walters RJ, Ebmeier SK, Wright TJ, Teanby NA, Lu Z (2015) Systematic assessment of atmospheric uncertainties for InSAR data at volcanic arcs using large-scale atmospheric models: application to the Cascade volcanoes, United States. Rem Sens Environ 170:102–114. doi:10.1016/j.rse.2015.09.003 Pierson TC, Pringle PT, Cameron KA (2010) Magnitude and timing of downstream channel aggradation and degradation in response to a dome-building eruption at Mount Hood, Oregon. Geol Soc A Bull 123(1–2):3–20. doi:10.1130/B30127.1 Poland M, Bawden G, Lisowski M, Dzurisin D (2004) Newly discovered subsidence at Lassen Peak, southern Cascade Range, California, from InSAR and GPS. Am Geophys Union Fall Meet, Dec. 13– 17, San Fran., Calif., Abstract G51A-0068. Poland MP, Bürgmann R, Dzurisin D, Lisowski M, Masterlark T, Owen S, Fink JH (2006) Constraints on the mechanism of long-term, steady subsidence at Medicine Lake volcano, northern California, from GPS, leveling, and InSAR. J Volcanol Geotherm Res 150(1– 3):55–78. doi:10.1016/j.jvolgeores.2005.07.007 Poland MP, Lu Z (2008) Radar interferometry observations of surface displacements during pre- and coeruptive periods at Mount St. Helens, Washington, 1992–2005. In: Sherrod DR, Scott WE, Stauffer PH (eds) A volcano rekindled: the renewed eruption of Mount St. Helens, 2004–2006. US Geol Surv Prof Pap 1750:361– 382 Pritchard ME, Simons M (2002) A satellite geodetic survey of large-scale deformation of volcanic centres in the central Andes. Nature 418: 167–171. doi:10.1038/nature00872 Pritchard ME, Simons M (2004) An InSAR-based survey of volcanic deformation in the Southern Andes. Geophys Res Lett 31(15). doi: 10.1029/2004GL020545 Riddick SN, Schmidt DA (2011) Time-dependent changes in volcanic inflation rate near Three Sisters, Oregon, revealed by InSAR. Geochem Geophys Geosys 12(Q12005). doi:10.1029/ 2011GC003826 Savage JC, Gan W, Svarc JL (2001) Strain accumulation and rotation in the Eastern California Shear Zone. J Geophys Res 106(B10):21, 995–22,007. doi:10.1029/2000JB000127 Scandone R, Malone SD (1985) Magma supply, magma discharge and readjustment of the feeding system of Mount St. Helens during 1980. J Volcanol Geotherm Res 23(3–4):239–262. doi:10.1016/ 0377-0273(85)90036-8 Schilling SP, Thompson RA, Messerich JA, Iwatsubo EY (2008) Use of digital aerophotogrammetry to determine rates of lava dome growth, Mount St. Helens, Washington, 2004–2005. In: Sherrod DR, Scott WE, Stauffer PH (eds) A volcano rekindled: the renewed eruption of Mount St. Helens, 2004–2006. US Geol Surv Prof Pap 1850:145–167

Bull Volcanol (2017) 79:59 Scott WE (1977) Quaternary glaciation and volcanism, Metolius River area, Oregon. Geol Soc Am Bull 88(1):113–124. doi:10.1130/00167606(1977)88<113:QGAVMR>2.0.CO;2 Scott WE (1987) Holocene rhyodacite eruptions on the flanks of South Sister volcano, Oregon. In: Fink JH (ed) The emplacement of silicic domes and lava flows. Geol Soc Am Spec Pap 212:35–54 Scott WE, Gardner CA, Sherrod DR, Tilling RI, Lanphere MA, Conrey RM (1997) Geologic history of Mount Hood volcano, Oregon—a field-trip guidebook. US Geol Surv Open File Rep 97–263 Scott WE, Sherrod DR, Gardner CA (2008) Overview of the 2004 to 2006, and continuing, eruption of Mount St. Helens, Washington. In: Sherrod DR, Scott WE, Stauffer PH (eds) A volcano rekindled: the renewed eruption of Mount St. Helens, 2004–2006. US Geol Surv Prof Pap 1850:3–22 Segall P (2013) Volcano deformation and eruption forecasting. In: Pyle DM, Mather TA, Biggs J (eds) Remote sensing of volcanoes and volcanic processes: integrating observation and modelling. Geol Soc Spec Pub 380:85–106. doi:10.1144/SP380.4 Sherrod DR, Smith JG (1990) Quaternary extrusion rates of the Cascade Range, northwestern United States and southern British Columbia. J Geophys Res 95(B12):19,465–19,474. doi:10.1029/ JB095iB12p19465 Sisson TW, Lanphere MA (2000) The geologic history of Mount Rainier volcano, Washington. Wash Geol 28(1/2):28 Sisson TW, Vallance JW (2009) Frequent eruptions of Mount Rainier over the last ~2,600 years. Bull Volcanol 71(6):595–618. doi:10. 1007/s00445-008-0245-7 Swanson DA (1992) The importance of field observations for monitoring volcanoes, and the approach of Bkeeping monitoring as simple as practical.^ In: Ewert JW, Swanson DA (eds) Monitoring volcanoes: techniques and strategies used by the staff of the Cascades Volcano Observatory 1980–90. US Geol Surv Bull 1966:219–223 Swanson DA, Lipman PW, Moore JG, Heliker CC, Yamashita KM (1981) Geodetic monitoring after the May 18 eruption. In: Lipman PW, Mullineaux DR (eds) The 1980 eruptions of Mount St. Helens, Washington. US Geol Surv Prof Pap 1250:157–168 Swanson DA, Casadevall TJ, Dzurisin D, Malone SD, Newhall CG, Weaver CS (1983) Predicting eruptions at Mount St. Helens, June 1980 through December 1982. Science 221(4618):1369– 1376. doi:10.1126/science.221.4618.1369 Thelen WA (2016) Seismicity of cascade volcanoes: characterization and comparison. Am Geophys Union Fall Meet, Dec. 12–16, San Fran., Calif., Abstract V53C-3106 Thompson RZ, Schilling SP (2007) Photogrammetry. In: Dzurisin D (ed) Volcano deformation: geodetic monitoring techniques. SpringerPraxis, Berlin, pp 195–221 Tucker D, Hildreth W, Ullrich T, Friedman R (2007) Geology and complex collapse mechanisms of the 3.72 Ma Hannegan caldera, North Cascades, Washington, USA. Geol Soc Am Bull 119(3/4):329–342. doi:10.1130/B25904.1 Vallance JW, Donaghue S (2000) Holocene eruptive history of Mount Rainier. Wash Geol 28(1/2):29 Vincent P, Cladouhos TT, Schultz A, Rose K, Urquhart S (2013) Combining satellite and ground-based radar interferometry data to measure surface deformation associated with EGS pumping activities at Newberry Crater, OR geothermal site. Am Geophys Union Fall Meet, Dec. 9–13, San Fran., Calif., Abstract G31A-0953 Walker GW (1974) Some implications of late Cenozoic volcanism to geothermal potential in the High Lava Plains of south-central Oregon. Ore Bin 36(7):109–118 Werner C, Evans WC, Poland M, Tucker DS, Doukas MP (2009) Longterm changes in quiescent degassing at Mount Baker Volcano, Washington, USA; evidence for a stalled intrusion in 1975 and connection to a deep magma source. J Volcanol Geotherm Res 186(3– 4):379–386. doi:10.1016/j.jvolgeores.2009.07.006

Bull Volcanol (2017) 79:59 Wessel P, Smith WHF, Scharroo R, Luis JF, Wobbe F (2013) Generic mapping tools: improved version released. EOS Trans AGU 94(45): 409–410. doi:10.1002/2013EO450001 Wicks CW Jr, Dzurisin D, Ingebritsen S, Thatcher W, Lu Z, Iverson J (2002) Magmatic activity beneath the quiescent Three Sisters volcanic center, central Oregon Cascade Range, USA. Geophys Res Lett 29(7). doi:10.1029/2001GL014205 Yamashita KM (1992) Single-setup leveling used to monitor vertical displacement (tilt) on Cascades volcanoes. In: Ewert JW, Swanson DA (eds) Monitoring volcanoes: techniques and strategies used by the staff of the Cascades Volcano Observatory 1980–90. US Geol Surv Bull 1966:143–149 Yamashita K, Doukas MP (1987) Precise level lines at Crater Lake, Newberry Crater, and South Sister, Oregon. US Geol Surv Open File Rep:87–293

Page 33 of 33 59 Yamashita KM, Wieprecht DE (1995) Bench mark descriptions and photographs for Global Positioning System (GPS) stations in the vicinity of Mt. Shasta and Medicine Lake, California. US Geol Surv Open File Rep 95–811, pp 27 Zucca JJ, Evans JR (1992) Active high-resolution compressional wave attenuation tomography at Newberry Volcano, central Cascade Range. J Geophys Res 97(B7):11,047–11,055. doi:10.1029/ 92JB00492 Zurek J, William-Jones G, Johnson D, Eggers A (2012) Constraining volcanic inflation at Three Sisters Volcanic Field in Oregon, USA, through microgravity and deformation modeling. Geochem Geophys Geosyst 13(10). doi:10.1029/2012GC004341