G43C-01 INVITED 13:40h
Moving Mountains and Deep Crustal Earthquakes: Evidence for Deep Magma Injection Beneath Lake Tahoe, Nevada-California
We recently reported [Smith et al., Science 305, 2004] an unusually deep swarm of 1611 earthquakes that occurred in late 2003 at Lake Tahoe, California (depth 29-33 kilometer; Richter Magnitude [ML] $< $2.2; sum of the moment magnitude of all events is Mw 3.1). This swarm was coeval with a GPS transient displacement of 6 $\pm$ 0.3 mm horizontally outward from the swarm and 7.9 $\pm$ 1.0 mm upward measured at the GPS station on Slide Mountain, Nevada (SLID) 18 km to the northeast. Station SLID is a part of the 53-station Basin and Range Geodetic Network (BARGEN) network, continuously operating since 1996 and one of the PBO Nucleus stations. Here we focus on the results of the geodetic analysis for SLID, and other BARGEN stations within 200 km of SLID, starting on 1 January 2000. The SLID transient displacement is 9.8 mm in a direction normal to the planar structure defined by the deep earthquake swarm, spanning the same time period of the swarm. The geodetic displacement here is too large to be explained by the elastic strain from the cumulative seismic moment of the sequence, suggesting an aseismic forcing mechanism. Aspects of the swarm and SLID displacements are consistent with lower-crustal magma injection under Lake Tahoe. During the first 23 days of the swarm, hypocentral depths migrated at a rate of 2.4 millimeters/second up-dip along a 40-km$^2$ structure striking N30$\deg$W and dipping 50$\deg$ to the northeast. Assuming a stress drop of 10 MPa (reasonable for upper crustal earthquakes), this event has an equivalent seismic moment to a magnitude (Mw) 6.0 earthquake with a displacement of ~1 meter at the source. Applying Okada's model for a tensile crack at 28 km depth in the source region, a potency equivalent to a volume of 3.7 $\times$ 10$^7$ m$^3$, or volumetric moment equivalent of Mw 6.4, fits the SLID observations. As there is no established evidence of recent volcanism ($<$1 Ma) in the Tahoe region, this discovery suggests the hypothesis that such deep magmatic events in extensional regimes may be of a "normal" tectonic process of lower crustal evolution, perhaps providing a mechanism to sustain crustal thickness and crustal strength in zones of extension. If indeed this represents one episode of a longer-duration steady-state process, similar geodetic transients might be commonplace in the Basin and Range Province, though the single event reported here does little to constrain the characteristic time scale. EarthScope is well-posed to provide further evidence. The geodetic and seismological data until June 2004 indicated that this event stopped around January 2004. Here we present an updated analysis using a GPS and microseismicity data through November 2004.
G43C-02 14:00h
Rapid Transient Deformation From a Shallow Magmatic Source at the Socorro Magma Body, NM, USA?
The Socorro Magma Body (SMB) lies within the central Rio Grande Rift (RGR) Valley and is one of the largest known magma bodies in the Earth's continental crust. Studies of local microseismicity and deep seismic soundings revealed an unusually strong reflector approximately 70 km wide at 19 km depth and identified it as a large active sill-like crustal magma intrusion. Using precision leveling (1912-80) and InSAR (1992-99), previous studies have found $\sim$2-4 mm/yr of averaged uplift centered near San Acacia, over the center of the reflector, and corresponding to about 10$^7$ m$^3$ of annual growth from an inflating sill at 19 km depth. We performed two GPS campaigns over the SMB on nine bedrock sites in 2002 and 2003. Vertical GPS velocities from six sites forming a transect over the central SMB are between $\sim$10 and 20 mm (1$\sigma \sim$10 mm) with the maximum measured surface uplift at two central stations near San Acacia. However, three sites forming a partial transect $\sim$12 km north show no uplift for this period. Additionally, continuous GPS 18 km south of the central transect shows 4-5 mm/yr uplift between 2001 and 2004. Collectively, these data suggest a significant and smaller body inflating between 5-10 km depth and corresponding to 0.5-5$\times 10^6$ m$^3$ between 2002 and 2003. Though horizontal velocities are all less than their individual errors ($\sim$5 mm) they generally radiate outward from the center of the SMB. These results indicate that the SMB may have considerable variation in the spatio-temporal pattern of deformation. This suggests that, though over several years to decades the SMB inflates at an average of 2-4 mm/yr, more frequent and widespread geodetic measurements are necessary to fully assess its complex sources. Additionally, because the southern portion of the SMB extends into the trilateration network of Savage et al. [1988], which found slow-to-no extension ($<$3 mm/yr) across the RGR, it may be that those results were contaminated by previously unknown transient effects of the SMB.
G43C-03 INVITED 14:15h
CALIPSO Borehole Monitoring Project at Soufriere Hills Volcano, Montserrat, BWI: Overview, and Response of Magma Reservoir to Prodigious Dome Collapse
Project {\bf CALIPSO} ({\bf C}aribbean {\bf A}ndesite {\bf L}ava {\bf I}sland {\bf P}recision {\bf S}eismo-geodetic {\bf O}bservatory) aims to investigate the magmatic system at the active Soufriere Hills Volcano (SHV), Montserrat. The collaborative project involves several institutions acting in partnership with the Montserrat Volcano Observatory (MVO), and is funded by NSF with assistance by NERC. SHV remains active after 9 years, displaying cyclic activity on several scales. Many aspects of andesite system dynamics remain poorly understood, and CALIPSO is expected to improve our understanding of SHV and andesite systems generally. Drilling was carried out Nov 02 to Mar 03. {\bf CALIPSO} comprises an integrated array of four strategically located 200-m boreholes, plus several shallower holes and surface installations. The borehole instruments are designed to have long life (decades). Each site includes a very broad-band Sacks-Evertson strainmeter, three-component seismometer, tiltmeter, and surface cGPS station. At one site a hot-hole strainmeter design, involving hydraulic sensors and no downhole electronics, has been used for the first time anywhere. FreeWave telemetry is coupled with Quanterra A/D converters. These instruments are intended to probe changes in the andesitic volcanic system and underlying mafic sources with unprecedented sensitivity. Early data from the July 2003 dome collapse suggest remarkable insights about the depth, shape and nature of the volatile-saturated magmatic reservoir, gleaned from the magnitude of dilatation pulses accompanying the collapse, and their change in sign of with radial distance.
G43C-04 14:35h
2004 Deformation of Okmok Volcano,Alaska, USA
Okmok Volcano is a basaltic shield volcano with a 10km diameter caldera located on Umnak Island in the Aleutian Arc, Alaska. Okmok has had frequent effusive eruptions, the latest in 1997. In 2002 the Alaska Volcano Observatory installed a seismic network and three continuous GPS stations. Two stations are located in the caldera and one is located at the base of the volcano at Fort Glenn. Because of instrumentation problems the GPS network was not fully operational until August 2003. A fourth GPS site, located on the south flank of the volcano, came online in September 2004. The three continuous GPS instruments captured a rapid inflation event at Okmok Volcano spanning 6 months from March to August 2004. The instruments give a wonderful time-series of the episode but poor spatial coverage. Modeling the deformation is accomplished by supplementing the continuous data with campaign surveys conducted in the summers of 2002, 2003 and 2004. Displacements between the 2002 and 2003 campaigns show a large inflation event between those time periods. The continuous and campaign data suggest that deformation at Okmok is characterized by short-lived rapid inflation interspersed with periods of moderate inflation. Velocities during the 2004 event reached a maximum of 31cm/yr in the vertical direction and 15cm/yr eastward at the station OKCD, compared with the pre-inflation velocities of 4cm/yr in the vertical and 2.5cm/yr southeastward. Using a Mogi point source model both prior to and during the inflation gives a source location in the center of the caldera and a depth of about 3km. The source strength rate is three times larger during the inflation event than the period preceding it. Based on the full time series of campaign and continuous GPS data, it appears that the variation in inflation rate results from changes in the magma supply rate and not from changes in the depth of the source.
G43C-05 14:50h
Volcano-Tectonic Deformation at Taal Volcano, Philippines
Taal Volcano, located in southern Luzon, Philippines, is an unusual, tholeiitic volcano situated within a calc-alkaline arc. It is one of the most active volcanic centers in the Philippines, with some 33 historic volcanic eruptions over the past four centuries. Volcanism at Taal is at least partly tectonically controlled, suggested by its location at the intersection of regional fault structures and by the location and shape of both Taal's caldera and Volcano Island. The alignment of modern eruption centers, are controlled by regional and local structures. Here, we review geomorphic and geodetic observations that constrain both tectonic and volcanic deformation in the vicinity of Taal volcano. We use GPS measurements from a 52-station GPS network measured from 1996 - 2001 to investigate overall plate interaction and microplate (intra-arc) deformation. The velocity field indicates that the majority of the Philippine Sea - Eurasia plate convergence is taking place west of Luzon, presumably largely by subduction at the Manila trench. A relatively small fraction of the convergence appears to be taking place within Luzon or across the East Luzon trough. The major intra-arc deformation is accommodated by strike-slip motion along the Philippine Fault, ranging from 25-40 mm/yr left-lateral slip. Detailed measurements in southern Luzon also indicate significant intra-arc deformation west of the Philippine Fault. GPS measurements in southwestern Luzon indicate significant motion within the arc, which could be explained by 11-13 mm/yr of left-lateral shear along the "Macolod Corridor", within which Taal Volcano resides. A dense network of continuous single- and dual-frequency GPS receivers at Taal Volcano, Philippines reveals highly time-variable deformation behavior, similar to that observed at other large calderas. While the caldera has been relatively quiescent for the past 2-3 years, previous deformation includes two major phases of intra-caldera deformation, including two phases of inflation and deflation in 1998-2000. The February-November 2000 period of inflation was characterized by approximately 120 mm of uplift of the center of Volcano Island relative to the northern caldera rim, at average rates up to 216 mm/yr. The source of deflation in 1999 was modeled as a contractional Mogi point source centered at 4.2 km depth beneath Volcano Island; the source of inflation in 2000 was modeled as a dilatational Mogi point source centered at 5.2 km depth beneath Volcano Island. The locations of the two sources are indistinguishable within the 95% confidence estimates. Modeling using a running four-month time window from June 1999-March 2001 reveals little evidence for source migration. We find marginal evidence for an elongate source whose long axis is oriented NW-SE, paralleling the caldera-controlling fault system. We suggest that the two periods of inflation observed at Taal represent episodic intrusions of magma into a shallow reservoir centered beneath Volcano Island whose position is controlled at least in part by regional tectonic structures.
G43C-06 15:05h
Geodetic Measurements and Modelling at Neapolitan Volcanoes(Southern Italy): Somma-Vesuvius and Campi Flegrei
We show the recent results about geodetic observations and modelling at two very explosive and densely populated volcanoes in Southern Italy, namely Somma-Vesuvius and Campi Flegrei caldera. The two areas, characterised by the highest volcanic risk in the World because of the density of population and exposed value, are among the best monitored ones in the World. Geodetic monitoring at these areas started more than 30 years ago, and was progressively improved in the last decade, including dense networks making use of both terrestrial and space techniques. The monitored period includes two strong unrests at Campi Flegrei caldera, not followed by eruptions, characterised by uplift of up to 3 m in few years, with rates up to 1 m/year, and intercurring subsidence with rates up to .08 m/year. Somma Vesuvius is on the contrary characterised, in the last 30 years, by a marked stability, except for a very localised subsidence at the young active center (Vesuvius) and a peculiar ring-like subsidence all around the volcanic edifice. The fast uplift and subsidence at Campi Flegrei has been modelled as due to shallow inflation sources and a dominant effect of passive slip along the ring faults bordering the collapsed area. Numerical modelling taking carefully into account the geometry of ring faults gives an accurate description of observed displacements. At Somma-Vesuvius, subsidence of Vesuvius cone is modelled in terms of gravitationally-induced slip along the contact limits between the older caldera and the younger active edifice. The ring-like subsidence around the whole edifice is modelled in terms of normal fault-like behaviour of the contacts among the loaded basement and the superimposed volcanic edifice, subject to the extensional tectonic stress of the area. Both models of ground deformations at the two areas appear very consistent with the behaviour of local volcano-tectonic seismicity, and enlighten the very important role played by volcano-tectonic structures in the genesis of ground deformations and seismicity.
G43C-07 15:20h
The shallow plumbing system of Stromboli volcano as imaged from 1 Hz instantaneous GPS positions
The Stromboli volcano (Aeolian Islands, Italy) erupted suddenly on 28 December 2002 after a 17-year period of typically persistent but moderate eruptive activity, followed two days later by a tsunamigenic landslide on its NW flank (Sciara del Fuoco) felt in the coastal areas of southern Italy. Three continuous GPS stations were quickly deployed near the volcano's rim sampling at 1 Hz, with instantaneous positions computed with a latency of 1 s relative to a fourth station on its flank. We report on two deformation episodes. A vent migration on 16-17 February 2003 caused significant displacements at only one site and contributed to the decision not to issue a warning of an impending tsunamigenic landslide. The second episode on 5 April 2003, a paroxystic explosion from the summit crater, allowed us to model, for the first time with geodetic data, the shallow magma chambers that give rise to Strombolian explosive activity.