Taiwan experiences high deformation rates, particularly along its eastern margin where a shortening rate of about 30mm/yr is experienced in the Longitudinal Valley and the Coastal Range. Four Sacks-Evertson borehole strainmeters have been installed in this area since 2003. Liu et al. (2009) proposed that a number of strain transient events, primarily coincident with low-barometric pressure during passages of typhoons, were due to deep-triggered slow slip. Here we extend that investigation with a quantitative analysis of the strain responses to precipitation as well as barometric pressure and the Earth tides in order to isolate tectonic source effects. Estimates of the strain responses to barometric pressure and groundwater level changes for the different stations vary over the ranges -1 to -3 nanostrain/millibar(hPa) and -0.3 to -1.0 nanostrain/hPa, respectively, consistent with theoretical values derived using Hooke's law. Liu et al. (2009) noted that during some typhoons, including at least one with very heavy rainfall, the observed strain changes were consistent with only barometric forcing. By considering a more extensive data set, we now find that the strain response to rainfall is about -5.1 nanostrain/hPa. A larger strain response to rainfall compared to that to air pressure and water level may be associated with an additional strain from fluid pressure changes that take place due to infiltration of precipitation. Using a state-space model, we remove the strain response to rainfall, in addition to those due to air pressure changes and the Earth tides, and investigate whether corrected strain changes are related to environmental disturbances or tectonic-original motions. The majority of strain changes attributed to slow earthquakes seem rather to be associated with environmental factors. However, some events show remaining strain changes after all corrections. These events include strain polarity changes during passages of typhoons (a characteristic that is not anticipated from our estimates of the precipitation transfer function) that are more readily explained in terms of tectonic-origin motions, but clearly the triggering argument is now weaker than that presented in Liu et al. (2009). Additional on-site water level sensors and rain gauges will provide data critical for a more complete understanding, including the currently unresolved issue of why, for some typhoons, there appears to be a much smaller transfer function for precipitation-induced strain changes.

Volumetric strain changes associated with the October 2013 M-w 6.2 Ruisui earthquake were recorded by a network made up with four borehole Sacks-Evertson dilatometers in eastern Taiwan. These instruments are located within 25-30 km of the seismic source providing also high-resolution near-field observations. Co-seismic offsets larger than a few 10(2) n epsilon were seen by most of the sensors. We relocated the 30 km x 30 km fault plane through a grid-search approach. The inferred fault parameters (217 degrees, 48 degrees, 49 degrees) are in reasonable agreement with those resulting from the inversions of long-period seismic waves (209 degrees, 59 degrees, 50 degrees) as well as from GPS data inversion (200 degrees, 45 degrees, 42 degrees). Moreover, analysis of the 100-Hz sampling data 10 s before seismic radiations indicate no pre-seismic strain change emergent from the instrumental noise level (from 10(-2) to 10(-1) n epsilon). Such an observation sets limits on any precursory change in a nucleation area, taken to have dimensions of about 250-300 m, seconds before the mainshock. Thus, the upper limit of any pre-seismic moment is about 10(-5) % of the total seismic moment of the Ruisui earthquake.

In March 2010 two borehole strainmeters and three Michelson tiltmeters within the Campi Flegrei volcanic system, Italy, registered an abrupt deformation signal that was followed 20min later by seismic slip on a pair of onshore normal faults. We demonstrate that the observed strain changes were caused by a small but rapid volume decrease in a previously identified offshore ellipsoidal magma source or part of it. Although the total deflation was below the detectability of interferometric synthetic aperture radar and GPS, deflation observed rates were briefly 2 orders of magnitude more rapid than decadal inflation rates. We conclude that this high dilatational contraction rate was responsible for triggering seismicity and that this process may be responsible for the normal faulting often observed in the Campi Flegrei region. Our study quantifies the crucial role played by a transient, minor reduction in dilatational stress, in triggering slip on a fault near critical failure. Our subsurface measurements of strain and tilt registered anomalous deformation three sigma above background noise levels 17min before the onset of microseismicity suggesting strain measurements have potential utility as an early warning system for the city of Naples.

Although analytical expressions for deformation due to simple reservoirs and planar dikes have been in use for decades, seldom are there sufficient data to provide well-constrained detailed models of volcanic activity. Deformation due to the eruption of Miharayama, Izu-Oshima, Japan, in November 1986 was captured by a wider array of monitoring than usual at that time. Here we show that the first eruptive stage requires a shallow non-spherical source, extended perpendicular to the axis of maximum tectonic tension, together with a deeper reservoir. The shallow reservoir was partially recharged, from a deeper reservoir, during the eruption, with a magma ascent rate of similar to 10 km/day; this has important implications for estimation of reservoir gas content. Precursory changes starting 2 h before the second phase require dike propagation from at least 10 km depth. Surface elevation changes are indicative of a long sub-surface dike. A model incorporating dikes and a deep (10 km) reservoir, extended in the same direction as the dikes, provides a very good match to the data. Such reservoir geometry may explain why the eruption was unexpected. (C) 2016 Elsevier B.V. All rights reserved.

A network of four borehole dilatometers has been installed on Etna in two successive phases (2010-2011 and 2014). The borehole dilatometers are installed in holes drilled at depths usually greater than 100 m, and they measure the volumetric strain of the surrounding rock with a nominal precision up to 10(-11) in a wide frequency range (10(-7) -25 Hz). Here we describe the characteristics of the network and the results of the in situ calibrations obtained after the installations by different methods. We illustrate short-term strain changes recorded during several lava fountains erupted by Etna during 2011-2013, and we also show signal changes recorded at all four stations during the lava fountain on 28 December 2014. Analytical and numerical computations constrained the eruptions source depth and also its volume change that is related to the magma volume emitted. Finally, we show the potential of the signal in the medium term to reveal strain changes related to different phases of the volcanic activity.

We analyze the high-resolution dilatation data for the October 2013 6.2 Ruisui, Taiwan, earthquake, which occurred at a distance of 15-20 km away from a Sacks-Evertson dilatometer network. Based on well-constrained source parameters (, , ), we propose a simple rupture model that explains the permanent static deformation and the dynamic vibrations at short period (3.5-4.5 s) for most of the four sites with less than 20 % of discrepancies. This study represents a first attempt of modeling simultaneously the dynamic and static crustal strain using dilatation data. The results illustrate the potential for strain recordings of high-frequency seismic waves in the near-field of an earthquake to add constraints on the properties of seismic sources.

Volcanic eruptions are typically accompanied by ground deflation due to the withdrawal of magma from depth and its effusion at the surface. Here, based on continuous high-resolution borehole strain data, we show that ground deformation was absent during the major effusion phases of the 1991 and 2000 eruptions of Hekla Volcano, Iceland. This lack of surface deformation challenges the classic model of magma intrusion/withdrawal as source for volcanic ground uplift/subsidence. We incorporate geodetic and geochemical observables into theoretical models of magma chamber dynamics in order to constrain quantitatively alternative co- and intereruptive physical mechanisms that govern magma propagation and system pressurization. We find the lack of surface deformation during lava effusion to be linked to chamber replenishment from below whilst magma migrates as a buoyancy-driven flow from the magma chamber towards the surface. We further demonstrate that intereruptive pressure build-up is likely to be generated by volatile ascent within the chamber rather than magma injection. Our model explains the persistent periodic eruptivity at Hekla throughout historic times with self-initiating cycles and is conceptually relevant to other volcanic systems.

Short-period deformation cycles are a common phenomenon at active volcanoes and are often attributed to the instability of magma flow in the upper plumbing system caused by fluctuations in magma viscosity related to cooling, degassing, and crystallization. Here we present 20-min periodic oscillations in ground deformation based on high-precision continuous borehole strain data that were associated with the 2003 massive dome collapse at the Soufriere Hills Volcano, Montserrat (West Indies). These high-frequency oscillations lasted similar to 80 min and were preceded by a 4-hour episode of rapid expansion of the shallow magma reservoir. Strain amplitude ratios indicate that the deformational changes were generated by pressure variations in the shallow magma reservoir and - with reversed polarity - the adjacent plumbing dike. The unusually short period of the oscillations cannot be explained with thermally induced variations in magma properties. We investigate the underlying mechanism of the oscillations via a numerical model of forced magma flow through a reservoir-dike system accounting for time-dependent dilation/contraction of the dike due to a viscous response in the surrounding host rock. Our results suggest that the cyclic pressure variations are modulated by the dynamical interplay between rapid expansion of the magma chamber and the incapacity of the narrow dike to take up fast enough the magma volumes supplied by the reservoir. Our results allow us to place first order constraints on the viscosity of crustal host rocks and consequently its fractional melt content. Hence, we present for the first time crustal scale in situ measurements of rheological properties of mush zones surrounding magmatic systems. (C) 2017 Elsevier B.V. All rights reserved.

We propose an approach for calibrating the horizontal tidal shear components [(differential extension () and engineering shear ()] of two Sacks-Evertson (in Pap Meteorol Geophys 22:195-208, 1971) SES-3 borehole strainmeters installed in the Longitudinal Valley in eastern Taiwan. The method is based on the waveform reconstruction of the Earth and ocean tidal shear signals through linear regressions on strain gauge signals, with variable sensor azimuth. This method allows us to derive the orientation of the sensor without any initial constraints and to calibrate the shear strain components and against tidal constituent. The results illustrate the potential of tensor strainmeters for recording horizontal tidal shear strain.

We report evidence for frictional afterslip at shallow depths (about 5 to 7km) during a small-magnitude seismic sequence (with M-L<5) along the Chihshang Fault, a main active structure of the Longitudinal Valley, in southeast Taiwan. The afterslip, which was recorded by a nearby borehole dilatometer, lasted about a month with a cumulative geodetic moment magnitude of 4.80.2. The afterslip comprised two stages and controlled the aftershock sequence. The first postseismic stage, which followed a M-L 4.6 earthquake, lasted about 6h and mostly controlled the ruptures of neighboring asperities (e.g., multiplets) near the hypocenter. Then, a 4week duration large afterslip event following a M-L 4.9 earthquake controlled the rate of aftershocks during its first 2days through brittle creep. The study presents a rare case of simultaneous seismological and geodetic observations for afterslip following earthquakes with magnitude lower than 5. Furthermore, the geodetic moment of the postseismic phase is at least equivalent to the coseismic moment of the sequence.