I have been recently awarded an SNF Ambizione project that will start in early 2015. The project is called
Earthquake-induced fault reactivation at oblique subduction zones: from cause to effect
The main goal of the project is to investigate how large-magnitude earthquakes occurring at oblique convergent margins reactivate strike slip faults in the overriding plate. More specifically, I will focus on hydrothermal systems developed upon strike slip fault zones characterised by near-critical conditions at depth. The hypothesis in the study suggests that the combined action of static and dynamic stresses imposed by the earthquake main-slip and after-slips alters the already near-lithostatic conditions at depth of the system promoting permeability enhancements, unclamping of locked faults and upwelling of deep-seated fluids. These processes mutually reinforce each other and may ultimately lead to volcanic and mud eruptions. The project aims to address three distinct and yet related questions:
1) What type of long-term tectonic deformation takes place in the volcanic arc after a large-magnitude earthquake? 2) How geological systems in near-critical conditions respond to incoming seismic waves? 3) How rapidly magmas migrate through unlocked fault planes?
Over the last few years I took part in many projects, and most of them are still ongoing:
• Southern Andes, Chile. Volcanic eruptions controlled by transient tectonic inversion following megathrust earthquakes.
The recent earthquakes of Sumatra 2004 (M=9.3) and 2005 (M=8.5), Chile 2010 (M=8.8), and Japan 2011 (M=9) provide the stimuli to test how megathrust earthquakes can promote faulting and trigger volcanic eruptions by means of a toggle switch mechanism in the volcanic arc (static stress triggering). Transtension resulting from co- and post-seismic slip along the subduction zone interface initiates transform faulting, which is proposed to be the primary mechanism for large-scale migration of deep fluids. The Maule region, Chile, was selected as the natural laboratory for this study and the three fold project consists in: i) deploying a broadband seismic network to highlight the focal mechanisms of the on-going seismicity in the arc, ii) revealing whether the regional-scale inversion of the stress regime from pre-earthquake transpression to post-earthquake transtension has already triggered large volcanic eruptions in the past iii) using the acquired seismic data to identify areas where future eruptions are most likely to occur.
• Snaefellsness, Iceland. Determination of the current state of the Snæfellsjoküll volcano, Iceland.
The SIL seismic network does not cover the Snaefellsnes volcanic peninsula. Hence, the magmatic and seismic activity of this region is still unknown. To shed light on the on-going dynamics we installed a broadband seismic network that highlighted the occurrence of seismic activity in this area. The data are scientifically relevant not only because of the discovery of a new seismic area but they also point out the necessity to add additional permanent seismic stations to the existing SIL network. We recently deployed additional short period and broadband stations to better constrain the seismic activity beneath the volcano.
• Triggering of the LUSI mud volcano, Indonesia.
Triggering of the (May, 2006) LUSI mud volcano in East Java, Indonesia segregates into two competing hypotheses. One posits that drilling operations in the area destabilised an over- pressured and uncased mud layer (Davies et al., 2008; Tingay et al., 2008), which then rose through the borehole and instigated the massive outflow of mud. The other posits that the M6.3 Yogyakarta earthquake, which occurred 47 hours prior to, and 250 km distant from, the mud eruption site undercompacted the sediments (Tanikawa et al., 2010) and initiated aseismic slip on a strike-slip fault (Mazzini et al., 2009). The first explanation is evidenced by drilling logs showing a pressure kick and supported by the argument that 250 km is further than the accepted distance for triggering phenomena from earthquakes of this size (Manga et al., 2006). The second explanation follows from geochemical and geological studies that show dextral slip on a tectonic fault (Mazzini et al., 2009), and isotopic signatures indicating deeply derived hydrothermal fluids in the exhaling mud (Mazzini et al., 2012). This hypothesis is also reinforced by numerical studies that show critical conditions (i.e. lithostatic pressures) in the mud layers (Tanikawa et al., 2010). Our numerical wave propagation study shows that the parabolically-shaped overall structure of the system, topped by a high velocity layer, is key to understand this system as it reflected, focussed, and amplified seismic energy into the mud layer. We show that amplification of the incoming seismic energy was sufficient for mud liquefaction, which then capitalised on the critical state induced by the drilling operations. Over the forty-seven hours time window high pore pressures rising through the well and induced by liquefaction build up in the mud layer and promoted slipping on a incipient fault system setting the catastrophe into motion. Since many volcanic and hydrothermal systems have similar domed-shaped structures with varying acoustic impedance, our model provides a new mechanistic explanation for triggering phenomena from distant earthquakes at volcanic and hydrothermal systems.