Investigating the 2010 Moro Gulf deep earthquake sequence using the continuous back-projection technique

Deep earthquakes make up approximately one-quarter of all earthquakes, yet current understanding of mechanisms for deep (300-700 km) earthquake generation fails to explain why deep range earthquakes occur at all. Various mechanisms, such as metastable phase change, have been proposed, however, there is a lack of observational constraints on the properties of deep earthquakes, which deter our understanding. In order to progress our knowledge of the deep earth, such as its chemistry, mineralogy, and dynamics, robust constraints on the mechanisms of deep earthquakes are required. The goal of this project is to explore the mechanism(s) behind deep earthquakes through studying the 2010 Moro Gulf deep earthquake sequence using a continuous back-projection technique. The Moro Gulf sequence features a 'triplet' of earthquakes with hypocentral depths between 585 and 640 km. The triplet earthquakes are particularly unusual, because they are large magnitude events (Mw7.3, 7.6, and 7.4) that occurred within an hour and a half of each other, which does not agree with the Gutenberg-Richter relationship. No other triplet sequences of this magnitude have been recorded within such a short time period. Another anomalous characteristic of these earthquakes is the emergent waveforms of the sequence. Typically, deep earthquakes have impulsive waveforms, which is thought to be associated with more rapid stress drop than shallow events, for which the stress drop occurs more gradually. The back-projection technique is an efficient method to constrain earthquake rupture properties, such as rupture direction, rupture speed, location, timing, and relative energy release of an earthquake. It requires high-quality data from a dense network of seismic stations covering a large area, and we use data from the High-Sensitivity Seismograph Network (Hi-net) in Japan and US Transportable Array (US-TA). The technique creates a grid of potential source locations around the hypocenter. The seismograms are time-shifted and stacked at each grid point. The data from the three earthquakes are filtered to a frequency range of 0.8-2 Hz in order to maximize the resolution from the back-projection, and to capture the first arriving P-waves in the waveforms. Another crucial step is aligning the P- and PKP(DF)-phases for Hi-net and US-TA, respectively. Since alignment has a strong influence on the back-projection, we aligned the US-TA phases using an aftershock because smaller magnitude earthquakes often improve alignment. The back-projection technique has been used in previous studies to investigate both shallow (0-100 km) and intermediate-depth (100-300 km) earthquakes. Using back-projection on deep earthquakes is expected to increase depth resolution results from the incorporation of depth phases, that, when combined with downward take-off phases, reveal a more accurate hypocentral depth. Also, better spatial resolution is achieved by the combination of the Hi-net and US-TA arrays, which overlap signals at a more accurate location of the rupture. Back-projection analysis of the Moro Gulf sequence is expected to provide a detailed rupture process of these deep events with high depth and spatial resolution.