[S09P-02] Machine-learning-based analysis: Spatiotemporal evolution of aftershocks following the 2019 Cotabato and 2019 Davao del Sur earthquakes
In the Cotabato and Davao del Sur area of the southern Philippines, there was an earthquake sequence involving five M~6 (Mw 6.4, 6.6, 5.9, 6.5, and 6.7) inland earthquakes with strike-slip focal mechanisms, which occurred during from October to December 2019. A deep-neural-network-based phase picker, PhaseNet [Zhu and Beroza, 2019], was applied to seismic waveform data from 18 stations within 200 km of the epicentral area in order to obtain the arrival time of P and S waves of the five largest events and their aftershocks. The obtained phases were first associated and initially located using Rapid Earthquake Association and Location (REAL) [Zhang et al., 2019]. Subsequently, the earthquakes were relocated using VELEST [Kissling et al., 1994], and the hypocenters were further refined using the relative hypocenter relocation method hypoDD [Waldhauser & Ellsworth, 2000]. Through these analysis procedures, we successfully determined ~5,000 earthquakes, comprising substantially more earthquakes than found in the original earthquake catalog (~3,000 events) by the DOST-PHIVOLCS which relied only on manually picked phases. The obtained relocated hypocenters of the earthquake sequence show a clear cross-fault geometry involving multiple conjugate faults having a cross-point of the initial large magnitude event (Mw 6.4), though the hypocenters by the DOST-PHIVOLCS only showed a very scattered hypocenter distribution. [RB1] The cross-fault hypocenter alignments in the SW-NE direction are generally parallel to mapped active faults, whereas the other alignment in the NW-SE direction looks to transfer seismic activity from Makilala-Malungon fault to Tangbulan fault. [RB2] This newly created earthquake catalog reveals spatial and temporal changes in seismicity following each significant event and uncovers detailed patterns in the clustering of the aftershocks, which are likely linked to complex fault system structures that may have contributed to the seismic activity[RB3] . [RB1]I am not sure what this is trying to describe. [RB2]This is again a bit hard to follow without a map. Does SSJ allow for including a figure? The fault orientations seem to be SW-NE and SE-NW, not EW & NS? [RB3]It seemed that the two geodetic modeling studies might have miss-identifed the orientation of one of the events (#2?), so the importance for establishing the geometry of the coseismic rupture asperities might be good to mention too?