9:45 AM - 10:00 AM
[SGD01-04] Inferring rheology of mud diapirs in SW Taiwan from postseismic deformation of the 2016 Mw 6.4 Meinong earthquake
Keywords:Mud diapir, Postseismic deformation, Meinong earthquake, Afterslip, viscoelastic deformation, GNSS measurements
Past geodetic studies highlighted that the mud diapirs of southwestern (SW) Taiwan could contribute anelastic deformation to the ground motion before and during the 2016 Mw 6.4 Meinong earthquake (Ching et al., 2016 Tectonophysics; Tsukahara & Takada, 2018 EPS). After the Meinong earthquake, GNSS observations from 2016 to 2019 reveal a substantial westward and uplift motion in several areas of SW Taiwan. This postseismic deformation may not only be explained by the continuous afterslip on the earthquake-source fault but also suggest the role of mud-diapir deformation (Tsai, 2019 JTEA). However, the rheology of the mud diapirs has yet to be investigated.
Here, we developed a 3-D numerical model of mud diapirs to explain the postseismic GNSS displacements. Our modeling incorporates two primary mechanisms: afterslip on earthquake-source fault and viscoelastic deformation of mud diapirs. The afterslip is modeled using the rate-strengthening friction law, where the postseismic slip rate is dependent on stress on the fault interface (Muto et al., 2019 Sci. Adv.; Dhar et al., 2022 GJI). The viscoelastic deformation of mud-diapir is modeled using the power-law Maxwell rheology, where the strain rate is proportional to stress raised to a power, n (i.e., stress exponent, Freed & Bürgmann, 2004 Nature). The fault interface and viscoelastic mud diapirs are discretized into triangular patches (surface elements) and cuboids (volume elements), respectively. The distributed stress field due to the deformation of patches and cuboids is calculated using semi-analytical Green's functions (see Dhar et al., 2022 GJI and references therein). We used the coseismic slip distribution proposed by Rau et al. (2022 Tectonophysics) to estimate the initial stress perturbation on the fault and mud diapirs. Our modeled mud diapirs have narrow crests (2 km width) and wider roots (6 km width), resembling the shape of mud-core anticlines (Chen et al., 2014 JAES). They are located at 1–5 km depth in the upper crust. We positioned numerous mud-diapir models following the spatial distribution of the mud-core anticlinal axis and Bouguer anomaly (Lo et al. 2016 Tectonophysics). We used pre-earthquake horizontal compression and vertical extension to simulate the diapiric stress regime. We used the grid search approach to optimize the uniformly assigned modeling parameters, including the fault friction parameter, power-law stress exponent, and Maxwell viscosity.
Our results suggest that the afterslip mainly contributes to the ground motion near the earthquake epicenter. However, the significant horizontal and uplift motion in the vast areas west of the Longchuan fault is mainly caused by the mud-diapir deformation. The effective viscosities (= strain rate/stress) of mud diapirs are in the order of ~1015 Pa s and 1020 Pa s at 1 day and 3 years after the earthquake, respectively. The transient effective viscosity of mud diapirs is one order less than that of the low-viscosity mantle (~1016 Pa s). Such low viscosity of mud diapirs and fast recovery over the postseismic period may correlate to the effect of over-pressurized mud in the mud diapirs. Our optimal model also preferred a higher order of power-law stress exponent (n = 5) compared to that of the viscoelastic lower crust-upper mantle (n = 3, Muto et al., 2019 Sci. Adv.; Dhar et al., 2022 GJI), which indicates the plastic behavior of mud in addition to their viscoelasticity. Understanding the mud-diapir rheology can provide valuable insights into the mechanisms of crustal deformation in SW Taiwan prior to, during, and after major earthquakes.
Here, we developed a 3-D numerical model of mud diapirs to explain the postseismic GNSS displacements. Our modeling incorporates two primary mechanisms: afterslip on earthquake-source fault and viscoelastic deformation of mud diapirs. The afterslip is modeled using the rate-strengthening friction law, where the postseismic slip rate is dependent on stress on the fault interface (Muto et al., 2019 Sci. Adv.; Dhar et al., 2022 GJI). The viscoelastic deformation of mud-diapir is modeled using the power-law Maxwell rheology, where the strain rate is proportional to stress raised to a power, n (i.e., stress exponent, Freed & Bürgmann, 2004 Nature). The fault interface and viscoelastic mud diapirs are discretized into triangular patches (surface elements) and cuboids (volume elements), respectively. The distributed stress field due to the deformation of patches and cuboids is calculated using semi-analytical Green's functions (see Dhar et al., 2022 GJI and references therein). We used the coseismic slip distribution proposed by Rau et al. (2022 Tectonophysics) to estimate the initial stress perturbation on the fault and mud diapirs. Our modeled mud diapirs have narrow crests (2 km width) and wider roots (6 km width), resembling the shape of mud-core anticlines (Chen et al., 2014 JAES). They are located at 1–5 km depth in the upper crust. We positioned numerous mud-diapir models following the spatial distribution of the mud-core anticlinal axis and Bouguer anomaly (Lo et al. 2016 Tectonophysics). We used pre-earthquake horizontal compression and vertical extension to simulate the diapiric stress regime. We used the grid search approach to optimize the uniformly assigned modeling parameters, including the fault friction parameter, power-law stress exponent, and Maxwell viscosity.
Our results suggest that the afterslip mainly contributes to the ground motion near the earthquake epicenter. However, the significant horizontal and uplift motion in the vast areas west of the Longchuan fault is mainly caused by the mud-diapir deformation. The effective viscosities (= strain rate/stress) of mud diapirs are in the order of ~1015 Pa s and 1020 Pa s at 1 day and 3 years after the earthquake, respectively. The transient effective viscosity of mud diapirs is one order less than that of the low-viscosity mantle (~1016 Pa s). Such low viscosity of mud diapirs and fast recovery over the postseismic period may correlate to the effect of over-pressurized mud in the mud diapirs. Our optimal model also preferred a higher order of power-law stress exponent (n = 5) compared to that of the viscoelastic lower crust-upper mantle (n = 3, Muto et al., 2019 Sci. Adv.; Dhar et al., 2022 GJI), which indicates the plastic behavior of mud in addition to their viscoelasticity. Understanding the mud-diapir rheology can provide valuable insights into the mechanisms of crustal deformation in SW Taiwan prior to, during, and after major earthquakes.