Estimation of the fault geometry of the Median Tectonic Line (MTL) in Eastern Shikoku, Southwestern Japan, via GNSS and InSAR time series analysis

Abstract The Median Tectonic Line (MTL) is one of the largest active fault systems in Japan. Whether the active MTL fault is north-dipping or near-vertical is still under debate. In this study, we estimated the MTL fault geometry in eastern Shikoku based on interseismic displacements obtained via Interferometric Synthetic Aperture Radar (InSAR) and Global Navigation Satellite System (GNSS). Detecting interseismic displacements in eastern Shikoku with InSAR is challenging because dense vegetation and steep topography increase InSAR observation noises. Therefore, we used Advanced Land Observing Satellite 2 data, which are less affected by vegetation. For InSAR processing, we applied state-of-the-art atmospheric correction methods and conducted small baseline subset -based InSAR time series analysis for two frames in eastern Shikoku. Obtained InSAR displacement velocities were consistent with GNSS velocities, with a root mean square error of 1.49 mm/year in line-of-sight direction. GNSS and InSAR velocity fields primarily reflected the plate coupling effect on the Nankai subduction zone. The displacement velocity discontinuity was not clear along the MTL, suggesting that the shallow part of the MTL was almost fully locked. We then conducted a parameter study for the MTL to estimate deep slip rate $${S}_{\text{MTL}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>S</mml:mi> <mml:mtext>MTL</mml:mtext> </mml:msub> </mml:math> , locking depth $${D}_{\text{MTL}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>D</mml:mi> <mml:mtext>MTL</mml:mtext> </mml:msub> </mml:math> , and dip angle $${\delta }_{\text{MTL}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>δ</mml:mi> <mml:mtext>MTL</mml:mtext> </mml:msub> </mml:math> using obtained velocity fields. Combinations of best-fit parameters ( $${S}_{\text{MTL}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>S</mml:mi> <mml:mtext>MTL</mml:mtext> </mml:msub> </mml:math> , $${D}_{\text{MTL}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>D</mml:mi> <mml:mtext>MTL</mml:mtext> </mml:msub> </mml:math> , $${\delta }_{\text{MTL}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>δ</mml:mi> <mml:mtext>MTL</mml:mtext> </mml:msub> </mml:math> ) which explained GNSS and InSAR observations with corresponding PH subduction models were (9.0 mm/year, 13 km, 30°) and (4.0 mm/year, 8 km, 35°), and corresponding root mean square values of GNSS and InSAR residual velocities were 1.20 and 0.57 mm/year, respectively. There was a trade-off between $${S}_{\text{MTL}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>S</mml:mi> <mml:mtext>MTL</mml:mtext> </mml:msub> </mml:math> and margin-parallel slip on the Nankai subduction zone, which prevented precise parameter estimation. Different combinations of the $${S}_{\text{MTL}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>S</mml:mi> <mml:mtext>MTL</mml:mtext> </mml:msub> </mml:math> , $${D}_{\text{MTL}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>D</mml:mi> <mml:mtext>MTL</mml:mtext> </mml:msub> </mml:math> and $${\delta }_{\text{MTL}}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>δ</mml:mi> <mml:mtext>MTL</mml:mtext> </mml:msub> </mml:math> produced spatially similar interseismic displacement fields due to the MTL, which made it difficult to find an optimal parameter combination. Nevertheless, our results generally supported the idea that the MTL is north-dipping rather than near-vertical, and shallow part of the MTL is almost fully locked in eastern Shikoku, which is consistent with a result from a previous study based on campaign GNSS observations. Graphical Abstract