Along the southeastern edge of Japan’s Honshu Island, the Philippine Sea plate is diving beneath the Eurasian plate at the Nankai subduction zone. Seismicity there has unleashed numerous devastating, megathrust earthquakes in a cycle typically characterized by pairs of closely spaced tremors that have occurred every 146 years, on average, since 1361. The region’s most recent earthquakes were the magnitude 8.1 Tonankai and the magnitude 8.4 Nankai events, which occurred in 1944 and 1946, respectively.
The seismic cycle is the process of repeatedly building stress on a fault over a long period of time that is rapidly released in a large earthquake. Previous studies have recognized that mantle flow plays a crucial role in the seismic cycle by relaxing postseismic stresses as well as straining the fault during the interseismic buildup to the next tremor. The surface deformation that accompanies this cycle can vary considerably depending on the mantle’s viscosity. But inferring this parameter has proven difficult because of a lack of geodetic data sets that span the centuries-long timescales of most seismic cycles.
The only place in the world where a record exists that is long enough to potentially record a complete, postseismic phase is in southern Japan, where leveling surveys as well as tide gauge and GPS measurements collectively span the period from 1890 to the present. Johnson and Tebo have harnessed this unique data set to model vertical deformation in the Nankai region from 1947 to 2015.
The results indicate that following the 1940s earthquakes, subsidence was centered about 250 kilometers inland of the Nankai trench, and its rate gradually decreased from 1947 until approximately 1995. By contrast, postseismic uplift occurred within a narrow belt stretching along Honshu’s southeastern coast, but it slowed and then switched to subsidence by the mid-1960s.
According to the authors’ two-dimensional modeling results, the uplift is best explained by afterslip along the coast, whereas the inland subsidence is strongly indicative of postseismic flow within the mantle wedge. The best fit to this pattern requires a viscosity on the order of 1019 Pascal seconds, which corresponds to a mantle relaxation time of 5 to 15 years. As the first study to clearly capture 5 decades of postseismic mantle flow, this research offers a crucial contribution toward improving our understanding of the cycle of seismicity that occurs in Earth’s subduction zones. (Journal of Geophysical Research: Solid Earth, https://doi.org/10.1029/2018JB016345, 2018)
—Terri Cook, Freelance Writer