Afterslip, mostly aseismic creep on and on the margins of the seismic source fault, occurs not only after a large subduction megathrust earthquake but also rarely on an onshore surface rupture such as the 1966 and 2004 Parkfield, California, earthquakes on the San Andreas fault. Here we present evidence for afterslip as post-seismic continuous creep along the part of the coseismic surface rupture of the 16 April 2016 Kumamoto earthquake of Mw 7.0 (MJMA 7.3). We describe evident surface afterslip on five sites along the northermost Hinagu fault that experienced 30-65 cm coseismic right-lateral slip. Even though the post-seismic tape measurements are not highly accurate and contains approximately±1-3 cm uncertainty depending on site condition, maximum right-lateral displacement have reached ~20 cm during the first one year after the mainshock. Additionally, up to 5-cm cumulative right-lateral slip of a wall newly built in February 2017 across the coseismic rupture proves that the post-seismic creep has been lasting for at least one year, probably longer than three years. Together with the field survey, we also mapped one-year postseismic ground displacement from analysis of interferometric synthetic aperture radar (InSAR) images. About 2.5 cm contraction across the Hinagu fault from InSAR is a little short but mostly consistent with ~3.5 cm contraction estimated from ~20 cm right-lateral post-seismic slip on the N10°E trending rupture zone. Furthermore, our field observation is also in agreement with the timeseries of horizontal movement of a GNSS station ~2 km southeast of the fault zone. To examine the contribution of local aftershocks to the afterslip, we examined the cumulative moment release of all the aftershocks that reached ~5.1 x 1024 dyne-cm. Assuming a 10 km by 10 km fault plane, one could expect ~16 cm slip that is roughly equivalent to the amount of measured surface afterslip. However, the numerous aftershocks are located probably off the subsurface Hinagu fault so that shallow aseismic creeping would play an essential role for surface afterslip. Although our field measurements, InSAR and aftershock analysis cannot simply judge that the afterslip has been continuously loading to the unruptured sections of the Hinagu fault zone, significant aftershock activity has been still occurring beyond the southern edge of the afterslip zone. The Hinagu afterslip gives us clues not only to understand the postseismic fault behavior but also how to prevent from postseismic damage of fault-crossing structures and better assess the timing of restoration.
