We document the 30 ka cumulative slip history and long-term slip vector azimuth on the northern Chelungpu fault based on a series of fault-bend folded alluvial terraces and draw quantitative relationships between geological structure, deformation observed from the geomorphology, and coseismic displacements during the 1999 M w = 7.6 Chi-Chi earthquake. In our study area, three main terrace levels show progressive folding by kink band migration in relation to the underlying fault geometry, forming a main N-S fold scarp up tõ 193 m high and secondary E-W scarps. Detailed analysis using 5 m resolution digital elevation model allows us to characterize the scarp morphology and quantify the deformation parameters, namely, terrace heights, fold scarp relief, and fold limb width and slope angle. The 3-D deformation of the highest terrace, dated by optically stimulated luminescence at 30.2 ± 4.0 ka, enables to simultaneously determine amplitude and azimuth of the long-term slip vector based on scarp relief. The long-term slip vector, oriented N338°± 6°, is found to parallel the Chi-Chi coseismic displacements in this area. Cumulative slip and dating results yield a constant slip rate of 17.7 ± 2.2 mm/a in the direction N338°± 6°. Late Quaternary shortening rates observed at four sites vary along strike in a similar way to Chi-Chi coseismic displacements. Together with the collinearity of long-term and coseismic slip vectors at our study site, this suggests that Chi-Chi earthquake is a characteristic earthquake for the Chelungpu thrust with recurrence interval~470 years. We also discuss implications for the regional and long-term distribution of shortening in the central Western Foothills.

Along the flanks of several valleys in the Swiss Alps, well-preserved fault scarps occur between 1900 and 2400 m altitude, which reveal uplift of the valley-side block relative to the mountain-side block. The height of these uphillfacing scarps varies between 0.5 m and more than 10 m along strike of the fault traces, which usually trend parallel to the valley axes. The formation of the scarps is generally attributed either to tectonic movements or gravitational slope instabilities. Here we combine field data and numerical experiments to show that the scarps may be of composite origin, i.e. that tectonic and gravitational processes as well as postglacial differential uplift may have contributed to their formation. Tectonic displacement may occur as the fault scarps run parallel to older tectonic faults. The tectonic component seems, however, to be minor as the studied valleys lack seismic activity. A large gravitational component, which is feasible owing to the steep dip of the schistosity and lithologic boundaries in the studied valleys, is indicated by the uneven morphology of the scarps, which is typical of slope movements. Postglacial differential uplift of the valley floor with respect to the summits provides a third feasible mechanism for scarp formation, as the scarps are postglacial in age and occur on the flanks of valleys that were filled with ice during the last glacial maximum. Finite-element experiments show that postglacial unloading and rebound can initiate slip on steeply dipping pre-existing weak zones and explain part of the observed scarp height. From our field and modelling results we conclude that the formation of uphill-facing scarps is primarily promoted by a steeply dipping schistosity striking parallel to the valley axes and, in addition, by mechanically weaker rocks in the valley with respect to the summits. Our findings imply that the identification of surface expressions related to active faults can be hindered by similar morphologic structures of non-tectonic origin.

Our study aims to characterize the post-glacial neotectonic activity by finding surface expressions of recently active tectonic faults. The central and western Swiss Alps were chosen as the study area because surface uplift rates are very high, indicating ongoing uplift of the external basement massifs. Moreover, the Valais area coincides with enhanced seismic activity. Active faults were searched by mapping lineaments on aerial photographs and subsequent field studies. Three main types of faults could be distinguished: gravitational faults (i.e. faults related to mass movements); tectonic faults; and composite faults (i.e. tectonic faults with a component of gravitational and post-glacial rebound-related reactivation). A large number of tectonic faults were found (over 1700), but only two unequivocally post-glacially active tectonic faults could be distinguished. Indications for their post-glacial (re-)activation are displaced Quaternary landforms or sediments. Large gravitational faults, as well as composite faults often correlate with deep-seated gravitational slope deformations (DSGSD). The latter occur mainly along valley slopes, particularly where a pervasive foliation strikes parallel to the valley. Fault orientations show correlations either with the regional main foliation (e.g. Aar and Gotthard massif), the orientation of valleys (e.g. Bedretto and Urseren valley), or pre-existing tectonic structures (e.g. faults parallel to joints that are perpendicular to the strike of major structures in the Helvetic nappes). Comparisons of fault orientations with orientations of nodal planes of earthquake focal mechanisms of the last 20 years show a poor indicative correlation. The central and western Swiss Alps host a large number of faults prone for reactivation in today's stress field. However, for most of these faults, no indications of their last phase of activity exist. The low number of unambiguously active tectonic faults suggests that the current strain is either predominantly aseismic or, alternatively, cumulated seismic moment is too low for producing surface rupture.