Evaluating Quaternary activity versus inactivity on faults and folds using geomorphological mapping and trenching: Seismic hazard implications

Abstract The incorporation of active faults in seismic hazard analyses may have a significant impact on the feasibility, design and cost of major engineering projects (e.g., nuclear facilities, dams), especially when located in the site vicinity. The regulatory definition of active versus inactive fault is generally based on whether the fault has ruptured or not after a specific chronological bound (i.e. fault recency). This work presents a methodology, mainly based on geomorphological mapping and trenching, for determining whether specific faults can be considered as active or inactive. The approach is illustrated through the analysis of several faults located in the Spanish Pyrenees (Loiti, Leyre, La Trinidad, Ruesta faults). The 29 km long Loiti Thrust was included in the Neotectonic Map of Spain as a probable neotectonic structure. Previous works, based on geomorphological investigations, incorporated the 28 km long Leyre Thrust as a significant seismic source in a probabilistic seismic hazard analysis, which challenged the seismic design of nearby large dams. The production of detailed geomorphological strip maps along the faults allowed the recognition of specific sites where the faults are covered by Quaternary deposits. The establishment of chronosequences (pediments-terrace sequences) and the available geochronological data helped identifying the most adequate morpho-stratigraphic units for satisfactorily evaluating fault activity vs. inactivity. The excavation of trenches at the selected sites provided unambiguous information on the presence or lack of deformation in the Quaternary cover overlying the fault, and the origin of scarps (tectonic versus erosional). Trenches were also useful for collecting samples and reliably measuring the relative height of terraces overlain by thick colluvium. The evidence gathered by these methods were complemented with the numerical dating of non-deformed slope deposits covering a fault, the analysis of the longitudinal profiles of old pediment surfaces located in the proximity of a fault, the examination of a cave situated next to a fault searching for speleoseismological evidence, and regional geodetic and seismotectonic data (GPS measurements, earthquake focal mechanisms). The integration of all the data, and especially the trenches dug in non-deformed old terrace deposits (>100 ka) truncating the faults, indicates that the analysed faults can be considered inactive and that previous neotectonic postulations were based on non-valid geomorphological interpretations.