Active tectonics from UAS-HR-DSM combined with PSInSAR: Case example along the Longitudinal Valley - Eastern Taiwan

Taiwan result in the active collision of both Eurasian and Philippine Sea Plates characterized by an annual average convergence rate close to 10 cm.y^-1. The Longitudinal Valley is parallel and eastward of the Central (Backbone) Range which is made of metamorphic rocks, and is also situated to the west of the Coastal Range (of volcanic affinity). In between both, lay the Longitudinal Valley (125km long and N020°E trending) which behave as the active crustal suture zone. The latter presents both inter-seismic creeping displacement (Champenois et al., 2013, Deffontaines et al., 2018) and was hit by 7 major earthquakes of magnitudes larger than 5 during the last 70 years which highlights its high seismic hazards.

We combine herein a preseismic UAS survey (May 20, 2015) with one done immediately after the last large earthquake on the eastern Central Range (Oct 07, 2022). We therefore study both (1) the differences from a quantitative point of view; and (2) from a morpho-structural qualitative analysis point of view.

We acquired so many high-resolution photographs using several drones flying at 350 meters above the ground. After photogrammetric processing, we calculate both (1) a high-resolution Digital Elevation Model (UAS-HR-DSM) that takes into account buildings and vegetations, and deduce (2) a Digital Terrain Model (UAS-HR-DSM) corresponding to the ground. Our ground validation (GCP’s) leads us to get a 7cm planimetric resolution (X, Y) and below 40cm vertical accuracy.

This UAS-HR-DSM combined with field work and the preliminary PSInSAR (PALSAR-JAXA) processing led us to better characterize the active tectonic features through a detailed morphostructural analysis. It also permit us to map into much details the active structures and consequently to up-date the pre-existing geological mappings (e.g. CGS geological maps, Lin et al., 2009; Shyu et al., 2005, 2006, 2007, 2008). Then we up-date and combined our new structural scheme with geodetic data (levelings, GPS…) and PALSAR PSInSAR results acquired during the same monitoring time period to locate, characterize and quantify the active tectonic structures, taking into account previous works (e.g. Yu et al., 1997; Lee et al., 2008; Hsu et al., 2009; Huang et al., 2010…). We then precise structural geometries and some geological processes as well as the location of active folds and active faults during the PSInSAR monitoring time-period.

This may lead us to better constrain the seismic hazards and the earthquake cycles of the place.