University of Twente Student Theses

Login

Unveiling earthquake legacy effects on hillslopes using InSAR

Sadhasivam, Nitheshnirmal (2022) Unveiling earthquake legacy effects on hillslopes using InSAR.

[img] PDF
Restricted to Restricted until 25 April 2024.

10MB
Full Text Status:Access to this publication is restricted
Abstract:Landslides, driven by seismic activities, have caused numerous fatalities and huge socio-economic losses globally. Many studies have suggested that intense seismic shaking not only triggers landslides co-seismically but also amplifies the post-seismic landslide activity, which is likely due to the decrease in shear strength of slope materials and/or disturbed hillslope geometry/hydrology. As a consequence, an elevated landslide susceptibility is observed in post-seismic periods. This practically means that a rainfall event with a given intensity and duration that does not trigger any landslides in pre-seismic periods becomes more effective after an earthquake and triggers landslides, which could be expressed as an increase in overall landslide susceptibility level of a given landscape. This concept is defined as earthquake legacy effect, which has been largely investigated merely by mapping inventories of rapidly failed slopes during the post-seismic phase. However, the increased post-seismic landslide activity has not been investigated in terms of deformation rates of slow-moving hillslopes, which exhibit for instance, millimeter-level deformation rate over time. In addition, understanding the evolution of hillslopes affected by intense seismic shaking helps to better evaluate post-seismic hazards and risks, as well as plan management and mitigation measures. This study aims to develop a novel systematic approach for detecting extremely slow-moving and very slow-moving hillslopes before and after the 2016 Mw 7.8 Kaikōura earthquake and monitoring their sub-meter evolution. In this research, I used freely available C-band Sentinel-1 Single Look Complex (SLC) Interferometric Wide (IW) mode dataset having a spatial resolution 5 × 20 m and polarisation of VV for Synthetic Aperture Radar Interferometry (InSAR) processing and deformation measurements extraction. Specifically, I examined 27 Sentinel-1 SAR scenes sensed before the earthquake and 63 images sensed following the event separately for extracting the deformation measurements in an area of about 2300 km2 using Persistent Scatterer Interferometry (PSI) approach. The analysis period of pre-Kaikōura phase is between 28 October 2014 and 10 November 2016, while the time window of post-earthquake phase is right after the earthquake mainshock from 16 November 2016 till 24 December 2018. A critical stability threshold value of ±10 mm/yr is defined on the extracted line-of-sight deformation velocity (VLOS) to detect active PS, which is further categorised into extremely slow-moving (±10 mm/yr ≥ VLOS < ±16 mm/yr) and very slow-moving (VLOS ≥ ± 16 mm/yr) hillslopes. Also, for the first time, I used Slope Units (SUs), which are terrain partitions associated with similar hydrological and geomorphological conditions, for the aggregation of active PS to identify extremely slow-moving and very slow-moving hillslopes. I then explored the dataset further and proposed a hillslope activity matrix for understanding the hillslope evolution after the impact of 2016 Kaikōura earthquake. Ultimately, I examined each category I defined in the proposed matrix via corresponding deformation time series in relation to daily precipitation. The results shows that in general there is an 130% absolute increase in the mean LOS deformation velocity during the post-Kaikōura phase compared to its pre-seismic counterpart. The regions that experienced higher ground shaking during the 2016 Kaikōura earthquake are observed to have larger deformations during the post-seismic period. In addition, most of the large negative deformations are observed to be associated with hillslope processes while high positive deformations are largely linked to the fluvial processes happening the study area. Comparing the pre-Kaikōura phase, there is a significant increase in the very slow-moving hillslopes, which chiefly concentrate around the rupture zone. Overall, I captured nine and 141 extremely slow-moving hillslopes during pre- and post- seismic phases, respectively. During the post- seismic phase I also identified 102 hillslopes showing very slow-movement. Based on these observations, this study proposed a hillslope activity matrix pointing out four hillslope evolution types: (i) inactive hillslope becoming active (Type I: SA), (ii) active hillslope remaining unaffected with changes in dynamics (Type II: AA), (iii) active hillslope that have become inactive (Type III: AS) and (iv) those hillslopes that are stable prior and following the earthquake (Type IV: SS). The hillslope activity matrix could be applied to other earthquake-affected areas to systematically and consistently examine hillslope evolution processes in post-seismic periods.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Subject:38 earth sciences
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:https://purl.utwente.nl/essays/93209
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page