Abstract
Landslides, masses of rock, earth, or debris moving down a slope, are geological events commonly observed in mountainous regions. Landslides can have profound societal impacts as they interact with both the natural environment and human settlements, often resulting in unexpected disasters. According to the United Nations office, between 1998 and
... read more
2017, landslides accounted for 18,414 fatalities and affected nearly 4.8 million individuals. To effectively reduce the damage caused by landslides, it is imperative to develop methodologies for spatial data collection and analysis of landslides. This includes the establishment of landslide inventories for purposes such as landslide dating (especially challenging in remote areas), the measurement of landslide surface displacement, and the assessment of landslide risks. To address these crucial aspects, this thesis aims to develop remote sensing-based methodologies and frameworks for landslide dating and landslide surface monitoring, and to provide spatial data for landslide risk assessment. These efforts are crucial for landslide prevention and mitigation for local communities and environmental resources.
Landslide dating is defined as achieving the most probable landslide occurrence date range through field measurements, global navigation satellite systems, terrestrial laser scanning, airborne laser scanning, photogrammetry or by analysis of remote sensing satellite imagery. In chapters 2 and 3 I developed novel landslide dating techniques using satellite image time series, and I explored their accuracy and potential for dating landslides.
Landslide monitoring involves the observation and recording of changes in landslide dynamics over a certain period, facilitated by installed equipment or through remote sensing technologies. In Chapter 4 I show how we can monitor the movement of slow moving landslides from very high resolution optical stereo imagery (Pléiades satellite), without the need for in-situ measurements.
In chapter 5 landslide risk is considered. Landslide risk comprises population risk and property risk. Population risk is the annual probability that individuals at risk will lose their life; and property risk is the potential economic damages to property. I conduct a semiquantitative landslide risk assessment at the community level at a scale of 1:10000 by combining multiple spatial data sets for a case study in western Hubei, China.
To conclude, landslide dating, monitoring and landslide risk assessment are complex research areas. In support of landslide prevention and mitigation, future efforts should focus on six key aspects: (1) landslide dating, (2) landslide monitoring, (3) landslide environmental impact, (4) landslide hazard and risk assessment (5) expert experience, and (6) remote sensing for multi-hazards.
show less