Landslide Detection
During my time as a research assistant in the Earth Observation and Remote Sensing group at ETH Zurich, I investigated possibilities for landslide detection with spaceborne remote sensing data and machine learning.
Quick Background
The three most common data types for landslide detection in remote sensing are: Synthetic Aperture Radar (SAR), Multispectral, and LiDAR data. Each data type comes with advantages and disadvantages.
SAR is not affected by clouds and DInSAR can even detect ground movement if the coherence between two acquisitions is high
Multispectral data consist of the sun’s reflection of the surface separated into bands by different wavelength intervals. This allows for the creation of spectral (change) profiles for each surface which helps in the detection process.
LiDAR data are often used to create high-resolution digital elevation models (DEM) which are then either used in susceptibility mapping or in DEM-differencing.