Landslide Susceptibility Mapping of Central and Western Greece, Combining NGI and WoE Methods, with Remote Sensing and Ground Truth Data

Abstract

The exploitation of remote sensing techniques has substantially improved pre- and post- disaster landslide management over the last decade. A variety of landslide susceptibility methods exists, with capabilities and limitations related to scale and spatial accuracy issues, as well as data availability. The Interferometric Synthetic Aperture Radar (InSAR) capabilities have significantly contributed to the detection, monitoring, and mapping of landslide phenomena. The present study aims to point out the contribution of InSAR data in landslide detection and to evaluate two different scale landslide models by comparing a heuristic to a statistical method for the rainfall-induced landslide hazard assessment. Aiming to include areas with both high and low landslide occurrence frequencies, the study area covers a large part of the Aetolia–Acarnania and Evritania prefectures, Central and Western Greece. The landslide susceptibility product provided from the weights of evidence (WoE) method proved more accurate, benefitting from the expert opinion and the landslide inventory. On the other hand, the Norwegian Geological Institute (NGI) methodology has the edge on its immediate implementation, with minimum data requirements. Finally, it was proved that using sequential SAR image acquisitions gives the benefit of an updated landslide inventory, resulting in the generation of, on request, updated landslide susceptibility maps.

Characterizing land deformations in Western Greece with satellite remote sensing observations and in situ ground monitoring

“…Substantial land displacements have been observed in Western Greece as a result of geophysical and environmental factors [3], [16]. The faults of the Corinthian Gulf contribute to these phenomena, as well as the occurrence of heavy rainfall [17] and possibly also human interventions (e.g., land cover changes) [18]. Several studies analyse land deformations in this region, using, for example, remote sensing data for PSI which are then cross-examined with measurements from permanent GNSS stations [19], [20].…”

“…Several studies analyse land deformations in this region, using, for example, remote sensing data for PSI which are then cross-examined with measurements from permanent GNSS stations [19], [20]. The use of LiCSAR (an automatic InSAR tool) for Small Baseline (SBAS) processing [21] has been also applied in the area, which helps analyzing displacement and rainfall patterns [18].…”

Complex analysis of earthquake deformations using SAR images: Examples from Croatia and Greece

“…The C-SAR instrument is able to measure land subsidence or uplifting through the formation of interferometric synthetic aperture radar (InSAR) images. SAR data processing is applied for many earthquakes [2][3][4] and landslides [5][6][7][8] in the Balkan Peninsula region.…”

Landslide Susceptibility Evaluation Based on Potential Disaster Identification and Ensemble Learning

“…To our knowledge, only one study carried out LSE using both potential and historical landslides. Kontoes et al [ 59 ] employed the characteristics of ground deformation detected by multi-temporal InSAR interferometry technique, geomorphology, slope angle, and land cover (vegetation deterioration) to determine the active landslides from 1992 to 2010. Both the past landslides and the identified active landslides constituted the landslide inventory, and the method of weights of evidence (WoE) was used to generate the LSE map.…”