Machine learning based landslide susceptibility mapping models and GB-SAR based landslide deformation monitoring systems: Growth and evolution

Ongoing landslides have wreaked havoc in various regions across the globe. This article presents a study of two forms of landslide monitoring viz; creation of Landslide Susceptibility Maps(LSMs) using machine learning and usage of Ground Based Synthetic Aperture Radar(GB-SAR). Landslide Susceptibility Mapping models generate an LSM for the given study area, which shows if the locations in the study area are prone to landslides or not. However, LSM is a post disaster management strategy. GB-SAR systems provide real-time data on the occurrence of landslides. Thus, a review of Ground based landslide deformation monitoring techniques is also presented in this article. Landslide deformation monitoring systems deal with identifying the changes occurring in a place that would trigger further landslides at the place where landslides have already occurred. Different techniques such as heuristic, analytic, and data driven statistical methods have been used in the existing literature for LSM creation. This study focuses mainly on the machine learning techniques used to create LSMs from the year 2000–2021. For each article in the literature, the metrics viz; region of study, country to which study area belongs, the spatial extent of the study area in square kilometers, principal triggering factors of landslide, sources used to collect data, type of landslide, number of landslide triggering factors used, number of landslide points in the landslide inventory, the algorithm used, evaluation parameter used to assess the performance of algorithms and values of these evaluation parameters have been noted. As a function of the type of landslides examined, the study region, and the fundamental triggering variables, we exhibit graphical depictions and discussions of similarities and contrasts discovered. The data analysis helps the researchers to identify future studies to be carried out in unexplored areas across the globe in the field of landslide monitoring. Furthermore, the study on GB-SAR technologies, which facilitates the formulation of better real-time techniques, than the state-of-the-art, has been discussed.