Landslide Geoanalytics Using LiDAR-derived Digital Elevation Models
MetadataShow full item record
Landslides are natural hazards that contribute to tremendous economic loss and result in fatalities if there is no well-prepared mitigation and planning. Assessing landslide hazard and optimizing quality to improve susceptibility maps with various contributing factors remain a challenge when working with various geospatial datasets. Also, the system of updating landslide inventories which identify geometry, deformation, and type of landslide with semi-automated computing processes in the Geographic Information System (GIS) can be flawed. This study explores landslide geoanalytics approaches combined with empirical approach and powerful analytics in the Zagros and Alborz Mountains of Iran. Light Detection And Ranging (LiDAR)-derived Digital Elevation Models (DEMs), Unmanned Aerial Vehicle (UAV) images, and Google Earth images are combined with the existing inventory dataset. GIS thematic data in conjunction with field observations are utilized along with geoanalytics approaches to accomplish the results. The purpose of this study is to explore the challenges and techniques of landslide investigations. The study is carried out by studying stream length-gradient (SL) index analysis in order to identify tectonic signatures. A correlation between the stream length-gradient index and the graded Dez River profile with slopes and landslides is investigated. By building on the previous study a quantitative approach for evaluating both spatial and temporal factors contributing to landslides for susceptibility mapping utilizing LiDAR-derived DEMs and the Probability Frequency Ratio (PFR) model is expanded. Furthermore, the purpose of this study is to create an algorithm and a software package in MATLAB for semi-automated geometric analysis to measure and determine the length, width, area, and volume of material displacement and flow direction, as well as the type of landslide. A classification method and taxonomy of landslides are explored in this study. LiDAR-derived DEMs and UAV images help to characterize landslide hazards, revise and update the inventory dataset, and validate the susceptibility model, geometric analysis, and landslide deformation. This study makes the following accomplishments and contributions: 1) Operational use of LiDAR-derived DEMs for landslide hazard assessment is estimated, which is a realistic ambition if we can continue to build on recent achievements; 2) While a steeper gradient could potentially be a signature for landslide identification, this study identifies the geospatial locations of high-gradient indices with potential to landslides; 3) An updated inventory dataset is achieved, this study indicates an improved landslide susceptibility map by implementing the PFR model compared to the existing data and previous studies in the same region. This study shows that the most effective factor is the lithology with 13.7% positive influence; and 4) This study builds a software package in MATLAB that can a) determine the type of landslide, b) calculate the area of a landslide polygon, c) determine and measure the length and width of a landslide, d) calculate the volume of material displacement and determine mass movement (i.e. deformation), and e) identify the flow direction of a landslide material movement. In addition to the contributions listed above, a class taxonomy of landslides is introduced in this study. The relative operating characteristic (ROC) curve method in conjunction with field observations and the inventory dataset are used to validate the accuracy of the PFR model. The validation of the result for susceptibility mapping accuracy is 92.59%. Further, the relative error method is applied to validate the performance of relative percentage of error of the selected landslides computing in the proposed software package. The relative percentage of error of the area, length, width, and volume is 0.16%, 1.67%, 0.30%, and 5.50% respectively, compared to ArcGIS. Marzan Abad and Chalus from Mazandaran Province of Iran and Madaling from Guizhou Province of China are used for validating the proposed algorithm.
Cite this version of the work
Saeid Pirasteh (2018). Landslide Geoanalytics Using LiDAR-derived Digital Elevation Models. UWSpace. http://hdl.handle.net/10012/12951