Landslide detection and inventory by integrating LiDAR data in a GIS environment

In this work, a simple methodology is presented for processing high-resolution topographical data over wide areas. It is based on digital elevation model of differences (DEMoD) among high-resolution digital models (HRDEM) produced from lightdetection and ranging (LiDAR) data. Because these qualitative approaches based on HRDEMs can be affected by errors related to misalignment between different passes of the airborne sensor and errors in classifying points, a simplified strategy was undertaken for their semi-automatic correction and supervision for analyzing geomorphological changes. Besides, it became possible to detect, delineate, and classify a total of 47 natural landslides and 50 slope-cut failures over an area of 234km2 on the basis of the analysis of the LiDAR products (DEMs and DEMoD) and the orthophotography imagery inspection integrated in a geographical information system (GIS). Most of the displacements detected were probably generated during the winter of 2009–2010 when a new record of cumulative rainfall was reached. The displacement rate of these movements cannot be known with precision, but the minimum velocity that can be obtained is 0.3m/year regarding the period between the two data acquisitions carried out in November 2008 and July 2010. On the other hand, a comparison was made of the existing susceptibility maps with respect to this new inventory, which indicated greater landslide frequency in areas of moderate susceptibility levels. The influence of treating inventories at different temporal scales is discussed.