Testing new methodologies for landslide features extraction from high resolution topography

In recent years new remotely sensed technologies, such as airborne and terrestrial laser scanner, have improved the detail and the quality of topographic data. These offer the opportunity to provide topographical high-resolution and high-quality data over larger areas better than other technologies. A new generation of high resolution (1m) Digital Terrain Models (DTMs) are now available for different areas, and widely used by researchers, offering new opportunities for the scientific community. These data call for the development of the new generation of methodologies for objective extraction of geomorphic features, such as channel heads, channel networks, bank geometry, landslide scars, service roads, etc. A high resolution DTM, for example, is able to detect in detail the divergence/convergence areas related to unchannelized/channelized processes respect to a coarse DTM. In these last few years different are the studies that use the landscape curvature from high resolution topography as a useful measure for the interpretation of dominant sediment transport processes on the landscape. McKean and Roering (2004) used landform curvature studying landslide morphology and distribution. Lashermes at al. (2007) introduced a new methodology to detect thresholds in topographic curvature and slope-direction change for defining valleys and probable channelized portions of the valley. They used the Quantile-Quantile plot of local curvature defining the threshold curvature for objective channel extraction. Tarolli and Dalla Fontana (2009) used the landform curvature to assess the capability of high resolution topography in the recognition of convergent hollow morphology of surveyed channel heads. They suggested an objective methodology to extract the channel network based on threshold range identified as n-times the standard deviation of curvature. In this work, we test the performance of a new methodology for the objectively extraction of geomorphic features related to landsliding processes in order to provide then a semi-automatic method to recognise landslides and classify morphometric features (landform elements) in a complex mountainous terrain. The methodology is based on the detection of thresholds of landform curvature based on the approaches suggested by the work of Lashermes et al. (2007) and Tarolli and Dalla Fontana (2009). The analysis was conducted using a high resolution DTM and different smoothing factors for the landscape curvature calculation in order to set the more suitable curvature maps for the recognition of selected features. The study was conducted on a specific area located in the Eastern Italian Alps, where an accurate field survey on shallow landsliding processes, and a high quality set of both terrestrial and airborne laser scanner elevation data are available. A recent campaign effort has also provided new detailed data of field-mapped channel heads and landslide activity. The results revealed the great capability of adopted methodology in recognition of landslide features highlight the opportunities but also challenges in fully automated methodologies of geomorphic feature extraction.