Application of Object-Based Image Analysis for Detecting and Differentiating between Shallow Landslides and Debris Flows

Abstract

Mass movement mapping is essential for susceptibility, vulnerability and risk assessments. Various mapping approaches based on Earth observation (EO) data have been used to identify different types of hazards. Object-based image analysis (OBIA) has been employed for EO-based landslide mapping worldwide. The development and application of efficient methods for recognition and mapping are essential to create standards for landslide inventory mapping, notably in Brazil where landslides are a frequent natural hazard. This study aims to detect landslide features and differentiate them into shallow landslides and debris flows using a semi-automated OBIA approach. RapidEye satellite images (5 m) were analysed and the Normalized Difference Vegetation Index (NDVI) was calculated. A Digital Elevation Model (DEM) (12.5 m) and its derived products were integrated into the analysis to support the OBIA landslide mapping. The results show that the method is suitable for the recognition of this type of hazard and are potentially of use for local stakeholders and decision-makers in disaster management.