Remote sensing techniques to assess badlands dynamics: insights from a systematic review

Badlands are typical landforms that develop on unconsolidated sediments or poorly consolidated bedrock, with bare or sparse vegetation, generally characterized by high rates of erosion. These landscapes are vulnerable to dynamic changes driven by natural processes such as rainfall and tectonic processes, as well as anthropogenic factors including deforestation and land reclamation. The evolution of their interaction significantly influences resource management, particularly soil and water, and informs sustainable land-use planning strategies. Monitoring and analyzing badlands dynamics is crucial for understanding their downstream effects and mitigating natural and environmental hazards such as landslides, debris flows, piping and sediment delivery to rivers. Remote sensing (RS) technologies, from ground- to satellite-based, have emerged as valuable tools for assessing these processes due to their ability to provide data at high spatial and/or temporal resolutions over complex terrains. This article provides a systematic overview of recent advancements in RS techniques applied to badlands, highlighting their respective contributions across various environmental contexts. Starting from 516 papers retrieved from Web of Science and Scopus databases, the review synthesizes the main findings of 96 peer-reviewed studies selected by the use of Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) process. The majority of these studies (59%) were conducted in Europe, with significant contributions from Italy, Spain and France. Ground-based methods like Terrestrial Laser Scanning (TLS) remain invaluable for site-specific studies that focus on fine-scale processes such as rill formation and micro-landslides, while airborne laser scanning and aerial photography and photogrammetry, offer broader spatial coverage, facilitating the creation of geomorphological maps and the analysis of large-scale erosional features. Unmanned Aerial Vehicles (UAVs), emerging since 2011, have bridged the gap between ground precision field studies and aerial scalability, becoming essential for 3D mapping and erosion monitoring in inaccessible terrain. Satellite imagery is a leading tool due to its extensive spatial and temporal coverage, enhancing land-use change monitoring and erosion modeling capabilities. The study also emphasizes the importance of well-known tools such as Geographic Information Systems (GIS) to support the analysis of data and the creation of thematic maps (e.g. erosion susceptibility, land use/land cover, geotourism), while also recognizing the increasing role of Machine Learning (ML) in handling large and complex datasets, identifying hidden patterns, and supporting predictive analyses in environmental research. By providing a structured comparison of RS approaches in relation to their spatial scale, resolution, and applicability, this study contributes to a better understanding of their potential and limitations in badlands research, and offers a useful reference for designing future monitoring strategies.