Landslide Activity Maps Generation by Means of Persistent Scatterer Interferometry

Remote Sensing

Silvia Bianchini³  Gerardo Herrera²  Rosa Maria Mateos²  Davide Notti²  Inmaculada Garcia²  Oscar Mora¹ 

[1] Altamira Information, C/. Còrsega 381-387, E-08037 Barcelona, Spain; E-Mail:;Geohazards InSAR Laboratory and Modeling group (InSARlab), Geological Survey of Spain (IGME), C/. Alenza 1, E-28003 Madrid, Spain; E-Mails:;Earth Sciences Department, University of Firenze, Via La Pira 4, I-50121 Firenze, Italy; E-Mail:

关键词: SAR (Synthetic Aperture Radar);    Persistent Scatterer Interferometry;    landslides;   

DOI  :  10.3390/rs5126198

来源: mdpi

【 摘 要 】

In this paper a methodology is proposed to elaborate landslide activity maps through the use of PS (Persistent Scatterer) data. This is illustrated through the case study of Tramuntana Range in the island of Majorca (Spain), where ALOS (Advanced Land Observing Satellite) images have been processed through a Persistent Scatterer Interferometry (PSI) technique during the period of 2007–2010. The landslide activity map provides, for every monitored landslide, an assessment of the PS visibility according to the relief, land use, and satellite acquisition parameters. Landslide displacement measurements are projected along the steepest slope, in order to compare landslide velocities with different slope orientations. Additionally, a ground motion activity map is also generated, based on active PS clusters not included within any known landslide phenomenon, but even moving, potentially referred to unmapped landslides or triggered by other kinds of geomorphological processes. In the Tramuntana range, 42 landslides were identified as active, four as being potential to produce moderate damage, intersecting the road Ma-10, which represents the most important road of the island and, thus, the main element at risk. In order to attest the reliability of measured displacements to represent landslide dynamics, a confidence degree evaluation is proposed. In this test site, seven landslides exhibit a high confidence degree, medium for 93 of them, and low for 51. A low confidence degree was also attributed to 615 detected active clusters with a potential to cause moderate damage, as their mechanism of the triggering cause is unknown. From this total amount, 18 of them intersect the Ma-10, representing further potentially hazardous areas. The outcomes of this work reveal the usefulness of landslide activity maps for environmental planning activities, being exportable to other radar data and different geomorphological settings.

【 授权许可】

CC BY   
© 2013 by the authors; licensee MDPI, Basel, Switzerland