Reactivation hazard mapping for ancient landslides in West Belgium

Several examples in western Europe have shown that, at least for deep-seated rotational slides, reactivation of formerly slipped masses is a more frequent phenomenon than the occurrence of new landslides, therefore representing a higher hazard. We selected a study area comprised of 13 landslides located in the Flemish Ardennes (West Belgium) and predicted the hazard related to scarp retreat. The scarp reactivations were identified from the comparison of DTMs produced for 1952 and 1996. Robust results were obtained with the Gamma operator of a fuzzy set approach and a combination of geomorphic,... Mehr ...

Verfasser: O. Dewitte
C.-J. Chung
A. Demoulin
Dokumenttyp: Artikel
Erscheinungsdatum: 2006
Reihe/Periodikum: Natural Hazards and Earth System Sciences, Vol 6, Iss 4, Pp 653-662 (2006)
Verlag/Hrsg.: Copernicus Publications
Schlagwörter: Environmental technology. Sanitary engineering / TD1-1066 / Geography. Anthropology. Recreation / G / Environmental sciences / GE1-350 / Geology / QE1-996.5
Sprache: Englisch
Permalink: https://search.fid-benelux.de/Record/base-27390977
Datenquelle: BASE; Originalkatalog
Powered By: BASE
Link(s) : https://doaj.org/article/1d3ac3d45b984480aa31be5268eeedc2

Several examples in western Europe have shown that, at least for deep-seated rotational slides, reactivation of formerly slipped masses is a more frequent phenomenon than the occurrence of new landslides, therefore representing a higher hazard. We selected a study area comprised of 13 landslides located in the Flemish Ardennes (West Belgium) and predicted the hazard related to scarp retreat. The scarp reactivations were identified from the comparison of DTMs produced for 1952 and 1996. Robust results were obtained with the Gamma operator of a fuzzy set approach and a combination of geomorphic, topographic and land use data. We built first a prediction model from the relations linking the 1952–1996 retreat events to the conditioning parameters of 1952. The prediction rate of the resulting susceptibility map is estimated by a cross-validation procedure. We then applied the statistics of this model to the data of 1996 in order to produce a susceptibility map responding to the present-day conditions. Finally, we estimated the conditional probabilities of occurrence of future reactivations for the period 1996–2036.