Landslide zoning over large areas from a sample inventory by means of scale-dependent terrain units

A procedure is proposed to produce landslide distribution zoning maps to be considered preparatory to susceptibility, hazard and risk zoning maps, based on 1) the results from a statistical multivariate analysis of a landslide inventory, which must be available for only a portion of the territory to be zoned, and 2) the use of appropriately defined terrain mapping units. The units are divided into terrain computational units (TCUs) and terrain zoning units (TZUs), whose size is related to the scale of zoning. The procedure comprises three phases: calibration, validation and prediction. The purpose of the prediction phase is the application of a calibrated and validated statistical model in a territory, previously recognized as viable on the basis of ‘a-priori applicability maps,’ for which no information is available regarding the distribution of landslides or where the information provided by the landslides inventory is unreliable or heterogeneous. The proposed procedure is applied to two case studies in southern Italy for the analysis and zoning of slow-moving landslides at 1:25,000 and 1:100,000 scales, respectively. The first case study illustrates the applicability of the procedure. The aim of the second case study is to address the part of the procedure related to the evaluation of the computational maps at the end of the calibration and validation phases.