A statistical method to predict debris flow deposited volumes on a debris fan

The present paper aims to give a useful methodology to predict the debris v olume G(s) associated to a debris flow and deposited on the basin's termina l fan in the alpine region. Predicting the occurrence and the magnitude of a debris flow is often a very difficult task, as it is necessary to conside r the variability of the geological, hydraulic and pluviometric factors con nected to the beginning, propagation and deposition of a debris flow. These factors are not easily quantifiable, as well as their influence on the deb ris volumes deposited on the debris fan. Even if some useful mathematical ( and deterministic) relations have been proposed by some authors to predict the likely debris volumes, they do not seem to take into account the high v ariability of the influencing factors. For these reasons a statistical inve stigation of the relationship between Gs and the basin area A has been unde rtaken. By means of well known statistical methodologies and by using a sam ple of about two hundred data (referred to the alpine region), a new criter ion is proposed, in the present paper, for the prediction of a probable Gs value. For a given basin, this value is obtained on the basis of a statisti cal risk degree arbitrarily chosen. A probabilistic methodology is also pro posed, which, when results of an accurate survey are not available for a pa rticular basin, may provide a rational basis for a first hazard zonation an d the design of mitigative measures. (C) 2001 Elsevier Science Ltd. All rig hts reserved.