Fjölrit RALA - 15.06.2004, Side 27

Landslide processes and andosols: the case study of the Campania region
F. Terribile1, A. Basile2, M. lammarino1, G. Mele2, S. Pepe1
1 Dipartimento di Scienze del Suolo, della Pianta e dell'Ambiente, Universitá di Napoli Federico II
2 Institute for Mediterranean Agricultural and Forest Systems, CNR, Ercolano (NA), Italy
The general goal of this research has been the study of soils connected to the initiation
mechanisms of debris flow landslide in the Campania region. More specifically the following
aims were investigated: 1) a regional scale analysis conceming the relationship between soil
types and landslide initiation mechanisms in Campania; 2) a local scale analysis on selected
sites where hydropedological landslide initiation processes were explored.
The study was conducted in the following step: (i) analysis of historical reports; (ii)
georeferencing of the main landslide detachment crown in Campania; (iii) morphological
description and sampling of selected profiles located in the detachment crown; (iv) chemical
and physical (hydrological) analysis; (v) hydrological physically based modelling for water
balance simulation at the time of landslides.
The analysis has clearly shown that the most important Campanian debris flow landshde
occurred in mountains very fertile forest (mostly chestnut) ecosystem with complex sequences
of Andosols both as surface and buried soils. Such soils have a high allophane and imogolite
content, high water retention properties and well developed tixotropy. These features along
with the high slope and/or the slope discontinuity (roads, cliff) seem to have played a very
important mle in the landslide initiation mechanisms.
From the outputs of the regional scale analysis, few selected sites (5 soil profiles) were
investigated with a detailed analysis of the hydropedological processes affecting landslide
initiation mechanisms. Bulk and undisturbed (steel cylinders) samples were collected in each
soil horizon for standard chemical analysis and for determining hydraulic properties by the
Wind method.
Physically based water balance models have been applied. More specifically, the
knowledge of water retention and hydraulic conductivity functions along with the knowledge
of the boundary conditions enable the solution of the Richard equation and therefore the
inference of water balance in the period of the landslide. Functional properties (i.e. water
content at specific depth, soil weight at specific depth) were then defined and obtained as an
integrated model output for estimating the susceptibility of the soil system to landslide
initiation processes.
16