Realizzazione di una rete di monitoraggio geodetico della frana di Scopello (Sicilia nord-occidentale)

The Scopello area (north-western Sicily; fig. 1) is affected by several landslide phenomena (AGNESI et alii, 1995), involving both surface mass movements and, mainly, Deep-seated Gravitational Slope Deformation (DsGSD) phenomena (see GOGUEL, 1978; SAVAGE & VARNES, 1987; DRAMIS & SORRISO-VALVO, 1994; CANCELLI & CASAGLI, 1995; HUTCHINSON, 1995; SORRISO-VALVO, 1995; CROSTA, 1996). The phenomena involve the external margin of an overthrust plane (ABATE et alii, 1998a), where a rigid carbonate tectonic unit (Trias-Oligocene) overlies a ductile clayey substratum (Middle-Upper Miocene). The upper carbonate bodies (fig. 2) are marked by an inner imbricate structure and are intensely faulted and fractured, determining the presence of several disjointed blocks of various size, constituting favorable conditions for DsGSD movements to occur. Within the carbonate relief the DsGSD phenomena are responsible for the existence of double crests, depression alignments and river valleys truncated at their head, due to differential settlements. The outermost sectors of the carbonatic slopes (figs. 2-3), affected by block movements, lateral spreads and large topples or falls, are the sites of rotated and/or rafted blocks, great escarpments and open or closed trenches and cracks (AGNESI et alii, 1989). Finally, the marlyclayey slopes show diagnostic landforms of flow and/or slide type surface movements (AGNESI et alii, 1987). The causes of the landslides (AGNESI et alii, 2000; AGNESI et alii, 2003) are: superposition of carbonate rocks over marly-clayey rocks; undercutting processes induced by coastal erosion along the landslide foot/tip; Quaternary tectonic movements. As a geomorphological model has already been established for the landslide, a multidisciplinary study has been implemented, in order to define best both its geometric and kinematic characteristics. The final aim of this study is to gain a mathematical model which allows the phenomena to be correctly described, so that fore- and back- projection can be implemented. The project involves geophysical surveys (resistivity prospecting), geodetic monitoring measurements (GPS) as well as geomorphological, hydrogeological and geotechnical analysis and modeling. In the present paper all the study activities needed in order to implement a monitoring network, requiring a multidisciplinary approach in order to optimize its geometry, are discussed: detailed geomorphological analysis and modeling; resistivity geophysical survey; design and setting up of a geodetic monitoring network. In order to define a monitoring network for the landslide, a detailed geomorphological analysis has been carried out, so that bodies having a different expected kinematic response have been discriminated at a detailed scale (fig. 2). A resistivity survey was also carried out to verify the geometry of the ductile substratum in the area (fig. 4). The GPS network (fig. 2) consists on 27 3D vertices, 7 of which (fig. 2): SC01, SC02, SC03, SC04, SC05, SC06 e SC07) are located outside the landslide area, so that these can constitute a reference system, in respect to which the displacements of the other 20 vertices can be calculated. The coordinates of SC05 point have been calculated by connections to two vertices (TRAP and VALD in fig. 1) of the IGM95 geodetic network. The non-fixed vertices are located in areas characterized by: a) outermost areas of the margin of the carbonate slab about to undergo disarticulation (SC26 e SC27); b) isolated carbonate bodies moving toward the sea (SC08, SC09, SC10, SC15, SC16, SC17, SC18, SC19, SC24 e SC25); c) large landslide cemented debris masses (SC14, SC20, SC21, SC22 e SC23); d) marly-clay substratum (SC11, SC12 e SC13). The expected movement ranges from: very slow (mm/year) rotational -translational (a); varying velocity (mm/year-cm/year) translational, rotational-translational or rotational, depending on the depth to which «the roots» of the carbonate blocks extend inside the substratum (b); slow (cm/year) translational (c); slow to quite fast (dm/year) rotational-translational (d). A preliminary survey has been carried out, in order to calculate the coordinates of the vertices SC02, SC04, SC05, SC06 that are not subject to the landslide movements and to individuate a polygon including all the vertices (fig. 5). Data have been collected in two 3 hours sessions, using geodetic double frequency receivers with a rate of 15s. Data collected have been processed using the software SKI-Pro, with standard parameters but using a cut-off angle of 20° to lower the noise and the average multipath and compensated fixing the SC05 vertex. The redundancy of the this survey is equal to 2, which is enough to give a good accuracy and reliability to the network (CARDINALI et alii, 1993). The uncertainty in 2D and 1D of the fixed vertices (at a confidence level of 95%) is shown in fig. 5, suggesting a precision of less than 1 cm for their location. The survey on the whole net has been carried out using GPS 7 receivers (4 on the fixed points and 3 on as many sites positioned on the landslide body) with sessions 6-9 hours long. A total of 92 baselines were measured having a redundancy of 4. The survey was carried out between 20th and 25th of June 2000, according to the time schedule of fig. 6 and the quality of data tested according to ESTER & MEERTENS (1999), verifying acceptable values for the multipaths, both in L1 and L2 (fig. 7). Sixteen independent recording sessions with duration of 3-4 hours have been completed and the frequency distribution of the 92 baseline lengths (fig. 8) is characterized by an average of 1.9 km. The residuals for the 138 resolved triangles is of about 3 p.p.m. The uncertainty in 2D and 1D (at a confidence level of 95%) is presented in fig. 9, showing error ellipses having semiaxes less than 6 mm long and the minor to major axis ratio greater than 0.8; the uncertainty for the heights is greater by a factor two. Residuals after compensation for the 138 triangles (figs. 10a-b) are much lower and generally don’t exceed 1 p.p.m., thanks to a good geometry both of the network and of the satellite configuration during the recording sessions. The research carried out allowed refinement of the geomorphological model that had been defined in the past for the gravitational phenomena at Scopello, on the basis of which a geodetic monitoring network was defined. Data of the zero survey have been processed showing that the monitoring system is able to describe the kinematic evolution of the area with a precision suitable for DsGSD phenomena.

L’area di Scopello (Sicilia Nord-Occidentale) è da tempo oggetto di studi che, sulla base di osservazioni e rilievi geomorfologici, hanno consentito sia di riconoscere la presenza di frane superficiali e di fenomeni di deformazione gravitativa profonda di versante (DGPV), sia di ipotizzare cause e tipologie dei movimenti in atto, giungendo alla definizione di un primo modello interpretativo. I movimenti franosi si sviluppano lungo i settori costieri dell’area di Scopello, dove è a giorno un piano di sovrascorrimento del Miocene superiore, che vede la sovrapposizione di unità rigide, costituite da successioni carbonatiche meso-cenozoiche, su unità duttili, costituite, al tetto, da argille-marnose del Miocene medio-superiore. Le unità argillo-marnose sono interessate da processi di scalzamento ad opera dell’erosione marina, che determinano condizioni di instabilità nei pendii. L’assetto strutturale ereditato, l’erosione del mare ed i movimenti tettonici plio-quaternari sono le cause che hanno consentito l’attivazione di fenomeni di DGPV, per le cui velocità non sono tuttavia disponibili valutazioni quantitative di riferimento. Il presente lavoro riferisce sulla realizzazione di una rete di monitoraggio GPS dei fenomeni gravitativi in atto nell’era di Scopello e sull’esecuzione del «rilievo di zero», evidenziando come la precisione sul posizionamento 3-D dei suoi vertici sia adeguata alle ampiezze di massimo spostamento attese e, quindi, come la rete possa consentire, con rilievi successivi, il monitoraggio del campo degli spostamenti.