Document Actions

ijege-14_01-deicas-et-alii.pdf

background image
51
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
DOI: 10.4408/IJEGE.2014-01.O-04
L
uca
DEI CAS, M
icheLe
aiLi, D
ennis
BOneTTi, F
rancescO
Ferrarini, F
rancescO
GiuDes
Arpa Lombardia - U.O. Centro Monitoraggio Geologico - Via del Gesù, 17 - 23100 Sondrio, Italy
Email: cmg@arpalombardia.it
VAL GENASCA LANDSLIDE (NORTH ITALY): AN EXAMPLE OF THE METHODOLOG
USED FOR THE IDENTIFICATION OF THE LANDSLIDE MAIN FEATURES
AND ITS MONITORING
EXTENDED ABSTRACT
Il Centro di Monitoraggio Geologico di ARPA si occupa del monitoraggio di 26 grandi frane individuate da Regione Lombardia come di
interesse regionale. In questo articolo ci si prefigge di descrivere la metodologia utilizzata per incrementare le conoscenza circa i movimenti
e poi giungere ad un controllo della frana denominata di Val Genasca. Il dissesto si trova nella parte inferiore della Valle Spluga e più pre-
cisamente all’interno del territorio comunale di San Giacomo Filippo in provincia di Sondrio.
Il metodo di lavoro è stato necessariamente modellato su step successivi che, a partire dalla realizzazione di un monitoraggio conoscitivo
(misure distometriche, periodiche campagne di misura topografiche e GPS), permettesse poi, ove si fosse manifestata l’esigenza, di svilup-
pare la rete sino a farla divenire un monitoraggio ai fini di allertamento e per l’attivazione dei piani di protezione civile. Tale gradualità di
intervento ha inoltre permesso che le conoscenze, via via acquisite sulla dinamica del dissesto, potessero meglio indirizzare gli approfon-
dimenti successivi anche in relazione alle disponibilità finanziarie annuali. In relazione a ciò, dopo una prima individuazione areale del
fenomeno (30.000 mq), si è provveduto ad uno studio finalizzato ad individuarne le principali caratteristiche cinematiche così da potervi
successivamente installare una adatta sensoristica che permettesse di seguire in tempo reale l’evoluzione del dissesto. A tal proposito la frat-
tura principale è stata monitorata mediante tre estensimetri a filo ed a monte dell’area di frana è stata installata una stazione meteorologica
dotata di pluviometro termometro e nivometro. In questa fase anche il monitoraggio topografico è stato automatizzato con rilevamento
dei movimenti di circa 20 mire ottiche poste sia all’interno che all’esterno dell’area in movimento. I dati così acquisiti sono quindi stati
trasmessi, con cadenza semi oraria, al fine dell’immediata verifica dei movimenti.
Per quanto riguarda i dati del sottosuolo, dopo una prima indagine geofisica, è stato possibile programmare una campagna di perforazioni
che, oltre a fornire i primi dati diretti sulla geologia del versante, ha permesso l’installazione di tubi inclinometrici e piezometrici. In det-
taglio a monte della nicchia è stato installato un tubo inclinometrico, spinto sino a 130 metri dal piano campagna, per misure periodiche
manuali ed un tubazione piezometrica strumentata con sensore per acquisizione in continuo del dato freatimetrico.
Un terzo sondaggio, ubicato nella parte medio alta dell’area in movimento (Fig. 6) , è stato spinto sino alla profondità di 103 metri da p.c. ed
è successivamente stato strumentato con colonna multiparametrica, modello DMS, con caratteristiche tali da poter sopportare deformazioni
elevate e rilevare oltre al movimento anche l’escursione di falda. Al fine di contenere al massimo i costi, ma sapendo di non poter escludere
alcuna profondità dalle possibili “zone di movimento”, la colonna è stata progettata con una alternanza di moduli DMS e moduli sterili.
Ovviamente gli uni collegati agli altri mediante “snodi” in grado di conservare e trasmettere il movimento. La colonna è stata alimentata
portando una linea in bassa tensione (12-24 V) fino al centro della frana.
Questa particolare configurazione (alternanza di modulo sensori con un modulo sterile), non ha permesso correlazioni dirette tra i movi-
menti di superficie e di profondità, ma ha comunque permesso l’identificazione precisa del fascia di taglio (con potenza di oltre 10 metri)
e la quantificazione dei movimenti. Per poter attivare un monitoraggio, con finalità di protezione civile, è infine stata commissionata
all’Università di Milano la modellazione del dissesto e lo studio delle prime soglie di allarme.
L’acquisizione e l’analisi dei primi 3 anni di dati della frana di Val Genasca ha permesso di individuare sette differenti periodi di accelera-
zione, che complessivamente hanno evidenziato movimenti di circa 9 metri (Fig. 13), intercalati da altrettanti periodi di stasi.
L’ultima accelerazione, durata circa 50 giorni fra la fine del 2013 e la prima metà del febbraio 2014, ha mostrato un movimento di circa 5
metri, sempre misurato lungo il vettore di massimo spostamento, in concomitanza di un significativo innalzamento della falda di versante.
L’analisi dei dati di profondità (Fig. 16-17) ha permesso di individuare come questi movimenti metrici si siano localizzati nelle prime de-
cine di metri da p.c., permettendo così di stimare in circa 500.000 metri cubi il volume di frana in movimento. I dati topografici (Fig. 22)
hanno infine permesso di definire un cinematismo di un tipico slittamento rotazionale con i valori di abbassamento maggiore nelle aree
altimetricamente più rilevate rispetto a quelli posizionati nella parte inferiore della frana.
background image
L. DEI CAS, M. AILI, D. BONETTI, F. FERRARINI & F. GIUDES
52
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
ABSTRACT
The Geological Monitoring Centre supervises deformations in
some major landslides in the Lombardy Region (Italy). This article
aims to present the methodology used to increase the knowledge
about the landslide of Genasca Valley, located in the municipality
of San Giacomo Filippo (Sondrio), and to monitor its evolution.
The approach to the problem has been developed through the im-
plementation of various activities in different stages.
The monitoring activities started in summer 2010 with a cog-
nitive monitoring based on distometric measurement, periodi-
cal topographic and GPS measurement. The slope areas (30.000
square meters) in movement were defined thanks to these activi-
ties and to the geological survey.
Second step was the set up of a real time monitoring with a me-
teorological station, three wire extensometers and one automatic
topographic system. A geophysical investigation and three drilling
boreholes were made to increase the knowledge of the subsurface.
Inclinometric casing and piezometric pipe for periodical
check were placed in the two drilling boreholes out of landslide
(above 10 meters upper the main scarp). A differential multipara-
metric system (DMS) for the depth monitoring inclinometric and
piezometric was installed in the borehole at the landslide centre.
In order to minimize the costs of the DMS column we designed
a particular configuration (alternation of sensors module with a ster-
ile module), that didn’t allow a direct correlations between the sur-
face and depth movements but it has however allowed the precise
identification of the shear band and the movements quantification.
In the meantime the University of Milan made a modeling of
the collapse in order to verify spread models and propose prelimi-
nary warning thresholds.
This monitoring network has permitted to acquire and ana-
lyze data collected for more than three years. During this period
Val Genasca landslide showed 7 different accelerations with
static period intervals. Last acceleration, dated winter 2013/14,
showed a 5 meters displacement, measured on the surface along
the maximum displaced vector.
The data analysis showed that the shear band affected a
thickness of about 15 meters and permitted to estimate in about
500.000 cubic meters the landslide volume in movement. The
topographic data allowed to define a kinematic motion of a typi-
cal rotational slip with the values of lowering higher in areas near
to the main scarp compared to those positioned in the lower part.
K
ey
words
: landslide, geological monitoring, Lombardy, Genasca
INTRODUCTION
The “Val Genasca” landslide is located in the hydrographic
right side of the valley in San Giacomo Filippo (Sondrio district).
The territory of San Giacomo Filippo delimits the Southern bor-
ders of the well known Vallespluga, the north-western side of Son-
drio district and, in general, the Lombardy entire area. (Fig. 1).
After some evidences of instability, in June 2010 the Local
Authorities asked ARPA Lombardia to verify and monitor the
situation on the slope. Below we will give account of the method
used in the study of the collapse and its first results.
GEOLOGICAL SETTING
The Spluga valley, where Val Genasca landslide is located, is
characterized by stacking of nappes belonging to the domain Pen-
nidic Superior. More specifically, the slope of the landslide is char-
acterized by rocks belonging to the Tambò formation stretching
with trend southeast-northwest for about 25-30 km with a thickness
of about 3.5 km (M
azzOLeni
, 2011). In the southern part of its base
the Tambò formation is intruded by granitoid Truzzo metagranite,
a Permian mass coeval to Roffna porphyry (M
arquer
et alii, 1998).
The lithologies on site are micaschist unit (at higher altitudes) and
the Unit of Truzzo Metagranite (in the lower part and on the valley
bottom). The morphology of the slope is mainly due to the instabil-
ity and deep gravitational movements (DGPV) of the area.
With particular reference to the collapse of Val Genasca, it was
possible to acquire some information covering the period 2003-2010
prior to the installation of the monitoring. More specifically, after
the rain event dated June 7
th
2003, the Lombardy Region, Direzione
Generale Territorio ed Urbanistica, carried out a survey (August 28
th
2003) well summarized by Dr. Dario Fossati: “some landslides with
a volume of some thousands of cubic meters moved in the lower part
of the river Genasca, Val Scura and Tettavacca about 650 m above
sea level”. Subsequently, with the report dated September 1st 2006,
the Lombardy Region updated the situation by informing that “an
area between the isoline 850 and the confluence with the river Liro
has been investigated. [...] In the range between 750 and 600 m above
sea level, long fractures were detected with heights up to 5-6 meters
which demonstrate the widening of a landslide (complex landslide
with falls, slides and debris flows) in constant evolution. The mate-
rial affected by the motion is on the whole estimated in 250,000-
350,000 cubic meters [...]” (c
eriani
, 2006). Further updates were
recorded in May 2009, when a new line of fracture is inspected: “it is
the reactivation of a landslip [...] it was possible to highlight the ret-
rogression of the landslide with the formation of a new crown from
altitude 824 m above sea level [...] The landslide is more marked on
Fig. 1
Map of the area
background image
VAL GENASCA LANDSLIDE (NORTH ITALY): AN EXAMPLE OF THE METHODOLOGY USED FOR THE IDENTIFICATION
OF THE LANDSLIDE MAIN FEATURES AND ITS MONITORING
53
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
the left side, with heights that reach two meters and involve a total
volume estimated in more than 200,000 cubic meters [...] Given the
steepness of the slope and the geomorphological situation, it is very
likely that the landslide may further evolve.” (c
eriani
et alii, 2009).
First report of ARPA (D
ei
c
as
, 2010) highlighted since the
beginning the presence of springs with variable flow, substan-
tially in line with the isoline of 630/650 meters above sea level,
located at the contact between bedrock and overlying deposit.
A summary of these events is attempted on the basis of aerial
photographs by the University of Milan (Prof. T. a
puani
), in rela-
tion to the agreement with ARPA Lombardia concerning activities of
geotechnical modelling and identification of critical thresholds, which
states that the first event of slippage has started between 1994 and
2000 (Fig. 2a-e). Subsequently, the instability has evolved involving
portions of the slope upstream and along the sides, showing in the
end a trend of evolution as retrogressive landslide and in enlargement.
Fig. 2a - Aerial view 1994: orthophoto Environment Ministry
Fig. 2b - Aerial view 1999-2000*: orthophoto IT 2000 (* photo Blom
CGR Spa)
Fig. 2c - Aerial view 2003*: orthophoto IT 2000 (* photo Blom CGR Spa)
Fig. 2d - Aerial view 2006*: orthophoto IT 2000 (* photo Blom CGR Spa)
Fig. 2e - Aerial view 2008: Aerial photo, DB Sondrio Municipality
background image
L. DEI CAS, M. AILI, D. BONETTI, F. FERRARINI & F. GIUDES
54
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
METHOD
As mentioned in the introduction, from summer 2010 ARPA
initiated a series of checks on the area aimed at deepening the
knowledge and then to the installation of a real-time monitoring.
The method of work was necessarily modelled on steps, start-
ing from the creation of a monitoring of knowledge (D
ei
c
as
,
2011), and then allow (steps 1 and 2), where it was revealed the
need to develop the network until to make it become a monitoring
for the purpose of alerting and for the activation of civil protection
plans (steps 3, 4, 5 and 6). This gradual approach has also allowed
gradually acquired knowledge on the dynamics of the collapse
could better address the following insights and allowed to develop
intervention in line with the available budget (the total amount has
been allocated over several years). In relation to that, after an initial
identification of areal phenomenon, it was decided to identify the
main kinematic characteristics so that can be possible to later install
a suitable sensors that allow to follow in real time the evolution
of the landslide. The knowledge of the subsurface, modeling and
comparative analysis of the main movements were the last step of
the project. In detail, the work can then be divided into six phases:
1. Geological survey and main findings of fractures
2. Installation of cognitive monitoring:
a. Control of movements of the main fractures through disto-
metric measures
b. Control of global movements through topographic measures
c. Control of back movements with GPS
3. Installation of real time monitoring:
a. Installation of a meteorological station
b. Installation of 3 extensometer
c. Installation of automatic topographic system (Figg. 3 e 4)
4. Advancing geological knowledge of the subsurface:
a. Geophysics
b. Drilling and monitoring implementation design
c. Drilling and coring
d. Analysis of samples taken
5. Landslide Modeling and first threshold values
1
:
a. Creation of detailed topography
b. Modeling of landslide evolution
c. First alert thresholds based on the instrumentation on the
surface
6. Installation of deep monitoring:
a. DMS, piezometer and inclinometers
A PTZ camera for visual inspection by remote area of the
landslide was installed in addition to what has been done so far.
The system was designed to activate and record when morphol-
ogy changes are detected (change detection), considering the im-
portance of having visual information as well.
1
It is evident that the process of modelling and definition of critical thresholds is by its
nature a procedure always updating as the sets of ever-increasing movement data are
acquired. In particular, an update of the modelling and thresholds will be compulsory
as soon as it’s possible to have a significant number of data acquired by the sensors.
SOLUTIONS IN PLANNING / IMPLEMENTATION OF THE MONI-
TORING NETWORK
STEPS 1 and 2: after a careful survey of the slope movement
it was possible to place 5 distometric chains to control the move-
ments of the upper side, located in the middle part of the main
scarp and on both sides, left and right parts of the landslide. These
distometric chains, positioned along the line of maximum move-
ment (substantially N-NE, orthogonally to the individual micro
areas isoline), have taken into account the possibility of measur-
ing in a different way both the movements of the main scarp (if
present) and the back fractures. At this stage, with the support of
Mountain Guides, a series (Fig. 4) of optical reflectors (both inside
Fig. 3 - Automatic topographic station
Fig. 4 - Topographic target in landslide
background image
VAL GENASCA LANDSLIDE (NORTH ITALY): AN EXAMPLE OF THE METHODOLOGY USED FOR THE IDENTIFICATION
OF THE LANDSLIDE MAIN FEATURES AND ITS MONITORING
55
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
and outside the estimated maximum movement area) were placed
to allow control from the opposite side. A concrete milestone has
been built on the opposite side for the execution of topographi-
cal measurements, it has been specifically anchored to the rock,
and made integral with the anchor plate for the positioning of the
measurement equipment. Two master references were placed out-
side the landslide, with the aim to achieve the maximum measure-
ment accuracy. Finally, some targets for recurrent manual meas-
urements with GPS dual frequency detector (rover) were placed in
the rock on the back of the main scarp, in order to quantitatively
assess back movement of the landslide. The GPS reference was lo-
cated on the opposite side, close to the topography measure build-
ing. The accuracy expected from these measures, albeit lower than
that of topographic measurements, allows, once a topographic re-
flector is also equipped for GPS measurements, a mutual verifica-
tion between the data acquired with different techniques (Fig. 5).
STEP 3: this network and 14 manual measurement campaigns
in the period August 2010-January 2011 allowed the technicians
of the CMG to precisely define the area affected by movements
and to estimate their speed. In connection with these results, it was
possible to identify the most significant points of the entire move-
ment and to implement a extensometric monitoring in real time.
In step 3 only 3 of the 8 distometric bases previously posi-
tioned were automated (this was meant to limit costs without los-
ing any significant information). The electrical signal of the exten-
someter, transmitted through cables to the data logger positioned
at the back of the scarp, is transmitted via GPRS to the ARPA
server. A sensor for the measurement of snow level on the ground,
a thermometer and a rain gauge were also installed on the same
station for a complete monitoring of meteorological variables.
The location of topographic measurement has been structured
to perform continuous measurements for a more extensive control
of the landslide surface. In this regard, after bringing the power line
to the point of measurement, a shelter for the station with an anti-
theft system (Fig. 3) was built. A series of measurement points on
the area of the landslide have been implemented so as to be able to
monitor, in addition to the whole landslide, even the areas nearby.
STEP 4: Fourth step, carried out in parallel with the third was
the planning and execution of a geophysical survey required to
identify the bedrock beneath the landslide body. The bibliography
didn’t permit to define any kind of geological data derived from
direct surveys (drilling) or indirect (geophysics), which allow to
produce a model of the slope.
In relation to this, and to the assumed considerable depth to
investigate, a prospecting seismic refraction tomography data
processing was chosen.
The analysis of the results obtained with the seismic survey
has allowed to plan a deep geological survey and subsequent
monitoring. This project was developed taking into account that
it had to meet multiple needs summarized as follows:
a) To have at least two stratigraphies on which to calibrate the
geoseismic survey;
b) To have three points of measurement to study the groundwa-
ter piezometric level;
c) To have three holes to check where the deep movements are
(inclinometer measures). Lacking any information about the
behaviour of the deposits identified by geoseismic, these ver-
ticals must reach the bedrock;
d) To have the ability to measure and transmit in real time the
most depth data acquired;
Fig. 5 - comparison between optical and GPS target
background image
L. DEI CAS, M. AILI, D. BONETTI, F. FERRARINI & F. GIUDES
56
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
e) To have the best guarantee of long-term accessibility to the
vertical drills taking into account that the landslide had deci-
metric movements in a short period (30 days);
f) To consider the nature of places and therefore the impossibi-
lity to reach the landslide with a street;
g) To consider the non-optimal slope sun exposure.
In relation to the above requirements, five drillings were
planned (two with soil coring and three with soil destruction) lo-
cated immediately upstream of the main scarp (the first two), in
the middle of the landslide (the third) and in proximity of the low-
er scarp (the last two). In order to bring the equipment needed for
the execution of the perforations to the landslide site, it is planned
to repeat the process of disassembly / reassembly borehole drill-
ing machine and subsequent transport on the survey points by
helicopter. The two vertical drills upstream of the main and lower
scarps are designed for the laying of the inclinometer tube and
piezometer tubes for manual measurements.
The drilling placed at the center of the landslide body was
instead designed with PVC pipe where to drop a multiparametric
column (DMS G
iuFFreDi
, 2003) with features so as to withstand
high strain. In order to limit the maximum costs, but knowing that
the movement can be at any depth, the column has been designed
with an alternation of DMS modules and empty modules. Obvi-
ously, each part is connected with the others in order to preserve
and transmit the movement. The column was connected to a low
voltage (12-24 V) electric line starting from upstream of the main
scarp to the center of the landslide.
STEP 5: simultaneously to the stages of investigation, the
University of Milan was asked to prepare a first model to iden-
tify the consequences of a possible landslide spread on the val-
ley. The University has also been in charge of a detailed topo-
graphical slope construction needed to run landslide models.
Prof. T Apuani, the University project manager, carried out a
modeling after the creation of a digital terrain model based on
the topographical construction of the slope, using the computer
code Flow2D. The choice of the University was to model the
movement of the masses potentially unstable as a debris flow
with a high concentration of solid, characterized by high viscos-
ity and a yield strength.
STEP 6: drillings showed an underestimation of geophysics
in relation to the depth of the bedrock. In relation to this, the incli-
nometer tube upstream of the high scarp has been realized up to a
depth of 130 meters and the DMS column to a depth of 102 me-
ters from ground level. These economically burdensome design
changes resulted in the abandonment of the holes placed at lower
altitudes; a total of 3 drillings have been carried out with the pos-
sibility of having two distinct measurement points for both the
water level and for the possible deep movements (respectively,
one manually upstream of the main scarp and one automatically
in the middle of the landslide).
Fig. 6 - Map of installed monitoring network
background image
VAL GENASCA LANDSLIDE (NORTH ITALY): AN EXAMPLE OF THE METHODOLOGY USED FOR THE IDENTIFICATION
OF THE LANDSLIDE MAIN FEATURES AND ITS MONITORING
57
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
DATA
After about 44 months of investigation and acquisition of mo-
tion data, a great number of useful information for the interpreta-
tion of the phenomenon have been collected. The principal will
be explained below:
GEOPHYSICS
Two measure lines located along the middle profile of the
landslide and the around median high isoline were carried out
in July 2011. The results of the two profiles (Fig. 7) allowed the
identification of four geoseismic units with different features.
STRATIGRAPHY
ARPA has developed a drilling project that would allow to in-
vestigate and subsequently equip with inclinometer tubes, piezo-
metric tubes and DMS column, some vertical drills upstream and
within the landslide. The stratigraphy, obtained from continuous
core drilling, relative to the survey site upstream of the main scarp
of the landslide (survey up to 130 meters from the ground level)
and within the landslide (survey up to 102 meters from ground lev-
el) showed a powerful non-consolidated coverage both upstream
and downstream of the scarp where a layer underlies, interpreted
as an ancient slide deposit, consisting primarily of crushed rock
and broken blocks with interbedded sand and silt. In both holes,
before reaching the weathered rock (consisting of mica schists and
altered paragneiss) and the bedrock, there is a level, with a bigger
thickness in the landslide, which consists mainly of silt (Fig. 8).
MOVEMENT DATA OUTSIDE MAIN SCARP
A series of GPS and optical targets for topographic measure-
ments were placed upstream of the main scarp from August 2010
in order to establish the area with greater movement and to quantify
any back movements of the landslide. GPS points 1, 2 and 3 and the
Fig. 7 - Seismic profile along the
slope line of the landslide
(CIS Geofisica s.r.l. 2011)
Fig. 9 - Stratigraphy of center landslide drilling (In.Co srl 2013)
background image
L. DEI CAS, M. AILI, D. BONETTI, F. FERRARINI & F. GIUDES
58
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
optical targets MO9, MO9A and MO13 (these last two positioned
in December 2012) were positioned upstream the upper scarp.
Downstream of the low scarp, between 610 m and 640 m
above sea level, optical targets MO1, MO2, MO3 and MO4 have
been positioned in order to monitor any movement of the out-
cropping bedrock (Truzzo Metagranite).
With reference to the upstream side of the landslide, after over
10 measurements GPS points have not shown displacements great-
er than the range of accuracy that can be expected from a periodic
manual measurement system located in this area with static process.
The optical targets positioned a few meters upstream of the
scarp show very low annual movements (Fig. 10).
Also points placed to control the outcropping bedrock below
the area in greater movement do not show any appreciable move-
ments (Fig. 11).
In December of 2012 some targets (MO10, MO12 and MO11)
were also placed to control any widening of the area affected by
the movement. Hereinafter the graph (Fig. 12).
MOVEMENT DATA INSIDE THE MAIN SCARP
Surface movements can be well explained by analysing the
topographical displacement vector determined for each target po-
sitioned in the landslide (Fig. 13).
Additional data are supplied by the distometric measures on
the opening of the main fracture delimiting, without interruption,
the landslide area underlying the upper scarp (30,000 sq. m).
In the figure 14, measures derived from the 3 main lines
(located in the altimetrically higher part of the main scarp and
Fig. 10 - Movement of topographic target upstream the main scarp
Fig. 8 - Silty layer found in the center landslide drilling at depth be-
tween 67.00 and 82.50 m
background image
VAL GENASCA LANDSLIDE (NORTH ITALY): AN EXAMPLE OF THE METHODOLOGY USED FOR THE IDENTIFICATION
OF THE LANDSLIDE MAIN FEATURES AND ITS MONITORING
59
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
symmetrically in the middle part of the right and left sides of the
fracture) are correlated with the cumulative rainfall as recorded
by the rain gauge, which is located upstream of the landslide, and
it is active since February 2011.
From the analysis of the chart above, would seem evident is
some correlation between the value of cumulated precipitation
within 30 days and the beginning of the events of acceleration.
The event of December 2013, as well as being the most par-
Fig. 11 - Movement of topographic target downstream the lower scarp
Fig. 12 - Movement of topographic target on the lower scarp sides
background image
L. DEI CAS, M. AILI, D. BONETTI, F. FERRARINI & F. GIUDES
60
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
oxysmal, is related to the passing of 200 mm of rain fallen in two
days prior to the start of acceleration.
FIRST INCLINOMETER AND PIEZOMETRIC DATA
Since August 2013 it was possible to acquire the first data
related to the depth movement and to the increase of the water
level. Although the period is very limited, it is interesting to make
Fig. 13 - Movement of topographic target inside landslide
Fig. 14 - Extensometric data (left ordinate axis) and rainfall in 30 days (right ordinate axis) vs. time
background image
VAL GENASCA LANDSLIDE (NORTH ITALY): AN EXAMPLE OF THE METHODOLOGY USED FOR THE IDENTIFICATION
OF THE LANDSLIDE MAIN FEATURES AND ITS MONITORING
61
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
some remarks. First of all, we observe that in a rather dry period
(in September and October), the water level is constant, close to
the DMS, at a depth of about -37 m to ground level. The heavy
rainfall in late October (205 mm in 7 days) had led to a rise (about
1 meter), which occurs the 6th day after the intense precipita-
tion. For more than 1 month the water level remains at a depth
of -36 m to ground level; when the water table lowers to share
-37 m to ground level, there are very serious precipitation on the
area, with around 210 mm of rain fallen between December 24
th
and 25
th
2013: next rise is about 2 meters. In this last period, the
delay between water level raising and rainfall appears almost
non-existent, with the water level which begins to rise during the
precipitation (unlike October when the elevation is with 6 days of
delay). Also in this case the water remains at an high level well
beyond the period of rain.
The rise of about 2 meters from October level is also con-
firmed by measurements made in the manual piezometric tube
placed upstream of the landslide.
As for the deep movement, the graph shows an acceleration
during the period October 2013 - February 2014 (Fig. 17); it is
interesting to notice that during the period December 26
th
2013
- February 8
th
2014 after a first phase (December 26
th
to January
4
th
), characterized by movements only on the top, down to -30 m
from the ground level, with a total displacement of about 35 cm,
starting from January 5th the deformation affected, albeit with
lower gradient, even the deepest part. In the first 8 days of Febru-
Fig. 15 - Water level measured (left ordinate axis) by DMS and rainfall in 30 days (right ordinate axis) vs. time
Fig. 16 - DMS polar diagram
background image
L. DEI CAS, M. AILI, D. BONETTI, F. FERRARINI & F. GIUDES
62
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
Fig. 17 - DMS Cumulative displacement DMS from October 2013 to
February 2014
Fig. 18 - I3 inclinometric cumulative displacement
background image
VAL GENASCA LANDSLIDE (NORTH ITALY): AN EXAMPLE OF THE METHODOLOGY USED FOR THE IDENTIFICATION
OF THE LANDSLIDE MAIN FEATURES AND ITS MONITORING
63
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
INTERPRETATION AND DISCUSSION
As it can be deduced from the data reported in the previous
paragraph, the moving area is estimated at 30,000 square meters
with the crown at 830 m above sea level and foot at 640 m above
sea level. In about 3 years of active monitoring, landslide showed
7 accelerations, which have affected simultaneously the entire
area of 30,000 square meters, with periods of stasis or even static.
In this interval the overall displacements measured at the sur-
face have been of about 9 meters along the surface vector of max-
imum displacement. Half of this movement has occurred within
50 days of the winter 2013/2014, while, if we consider a longer
interval, we can say that 80% of the movement (7 meters) was
observed between May 2013 and February 2014.
As shown in previous sections, all seven accelerations oc-
curred as a result of large amounts of rain fallen on the area. The
accelerations have had a velocity increasing over time: while the
first 4 accelerations and the sixth had comparable velocity, the
fifth and the seventh have been completely out of scale (Fig. 21).
From the analysis so far, the motion is detected, at the DMS,
at a depth of 13 to 27 meters from ground level (Fig. 17).
It is not a conventional motion surface with a thickness of 2
or 3 meters, but in this case the shear band affects a thickness of
about 15 meters (Fig. 23).
The topographic data allow to define a kinematic motion of
a typical rotational slip (c
ruDen
& V
arnes
, 1996): the values
ary, while the remaining part of the column showed almost zero
movements, a movement of about 13 cm has involved a band be-
tween 90 and 95 meters from ground level. The movement of this
small section of the column has assumed a direction orthogonal to
the movement (55° North) found by the whole column so far and,
also in this period, from the top of the column (Fig. 16). It is pos-
sible that the deepest movement observed from February involve
the lower part of the inclinometric column, installed in bedrock.
It is important to report the graphs produced on the basis of
the first inclinometer measures carried out by technicians of ARPA
CMG (Fig. 18) in the hole called I3, upstream of the upper scarp.
During the acceleration of November 2013 piezometric data
were also available. It is observed that the movement (both in
surface and in depth) occurs, after a sudden acceleration during
the precipitation, with a delay of about 6/7 days from more in-
tense precipitations (Fig. 14) and continues for a period of about
25 days compared to decrease of cumulative rainfall values. The
piezometric data shows that the beginning of the movement coin-
cides with the raising of the water level (Fig. 19).
Even the acceleration of December 2013-February 2014
is observed in conjunction to the raising of water level. The
slowdown began in the second half of February (after the break
of the piezometric module of DMS); the piezometer PZ3 up-
stream of the landslide shows the lowering of the water level at
deceleration (Fig. 20).
Fig. 19 - Water level (left ordinate axis) vs. extensometric data (right ordinate axis) from October 19 to December 18 2013
background image
L. DEI CAS, M. AILI, D. BONETTI, F. FERRARINI & F. GIUDES
64
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
The specific configuration (alternation of a module DMS with
a sterile module), used to minimize the costs of the DMS column,
does not allow direct correlations between the surface movement,
detected from the target optical MO8D, and the cumulative shift
detected in depth (Fig. 24). The alternation of the modules has,
however, allowed us to identify with precision the shear band and
to quantify movements.
Despite this direct correlation is not possible, and it is evi-
dent an underestimation of the displacement given by the DMS
column in relation to the target placed at the head of the hole
(MO8d), however, it is currently possible to exclude movements
involving the bedrock at a depth greater than that of the incli-
nometer column, as the data of the optical targets, on the sides
and upstream of the main fracture, show no significant movement
(only MO10 shows recall surface movements - Fig. 12).
DMS data show deep movements with displacements of few
meters (the column was interrupted in conjunction with a shift in
the measured area of approximately 4.5 meters).
The analysis of the collected data, in addition to the geophysi-
cal and stratigraphic available data, has allowed the construction
of a geological model that can be well outlined by the section of
Fig. 23. An area of 30,000 square meters, bounded upstream and
on the sides by the main fracture and downstream by the emer-
gence of some springs, moves with subsequent gradually higher
accelerations. A sliding surface, easily recognizable in drilling
of lowering are higher in the areas closer to the crown (MO8a,
MO8b, MO8c, MO8d) compared to those positioned in the lower
part (MO5, MO5a, MO6, MO6a) (Fig. 22).
Fig. 20 - Water level (left ordinate axis) (measured by DMS, after his break by PZ3) vs. extensometric data (right ordinate axis) from December 24
2013 to February 20 2014
Fig. 21 - E3 gradient in the 7 measured events
background image
VAL GENASCA LANDSLIDE (NORTH ITALY): AN EXAMPLE OF THE METHODOLOGY USED FOR THE IDENTIFICATION
OF THE LANDSLIDE MAIN FEATURES AND ITS MONITORING
65
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
Fig. 22 - Topographic values of lowering
Fig. 23 - Interpretative geological section
background image
L. DEI CAS, M. AILI, D. BONETTI, F. FERRARINI & F. GIUDES
66
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
estimated at 500,000 cubic meters that, in conjunction with a rise
in the water table of about 2 m, showed movement during the
period December 2013- February 2014. At depths greater than
30 m and up to the bedrock, substantial movements were how-
ever observed (Fig. 17) albeit considerably lower than the upper
part. Also in relation to the observation on the different direction
shown by the deepest part of the column, it will be important to
continue with deep monitoring activities.
The evolution of the instability can occur with subsequent
acceleration cycles (concurrent with rainfall events that cause a
rise in the water level), which, without leading to the sudden de-
tachment of the entire mass, cause a progressive slide of the lower
crown as happened in May 2013 and in the winter of 2013/2014.
One paroxysmal scenario, but that cannot be excluded, is the si-
multaneous sliding of the entire mass (estimated, according to the
most recent event, in about 500,000 cubic meters) with subse-
quent formation of a dam (the model proposed by the University
shows thickness of accumulation of 30 meters) in the bed of Liro
torrent. This scenario needs to be deeply examined for the inter-
actions with civil protection planning.
ACKNOWLEDGEMENTS
Thanks to Mario Lovisolo for the DMS’s planning and the
discussion on the DMS data.
stratigraphy, bounded downstream from contact with the bedrock
close to the water springs (Fig. 25) is present on the bottom. On
the basis of this model, the volume of paleo-landslide has been
Fig. 24 - MO8d topographic target movement vs. DMS total shift (from 12.18.13 to 01.16.14)
Fig. 25 - Frontal view of landslide; snow highlights springs in the
lower sector
background image
VAL GENASCA LANDSLIDE (NORTH ITALY): AN EXAMPLE OF THE METHODOLOGY USED FOR THE IDENTIFICATION
OF THE LANDSLIDE MAIN FEATURES AND ITS MONITORING
67
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
REFERENCES
a
puani
T., M
aseTTi
M. & T
anTarDini
D. (2013) - Attività di modellazione geotecnica e individuazione delle soglie di criticità nelle aree di Gera Lario e Val
Genasca Milano, Italia.
c
ruDen
D.M. & V
arnes
D.J. (1996) - Landslides Types and Processes. T
urner
a.K. & s
chusTer
R.L. (e
Ds
.). Landslides: Investigation and Mitigation.
Transportation Research Board Special Report 247. National Academy Press, WA, 36-75.
D
unnicLiFF
J. (1993) - Geotechnical instrumentation for monitoring field performance .Wiley-Interscience Publication, New York, N.Y.
D
ei
c
as
L. (2014) - Guidelines for designing, management and maintenance of the geologic monitoring systems. Rend. Online Soc. Geol. It., 30: 16-18.
G
iLi
J.a., c
OrOMinas
J. & r
ius
J. (2000) - Using Global Positining System techniques in landslide monitoring. New technologies for landslide assessment
and management in Europe. Engineering geology, 55(3): 167-192.
G
iuFFreDi
F., z
anOLini
L. & F
OGLinO
L. (2003) - Application of a New Integrated Multi-Parametric Monitoring System for the Analysis of Micro-Movements in
Civil Protection Activities. An Example Proceedings of the 6
th
International Symposium on Field Measurements in Geomechanics: 61-66, Oslo, Norway,
M
azzOLeni
G. (2011) - Geologia e geomorfologia: obbiettivi e prodotti cartografici realizzati. Quaderni di dinamica alpina e quaternaria, 10(2): 11-50.
M
arquer
D., c
haLLanDes
n. & s
chaLTeGGer
u. (1998) - Early Permian magmatism in Brianconnais terranes: Truzzo granite and Roffna Rhyolite (eastern
Penninic nappes, Swiss and Italian Alps). Schweiz. Mineral. Petrogr. Mitt, 78: 397-414.
r
eGiOne
L
OMBarDia
D
Gr
n. X/2684 (2011) - Linee guida per la progettazione, esecuzione, gestione e manutenzione dei sistemi di monitoraggio nell’ambito
della prevenzione del rischio idrogeologico in Lombardia. BURL Serie Ordinaria 51: 25-51.
Received February 2014 - Accepted May 2014
background image
Statistics