Document Actions

ijege-13_bs-bossi-et-alii.pdf

background image
Italian Journal of Engineering Geology and Environment - Book Series (6) www.ijege.uniroma1.it © 2013 Sapienza Università
Editrice
221
DOI: 10.4408/IJEGE.2013-06.B-19
HAZARD ASSESSMENT OF A POTENTIAL ROCK AVALANCHE
IN SOUTH TYROL, ITALY: 3D MODELING AND RISK SCENARIOS
G
iulia
BOSSI, S
imone
FRiGeRio, m
atteo
mantoVani,
l
uca
ScHenato, a
leSSandRo
PaSuto & G
ianluca
maRcato
CNR-IRPI – National Research Council of Italy, Research Institute for Hydro-Geological Hazard Protection - Padova, Italy
INTRODUCTION
The prediction of the landslide runout is of crucial
importance for risk assessment, especially in densely
populated valleys. As the mechanisms of fast landslide
such as rock avalanches is not completely understood,
mostly because of the numerousness of the variables
involved in the process (e
VanS
et alii, 2006), the task
of defining the related hazard is still problematic.
Moreover, especially in Alpine environments, river
damming is a potential scenario which implies consid-
erable threats for the nearby population due to back-
water impounding and catastrophic dam-brake waves
(c
oSta
& S
cHuSteR
, 1987). To properly estimate this
consequential hazard (K
oRuP
, 2005) an hypothesis on
the geometry of the deposit should be made; several ap-
proaches both empirical (n
icoletti
& S
oRRiSo
-V
alVo
,
1991; c
oRominaS
, 1996) and analytical (S
aVaGe
&
H
utteR
, 1989; H
unGR
, 1995) have been proposed in the
past literature.
Runout models are regularly used for debris flow
but modelling potential rock avalanches is far from
current practice (c
RoSta
, 2006). The selection of the
correct soil parameters to use in the model is the main
constraint; in doing so the best approach seems to be
the retroactive simulation of a past event in the study
area (H
unGR
, 1996; W
illenbeRG
, 2009).
The present paper deals with the assessment
via numerical model of a potential rock avalanche
in South Tirol. A similar phenomenon, dating 1401,
occurred in the same site forming a natural dam and
ABSTRACT
A large DSGDS (Deep-Seated Gravitational
Slope Deformation) which extends on an area of
3.75 km
2
affects the southwest flank of Mount Gan-
derberg some 40 km north of Bolzano. The gravita-
tional sagging of the slope is generating a wide field
of tension cracks in the crown area where a rock slab
of about 800.000 m
3
shows clear precursory signs of
possible detachment.
In 1401 a rock avalanche detaching from the same
area dammed the Passer River thus creating a 35 m
deep lake which through sequential breaching caused
many casualties in the villages downstream. Using
geomorphological survey and historical analysis it
was possible to estimate the volume of the landslide
deposit in 2 × 10
6
m
3
.
In this study the historical event is back-analyzed
using the 2D code DAN-W in order to select the
proper soil parameters to use in the modelling of the
present potential rock avalanche. Then a 3D model-
ling is carried out using DAN3D software to evaluate
the run-out scenario and the shape of the natural dam.
Owing to the relevant discharge of the Passer River, a
2 × 10
6
m
3
lake might be quickly filled. This can cause
the breaching and subsequent collapse of the landslide
dam with great risk and potentially heavy damages for
the downstream population.
K
ey
words
: landslide hazard, 3D modeling, runout, risk sce-
narios, Ganderberg
background image
G. BOSSI, S. FRIGERIO, M. MANTOVANI, L. SCHENATO, A. PASUTO & G. MARCATO
222
International Conference Vajont 1963-2013. Thoughts and analyses after 50 years since the catastrophic landslide Padua, Italy - 8-10 October 2013
PAST RECURRENCE OF LANDSLIDES
Historical records (W
alcHeR
, 1773) report that
in 1401, due to the collapse of a rock slab from the
north-west ridge of Mt Ganderberg, the Passer river
was dammed forming a 35 m deep lake named Kum-
mersee (Lake of Grief).
During the first years the lake overtopped the dam
but the stability of the deposit was not endangered.
Then in 1419 the dam was breached for the first time
causing rapid erosion of the deposit and a sudden sed-
iment-laden flood that reached the town of Merano,
which lies 26 km downstream, causing 400 casualties.
Other catastrophic bursts followed several times in the
ensuing years, mostly after autumnal heavy rainfall
events; in particular in 1503 the flood was so severe
that chronicles report that the city walls of Merano
were shattered (e
iSbacHeR
& c
laGue
, 1984).
The lake eventually disappeared in the 18
th
cen-
tury but the 1.5×10
6
m
3
remnant of the former dam
still lays on the opposite slope. The dam and the sedi-
mentation in the lake have significantly influenced
the longitudinal profile of the Passer River, inducing
locally a steep slope that has been regulated in recent
years by several check dams.
GEOLOGICAL AND GEOMORPHOLOGICAL
SETTING
In the study area mainly metamorphic units of the
Australpine basement complex outcrop, that is a pre-
disposing geological framework for DSGSD (H
utcH
-
inSon
, 1988). Besides, the unremitting movements of
the DSGSD lead to the progressive weakening of the
rock mass inducing collateral phenomena like rock ava-
lanche, debris flow or secondary slides (a
GliaRdi
, 2001).
Silver-gray shining mica-schist outcrops at the
crown of the landslide in the north-east ridge where
minor signs of deformation are found. However in
Mt Ganderberg paragneiss ridge transverse cracks
and trenches are extensively widespread. The sag-
ging of the DSGSD generates locally steep slope gra-
dients with high potential energy; distinct secondary
movements are recognizable and mostly influenced
by the attitude of the several joints affecting the bed-
rock. In particular the rock slab in Fig. 3 is back-
tilted and widely fractured. The contour of the slab
has been outlined through a GPS survey; thus, con-
sidering the intersection of the rock discontinuities
of 230/80 and 165/45 dip direction, it was possible
consequently a large lake, about 1 km long. Since the
downstream settlements were repeatedly affected by
outburst debris flows due to the collapse of sections
of the dam until the 18
th
century (S
cHuSteR
, 2000), our
goal is to evaluate the possibility of the formation of
a new landslide dam and to assess the associated risk
GEOGRAPHICAL SETTING
A large DSGDS (Deep-Seated Gravitational
Slope Deformation) of 3.75 km
2
affects the southwest
flank of Mt. Ganderberg (46°51’27’’N, 11°10’26’’E)
some 40 km north of Bolzano, in the Passeiertal, near
the border with Austria (Fig. 1).
With an estimated 150 m deep shear surface, it
stretches from the 2330-2450 m a.s.l. of Mt Ganderberg
and Mt Kreuzjoch down to the Passer River at 1170 m
a.s.l (Fig. 1). The DSGSD is concave in the upper part,
where a huge ridge borders the landslide crown, flat in the
center and bowed at the toe; therefore it can be considered
in the final stage (P
aSuto
& S
oldati
, 1990) (Fig. 2).
The upper Passer River basin, closed at the toe of
the Ganderberg landslide, covers an area of 85 km
2
with maximum altitude of 3400 m. The river is char-
acterized by an alpine regime with high rate of flow
in spring during snowmelt that has reached several
times a peak discharge of 42.5 m
3
/s.
The elements at risk in the study area are the
National Road n. 44bis, which crosses the landslide,
and the small settlement of Hahnebaum which lays,
in the southern sector, near the toe of the DSGSD.
Moreover, in case of landslide damming of the Pas-
ser River, the risk would spread along the stream for
several kilometres as resulted from the past events.
Fig. 1 - Location of the study area
background image
HAZARD ASSESSMENT OF A POTENTIAL ROCK AVALANCHE IN SOUTH TYROL, ITALY: 3D MODELING AND RISK SCENARIOS
Italian Journal of Engineering Geology and Environment - Book Series (6) www.ijege.uniroma1.it © 2013 Sapienza Università
Editrice
223
former talweg was likely to flow in the middle of the
valley as it is in its other reaches, even though nowa-
days it circles the 1401 landslide dam remnant. Fur-
thermore in the area a small debris slide was found
on the left bank clearly denoting a relatively recent
modification of the valley morphology, the trigger of
such landslide being evidently toe erosion since the
slope is steeper than nearby.
to delineate a 800,000 m
3
unstable rock wedge. In
case of detachment, this block is likely to evolve into
a rock avalanche, as it happened in the past, since
there is no confinement or topographic discontinuity
that could stop it before it reaches the valley bottom.
MODELLING
BACK ANALYSIS OF THE 1401 EVENT
A dynamic analysis of the runout for the 1401
event was carried out in order to determine the pa-
rameters to use in the 3D modeling of the present
potential rock avalanche (b
oSSi
et alii, 2011).
Considering the orientation of the discontinuities
and the data from a GPS survey through a geomorpho-
logical analysis it was possible to reconstruct the contour
of the former rock slab with a 3D software (Surfer 9,
Golden Software, Golden, CO, USA). The total volume
of the detached rock mass was estimated as 2×10
6
m
3
.
The model was calibrated on the historical
records of the event and on the basis of geomorpho-
logical surveys of the deposition area. The former
river bed profile has been derived appraising the av-
erage slope of the Passer River between the stretches
that were not affected by the event. Moreover the
Fig. 3 - View of the unstable rock slab. The arrangement
of the joint planes and the former vertical move-
ment of the rock block (beside the lesson learnt
from previous events) pose a distinct hazard due
to the potential triggering of a rock avalanche
Fig. 2 - Shaded relief of the DSGDS with highlighted the remnant of the 1401 landslide dam
background image
G. BOSSI, S. FRIGERIO, M. MANTOVANI, L. SCHENATO, A. PASUTO & G. MARCATO
224
International Conference Vajont 1963-2013. Thoughts and analyses after 50 years since the catastrophic landslide Padua, Italy - 8-10 October 2013
3D MODELLING, RISK SCENARIOS
The 3D simulation was performed with DAN3D
software (H
unGR
, 2009). DAN3D is a meshless
Smoothed Particle Hydrodynamics (SPH) code which
use an integrated two-dimensional Lagrangian solu-
tion. The rock mass is discretize in numerous particles
that flow forced by topography on the basis of a se-
lected rheology.
The shape of the unstable slab was reconstructed
geometrically in 3D, considering the orientations of the
joints and the internal friction angle of the rock mass
(Fig. 4). The Digital Terrain Model with a resolution of
2.5 m was acquired from the regional authorities and it
was derived by LiDAR survey performed via helicop-
ter. The data from the best-fit back-analysis were used
for the soil input parameters.
The outputs show that if the unstable rock slab
collapse it would form a 20 meters high deposit on
the valley bottom already narrowed by the 1401 land-
slide dam remnant (Fig. 5). The accumulated material
would form a dam which could generate a 15 meters
deep lake; it has been calculated that such lake would
have a maximum volume of 2×10
6
di m
3
, reaching the
length of 1.5 km (Fig. 6). The lake would inundate a
quarry and a long portion of a local road which is the
only way to reach the village of Corvara in Passiria.
Nevertheless the major risk would be for the popula-
tion living downstream since the dam failure could
result in a dramatic outburst flood.
As the potential deposit is large and elongated the
hazard of failure by piping is negligible. On the other
Due to the narrowness of the alpine valley the
shape of the deposit can be classified as “strong im-
pact” (H
eWitt
, 2002): the avalanche material climbed
upward on the opposite slope, like backwash. The
phenomenon is called “brandung” in German (H
eim
,
1932) and it causes great energy dissipations. There-
fore the typical long runout of rock avalanches, due
to the impact on the facing mountainside, resulted
in a concentrated accumulation of material. In the
surroundings of the rock avalanche deposit, regions
altered by the pressure and temperature (frictional
breccia, partial melting) are found. On the base of
petrographic investigations these dark beds came
probably to being by heating due to the friction of
the strongly mylonitised and at least partially liquated
rocks (e
GGeR
, 2001).
Considering the usual geometry of “brandung”
the volume of the former deposit has been estimated
in 3.5×10
6
m
3
; the increasing in volume (up to 75%)
since detachment is ascribed to erosion and to the
fragmentation of the rock mass.
The modelling was performed using DAN-W (OH-
GRI, West Vancouver, BC, Canada), a pseudo-3D code
in which the flow depth reflects the amount of spread-
ing expected for the flow path. A frictional model was
selected since the Voelmmy rheology did not reproduce
well the “brandung” dynamic. After several simulations
a good fit was established with a bulk friction angle of
16° for the material, which is low but it reproduces the
peculiar mobility of rock avalanches and it is in accord-
ance with several other authors findings (H
unGR
, 1995).
Fig. 4 - Profile of the landforms on the Passer River banks at the cross-section with the former deposit
background image
HAZARD ASSESSMENT OF A POTENTIAL ROCK AVALANCHE IN SOUTH TYROL, ITALY: 3D MODELING AND RISK SCENARIOS
Italian Journal of Engineering Geology and Environment - Book Series (6) www.ijege.uniroma1.it © 2013 Sapienza Università
Editrice
225
Fig. 5 - Area and thickness of the landslide dam obtained through the DAN3D simulation
Fig. 6 - Flooded area
background image
G. BOSSI, S. FRIGERIO, M. MANTOVANI, L. SCHENATO, A. PASUTO & G. MARCATO
226
International Conference Vajont 1963-2013. Thoughts and analyses after 50 years since the catastrophic landslide Padua, Italy - 8-10 October 2013
side the threat of headward and downward erosion is
significant as the deposit would likely be composed by
loose materials owing to the mechanics of rock ava-
lanches. Moreover, due to the size of the impounded
area and in high water discharge periods the collapse
of the landslide dam could be very fast.
A statistical analysis of the historical records of the
discharge of the Passer River show that in the basin
the outflow reaches 500 l/s km
2
every 10 years. This
happens especially in springtime when to the seasonal
rains the hydric contribution of snowmelt is added. At
the Ganderberg outlet the discharge in these conditions
is 42.5 m
3
/s, therefore the lake could be filled in less
than a day leaving a small span of time to implement
civil defence emergency plans.
DISCUSSION
The main purpose of the present study was to
provide to the public administration in charge of de-
velopment the Civil Defence Plan an effective tool
to evaluate the hazard and manage the risk induced
by the DSGSD affecting Mt. Ganderberg. In doing
so an integrated approach combining modelling and
monitoring activities has been selected. The analy-
sis carried out indicates that the hazard related to the
possible detachment of an 800.000 m
3
unstable rock
slab from Mt. Ganderberg is relevant. The rock mass
could collapse and evolve as a rock avalanche damn-
ing the Passer River and creating a 2×10
6
di m
3
lake.
To mitigate and manage the risk for the popula-
tion of the valley a real time monitoring system has
been designed and installed in order to assess the
magnitude of displacement and the rate of move-
ment. A control unit able to collect and storage the
data as well as to transfer them via GPRS to the
headquarter of CNR-IRPI in Padova, is located in a
safe position outside the detached area. Across the
most evident cracks six extensometers (± 0.01 mm
accuracy) were installed along with a geophone for
acoustic emissions detection. This system helps in
assessing the trigger processes that can induce sig-
nificant movements and allows the definition of the
dynamic of the rock mass therefore controlling the
possibility of secondary movements like partial top-
pling. Besides, a long-base wire extensometer was
installed because it could be useful in measuring
large displacements before the final collapse. In order
to forecast the failure occurrence the collected data
will be analysed through a Saito & Fukuzono equa-
tion analysis (F
ujiSaWa
et alii, 2010). This procedure
will be automatically implemented in the system and
will represent an early warning system thus mitigat-
ing the risk associated with the phenomenon.
Once the back-analysis has been carried out the
modelling activities provided a range of possible risk
scenarios to be included in the civil defence emer-
gency plan. The data gathered from the monitoring
and the modelling are therefore mutually dependant
and their integration allow the local authorities to
periodically update the plan, according with possi-
ble new risk scenarios. In fact, if the monitoring data
indicate the possibility of partial collapse the model
will be easily updated because the soil parameters to
use are already known though the back-analysis. The
flexibility of the approach is a crucial element since
the modelling activity results in a dynamic tool for
the generation of future risk scenarios. This seems to
be of paramount importance in creating an effective
risk management plan helping the public authorities
in emergency phases.
ACKNOWLEDGEMENTS
The research has been founded by the Interreg
project MONITOR II, financed by the South East
Europe Programme 2007-2013. This work has been
also partially funded by the Fondazione Cassa di
Risparmio di Padova e Rovigo within the Research
Project “SMILAND”.
REFERENCES
a
GliaRdi
F., c
RoSta
G. & Z
ancHi
A. (2001) - Structural constraints on deep-seated slope deformation kinematics. Engineering
Geology, 59: 83-102.
b
oSSi
G., m
antoVani
m., m
aRcato
G., P
aSuto
a., m
aiR
V., n
öSSinG
l., & S
teFani
m. (2012) - The Ganderberg landslide
(South Tyrol, Italy): Residual hazard assessment and risk scenarios. Proc. of the 11
th
Int. and 2
nd
North American Symp. on
Landslides and Engineered Slopes. Landslides and Engineered Slopes: Protecting Society through Improved Understanding,
Banff, Canada, 763-768.
background image
HAZARD ASSESSMENT OF A POTENTIAL ROCK AVALANCHE IN SOUTH TYROL, ITALY: 3D MODELING AND RISK SCENARIOS
Italian Journal of Engineering Geology and Environment - Book Series (6) www.ijege.uniroma1.it © 2013 Sapienza Università
Editrice
227
c
oRominaS
J. (1996) - The angle of reach as a mobility index for small and large landslides. Canadian Geotechnical Journal,
33: 260-271.
c
oSta
j.e. & S
cHuSteR
R.L. (1987) - The formation and failure of natural dams. US Geological Survey, Open-File Report 87-392.
c
RoSta
G.b., c
Hen
H. & F
Rattini
P. (2006) - Forecasting Hazard Scenarios and implications for the evaluation of countermeasure
efficiency for large debris avalanches. Engineering Geology, 83: 236-253.
e
GGeR
H. (2001) - Ingenieurgeologische Untersuchung des Gan-derberg Talzuschubs und der Seehofer Bergsturzmasse.
Diplomarbeit, Institute of Geology and Mineralogy Friedrich-Alexander University.
e
iSbacHeR
G.H. & c
laGue
j.j. (1984) - Destructive mass movements in high mountains: hazard and management. Geological
Survey of Canada, 84-16
e
VanS
S.G., S
caRaScia
m
uGnoZZa
G., S
tRom
a.l., H
eRmannS
R.l., i
SHuK
a., & V
innicHenKo
S. (2006) - Landslides from massive
rock slope failure and associated phenomena. Landslides from Massive Rock Slope Failure, Springer, Netherlands, 3-52.
F
ujiSaWa
K., m
aRcato
G., n
omuRa
Y. & P
aSuto
A. (2010) - Management of a typhoon-induced landslide in Otomura (Japan).
Geomorphology, 124: 150-156.
H
eim
A. (1932) - Dersturz und Menschenleben (Landslides and Human Lives). BiTech Publishers Ltd, Vancouver.
H
eWitt
K. (2002) - Styles of rock avalanche depositional complex in very rugged terrain, Karakoram Himalaya, Pakistan.
In: e
VanS
S.G. & d
eGRaFF
J.V. (2003, eds.) - Catastrophic Landslides: effects, occurrence and mechanisms, Reviews in
Engineering Geology. Geological Society of America, Boulder, Colorado, 345-78.
H
unGR
O. (1995) - A model for runout analysis of rapid flow slides, debris flows and avalanches. Canadian Geotechnical Journal,
32: 610-623
H
unGR
o. & e
VanS
S.G. (1996) - Rock avalanche runout prediction using a dynamic model. Proc. 7
th
Int. Symp. on Landslides,
Balkema, Rotterdam, 1: 233-238.
H
unGR
o. & m
c
d
ouGall
S. (2009) - Two numerical models for landslide dynamic analysis. Computers & Geoscience, 35:
978-992.
H
utcHinSon
J.N. (1988) - General Report: Morphological and geotechnical parameters of landslides in relation to geology and
hydrogeology. Proc. 5
th
Int. Symp. on Landslides, Rotterdam: Balkema 1: 3-35.
n
icoletti
P.G. & S
oRRiSo
-V
alVo
m. (1991) - Geomorphic controls of the shape and mobility of rock avalanches. Geological
Society of American Bulletin, 103: 1365-1373.
P
aSuto
a. & S
oldati
M. (1990) - Some cases of deep-seated gravitational deformations in the area of Cortina d'Ampezzo
(Dolomites). Implications in environmental assessment. Proc. of the European experimental course on geomorphology
applied to environmental risk in man's impact on the Dolomites, 2: 91-104.
S
aVaGe
S.b. & H
utteR
K. (1989) - The motion of a finite mass of granular material down a rough incline. Journal of Fluid
Mechanics, 199: 177-215.
S
cHuSteR
R.L. (2000) - Outburst debris-flows from failure of natural dams. In: W
iecZoReK
G.F. & n
aeSeR
n.D. (eds.) - Debris-
Flow Hazard Mitigation: Mechanics, Prediction and Assessment. Bakelma, Rotterdam, 29-42.
W
alKeR
J. (1773) - Neue Nachrichten von den Eisbergen in Tyrol. Kurzböck, Wien.
W
illenbeRG
H., e
beRHaRdt
e., l
oeW
S., m
c
d
ouGall
S. & H
unGR
O. (2009) - Hazard assessment and runout analysis for an
unstable rock slope above an industrial site in the Riviera valley, Switzerland. Landslides, 6(2): 111-119.
background image
Statistics