Document Actions

IJEGE-11_BS-Cascini-et-alii

background image
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
635
DOI: 10.4408/IJEGE.2011-03.B-069
NUMERICAL MODELLING OF THE DECEMBER 1999 CERVINARA
FLOW-LIKE MASS MOVEMENTS (SOUTHERN ITALY)
l
eonaRdo
CASCINI
(*)
, s
abatino
CUOMO
(*)
& a
ndRea
DE SANTIS
(**)
(*)
Università degli Studi di Salerno - Dipartimento di Ingegneria Civile - Via Ponte don Melillo - 84084 Fisciano (ITALY)
Email: l.cascini@unisa.it, scuomo@unisa.it, a.desantis85@gmail.com
factors is outlined for the propagation stage of the ana-
lysed phenomena. Finally, the possibility to extend the
obtained results to other similar contexts is discussed.
K
ey
words
: modelling, landslides, flow-like, flood, propagation
INTRODUCTION
Rainfall events can trigger different types of
flow-like mass movements (H
utCHinson
, 2004) in-
side the same territory depending on either the slope
morphology or solid/water percentages of the propa-
gating mass. These phenomena can be classified as
debris flows (H
unGR
et alii, 2001) when “a very rapid
to extremely rapid flow of saturated non-plastic de-
bris in a steep channel“ occurs. If a smaller amount
of solids is transported, these phenomena are usually
referred to as hyperconcentrated flows (C
oussout
&
m
eunieR
, 1996). Finally, when water prevails over
solids, the phenomenon is usually called either debris
floods or water floods (C
osta
, 1988).
The run-out distances and consequences associ-
ated to debris flows, hyperconcentrated flows and de-
bris floods are extremely different. Thus, it is impor-
tant to distinguish these phenomena for risk analysis
and zoning. The latter, in turn, necessarily requires an
appropriate evaluation of the propagation stage.
Current scientific literature gives either empirical
or numerical models for the analysis of the propa-
gation stage as recently reviewed by H
unGR
et alii
(2005). The empirical models generally provide an
ABSTRACT
This paper deals with the flow-like mass move-
ments which occurred on 15-16 December 1999 in
Cervinara (Southern Italy). During this event, a huge
amount of water, debris and boulders was transported
from the hillslopes towards the piedmont areas caus-
ing victims, damage to buildings as well as the flood-
ing of a large part of the urban area of the municipal-
ity. In this paper, a study is carried out with reference
to the modelling of the propagation stage of the oc-
curred phenomena which has yet to be satisfactorily
addressed in current scientific literature.
To this aim, the geological, geomorphological and
geotechnical settings are firstly drawn up based on the
available advanced data-set. Then, the event scenarios
are reconstructed referring to rainfall data, informa-
tion on damage and eyewitness accounts. It is high-
lighted that three mountain basins were affected by
three debris floods and, three hours later, a huge debris
flow occurred. For all the phenomena, the propagation
areas are characterised and the rheology of the propa-
gating masses assessed; subsequently, the numerical
modelling of the propagation stage is carried out. For
the numerical analyses, a commercial FLO-2D code
is used and different scenarios are considered with
reference to different digital elevation models, inflow
hydrographs and rheological parameters. The results
obtained for the propagation areas match either the
in-situ evidence or the eyewitness accounts of the De-
cember 1999 events. Moreover, the role of the major
background image
L. CASCINI, S. CUOMO & A. DE SANTIS
636
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
similar events which occurred in the Campania re-
gion. For the same landslide, R
evellino
et alii (2004)
provide a 1D numerical modelling of the propagation
stage which outlines the important role of erosion
phenomena along the propagation path. However, a
comprehensive analysis of all the occurred December
1999 flowlike mass movements has yet to be provid-
ed in current scientific literature.
Due to the availability of an advanced data-set, the
paper firstly outlines the geo-environmental context af-
fected by the 1999 events which are reconstructed and
characterised. Then, the numerical modelling of all the
recognised flow-like mass movements is provided with
reference to their propagation stage. The obtained re-
sults are compared with both the eyewitness accounts
and results of the previous studies. Finally, the role of
the key factors is outlined in relation to the propagation
stage of the occurred flow-like phenomena.
THE 15-16 DECEMBER 1999 CERVINARA
FLOWLIKE MASS MOVEMENTS
GEO-ENVIRONMENTAL SETTING
Cervinara is a village (about 10,000 inhabitants)
located at the toe of Mount Partenio (Fig. 2) where the
mountain basins are characterised by high-order drain-
age networks and extensions varying from 0.62 to 3.22
km
2
. Inside the mountain basins (located 320 - 1300 m
a.s.l.), the main geomorphological units are represented
by either zero order basins or open slopes (Cascini et
alii
, 2008a), where shallow deposits of pyroclastic soils
have depths lower than 2 - 3 m. These deposits gener-
ally lie on steep slopes (30° - 40°) constituted by a frac-
tured carbonate bedrock (o
livaRes
& P
iCaRelli
, 2003).
estimation of the run-out distance which mostly de-
pend on the amount of the unstable volume (C
oRomi
-
nas
, 1996) as well as the features of the source areas
and slope morphology (C
asCini
et alii, 2008a, 2010).
Numerical modelling (P
astoR
et alii, 2009; P
iRulli
& s
oRbino
, 2008; H
unGR
& m
C
d
ouGall
, 2009) also
provides the velocity and height of the propagating
mass, which are important inputs for risk analysis.
However, literature does not satisfactorily address
case studies concerning an almost simultaneous oc-
currence of different types of flow-like phenomena
that, quite often, are not distinguished. Thus, creating
a significant misunderstanding of the related effects.
This is the case of the 15-16 December 1999
Cervinara events (Southern Italy) which caused
6 victims as well as a huge amount of damage to
buildings and facilities.
The area threatened by the 1999 events is located
75 km from the Vesuvius volcano, in the eastern part
of the Campania region, which is one of the most at
risk landslide areas of Europe (C
asCini
et alii, 2008b)
(Fig. 1). In this region, unsaturated pyroclastic soils de-
rive from the explosive eruptions of Vesuvius volcano
and are mostly sands/gravels (pumice soils) and silty
sands/sandy silts (ashy soils) (b
ilotta
et alii, 2005).
Pyroclastic soils are extensively widespread over
carbonate bedrocks (C
asCini
et alii, 2008a) and are
frequently affected by rainfall-induced shallow land-
slides of the flow-type (C
asCini
et alii, 2008a).
The 1999 Cervinara event has already been ana-
lysed in previous studies. Particularly, b
udetta
&
de
R
iso
(2004) outline that the reach angle of the major
1999 landslide is comparable with those of previous
Fig. 1 - Cervinara study area: a) loca-
tion, b) 3D view
background image
NUMERICAL MODELLING OF THE DECEMBER 1999 CERVINARA FLOW-LIKE MASS MOVEMENTS (SOUTHERN ITALY)
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
637
“Valle” (point 1); ii) on 16
th
December at 00:15 another
flooding was recorded (point 2) due to a local break
of the levees of the Castello torrent; iii) afterwards (at
00:35) both water, debris and trees propagated downhill
(point 3); iv) at the same time (about 00:40), in Piazza
Ioffredo (point 4), the cross section of a bridge over the
Ioffredo torrent was completely filled with debris and
trees coming from the upslope Ioffredo basin; v) finally,
from the same mountain basin, a huge amount of debris
flooded Via Ioffredo (point 5) at 01:20.
Figure 2 shows the main watersheds of the Cer-
vinara municipality (data provided by the River Ba-
sin Authority “Liri- Garigliano-Volturno rivers”).
In the same figure, an area of about 6.40 km
2
(box
named “f”) is also shown, which was mostly affect-
ed by the 1999 events. This area will be referred to
in the following sections of the paper as the study
area for the numerical modelling.
TEMPORAL AND SPATIAL OCCURRENCE
Between 14-16 December 1999, starting from 14
th
December 12 a.m, a cumulated rainfall of 264 mm in
38 hours (f
ioRillo
et alii, 2001) was recorded at the
S. Martino Valle Caudina rain-gauge (Fig. 3). Hydro-
logical analyses show that the return period of the cu-
mulated rainfall over the previous 24 hours was equal
to 10-20 years on 15
th
December at 6:00 p.m. and it
rapidly increased up to values of 50-100 years on 16
th
December at 00:00 a.m. (C
asCini
et alii, 2005).
During the rainstorm, multiple flow-like mass
movements threatened the Cervinara municipality in
about three hours. From the available data-set (source:
River Basin Authority “Liri-Garigliano- Volturno riv-
ers”), eyewitness accounts of the inhabitants of Cer-
vinara (personal communication) and in reference to
table 1 and figure 4, it can be obsrved that: i) on 15
th
De-
cember at 22:15, a street was flooded at location named
Fig. 2 - Geo-environmental setting of the Cervinara study
area and 1999 flow-like phenomena: a) bound-
ary of Cervinara municipality, b) 50 m elevation
contour lines, c) main watersheds, d) main draw-
ing network, e) buildings, f) mostly affected area
by the propagation stage of the December 1999
events, g) debris flow, h) debris flood
Fig. 3 - Rainfall data recorded at S. Martino Valle Cau-
dina rain-gauge (288 m a.s.l., 2 km far from Cer-
vinara) (modified from c
ASciNi
et alii, 2005)
Tab. 1 - Temporal occurrence of the events
Fig. 4 - Spatial occurrence of the events
inside the area “f” of figure 2
background image
L. CASCINI, S. CUOMO & A. DE SANTIS
638
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
efficacy has been extensively proven (b
ello
et alii,
2003; P
iRulli
& s
oRbino
, 2008).
The FLO-2D model assumes the propagating mass as
a continuum “equivalent” fluid whose rheological behav-
iour must approximate the behaviour of the real mixture
of solid and fluid phases. The dynamic behaviour of the
equivalent fluid is described by the mass balance equa-
tion (eq. 1) and momentum balance equation (eq. 2-3),
which written in Eulerian form, depth-averaged and im-
plemented in a finite difference scheme are the following:
where S
f
denotes the depth-averaged flow velocity, h
is the flow depth, v the velocity, i is the rainfall inten-
sity on the flow surface and g is the gravity constant.
The friction slope components S
fx
and S
fy
are written as
functions of the bed slope S
0x
and S
0y
, the pressure gra-
dient and the convective and local acceleration terms.
Regarding the rheological characteristics of the
flowing mass, the quadratic rheological approach
proposed by J
ulien
& l
an
(1991) is adopted. In par-
ticular, the friction slope components S
fx
and S
fy
are
provided by the following expression:
where τ is the shear stress at the contact between the
flowing mass and the bed load, τ
y
is the Bingham yield
stress, η is the Bingham viscosity, k is the flow re-
sistance parameter and n
td
is the equivalent Manning
From the reconstruction of the events, it can be
outlined that: i) firstly three debris floods occurred at
the San Gennaro, Ioffredo and Castello watersheds
(Fig. 4), ii) 3 hours later, a debris flow was triggered
at the Ioffredo mountain basin (Fig. 4). The three de-
bris floods travelled distances up to 2.50 km inside the
piedmont zone involving an area of about 4.70 km
2
.
The Ioffredo debris flow propagated 1.50 km downs-
lope from the crest of the source area, affecting an
area of about 0.20 km
2
in the piedmont zone.
Regarding the observed damage, debris floods
mostly caused the inundation of streets and bridges
with debris and trees (Fig. 5a, 5b) as well as minor
damage to buildings (Fig. 5c, 5d). The Ioffredo debris
flow, characterised by a much greater amount of de-
bris, caused the collapse of either non-structural (Fig.
6a, 6b) or structural parts of buildings (Fig. 6c, 6d).
Considering the differences among the 1999 flow-
like phenomena and the time delay in their occurrence,
it can be consistently assumed that their propagation
patterns were independent. Therefore, they are here-
after separately back-analysed. To this aim, numerical
models are used to assess the propagation areas of all
the types of the flow-like mass movements.
NUMERICAL MODELLING OF THE
PROPAGATION STAGE
THE FLO-2D MODEL
Numerical modelling of the propagation stage of
flow-like mass movements can be carried out through
several numerical codes available in current literature
(P
astoR
et alii, 2009; P
iRulli
& s
oRbino
, 2008; among
others). In this paper, the commercial numerical code
FLO-2D (o’b
Rien
et alii, 1993) was used, since its
Fig. 5 - Typical examples of damage caused by debris
floods occurred at San Gennaro, Ioffredo and
Castello watersheds
Fig. 6 - Typical examples of damage caused by the Iof-
fredo debris flow
(1)
(3)
(4)
(2)
background image
NUMERICAL MODELLING OF THE DECEMBER 1999 CERVINARA FLOW-LIKE MASS MOVEMENTS (SOUTHERN ITALY)
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
639
coefficient for the turbulent and dispersive shear stress
components. Particularly, the first and second terms
on the right hand side of Eq. (4) are, respectively, the
yield term and the viscous term as defined in the Bing-
ham equation. The last term represents the turbulence
contribution (o’b
Rien
et alii, 1993).
INPUT AND METHODS
In order to simulate the propagation stage of
the 1999 events, different digital elevation models
(DEM), available from 1:25,000 and 1:5,000 maps,
were used. Aimed at solving the governing system of
equations (1-4), for a given DEM, the FLO-2D code
overlays the topographic surface with a square finite
difference grid system and the flow is routed in eight
possible flow directions (the four compass directions
and the four diagonal directions). In the analyses car-
ried out, either 25 m or 5m topographic grids were
used depending on both the type of phenomena and
analysis purposes. As input data, the location of the
source areas were assigned as well as the amount of
the unstable mass which was introduced as an inflow
hydrograph. The latter, in turn, requires the evaluation
of the mobilised volume and/or the flow discharge.
Particularly, for the debris floods occurred inside
the San Gennaro, Ioffredo and Castello watersheds
(Fig. 2), the propagating mass was assumed as a New-
tonian equivalent fluid eventually considering low de-
bris concentrations (by volume) not larger than 0.2.
Accordingly, the hydrograph was assumed triangular
with the maximum corresponding to half the duration
of the hydrograph, in agreement with current litera-
ture. From the data-set (source: River Basin Authority
“Liri-Garigliano-Volturno rivers”) the peak discharg-
es were available, evaluated explicitly considering the
morphometric features of the watersheds. The water-
shed concentration times (C
How
et alii, 1988) were as-
sumed as the total duration of the inflow hydrographs
(Tab. 1). Particularly, two hydrographs were avail-
able, respectively computed referring to: i) a rainfall
characterised by a return period T = 100 years, ac-
cording to VAPI procedure (R
ossi
& v
illani
, 1994)
(hydrographs ”a”, later on), ii) the rainfall intensity
measured during the 15-16 December 1999 rainstorm
(hydrographs “b”, later on). Hydrographs “a” and “b”
are reported respectively in figure 7a and 7b, being the
peak discharge and duration of the former hydrograph
much larger than those of the latter one.
For the debris flow occurred at the Ioffredo wa-
tershed, the volume mobilised inside the source area
was initially evaluated. In particular, assuming a mean
depth of the slip surface from the ground surface equal
to 1.5 m (d
amiano
, 2003), the mobilized volume was
estimated equal to 31,000 m
3
. This value is in agree-
ment with that estimated by d
amiano
(2003), while
it is quite smaller than those outlined by R
evellino
et alii (2004) and b
udetta
& d
e
R
iso
(2004) respec-
tively equal to 120,000 m
3
and 240,000 m
3
.
Regarding the rheology of the propagating mass,
the quadratic rheological approach proposed by J
u
-
lien
& l
an
(1991) was used adopting values of the
rheological parameters taken from current literature
(P
iRulli
& s
oRbino
, 2007, 2008). Particularly, it is as-
sumed that the unit weight of the propagating mass
is equal to 12 kN/m
3
; the sediment concentration (by
volume) ranges from 20% to 40%; the Bingham yield
stress, τ
y
is in the range from 0.5 kPa to 0.7 kPa; the
Bingham viscosity
η
is equal to 10 - 20 Pa·s. Moreo-
ver, as suggested in literature (P
iRulli
& s
oRbino
,
2007, 2008), a triangularshaped hydrograph was re-
ferred to with the peak discharge Q
p
corresponding to
1/3 of the duration of the hydrograph. The peak dis-
Tab. 2 - Morphometric and hydrological characteristics of
the watersheds (z: elevation, tc: watershed con-
centration time)
Fig. 7 - Input hydrographs (a) and (b) for debris floods
(data source: River Basin Authority “Liri-
Garigliano-Volturno rivers”)
background image
L. CASCINI, S. CUOMO & A. DE SANTIS
640
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
ferent, with higher depths of the propagating mass as
well as smaller areal extent. Therefore, the December
1999 events must be considered debris floods rather
than hyper-concentrated flows since the back- ana-
lysed debris concentrations did not exceed 10-20%.
The role of the buildings on the propagation stage
of the occurred debris floods was also analysed. The
charge was estimated through the empirical relation-
ship proposed by R
iCkenmann
(1999) which follows:
where Q
p
(m
3
/s) is the peak discharge and V (m
3
) is the
volume of the debris flow. The computed values are
reported in Figure 8.
NUMERICAL RESULTS FOR DEBRIS FLOODS
For the numerical modelling of the debris floods
occurred in the three watersheds of table 2, a 25 m x
25 m Digital Elevation Model was used and different
event scenarios were analysed by changing: i) the in-
flow hydrograph, ii) the debris concentration and iii)
considering or not the presence of buildings (Fig. 9).
Figure 10 shows the flow depths simulated consid-
ering the hydrographs of figure 7. In particular, assum-
ing the hydrograph “a” of fig. 7a, the propagation area
is overestimated both for the watersheds 1 and 2-3 (Fig.
10a). Conversely, assuming the hydrograph ”b” of fig.
7b, a satisfactory matching is obtained among the simu-
lated and observed propagation areas in watershed 1,
while in watersheds 2-3, an underestimation is obtained
(Fig. 10b). However, with reference to this last area, the
observed local breaks of the levees of the Ioffredo tor-
rent are not considered in the numerical analyses, while
these phenomena certainly contributed to enlarging the
propagation areas, even if with low depths.
In order to assess the rheology of the propagat-
ing masses, different debris concentrations (by vol-
ume) were considered up to 20 %, referring to the
hydrograph ”b” of figure 7b, with figure 11 showing
the simulated propagation areas. It is evident that de-
bris concentrations up to 10% negligibly modify the
propagation patterns and the simulated depths, as
highlighted by the comparison with figure 10b. Con-
versely, if the considered debris concentration is high-
er than 20%, the simulated scenario is completely dif-
Figure 8. Input hydrographs for the Ioffredo debris flow as-
suming different mobilised volumes: 30,000 m
3
(a), 120,000 m
3
(b), 240,000 m
3
(c)
(5)
Fig. 9 - Topographic grids (25m x 25m ) used as input for
the FLO-2D code: a) not considering and b) con-
sidering the presence of buildings
Fig. 10 - Computed propagation areas of debris floods obtained
assuming the hydrographs “a” and “b” of figure 7
Fig. 11 - Computed propagation areas of debris floods as-
suming the hydrographs “b” of figure 7b and de-
bris concentrations equal to 0.1 (a) and 0.2 (b)
background image
NUMERICAL MODELLING OF THE DECEMBER 1999 CERVINARA FLOW-LIKE MASS MOVEMENTS (SOUTHERN ITALY)
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
641
fundamental step for an adequate assessment of the af-
fected areas, ii) a 25 m x 25 m DEM makes it possible
to obtain only a rough estimation of the kinematic and
rheological features of the debris flow.
Therefore, further analyses were carried out
aimed at adequately assessing: i) the rheology of the
propagating mass, ii) the effect of the presence of
buildings on the propagation pattern, iii) the rela-
tionships between the kinematic features of the de-
bris flow and damage to buildings.
Accordingly, a more detailed 5m x 5m Digital El-
evation Model was used and different scenarios were
outlined, based on different values of the rheologi-
cal parameters of the mass (Fig. 14a). The obtained
results show that a value of t
y
ranging from 0.5 kPa
to 0.7 kPa allows for a satisfactory estimation of the
run-out distance of the debris flow, with the last value
being the more suitable for adequately reproducing
the in-situ evidence (Fig. 15). On the contrary, negli-
gible differences were obtained by changing the pa-
rameter h in the range 10 - 20 Pa·s.
In order to further investigate the kinematic
features of the Ioffredo debris flow, the presence of
buildings was also considered (Fig. 14b). The ob-
tained results show the relevant role played by the
obstacles in determining both the flowing directions
and the depths of propagating mass (Fig. 15 - 16),
especially in the densely urbanized area.
cells of the topographic grid of figure 9 corresponding
to buildings were therefore assumed to be not floodable
and the obtained results are given in figure 12. Gener-
ally speaking, the obtained scenario is quite similar to
that of figure 9b even though some local differences can
be noted. Specifically, in the simulation of figure 12, a
greater correspondence with the in-situ evidence is ob-
tained on the left side for watershed 1 and on the right
side for watersheds 2 and 3. However, the presence of
buildings does not significantly modify the propagation
patterns and depths of this type of phenomena.
NUMERICAL RESULTS FOR THE IOFFREDO
DEBRIS FLOw
The propagation stage of the Ioffredo debris flow
was initially simulated referring to the 25 m x 25 m
Digital Elevation Model of figure 9a in order to: i) ob-
tain a preliminary evaluation of the mobilised volume,
ii) investigate the potentialities to reproduce this type
of phenomenon with a 25 m x 25 m DEM.
Numerical simulations (Fig. 13) were carried out
with three volumes of the inflow hydrograph, respec-
tively equal to 240,000 m
3
, 120,000 m
3
and 31,000 m
3
,
with the first two values being available from literature
and the last one assumed in this study. The obtained
results clearly show that this last estimated volume is
the only one that allows for a better reproduction of
the observed propagation area as well as the depths
of the debris flow inside the propagation area (Fig.
13). From the obtained results, it is also evident that:
i) the accurate evaluation of the mobilised volume is a
Fig. 12 - Computed propagation areas of the debris
floods assuming the hydrographs “a” of figure
7a, absence of sediments and considering the
presence of the buildings
Fig. 13 - Computed propagation areas of the Ioffredo de-
bris flow assuming the same rheological param-
eter (τ
y
= 0.7 kPa,
η
= 20 Pa·s) while different
volumes for the inflow hydrograph available from:
a) B
uDettA
&
De
r
iSo
(2004), b) r
evelliNo
et alii
(2004) and c) present study
background image
L. CASCINI, S. CUOMO & A. DE SANTIS
642
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
The computed depths and velocities of the debris
flow (Fig. 17) also make it possible to highlight sev-
eral insights relating to the observed damage.
Particularly, the most damaged buildings “a”, “b”,
“c” and “e” correspond to simulated velocities 12 m/s
< V < 21 m/s, while the buildings “d” and “f” are as-
sociated to 8 m/s < V < 12 m/s. These results match
the results of f
aella
& n
iGRo
(2003) which estimated
velocities ranging from 14 m/s to 21 m/s as necessary
for the column collapse of concrete buildings, while
lower velocities, within the range 3 - 19 m/s, may
cause the failure of external walls.
CONCLUDING REMARKS
This paper deals with flow-like mass movements
which occurred in December 1999, causing 6 victims
inside an area of about 4.70 km
2
of the municipality of
Cervinara (Southern Italy).
For this event, the temporal occurrence was firstly re-
constructed based on the available information on damage
to structures and infrastructures as well as the eyewitness
accounts of local inhabitants and the available dataset.
The analysis of the available data outlines dif-
ferent types of flow-like movements which occurred
during the 1999 disaster: i) debris floods originated
Fig. 14 - Topographic grids (5m x 5m ) used as input for the
FLO-2D code: a) not considering and b) consid-
ering the presence of buildings
Fig. 15 - Computed propagation areas assuming
η
= 20
Pa·s, τ
y
= 0.7 kPa (a) and τ
y
= 0.5 kPa (b), disre-
garding the presence of buildings
Fig. 16 - Computed propagation areas assuming (τ
y
= 0.7 kPa,
η
= 20 Pa·s) and considering the presence of buildings
Fig. 17 - Buildings destroyed (I) and severely damaged (II)
compared to the computed velocities of the propa-
gating mass using the same rheological param-
eters of figure 16
background image
NUMERICAL MODELLING OF THE DECEMBER 1999 CERVINARA FLOW-LIKE MASS MOVEMENTS (SOUTHERN ITALY)
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
643
strated that the buildings play a negligible role on the
extent and depths of the propagation area of the ob-
served and modelled debris floods.
For the occurred debris flow, the back-analysed
volume matches the in-situ evidence, with it being in
agreement with one of the studies in current literature.
Moreover, it has been highlighted that a quadratic
rheological law is appropriate in order to obtain a sat-
isfactory simulation of the propagation area. Finally,
taking into account the role played by the buildings
on the kinematics of the debris flow, the simulated
depths of the debris flow match those observed in-situ;
moreover, velocities seem to be strictly related to the
type and severity of the recorded damage to buildings.
ACKNOWLEDGEMENTS
The Authors kindly acknowledge the National
River Basin Authority “Liri-Garigliano and Volturno
rivers” which provided the geo-environmental data-
set regarding the Cervinara study area as well as in-
formation on the December 1999 events.
in three mountain basins that affected a large propa-
gation area, ii) three hours later, a huge debris flow
originated in a small watershed and propagated inde-
pendently from the previous phenomena.
For both the aforementioned debris floods and
debris flow, a numerical modelling of the propagation
stage was carried out using a commercial FLO-2D code,
referring to i) inflow hydrographs evaluated by using
different methods depending on the typology of phe-
nomena, ii) rheological parameters assessed on the basis
of the back-analysis of the observed propagation areas.
In the three analysed watersheds, the numerical
results confirm that debris floods rather than hyper-
concentrated flows occurred, since the debris concen-
trations is lower than 10 - 20%. For these phenom-
ena, the inflow hydrographs estimated referring to
the measured rainfall of the 15-16 December 1999
rainstorm allow for a satisfactory back-analysis of the
propagation area. On the contrary, the hydrographs
computed referring to a rainfall characterised by a re-
turn period equal to 100 years give an overestimation
of the propagation areas. Finally, it has been demon-
REFERENCES
b
ello
m. e., o’b
Rien
J. s., l
oPez
J. l., & G
aRCia
-m
aRtinez
R. (2003) - Simulation of flooding and debris flows in the Cerro
Grande River, In: Proc. 3rd Int. Conf. on Debris flow hazards mitigation: mechanics, prediction, and assessment, edited by:
R
iCkenmann
d. & C
Hen
, C. Millpress, Rotterdam.
b
ilotta
e., C
asCini
l., f
oResta
v. & s
oRbino
G. (2005) - Geotechnical characterization of pyroclastic soils involved in huge
flowslides. Geotechnical and Geological Engineering, Springer Ed, 23: 365-402.
b
udetta
P. &
de
R
iso
R. (2004) - The mobility of some debris flows in pyroclastic deposits of the northwestern Campanian region
southern Italy. Bull Eng Geol Environ, 63: 293-302.
C
asCini
l., b
onnaRd
C
H
., C
oRominas
J., J
ibson
R., m
onteRo
-o
laRte
J. (2005) - Landslide hazard and risk zoning for urban
planning and development. - State of the Art report. Atti della International Conference on Landslide Risk Management,
H
unGR
, f
ell
, C
outuRe
& e
beRHaRdt
(eds):199-235, A.A. Balkema Publishers
C
asCini
l., C
uomo
s. & G
uida
d. (2008a) - Typical source areas of May 1998 flow-like mass movements in the Campania region,
Southern Italy. Engineering Geology, 96: 107-125.
C
asCini
l., C
uomo
s. & d
ella
s
ala
m. (2010) - Spatial and temporal occurrence of rainfall-induced shallow landslides of flow
type: A case of Sarno-Quindici, Italy. Geomorphology, DOI: 10.1016/j.geomorph.2010.10.038 (in press.)
C
asCini
l., f
eRlisi
s. & v
itolo
e. (2008b) - Individual and societal risk owing to landslides in the Campania region (southern
Italy), Georisk: Assessment and Management of Risk for Engineered Systems and Geohazards, 2(3): 125-140.
C
asCini
L. (2004) - The flowslides of May 1998 in the Campania region, Italy: the scientific emergency management. Italian
Geotechnical Journal, 2: 11-44.
C
How
v.t., m
aidment
d.R. & m
ays
w.l. (1988) - Applied Hydrology, McGraw-Hill Book Company.
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. (1988) - Rheologic, geomorphic, and sedimentologic differentiation of water floods, hyperconcentrated flows, and
debris flows. Flood geomorphology. In: b
akeR
v.R., k
oCHel
R.C. & P
atton
P.C. (eds.), Flood Geomorphology, John Wiley
and Sons, New York: 113-122.
background image
L. CASCINI, S. CUOMO & A. DE SANTIS
644
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
C
oussot
P. & m
eunieR
m. (1996) - Recognition, classification and mechanical description of debris flow. Earth-Science Reviews
40: 209-227
d
amiano
(2003) - Meccanismi d’innesco di colate di fango in terreni piroclastici. PhD thesis “Seconda Università Degli Studi
Di Napoli” (in Italian): 275.
f
aella
C. & n
iGRo
e. (2003) - Dynamic impact of the debris flows on the constructions during the hydrogeological disaster in
Campania 1998: failure mechanical models and evaluation of the impact velocity. Proc. Int. Conference on “Fast Slope
Movements - Prediction and Prevention for Risk Mitigation”. Patron Editore, Napoli, 179-186.
f
ioRillo
f., G
uadaGno
f.m., a
Quino
s. & d
e
b
lasio
a. (2001) - The December 1999 Cervinara landslides: further debris flows
in the pyroclastic deposits of Campania (Southern Italy). Bull Eng Geol Env, 60: 171-184.
H
unGR
o., e
vans
s.G., b
ovis
m.J. & H
utCHinson
J.n. (2001) - A review of the classification of landslides of the flow type.
Environ & Eng. Geoscience, VII 3): 221-238.
H
unGR
o., C
oRominas
J. & e
beRHaRdt
(2005) - Estimating landslide motion mechanism, travel distance and velocity. Proc. of the
International Conference on Landslide Risk Management, May 31 - June 3, 2005, Vancouver, Canada. Ed. H
unGR
o., f
ell
R., C
outuRe
R., e
beRHaRdt
e., 99-128, Balkema.
H
unGR
o. & m
C
d
ouGall
s. (2009) - Two numerical models for landslide dynamic analysis. Computers & Geosciences, 35
(5): 978-992.
H
utCHinson
J.N. (2004) - Review of flow-like mass movements in granular and fine-grained materials. Proc. of the Int. Workshop
“Flows 2003 - Occurrence and Mechanisms of Flows in Natural Slopes and Earthfill”: 3-16.
J
ulien
P. & l
an
y. (1991) - Rheology of hyperconcentrations, J. Hydrol. Eng., 117(3): 346-353.
o’b
Rien
J. s., J
ulien
P. y. & f
ulleRton
w. t. (1993) - Two-dimensional water flood and mudflow simulation. J. Hydrol. Eng.,
119(2): 244-261.
o
livaRes
l. & P
iCaRelli
(2003) - Shallow flowslides triggered by intense rainfalls on natural slopes covered by loose unsaturated
pyroclastic soils. Géotechnique, 53(2): 283-287.
P
astoR
m., H
addad
b., s
oRbino
G., C
uomo
s. & d
RemPetiC
v. (2009) - A depth-integrated, coupled SPH model for flow-like
landslides and related phenomena. International Journal for Numerical and Analytical Methods in Geomechanics, 2009,
33: 143-172.
P
iRulli
m. & s
oRbino
s. (2007) - The runout of debris flows: application of two numerical models and comparison of results. In:
Proceedings of the 1
st
North American Landslides Conference, 3-8 June 2007, 1, 1542-1551). ISBN/ISSN: 978-0-975-4295-
3-2. : Editors: s
CHaefeR
v.R., s
CHusteR
R.l. & a.k. t
uRneR
A.K.
P
iRulli
m. & s
oRbino
G. (2008) - Assessing potential debris flow runout: a comparison of two simulation models. Nat. Hazards
Earth Syst. Sci., 8: 961-971.
R
evellino
P., H
unGR
o., G
uadaGno
f.m. & e
vans
s.(2004) - Velocity and runout simulation of destructive debris flows and
debris avalanches in pyroclastic deposits, Campania region, Italy. Environmental Geology, 45: 295-311
R
ossi
f & v
illani
P. (1994) - Valutazione delle piene in Campania. Report CNR-GNDCI (in Italian).
R
iCkenmann
D. (1999) - Empirical relationships for debris flows. Natural Hazards, 19(1): 47-77.
Statistics