Document Actions

IJEGE-11_BS-Tessitore-et-alii

background image
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
513
DOI: 10.4408/IJEGE.2011-03.B-057
COMPARISON OF 2D MODELS FOR THE SIMULATION OF THE
OCTOBER 1954 DEBRIS FLOW AND FLOOD EVENT AT MAIORI
(CAMPANIA REGION, ITALY
s
eRena
TESSITORE, d
ieGo
DI MARTIRE, R
iCCaRdo
MARTINO
& d
omeniCo
CALCATERRA
(*)
(*)
Federico II University of Naples, Department of Hydraulic, Geotechnical and Environmental Engineering, Naples, Italy
INTRODUCTION
On October 25-26, 1954 a wide territory in the
Campania region (Italy) including the municipalities
of Salerno, Vietri sul Mare, Maiori, Minori, was struck
by a huge flood event which caused about 300 victims
and severe damage to urban centres and primary in-
frastructures. The event, triggered by a short duration,
high intensity rainfall was also characterized by the
detachment of a great number of mass movements,
mainly of the slide-flow type, which mobilized loose
pyroclastic deposits. The 1954 is, up to now, the worst
episode of a long series, which, from that year on-
ward, hit the Campania region and which has known a
further, tragic peak on May 5, 1998. On that occasion
five foothill towns (Sarno, Bracigliano, Quindici, S.
Felice a Cancello and Siano) were invaded by some
extremely rapid flow-like movements, again involv-
ing the local volcaniclastic covers: as main conse-
quence, 160 persons lost their lives.
The events occurred in the last decades clearly
evidenced the intrinsic vulnerability of the Campanian
territory, where more than 3000 km
2
(ca. 22% of the
regional extension) are exposed to a very high hazard
related to fast-moving (> 10 m/s), long-runout (3-4
km) slide-flows in pyroclastic sediments.
In the case of fast-moving landslides such as flow-
like movements, the assessment of their runout is of
paramount importance for both researchers and ad-
ministrators, when a thorough evaluation of the related
hazard is required, e.g. for the implementation of urban
ABSTRACT
The Campania Region (southern Italy) is charac-
terized by a frequent occurrence of volcaniclastic de-
bris flows causing extreme loss of life and property
damage where a large population occupies alluvial
fans. In 1954 the small town of Maiori was struck by
several debris flows initiated as soil slips triggered
by prolonged rainfalls. Historical sources report sea-
ward shift of the coastline of some tens of metres
during major flood events, also documented by air
photographs taken soon after the disaster. The 1954
event has been simulated using two commercially
available models, DAN-W and FLO-2D. DAN-W
is a program used to model the post-failure motion
of rapid landslides such as debris flows and suitable
for estimating their runout behaviour. The same 1954
event has been further simulated by means of FLO-
2D, whose sensitivity to different routing conditions
was tested by varying several critical input data as
the shape of flow hydrograph and the volume mobi-
lised. The rheological parameters of the model have
been assigned assuming a Bingham behaviour for
the debris flow, based on other works available in
literature. The results obtained with both programs
have been then critically commented, with the aim
of assessing their capability to reproduce the studied
event and, more generally, to help in the specific haz-
ard zoning and mapping.
K
ey
words
: debris flows, flood event, back-analysis
background image
S. TESSITORE, D. DI MARTIRE, R. MARTINO & D. CALCATERRA
514
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
verify the feasibility of two well-known commercial
codes (DAN-W - Hungr, 1995; FLO-2D - o’b
Rien
et
alii., 1993) in modelling the October 1954 event, with
respect to the Reginna Maior basin, which has been
one of the focus areas in terms of damage caused.
Such event, although triggered by an extreme rainfall,
showed features common to vast areas of the Campa-
nia region, where slide-flows in volcaniclastic terrains
have repeatedly occurred.
Both models discussed below solve the continuity
and momentum equations in both orthogonal flow di-
rections. Only the main features are summarized here,
and the appropriate literature is cited for each model.
The debris-flow mixture is assumed to be a continuous,
homogeneous, and incompressible fluid.
GENERAL SETTING
The Regina (or Reginna) Maior torrent drains a
basin of about 33 km
2
, which shows a prevailing N-S
direction. It crosses the territory of Maiori, a town of
about 5700 inhabitants, located some 20 km west of
Salerno, the province capitol (Fig. 1). The basin, which
culminates at about 1300 m a.s.l. with Mt. Cerreto, is
of fifth order, following the hierarchical system pro-
posed by Strahler (1952); the main channel reaches the
Tyrrhenian sea after about 6 km.
The geological setting is characterized, on the
whole, by the presence of a carbonate bedrock over-
lain by loose pyroclastic deposits. The bedrock is in
fact made up of the basal terms of a carbonate plat-
form sequence, Mesozoic to Tertiary in age, mainly
represented by dolostones and dolomitic limestones
ascribed to Triassic. Some of these rocks are so per-
vasively jointed to be termed, in the local geological
literature, as floury (farinose).
The bedrock crops out mainly along high-angle
scarps or cut-slopes; in the remaining areas, it is man-
tled by a cover made up of pyroclastic deposits. In
the area, such deposits belong to the 79 a.D. Vesuvian
eruption, famous for having destroyed Pompeii and
Herculaneum. Volcanological studies (l
iReR
et alii,
1973) have evidenced that the 79 a.D. eruption was
characterized by a dispersion axis SE oriented, which
caused the onset of up to 2.5 m of volcaniclastic prod-
ucts in the Sorrento Peninsula.
For the study area, a 1:5000 scale map is available
which shows the areal distribution of the pyroclastic
cover and the related thickness (Fig. 2). As evident
planning strategies. Multiple approaches can notori-
ously be applied to the assessment of the maximum
travel distance of phenomena such as debris flows, de-
bris avalanches, hyperconcentrated flows: empirical,
statistical, process-based methods can in fact be used,
whose drawbacks are tackled by culturally competent
researchers. As concerns debris flow modelling, two
approaches are possible: two-phase model and single-
phase model. A two-phase model treats separately
the solids and the fluid. It is very useful to simulate
the entrainment and deposition processes through the
movable bottom line, and, therefore, is suitable to face
problems where the morphological evolution is to be
determined. The state of art for describing the inter-
action effects between the two phases is insufficient
(t
akaHasHi
, 2007). As an alternative to separating the
stress contributions of solid and fluid constituents and
their interactions, many investigators use lumped-rhe-
ology models or calibrated resistance formulas to mod-
el the effects of the stress on a debris flow (i
veRson
,
2005). The mixture of fluid and solids is considered as
a kind of continuous fluid whose properties implicitly
reflect the fluid-particles interaction effects.
From a scientific point of view the one-phase mod-
el appears obsolete and recent debris flow theories tend
to select the two-phase model. However, for practical
applications one-phase approach is usually preferred.
Therefore we believe that systematic comparisons of
debris flow models with well-documented field cases
are of value.
In recent years, several runout prediction meth-
ods have been developed (e.g. RAMMS, RASH3D,
SHWCIN, TITAN2D, etc.), including innovations that
have significantly advanced our ability to mimic debris
flows’ behaviour (m
C
d
ouGall
et alii, 2008). However,
while studies are promoted to improve the existing
approaches, also mutuating hints from different fields
(cellular automata, fractal logics, neural network, etc.),
there is an urgent need to make assignment on robust
and trustworthy models, essential to perform a land-
slide hazard and risk assessment both at region- and
slope-scale. Accordingly, a major effort has to be de-
voted not only to search more reliable predictive mod-
els, but also to validate the existing procedures, trying
to overcome uncertainties and shortcomings which
actually limit our capability to diagnose a basin or a
slope behaviour.
Main aim of the present contribution is to cross-
background image
COMPARISON OF 2D MODELS FOR THE SIMULATION OF THE OCTOBER 1954 DEBRIS FLOW AND FLOOD EVENT AT MAIORI (CAM-
PANIA REGION, ITALY)
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
515
hurricane-like meso-cyclonic vortex, characterized by
a return period of about 59.000 years.
Among the main consequences, a huge number
of mass movements occurred (l
azzaRi
, 1954; P
enta
et alii, 1954), initially detached as slides or slumps
and eventually evolved as debris avalanches along
open slopes or channelled debris flows: within the
Reginna Maior basin, more than 110 landslides have
been mapped (Fig. 2). The landslides were fed by the
pyroclastic cover mantling the carbonate bedrock and
mainly ascribable to the 79 a.D. products.
An extensive collection of eyewitnesses accounts is
available for the 1954 event (see e
sPosito
et alii, 2004
and references therein). According to such accounts, it
has been possible to extract the following information,
valuable with respect to the present study:
- on October 26, the Reginna Maior torrent passed
through Maiori with three successive flood waves,
each of which higher than the preceding one; the
waves occurred at 20:30, 21:00 and 22:00, respective-
ly. The last wave was responsible for the major urban
damage, and for the sudden opening of a sinkhole;
- the debris invaded seven buildings, located along the
from Figure 2, a cover thick up to 5 m is present on
the slopes, which is the key-element, along with slope
angle, accounting for the high susceptibility to mass
movements such as slides and slumps, which, albeit
initially surficial, can eventually evolve into debris
flows. In the footslope areas, along the main draining
channels and on the planar surfaces, the cover reaches
the highest values, comprised in the 5 to 20 m class.
THE 25-26 OCTOBER 1954 EVENT
The 25-26 October 1954 event, characterized by
extreme rainfall values clustered in a few hours, hit
an area of more than 500 km
2
, mainly falling within
the Salerno province. The most notable figures which
dramatically synthesize what happened are reported in
Table 1: 318 victims, more than 11.000 homeless, 320
buildings destroyed; the overall damage was estimat-
ed at about 40 billions of Italian Liras, corresponding
to a present-day value of 550 millions of Euro (e
s
-
Posito
et alii, 2003b) (Fig. 3).
The municipalities which suffered the worst con-
sequences were all located in the so-called Amalfi
coast, west of Salerno, and included the towns of
Vietri sul Mare, Maiori, Cava de' Tirreni, Tramonti,
Minori. Several hydrographic basins were interested
by phenomena such as areal slope denudation, mass
movements, overflowing, shore progradation: among
them, the Reginna Maior, Regina Minor, Bonea, Ce-
tus, Fusandola and Rafastia basins were the focus of
the main ground effects.
On October 25 a weather disturbance which had
already caused huge rainfall in northern Italy moved
southward, reaching the Campania region and the prov-
ince of Salerno (Fig. 4). The rainfall lasted about 16
hours, from 1:00 PM on October 25, to 5:00 AM on
October 26, with a maximum value of 504 mm and
maximum intensity of 150 mm/hour (f
Rosini
, 1954).
d
e
l
uCa
et alii (2010) defined the 1954 event as a
Fig. 1 - Location of study area
Fig.2 - Reginna Maior basin: Pyroclastic cover thickness
map (on the left); 1954 Landslide-inventory map
(on the right). Data provided by the Destra Sele
Regional Basin Authority
Fig. 3 - Effects of the October 1954 event at Maiori
background image
S. TESSITORE, D. DI MARTIRE, R. MARTINO & D. CALCATERRA
516
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
been carried out using DAN (Dynamic ANalysis -
H
unGR
, 1995), in its latest version DAN-W, a numeri-
cal model which allows to estimate the post-failure
motion for debris flows and avalanches.
The model is based on an explicit Lagrangian
solution of equations of unsteady non-uniform flow
moving in a shallow open channel (Saint Venant’s
equations). The input data to start the model are: a)
trigger volume, b) topographic profile of the slope
(2D), c) width of the channel, and d) a constant ero-
sion thickness. Moreover, the program includes an
open rheological kernel, so that a variety of constitu-
tive behaviours can be implemented (H
unGR
, 2003).
The rheological relationship selected for the
analysis is the v
oellmy
(1955) model, as modified
by H
unGR
(1995):
where τ:is the resisting stress at the base of the flow,
α is the slope angle,γ is the unit weight of the flow-
ing material (approximately 15 kN/m
3
), H is the flow
depth, ξ the dynamic friction coefficient of the materi-
al, ac is the centrifugal acceleration resulting from the
vertical curvature of the flow path, μ is a turbulence
coefficient with dimensions of m/s
2
, g is the gravity
acceleration and v is the velocity of the flow.
The choice of the Voellmy rheology was suggest-
ed by the good results obtained by previous authors
in similar settings (R
evellino
et alii, 2004; s
Cotto
main street of Maiori, destroying them completely;
- along the torrent stretch which crosses Maiori
the alluvial bed raised of about 3 m, while in a
couple of sections located about 2.3 (Ponte Pri-
mario) and 1.5 km (Ponte Vecite) upstream of
Maiori, debris deposition caused a bed elevation
of about 14 m;
- the Maiori beach prograded seaward of about 50 m
and, at the mouth of the Reginna Maior, a tempo-
rary fan delta was built, whose maximum width
reached 100 m from the pre-existing shoreline. It
is interesting to notice that the 1954 fan-delta was
still visible from airphotos taken about one year
after the 1954 event (Fig. 5);
- the beach of the nearby town of Minori, a small
pocket beach normally constrained by headlands,
prograded to such an extent that it was possible to
reach Maiori by walking along the shore.
BACK-ANALYSES OF THE 1954 EVENT
DAN-w: MODEL AND INPUT DATA
The maximum runout of single debris flows mov-
ing along channels tributary to the Reginna Maior has
Tab. 1 - Damages and victims caused by the 1954 event
(modified after e
SPoSito
et alii, 2003a)
Fig. 4 - Rainfall field of the October 1954 event (D
e
l
ucA
et alii, 2010)
Fig. 5 - Airphoto of the study area taken on July 1955.
Blue line: Reginna Maior Torrent; yellow line:
submarine fan-delta
background image
COMPARISON OF 2D MODELS FOR THE SIMULATION OF THE OCTOBER 1954 DEBRIS FLOW AND FLOOD EVENT AT MAIORI (CAM-
PANIA REGION, ITALY)
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
517
where: t
in
is the central value of the thickness class of
the n-th element in which the buffer area, of the i-th
landslide, is divided; ain is the area of the n-th ele-
ment in which the buffer area, of the i-th landslide, is
divided; a
i
is the area of the whole buffer area (∑a
in
).
After calculating the source volumes for each
landslide, a total value of about 1.3 million cubic me-
tres was obtained (Tab. 2). Owing to the high number
of individual mass movements occurred (about 110
within the Reginna Maior basin), the source volumes
introduced in the DAN-W back-analyses were con-
sidered as the sum of all landslides occurred in every
tributary basin of the Reginna Maior torrent (Tab. 2).
The resulting source volumes were then transformed
into a debris slab of constant width, length and thick-
ness (the latter ranging from 2 to 10 m), for each
tributary basin, numbered from 1 to 15 (Fig. 7 and
Tab. 2). Moreover, an average constant thickness of
pyroclastic deposits (from 1 to 2 m, depending upon
the map of Figure 2) was assumed to be eroded along
each flow path.
As shown in Table 2, a final detached volume of
about two million cubic metres has been calculated,
with runout distances comprised between 900 and
2800 m from source. In nearly all the analyzed pro-
files the simulated flow passed over the confluence
between the tributary basin and the main stream,
coming to a halt from 100 to 500 m downstream,
along the Reginna Maior. The predicted maximum
flow velocities were in the order of 25-30 m/s, which
decrease down to 2-4 m/s in the terminal stretches of
the streamlines. Here, the front height showed values
from 1 to 8 m.
di
s
antolo
& e
vanGelista
, 2009; C
alCateRRa
et alii.
2010). In fact, in the Campanian Apennines, the ve-
locities obtained with the frictional as well as Her-
schel-Bulkley models proved to be excessively high,
even in the zone close to the arrest point, with re-
spect to the data available from the literature (f
aella
& n
iGRo
2003; s
Cotto
di
s
antolo
& e
vanGelista
,
2009). The Voellmy model differs significantly, with
considerably lower and more realistic velocities than
for the other two models. Such observations confirm
what was observed by H
unGR
(1995), who noted that,
by using the frictional model, the velocities were
significantly overestimated. Moreover, the Voellmy
model has been widely confirmed in similar contexts
(H
unGR
& e
vans
1996; f
ioRillo
et alii. 2001; R
evel
-
lino
et alii. 2004) by also using three-dimensional
geometric models (m
C
d
ouGall
& H
unGR
2005) in
terms of both distances travelled and velocities. In
the papers by R
evellino
et alii (2004)
and
s
Cotto
di
s
antolo
& e
vanGelista
(2009), in particular, by
means of Voellmy-based back-analyses, a calibration
of relevant input parameters (dynamic friction coef-
ficient of the material, μ, and turbulence coefficient,
ξ was achieved. The obtained values of μ= m/s
2
and
ξ= 0.07 allowed to properly simulate the behaviour
of the real events.
The flow paths have been reconstructed on a
1:5.000 scale map. No observed data were available
as regards runout distances and final landslide debris
thicknesses, so the input parameters were not adjust-
ed to the selected rheology.
DAN-w: RESULTS
The volumes of pyroclastic deposits initially de-
tached for each of the 1954 landslides were assessed
by means of a GIS-based procedure (Fig. 6), through
consecutive overlay operations. The procedure con-
sisted in setting a buffer area (1 m) for the landslide
perimeter, giving a weighted average thickness (T
wa
) of
sediments for each landslide (based on the pyroclastic
cover thickness map of Figure 2), and then in multiply-
ing the obtained value by the areal extent of the land-
slide.The Twa value was hence calculated by means of
a weighted average of the different thickness classes of
pyroclastic deposits included in the buffer area:
Fig. 6 - Graphical procedure adopted to estimate the initial
volume detached for each 1954 landslide (see text)
background image
S. TESSITORE, D. DI MARTIRE, R. MARTINO & D. CALCATERRA
518
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
FLO-2D: MODEL AND INPUT DATA
In order to predict and to compare the runout dis-
tance and the velocity of the debris flows observed in
the study area, the two-dimensional FLO-2D model
(o’b
Rien
et alii., 1993) has been applied. The FLO-2D
simulation code solves the governing equations using a
finite difference method on a fixed rectangular grid. The
governing equations include the continuity equation
and the two-dimensional equations of motion:
where h is the flow depth, V
x
and V
y
are the velocity
components, S
fx
and S
fy
are the friction slope compo-
nents and Sox and Soy are the bed slopes.
The basic equation for the total friction slope S
f
considers a combination of yield, viscous, collision and
turbulent stress components. Based on the so-called
quadratic rheological model of J
ulien
& l
an
(1991),
the total friction slope Sf is expressed as:
where τ
B
is the Bingham yield stress, ρ is the mixture
density, g is the gravitational acceleration, μ
B
is the
Bingham viscosity, V is the mean flow velocity, k is the
laminar flow resistance coefficient.
The laminar flow resistance coefficient K equals
4000 has been considered in relation with roughness
and irregular cross section geometry (as suggested
in FLO-2D Reference Manual, 2009). The Bing-
ham parameters τ
B
and μ
B
are defined as exponen-
tial functions of sediment concentration (C
oussot
,
1994; m
aRtino
, 2003) which may vary over time;
they have been inferred from the rheology of debris
samples collected in another basin of Campania Re-
gion, measured by a coaxial viscometer (m
aRtino
,
2003). The pseudo-Manning’s resistance coefficient
(n=0.02) accounts for both collisional (inertial grain
shear) and turbulent frictional losses.
The simulations have been developed for several
values of solid concentration (0.3-0.6) and for two
different variations of input discharge with respect to
time: single surge input hydrograph or a succession
of hydrographs generated from multiple events of
debris flow (multisurge input hydrograph). The input
hydrograph was reconstructed on the basis of debris
volumes reported in Table 2. The surface topography
was discretised into square-grid elements, and each
one was assigned an elevation and a roughness factor.
The detail and accuracy of a simulation are related to
grid size: the sensitivity analysis showed that the qual-
ity of the simulations is higher when using the finer
grid (small-sized cells).
FLO-2D: RESULTS
The debris flow event occurred on 1954 had a du-
ration of 16 hours after an intense rainfall and reached
the sea depositing there a volume of solid materials
that permitted to reach by walking the village of Mi-
Fig. 7 - Cross sections and points of maximum runout for
the analyzed streamlines
Tab 2 - Source and final volume, confluence and final ru-
nout for each tributary basin
background image
COMPARISON OF 2D MODELS FOR THE SIMULATION OF THE OCTOBER 1954 DEBRIS FLOW AND FLOOD EVENT AT MAIORI (CAM-
PANIA REGION, ITALY)
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
519
Moreover, C
alCateRRa
et alii. (2010), who simulated
a landslide-flood event occurred on November 2009
on the Ischia Island, Campania region, Italy, sug-
gested a potential complementarity between DAN-W
and FLO-2D in allowing to simulate a mixed land-
slide-flood event. In fact, the differences in the results
obtained (DAN-W runout distance shorter than FLO-
2D one) can be considered as a proof of the different
behaviour of the unstable masses during their downs-
lope movement. The Ischia event, in fact, started as a
gravity-driven phenomenon, represented by some soil
slides evolved into a channelled debris flow, whose
mobility was strongly controlled by the local slope
breaks (max runout distance = 1100 m). The debris
flow eventually underwent a further evolution into a
low-viscosity debris mixture (an hyperconcentrated
flow, according to P
ieRson
& C
osta
, 1987), which,
favoured by the relevant water discharge, was able
to reach the shoreline after a further travel of about
1500 m.
The results deriving from the simulation of the
debris flow – flood event occurred at Maiori in 1954,
obtained with the DAN-W and FLO-2D models, were
here compared. In our case-study, an extreme, mixed
event was back-analyzed, characterized by very long
runout distances (5-6 km from source) and by the mo-
bilization of about 2 million cubic metres of pyroclas-
tic deposits, values unusually large for the Campanian
Apennines (C
alCateRRa
et alii, 2004).
As regards runout distances, the noticed disagree-
ment between the two models’ output is not necessar-
ily a critical evidence. In fact, the lower values ob-
tained with DAN-W (900-2800 m) can be explained,
nori “almost without getting wet” as reported by eye-
witnesses. The flow spread on the alluvial fan where
several buildings were damaged. The procedure to op-
timize the model parameter consisted in obtaining the
best agreement with the field observations: the com-
parison between field and numerical results has been
developed taking into account the volume deposited
in the sea and the velocities reached in correspond-
ence of the buildings destroyed, where presumably
they reached values higher than 6-7 m/s (f
edeRiCo
&
a
moRuso
, 2008) (Fig. 8).
Starting from these event data, a reasonable
agreement of the maximum depositional extent can be
obtained for a value of solid concentration equal to 0.6
and a multiple hydrograph: these choices more closely
match the depositional area in the sea and the high
maximum velocities in the cells with the buildings de-
stroyed (indicated with circles in Figure 8).
DISCUSSION AND CONCLUSIONS
Previous authors have already performed a com-
parison between DAN-W and FLO-2D, generally
achieving good and intercomparable results, which, in
turn, resulted in good agreement with the findings of
the present study, considering the differences among
the simulated events.
Among the most recent papers, a
Rmento
et alii
(2008), b
eRtolo
& b
ottino
(2008), C
alCateRRa
et
alii (2010) found that DAN-W gives a more accurate
representation of the documented events in terms of
velocity and runout distances using the Voellmy rheol-
ogy, while FLO-2D, requiring a high-resolution DEM,
represents more precisely the terminal flooded area.
Fig. 8 - Distribution of deposit (on the
left); maximum velocities predict-
ed (on the right)
background image
S. TESSITORE, D. DI MARTIRE, R. MARTINO & D. CALCATERRA
520
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
REFERENCES
ARMENTO M.C., GENEVOIS R. & TECCA P.R. (2008) - Comparison of numerical models of two debris flows in the Cortina
d’Ampezzo area, Dolomites, Italy. Landslides, 5, 143-150, Springer-Verlag, Berlin-Heidelberg, Germany.
BERTOLO P. & BOTTINO G. (2008) - Debris-flow event in the Frangerello Stream-Susa Valley (Italy) - calibration of nu-
merical models for the back analysis of the 16 October, 2000 rainstorm. Landslides, 5, 19–30, Springer-Verlag, Berlin-
Heidelberg, Germany.
with FLO-2D is the size of inundation in the sea that
permitted to reach by walking the village of Minori.
The values of the maximum predicted velocities re-
vealed that they are compatible with the damage oc-
curred on the buildings during the event. Moreover,
while DAN-W cannot reproduce the contemporary
trigger of multiple events, this effect can be simulated
with FLO-2D assigning, in the hydrographical inputs,
a succession of hydrographs, each one generated from
a single event.
In conclusion, our study evidenced that, despite the
noticed differences, the combined application of DAN-
W and FLO-2D is useful to back-analyze with a rea-
sonably accurate representation debris flow and flood
events, even in the case of “extreme” episodes such as
that occurred at Maiori and surrounding areas in 1954.
The above differences, especially as regards the maxi-
mum runout, can be explained by means of the presum-
ably different behaviour of the simulated phenomena,
which, in their initial stages, have been prevailingly
controlled by gravity (surficial slides evolved into
channelled debris flows), and eventually transformed
into low-viscosity debris mixtures (hyperconcentrated
flows), owing to the significant water discharge. In any
case, for both models, it is of paramount importance
the definition of the rheological model, notoriously dif-
ficult to be a-priori defined if not by means of accu-
rate laboratory experiments. Hence, both DAN-W and
FLO-2D can be considered as useful tools to predict
debris flow and flood hazard especially when detailed
input data are not available, and their back-analysed pa-
rameters can be conveniently adopted for other settings,
where similar boundary conditions can be found.
ACKNOWLEDGEMENTS
This research was partly supported by the Federi-
co II University of Napoli (funds granted to d. Calca-
terra). Sincere thanks are due to Destra Sele Regional
Basin Authority for providing cartographic data. Valu-
able advice from two anonymous reviewers is grate-
fully acknowledged.
following C
alCateRRa
et alii (2010), as essentially
controlled by the morphology of the flowpaths, where
the main slope breaks found along the profiles and the
low-angle channel of the Reginna Maior are respon-
sible for the sudden slowdown of the gravity-driven
flow. To this respect, it must be evidenced that indi-
vidual soil slips moved from source areas showing
slope angles between 29° and 43°. The main break at
the slope base caused a reduction of 10° to 30°. Slope
angles further decrease at the confluence between
tributary channels and the main course of the Reginna
Maior: the latter shows an average seaward dip of
about 12°. In addition, if the 1954 landslide events
are compared with similar phenomena occurred in
the same geographical setting (Sorrento Peninsula),
it can be evidenced that the debris flows known for
the above area have displayed a maximum runout in
the order of 800-900 m (C
alCateRRa
et alii, 2004).
Runout distances comparable to those calculated by
DAN-W for the 1954 events are typical of different
Campanian settings, such as Mt. Pizzo d’Alvano, the
carbonate relief involved in the May 1998 landslide
event (d
el
P
Rete
et alii, 1998; C
alCateRRa
et alii,
1999): on that occasion the foothill towns of Sarno,
Bracigliano, Quindici and Siano were invaded by
some channelled debris flows, which attained a maxi-
mum travel distance of 3500-4000 m (C
alCateRRa
et alii, 2004). Hence, the 1954 runout distances ob-
tained with DAN-W, even though underestimated
with respect to the total distance travelled by the low-
viscosity flow along the Reginna Maior streamline, at
the same time should be considered as the debris flow
characterized by the highest mobility in the Sorrento
Peninsula, thus confirming the exceptional conditions
under which the 1954 landslides have occurred.
Flow velocities obtained with DAN were, as al-
ready pointed out by previous authors, higher than
reasonably expected. In our case, following C
alCa
-
teRRa
et alii (2010), it could be explained with the
high source volumes assumed in the simulations.
The key to the accuracy of simulation obtained
background image
COMPARISON OF 2D MODELS FOR THE SIMULATION OF THE OCTOBER 1954 DEBRIS FLOW AND FLOOD EVENT AT MAIORI (CAM-
PANIA REGION, ITALY)
Italian Journal of Engineering Geology and Environment - Book www.ijege.uniroma1.it © 2011 Casa Editrice Università La Sapienza
521
CALCATERRA D., DE RISO R., EVANGELISTA A., NICOTERA M.V., SANTO A. & SCOTTO DI SANTOLO A. (2004) –
Slope instabilities of the pyroclastic deposits in the Phlegraean district and in the carbonate Apennine (Campania, Italy).
Proc. Intern. Workshop on Occurrence and Mechanisms of Flows in Natural Slopes and Earthfills, Sorrento, Italy, May
14-16, 2003: 61-75, Pàtron Editore, Bologna.
CALCATERRA D., DI CRESCENZO G., DI MARTIRE D., DI NATALE M., GISONNI C., SANTILLO A. & SANTO A.
(2010) - Confronto tra modelli di propagazione di flussi iperconcentrati: il caso di Casamicciola Terme (Ischia). Proc.
XXXII Italian Conf. of Hydraulics and Hydraulic Constructions, IDRA 2010, Palermo, Italy, September 14-17, 2010, CD-
ROM, 10 pp., ISBN: 978-88-903895-2-8.
CALCATERRA D., PARISE M., PALMA B. & PELELLA L. (1999) - The May 5th 1998, landsliding event in Campania (south-
ern Italy): inventory of slope movements in the Quindici area. Proc. Int. Symp. on Slope Stability Engineering, Matsuyama,
Japan, November 8-11, 1999, 1361-1366, Balkema, Rotterdam, The Netherlands.
COUSSOT P. (1994) - Steady, laminar, flow of concentrated mud suspensions in open channel. Journal of Hydraulic Research,
32(4): 535-559.
DE LUCA C., FURCOLO P., ROSSI F., VILLANI P. & VITOLO C. (2010) - Extreme rainfall in the Mediterranean. Proc. Intern.
Workshop on “Advances in statistical hydrology”, STAHY 2010, Taormina, Italy, May 23-25, 2010, 11 pp.. Available at
http://www.risorseidriche.dica.unict.it/Sito_STAHY2010_web/proceedings.htm
DEL PRETE M., GUADAGNO F.M. & HAWKINS A.B. (1998) - Preliminary report on the landslides of 5 May 1998, Campa-
nia, southern Italy. Bull. Eng. Geol. Env., 57(2), 113–129, Springer-Verlag, Berlin-Heidelberg, Germany.
ESPOSITO E., PORFIDO S. & VIOLANTE C. (2003a) - Reconstruction and recurrence of flood-induced geological effects:
the Vietri sul Mare case history (Amalfi Coast, southern Italy). Proc. Int. Conf. on Fast Slope Movements: Prediction and
Prevention for Risk Mitigation, Naples, Italy, May 11-13, 2003, 169-172, Pàtron Editore, Padova, Italy.
ESPOSITO E., PORFIDO S. & VIOLANTE C. (2004) - Il nubifragio dell’ottobre 1954 a Vietri sul Mare-Costa di Amalfi,
Salerno. Scenario ed effetti di una piena fluviale catastrofica in un’area di costa rocciosa. Pubbl. CNR- GNDCI n. 2870,
Naples, Italy.
ESPOSITO E., PORFIDO S., VIOLANTE C. & ALAIA F. (2003b) - Disaster induced by historical floods in a selected coastal
area (southern Italy). Proc. Intern. Workshop on Palaeofloods, Historical Data & Climatic Variability: Applications in Flood
Risk Assessment, Barcelona, Spain, October 16-19, 2002, 143-148, Centro de Ciencias Medioambientales, Madrid, Spain.
FAELLA C. & NIGRO E. (2003) - Dynamic impact of debris flows on the constructions during the hydrogeological disaster
in Campania 1998: failure mechanical models and evaluation of the impact velocity. Proc. Int. Conf. on Fast Slope Move-
ments: Prediction and Prevention for Risk Mitigation, Naples, May 11-13, 2003: 179–186, Pàtron Editore, Padova, Italy.
FEDERICO F. & AMORUSO A. (2008) -
Simulation of mechanical effects due to the impact of fluid-like debris flows on
structures. Italian Journal of Engineering Geology and Environment, 1(2008): 5-24, Casa Editrice Università La Sapienza,
Rome, Italy.
FIORILLO F., GUADAGNO F.M., AQUINO S. & DE BLASIO A. (2001) -
The December Cervinara landslides: further debris
flows in the pyroclastic deposits of Campania (southern Italy). Bull. Eng. Geol. Environ., 60, 171–184, Springer-Verlag,
Berlin-Heidelberg, Germany.
FLO-2D REFERENCE MANUAL (2009) – Available at http://www.flo-2d.com/wp-content/uploads/FLO-2D-Reference-Man-
ual-2009.pdf
FROSINI P. (1955) - Il nubifragio di Salerno del 25-26 Ottobre 1954. Giornale Genio Civile, 3-4, 179-188, Rome, Italy.
HUNGR O. (1995) - A model for the runout analysis of rapid flow slides, debris flows, and avalanches. Can. Geotech. J., , 32,
610–623, New Brunswick, Canada.
HUNGR O., Geotechnical Research (2003) - DAN-W – Release 4, User’s Manual. West Vancouver, BC, Canada.
HUNGR O. & EVANS S.G. (1996) - Rock avalanche runout prediction using a dynamic model. Proc. 7th Intern. Symp. on
Landslides, Trondheim, Norway, 17-21 June, 1996, 233–238, Balkema, Rotterdam, The Netherlands.
IVERSON R.M. (2005) - Debris-flow mechanics. In J
akob
m. & H
unGR
o. (
eds
.) - Debris-Flow Hazards and Related Phenom-
ena. 105-134, Springer Praxis Books, Berlin, Germany.
JULIEN P.Y. & LAN Y. (1991) - Rheology of hyperconcentrations. J. Hydraul. Eng., 117(3), 346-353, American Society of Civil
Engineers, Baltimore, MD, U.S.A..
background image
S. TESSITORE, D. DI MARTIRE, R. MARTINO & D. CALCATERRA
522
5th International Conference on Debris-Flow Hazards Mitigation: Mechanics, Prediction and Assessment Padua, Italy - 14-17 June 2011
LAZZARI A. (1954) - Aspetti geologici dei fenomeni verificatisi nel Salernitano in conseguenza del nubifragio del 25-26 Ot-
tobre 1954. Boll. Soc. Natur. Napoli, 63: 131-142, Naples. Italy.
LIRER L., PESCATORE T., BOOTH B. & WALKER G.P.L. (1973) - Two plinian pumice-fall deposits from Somma-Vesuvius,
Italy. Bull. Geol. Soc. Am., 84: 759-772, Boulder, CO, U.S.A..
MARTINO R. (2003) - Experimental analysis on the rheological properties of a debris-flow deposit. In R
iCkenmann
d. & C
Hen
C.l. (
eds
.) - Debris-flow Hazards Mitigation: Mechanics, Prediction, and Assessment. Proc. 3rd Intern. DFHM Conference,
Davos, Switzerland, September 10-12, 2003, 363-373, Millpress, Rotterdam, The Netherlands.
McDOUGALL S.D. & HUNGR O. (2005) - Dynamic modelling of entrainment in rapid landslides. Can. Geotech. J., 42:
1437-1448.
MCDOUGALL S.D., PIRULLI M., HUNGR O. & SCAVIA C. (2008) - Advances in landslide continuum dynamic modeling.
Proc. 10th Intern. Symp. on Landslides and Engineered Slopes, Xi’an, China, 30 June-4 July, 2008, 835-841, Taylor &
Francis Group, London, UK.
O’ BRIEN J.S., JULIEN P.Y. & FULLERTON W.T. (1993) - Two-dimensional water flood and mudflow simulation. J. Hydraul.
Eng., 119(2); 244-259, American Society of Civil Engineers, Baltimore, MD, U.S.A..
PENTA F., LUPINO R., CAPOZZA F. & ESU F. (1954) - Effetti dell’alluvione del 26 ottobre 1954 nel salernitano. Geotecnica,
6: 3-16, IPI, Milan, Italy.
PIERSON T.C. & COSTA J.E. (1987) - A rheologic classification of subaerial sediment-water flows. In: C
osta
J.e., w
ieCzoRek
G.F. (
eds
.) - Debris flows/avalanches: process, recognition, and mitigation. Geol. Soc. America, Reviews in Eng. Geol., 7:
pp. 1-12. Boulder, CO, U.S.A..
REVELLINO P., HUNGR O., GUADAGNO F.M. & EVANS S.G. (2004) - Velocity and runout simulation of destructive de-
bris flows and debris avalanches in pyroclastic deposits, Campania region, Italy. Environmental Geology, 45: 295–311,
Springer-Verlag, Berlin-Heidelberg, Germany.
SCOTTO DI SANTOLO A. & EVANGELISTA A. (2009) - Some observations on the prediction of the dynamic parameters of
debris flows in pyroclastic deposits in the Campania region of Italy. Natural Hazards, 50: 605–622, Springer, Dordrecht,
The Netherlands.
STRAHLER A.N. (1952) - Hypsometric (area-altitude) analysis of erosional topography. Bull. Geol. Soc. Am., 63, 1117-1142,
Boulder, CO, U.S.A..
TAKAHASHI T. (2007) - Debris flow: mechanics, prediction and countermeasures. Taylor & Francis, Leiden, The Netherlands.
VOELLMY A. (1955) - Über die Zerstörungskraft von Lawinen. Schweizerische Bauzeitung, 73: 159-165, 212-217, 246-249,
280-285, Zurich,Switz.
Statistics