Document Actions

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

background image
15
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
DOI: 10.4408/IJEGE.2014-01.O-02
D
ario
aLESSi
(*)
, F
rancESca
Bozzano
(*)
, a
nDrEa
Di Lisa
(*)
, c
arLo
Esposito
(*)
,
a
nDrEa
Fantini
(**)
, a
Driano
LoFFrEDo
(*)
, S
aLvatorE
Martino
(*)
, F
rancESco
MELE
(***)
,
S
ErEna
MorEtto
(*)
, a
LESSanDra
noviELLo
(*)
, a
LbErto
prEstininzi
(*)
, P
aoLo
saranDrEa
(**)
,
G
abriELE
scarascia Mugnozza
(*)
, L
uca
schiLirò
(****)
& c
hiara
varonE
(****)
(*)
Sapienza Università di Roma - Dipartimento di Scienze della Terra e Centro di Ricerca per i Rischi Gelogici (CERI) - P.le A. Moro, 5 - 00185 Roma, Italy
(**)
Tecnostudi Ambiente s.r.l. - Piazza Manfredo Fanti, 30 - 00185 Roma, Italy
(***)
Regione Lazio - Dipartimento Istituzionale e Territorio - Direzione Regionale Infrastrutture, Ambiente e Politiche abitative
Centro Funzionale Regionale - Via Monzambano, 10 - 00185 Roma, Italy
(****)
PhD student -
Sapienza Università di Roma - Dipartimento di Scienze della Terra - P.le A. Moro, 5 - 00185 Roma, Italy
GeoloGical risks in larGe cities: the landslides triGGered in the city
of rome (italy) by the rainfall of 31 January-2 february 2014
(°)
eXtended abstract
La città di roma sorge in un contesto geologico risultante dall’azione combinata di diversi processi geologici, quali la strutturazione
della catena appenninica, il vulcanismo della provincia comagmatica romana, l’impostazione della valle alluvionale del Fiume tevere.
a tale assetto sono, di fatto, legati i diversi tipi di rischio geologico ai quali la città è esposta ovvero: sismico, vulcanico e idrogeologico,
quest’ultimo comprensivo del rischio per frane, alluvioni ed inondazioni.
il rischio sismico nella città di roma è, invece, riconducibile a tre livelli di sismicità ciascuno dei quali è associato ad una diversa area
sismo-genetica, caratterizzata da una specifica pericolosità. Questi livelli corrispondono:
- alla sismicità regionale connessa alle aree sismogenetiche dell’appennino centro-Meridionale (massima magnitudo attesa pari a 7);
- alla sismicità legata all’evoluzione del Distretto vulcanico dei colli albani, (massima magnitudo attesa pari a 5);
- alla sismicità connessa all’area urbana (massima magnitudo attesa pari a 4).
il rischio vulcanico della città di roma è essenzialmente connesso alla presenza del vicino Distretto vulcanico dei colli albani, oggi
ritenuto in stato di “quiescenza”, a cui è da aggiungere l’attività collaterale tardo-vulcanica presente su ampie zone di roma (tra cui ciam-
pino) dove sono ampiamente documentate emissioni di co
2
, Metano e radon.
il rischio idrogeologico è, invece, principalmente connesso agli eventi di piena dei due fiumi che attraversano la città, il tevere ed
l’aniene che già in passato sono stati causa di storiche inondazioni. per fare riferimento al solo ultimo secolo, esemplare è stata l’esonda-
zione del F. tevere del dicembre 1937 che ha portato all’allagamento di ampi settori della città, tra cui la zona a monte di ponte Milvio,
l’isola tiberina e il Lungotevere di ripa, all’altezza di san Michele.
per ciò che attiene i rischi idrogeologici, i numerosi rilievi collinari sui cui sorge roma sono stati in passato interessati da numerose
frane che, fatta eccezione per alcuni crolli nei depositi vulcanici, sono quasi sempre state associate a intense precipitazioni che, comune-
mente, sono anche responsabili di cospicue inondazioni come recentemente avvenuto nell’ottobre 2008 e 2011.
Le piogge eccezionali che hanno interessato roma nei primi mesi del 2014 hanno riattivato alcune frane storiche, come quelle di via
a.Labriola e di viale tiziano, ed innescato oltre 60 nuovi fenomeni franosi. L’evento meteorologico è avvenuto tra il 31 gennaio ed il 2
febbraio 2014 ed e’ stato caratterizzato da un picco di precipitazioni alle 04:00 a.m. del 31/01/2014; esso, tuttavia, ha interessato la città di
roma con intensità e durata variabili a seconda delle aree geografiche.
un gruppo di lavoro del centro di ricerca per i rischi geologici della sapienza (cEri) in collaborazione con il comune di roma (con
cui è ora in essere una specifica convenzione) e con Roma Natura, ha studiato l’evento meteorologico e le frane da esso innescate. a tal fine,
è stato effettuato un censimento delle frane e dei danni da esse causati a strutture ed infrastrutture, una classificazione per tipologia, litologia
coinvolta e parametri morfometrici dei versanti interessati. L’attività di ricerca ha permesso, inoltre, di realizzare un “catalogo Webgis”,
consultabile pubblicamente on-line dal 9 maggio 2014 sul sito del cEri (www.ceri.uniroma1.it). Le massime intensità di pioggia causate
dall’evento pluviometrico sono state registrate nel settore nord-occidentale dell’area urbana, con valori di circa 105 mm di pioggia cumulata
dalle 03.00 del 31 gennaio alle 03:00 del 1 febbraio. Questa intensità di pioggia è stata responsabile dell’innesco di 68 frane, perlopiù concen-
trate nel settore nord-occidentali di roma (ed in particolare rilievi di Monte Mario e Monte ciocci), dove hanno coinvolto diffusamente i depo-
siti sabbiosi e sabbioso-limosi delle Formazioni di Monte Mario, ponte galeria e valle giulia. in base alle considerazioni qui tratte, le anomalie
nella distribuzione spaziale e nella statistica per litologia degli eventi franosi censiti, rilevate rispetto agli eventi storicamente documentati,
sono da mettere in relazione: i) alla distribuzione delle piogge cumulate in 6 ore, che hanno rappresentato un evento eccezionale caratterizzato
da un tempo di ritorno superiore ai 50 anni; ii) al prevalente coinvolgimento del settore nord-occidentale dell’area urbana della città, dove sono
presenti i rilievi collinari con maggiore dislivello; iii) al prevalente affioramento, in tale settore, di depositi clastici non vulcanici.
L’elaborato cartografico allegato al presente lavoro, riporta una sintesi iconografica delle attività di rilevamento, censimento, cataloga-
zione ed analisi statistica condotte nel presente studio.
(°) This study was performed with the contribution of “Roma Capitale” (agreement between Roma Capitale and CERI - scientific co-ordinator Prof. A.Prestininzi)
background image
d. alessi, f. bozzano, a. di lisa, c. esposito, a. fantini, a. loffredo, s. martino, f. mele, s. moretto,
a. noviello, a. prestininzi, p. sarandrea, G. scarascia muGnozza, l. schilirò & c. varone
16
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
abstract
an exceptional rainfall battered the city of rome (italy) from
31 January to 2 February 2014. the event had variable intensity
and duration in the different parts of the city. the exceptionality
of the event lies in the intensity of rainfall cumulated in 6 hours
(return period > 50 years) and in its uneven distribution over the
urban area. the event triggered a number of landslides of differ-
ent type, which caused substantial damage. researchers from the
Centro di Ricerca per i Rischi Geologici (research centre on
prediction, prevention and control of geological risks - cEri)
of the university of rome “sapienza” carried out field surveys
and assessments immediately after the event. the team detected
and inventoried 68 landslides, mostly occurring in the sandy and
sandy-silty deposits of the Monte Mario, Ponte Galeria and Valle
Giulia Formations
. the complete inventory of the landslides
is accessible via Webgis on cEri’s website http://www.ceri.
uniroma1.it/cn/landslidesroma.jsp. the spatial distribution of
the landslides evidences that 69% occurred in clastic deposits of
sedimentary origin and only 6% in volcanic deposits. this finding
disagrees with more general statistical data, based on the inven-
tory of rome’s historical landslides, indicating that almost 41%
of slope instabilities occur in volcanic deposits and almost 12% in
sedimentary ones. in the data reported here, this apparent contra-
diction is justified by the fact that most the rainfall under review
was concentrated in the north-western portion of rome’s urban
area, whose hills accommodate outcrops of dominantly sedimen-
tary deposits from plio-pleistocene marine and continental cycles.
K
ey
words
: Rome, exceptional rainfall in 2014, landslides, inventory,
WebGIS
introduction
an exceptional rainfall battered the city of rome from 31 Janu-
ary to 2 February 2014. the event, with peak rainfall at 4:00 a.m.
of 31 Jan. 2014, had variable intensity and duration in the differ-
ent parts of the city. as reported by L
EonE
(2014) with reference to
the collegio romano rainfall station, the intensity cumulated in the
month of January 2014 can be regarded as exceptional with respect
to the monthly average. the nW portions of the city recorded the
highest intensity: about 105 mm of average rainfall cumulated from
3.00 a.m. of 31 Jan. to 3:00 a.m. of 1 Feb. the extremely intense
rainfall triggered a number of landslides of different type in rome’s
urban area, resulting into damage of variable extent. several land-
slides, triggered by the intense rainfall event, were also inventoried
in other areas of the Municipality, located a few kilometres away
from rome (c
aStiGLionE
et alii, 2014). the research centre on pre-
diction, prevention and control of geological risks (cEri) of the
university of rome “sapienza” set up a working group, which coop-
erated with the municipality of rome and roma natura in drawing
up a specific inventory. sixty-eight landslides, mainly affecting the
sandy and sandy-silty deposits of the Monte Mario, Ponte Galeria
and Valle Giulia Formations, were identified and mapped. Most of
the landslides did not cause major damage to people and property.
however, 7 slope instabilities (via della Marcigliana, via trionfale,
via cavalieri di vittorio veneto, via del Foro italico, carreggiata
interna del grande raccordo anulare (gra) - inside lane of the
gra ring road, via della Maglianella, piazza dei giuochi Delfici)
disrupted the road network, with consequent vehicle traffic bans and
inconveniences for citizens. given the severity of the event and the
number of related gravitational instabilities, technical authorities
deemed it necessary to conduct a more thorough study (also as part
of an agreement signed with the Municipality of rome).
the exceptional rainfall that occurred in the winter of 2014 and
the damage caused by the numerous landslides that it triggered tes-
tify, once again, rome’s susceptibility to geological hazards, which
can cause substantial damage and significant levels of risk.
the city of rome not only hosts an invaluable historical-
cultural heritage, but is also the administrative hub of the italian
republic. its geological setting originates from a combination of
processes, both geodynamic (e.g. the evolution of the apennine
chain and the volcanism of the roman comagmatic province) and
depositional (e.g. fluvial and marine sedimentation, both signifi-
cantly affected by glacio-eustatic sea level changes in the Quater-
nary) (F
accEnna
et alii, 1995; F
uniciELLo
et alii, 1995; a
LvarEz
et alii, 1996; K
arnEr
& M
arra
, 1998; F
uniciELLo
&
GiorDano
,
2008a; F
uniciELLo
& G
iorDano
, 2008
b
; P
arotto
2008).
the diversified geological-stratigraphic and geomorphologi-
cal setting (a
Scani
et alii, 2008) of the roman area is actually
associated with different types of geological risk. the need thus
arises for carrying out technical-scientific investigations to collect
detailed data on critical natural events, analyse their damage sce-
narios, as well as their role in setting off other natural phenomena.
these investigations play a crucial role in addressing and mitigat-
ing future geological risks. indeed, they represent the groundwork
for developing strategic plans aimed at preventing these risks in
sensitive and critical urban areas, such as the one of the monu-
mental “Eternal city”, through emergency planning in the imme-
diate/short term and land-use planning in the medium/long term.
Mitigating these risks has become an undeferrable priority to a
critical urban area like rome, with a precious archaeological and
monumental heritage and a very high population density.
in the case of rome’s urban area, the above-mentioned proc-
esses include volcanic and seismic activity, as well as floods, which
should be expressed in increasing order of frequency of occurrence
and damage to the city over the historically documented period.
volcanic risk in the city of rome is mainly related to the oc-
currence of the nearby alban hills volcanic district, whose activ-
ity began in the middle pleistocene (K
arnEr
et alii, 2001
b
) and
which now appears to be quiescent (G
iorDano
, 2008). the current
activity is concentrated in lake albano, where strong emissions of
background image
GeoloGical risks in larGe cities: the landslides triGGered in the city of rome (italy)
by the rainfall of 31 January-2 february 2014
17
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
co
2
(c
araPEzza
et alii, 2003; a
nziDEi
et alii, 2008) give evidence
of confined geothermal aquifers (G
iorDano
, 2008; c
iotoLi
et alii,
2013; S
ELLa
et alii, 2013). Furthermore, co
2
, methane and radon
emissions are also documented in many peripheral areas of rome
(i.e. the ciampino plain).
conversely, rome’s seismic risk is related to 3 levels of
seismicity (M
oLin
et alii, 1995; DbMi, 2004; cPti, 2004), each
associated with a different seismogenetic area with a maximum
expected magnitude. these levels are:
- far-field seismicity, connected with the seismogenetic are-
as of the central-southern apennines, (maximum expected
6≤M≤7), with a macroseismic intensity in rome of up to vii-
viii Mcs (P
rEStininzi
et alii, 2005);
- near-field seismicity, linked to the evolution of the alban hil-
ls volcanic district (maximum expected M@ 5);
- urban seismicity, whose epicentres are located within the city
area (maximum expected M@ 4).
seismic events from these seismogenetic areas (b
aSiLi
et
alii, 2008; DiSS W
orKinG
G
rouP
, 2010) may generate different
seismic response effects, depending on the local morphological
and stratigraphic conditions of the urban area. a case in point
is the recent seismic crisis of L’aquila, associated with regional
seismicity: a main shock of M
W
=6.3 on 6 apr. 2009, followed
by about one hundred aftershocks, induced a felt macroseismic
intensity in rome of up to v Mcs (ingv, 2009). the shaking
caused damage of low, average and high degree to approximately
1,019 buildings. the damage distribution was in part controlled
by the local geological setting (b
ozzano
et alii, 2011), as in the
case of the felt umbria-Marche earthquake in 1997 and by the
Fucino one in 1915 (c
iFELLi
et alii, 1997).
the last type of natural geological risk to which rome is ex-
posed is the hydrogeological one; its level of hazard is related to
the occurrence of two rivers, the tiber and aniene, which flow
into the urban area, draining part of its municipal territory and
giving rise to floods after particularly intense rainfall, as hap-
pened in the past (L
E
G
aLL
1953; F
roSini
1977; L
uciani
1985;
b
EncivEnGa
et alii, 1995; r
EMEDia
et alii, 1998) .
a case in point is the latest flood of the tiber (December
1937), which invaded some sectors of the city, e.g. the area uphill
of ponte Milvio, the isola tiberina (tiber island) and the Lungote-
vere di ripa (ripa tiber embankment) near san Michele. this
event led technical authorities to improve flood protection and
control systems by building the corbara dam (completed in 1963).
the landslide risk is strongly linked to the aforementioned hy-
draulic risk. the numerous, dominantly sandy, sandy-clayey and
clayey hills on which rome rises (h
EiKEn
, 2006) historically ex-
perienced landslides of different type, intensity and recurrence. as
reported by multiple historical and bibliographic sources (c
oLoSi
-
Mo
, 1974; S
ciotti
, 1986; P
rEStininzi
, 2000; c
orazza
et alii, 2002;
b
ozzano
et alii, 2006; a
Manti
, 2008), these landslides (except for
rock falls involving volcanic deposits) were often associated with
intense precipitation; in this case, the landslides were induced by an
independent natural phenomenon. rome has never had particularly
severe landslides in terms of mobilised volumes. however, there is
plenty of historical records about the falls of the rupe tarpea (tar-
peian rock) on the capitoline hill, the landslide of viale tiziano at
the foot of the Monti parioli hill (1972), the century-old instability
of the foot of the gianicolo hill (via aurelio saffi - via ugo Bassi)
and the more recent instabilities of via a. Labriola, viale pilsudsky
(Flaminio) and via teulada, at the foot of Monte Mario (a
Manti
et alii, 2008). it is worth noticing that most of these landslides dis-
placed limited volumes of soil and that their intensity was low to
average due, among others, to their low velocity.
the exceptional rainfall of 2014 reactivated the historical
landslides of via a. Labriola and viale tiziano. this paper is a
technical report on the event and associated instabilities, based
on field and remote surveys and assessments that inventoried
the landslides and investigated the possible correlations between
their type of movement and the geological and geomorphological
setting of the affected areas. the field and remote surveys and
assessments led, among others, to the compilation of a Webgis
inventory, which has been posted on cEri website (www.ceri.
uniroma1.it) since 9 May 2014. the inventory makes the findings
from the project available on line to all interested users. a work-
flow diagram, describing the inventory and the statistical analyses
conducted as part of the project, is enclosed hereto.
analyses of the rainfall event
Reconstruction of the rainfall event of 31 Jan.-2 Feb. 2014
in the early hours of 31 Jan. 2014, a high-potential storm
system blew through most of the Latium region (including the
area of the municipality of rome). the storm system, induced
by a trough centred on the western Mediterranean, was fed by
strong sE sirocco winds and by a jet stream of maritime polar air.
the event caused downpours with extremely high intensity peaks
(c
Entro
F
unzionaLE
r
EGionE
L
azio
, 2014).
a more detailed analysis of the event (personal communica-
tion by andrea Luzzi Franzoni) highlights its connection with a
minimum bar pressure (998 hpa) E of Majorca on 30 Jan. 2014.
the event resulted from the occlusion between a cold front from
the atlantic and a warm and moist air stream from the african
continent. the occlusion gave rise to a cyclonic depression,
whose eye was rising and subsiding towards the gulf of Liguria
(Fig. 1a). at the same time, the warm and moist air stream from
africa, no longer attracted by the cyclonic vortex, was interact-
ing with a very moist and cold nW air stream, thereby fuelling
an organised storm system, called v-shape cell, in the heart of the
tyrrhenian sea (Fig. 1b). the cell, blown northwards by south-
ern winds, unleashed heavy and persistent rain over the coast
and hinterland of rome and, in particular, its western and south-
background image
d. alessi, f. bozzano, a. di lisa, c. esposito, a. fantini, a. loffredo, s. martino, f. mele, s. moretto,
a. noviello, a. prestininzi, p. sarandrea, G. scarascia muGnozza, l. schilirò & c. varone
18
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
Fig. 1 - Satellite image from Meteosat MSG3, IR 8.7 channel (infrared), with filter-enhanced colour; filter applied to estimate cloud top temperature
(in °C) and height. Lower temperatures are indicative of a higher cloud top, as in the case of congested cumulonimbuses with severe storms. In
particular, temperature distribution can give an improved understanding of the “physics” of the phenomenon (e.g. the probability of formation of
ice at high altitude ). The number-colour scale only applies to clouds: a) 10:45 a.m. UTC, 30 Jan. 2014; b) 9:00 p.m. UTC, 30 Jan. 2014; c) 2:15
a.m. UTC, 31 Jan. 2014; d) 5:30 a.m. UTC, 31 Jan. 2014 - Source: EUMETSAT
Fig. 2 - Weather forecast at 6:00 a.m. UTC, 31 Jan 2014: a) wind map; b) precipitation map. (Archivio Previsioni Meteorologiche CNR-ISAC)
background image
GeoloGical risks in larGe cities: the landslides triGGered in the city of rome (italy)
by the rainfall of 31 January-2 february 2014
19
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
western metropolitan area. the cell was then deviated towards
nE by southern winds (Figs. 1c-1d; 2a), battering the coast of
rome (Fig. 2b) and its hinterland and, in particular, its western
and sW metropolitan area with intense and persistent rain.
Based on the data recorded by 30 rainfall stations operating in
rome upon the event (Fig. 3a and plate i, section 2), the rainfall
cumulated from 31 Jan. to 3 Feb. 2014 was equal to about 148
mm on average, but much above 200 mm at some stations (e.g.
about 252 and 245 mm at the ottavia and roma Monte Mario
stations, respectively).
however, in this time interval, two main sub-events may be
distinguished. the first (and most severe in terms of amount of
rainfall) started in the early hours of 31 Jan. and ended after about
24 hours, with an average cumulated rainfall of about 105 mm and
hourly intensity peaks exceeding 40 mm in some cases (e.g. 46
mm at the roma Monte Mario station). the second sub-event, of
shorter duration and with less intensity peaks, took place in the af-
ternoon of 2 Feb., from about 1:00 to 6:00 p.m., with a total average
cumulated rainfall of approximately 25 mm and maximum hourly
peaks slightly above 10 mm (e.g. 11.2 mm at the ottavia station).
use was made of the thiessen polygon method in a gis en-
vironment to estimate the areal distribution of the rainfall associ-
ated with the above-mentioned two events. rainfall, especially
on 31 Jan., was clearly concentrated in north-western rome, in
terms of both maximum hourly peak (at 4:00 a.m., Fig. 3b) and
cumulated values for the entire event (Fig. 3c). the areal rainfall
associated with the event of 2 Feb. (Fig. 3d) shows the same pat-
tern, albeit less clearly.
Fig. 3 - a) location of the 30 rainfall stations operating in the area of the municipality of Rome during the event of the winter of 2014; b) areal rainfall
cumulated from 3:00 to 4:00 a.m. UTC, 31 Jan 2014; c) areal rainfall cumulated during the event of 31 Jan. 2014 (from 3:00 a.m. UTC of 31 Jan.
2014 to 3:00 a.m. UTC of 1 Feb. 2014); d) areal rainfall cumulated during the event of 2 Feb. 2014 (from 1:00 p.m. UTC of 2 Feb. 2014 to 6:00
p.m. UTC of 2 Feb. 2014)
background image
d. alessi, f. bozzano, a. di lisa, c. esposito, a. fantini, a. loffredo, s. martino, f. mele, s. moretto,
a. noviello, a. prestininzi, p. sarandrea, G. scarascia muGnozza, l. schilirò & c. varone
20
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
Assessment of the event return period
Estimating the return period for an event similar to the one de-
scribed here requires analysing the time series of daily rainfall in
the study area, in order to check, first of all, the continuity of the
data. tab. 1 displays the recording time interval of each station that
was and/or is operational in the area of the municipality of rome.
unfortunately, out of the 33 stations identified, very few had
statistically significant datasets. Many stations had been in place for
relatively few years (since the early 1990s) and had recorded data in
a discontinuous way. therefore, to build the statistical hydrological
model of the study area, use was made of data from the following 5
stations: collegio romano, roma Eur, roma Macao, roma Flamin-
io and tevere castel giubileo. Even if the latter two stations had
more discontinuous records than the other three, they were included
in the analysis, taking into account the long timespan investigated.
therefore, consideration was given to the daily rainfall data
collected since 1952 (collegio romano, roma Eur and tevere
castel giubileo stations), 1953 (roma Macao station) and 1957
(roma Flaminio station). the probability model relied on the
generalized Extreme value (gEv) distribution introduced by
J
EnKinSon
(1955), whose probability function is:
where µ and α are the location and scale parameters, respectively,
whereas k (shape parameter) expresses the type of distribution.
this distribution is a generalised version of the more known gum-
bel distribution (whose function coincides with the gEv one, when
k is equal to 0), largely used in the study of extreme events.
the variables of “rainfall cumulated” in 2, 5, 10, 30, 60, 90,
120 and 180 days were computed from daily rainfall data. the
maximum values of each variable were extracted, year by year,
from the datasets so generated and the parameters k, α and µ of
the gEv function were determined from the above values, by
applying the probability Weighted Moments (pWM) method in-
troduced by G
rEEnWooD
et alii (1979) and subsequently modified
by h
oSKinG
et alii (1984). Finally, the inversion of the probability
function yielded the values of cumulated rainfall x for each of the
variables (2, 5, 10..… 180 days) and for 6 different return periods
(2, 4, 10, 20, 50 and 100 years). then, these values were interpo-
lated with a view to building the rainfall probability curves.
the comparison of the different curves obtained for the 5 sta-
tions (Fig. 4) reveals that, return period remaining equal, the curves
of tevere castel giubileo and roma Flaminio are those with the
highest rainfall values. this finding emphasises that, in the past, the
two stations (the most representative of the sector most severely hit
by the event of 2014, i.e. the one of the nW quadrant of the city)
had recorded more intense and severe rainfall than the other three.
tab. 2 exhibits the estimated return periods, based on the
above curves, for rainfall cumulated until 30 Jan. (the day prior
to the peak event) and until 31 Jan., respectively. the values infer
that rainfall cumulated until 30 Jan., at all the stations, is far from
exceptional (estimated return periods of 1-2 years); thus, rain-
fall prior to the critical event practically lies within the standard
range, in contrast with rainfall cumulated until 31 Jan.
in this case, while rainfall recorded at roma Eur and roma Ma-
cao continues to be unexceptional, rainfall cumulated in 2 days (30
and 31 Jan.) at roma Flaminio and tevere castel giubileo has a re-
turn period of 15 and 20 years, respectively. conversely, the collegio
romano station, with an estimated return period of 9 years, ranks in
an intermediate position. this means that the event under review was
not only strongly localised in space, but also particularly severe in that
specific sector of the city. this finding is also substantiated by what
has been previously pointed out, i.e. the highest return periods were
obtained for the stations with the highest rainfall probability curves.
considering that only the event of 31 Jan. had a seemingly
exceptional nature, the same statistical analysis was carried out on
the historical data of maximum intensity rainfall (measured in 1,
3, 6, 12 and 24 hours) recorded by the same stations. the curves
so built (Fig. 5) have a different pattern with respect to daily ones.
Tab. 1 - Time series continuity distribution for the daily rainfall data recorded by the stations located in the area of the municipality of Rome
background image
GeoloGical risks in larGe cities: the landslides triGGered in the city of rome (italy)
by the rainfall of 31 January-2 february 2014
21
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
their comparison suggests that, in this case, the rainfall values
recorded by the roma Flaminio and tevere castel giubileo sta-
tions are lower than the ones of the other stations, especially in the
range from 1 to 6 hours. this infers that, in the past, the intensity
of hourly rainfall measured by the first two stations was lower
than the one recorded by the other stations, contrary to what was
observed on a daily basis. it is worth stressing, however, that data
about intense precipitation are generally scantier than daily ones
and that the resulting statistical analyses are usually less reliable.
For instance, in this case, the stations of roma Flaminio and te-
vere castel giubileo recorded these data only for 26 and 25 years,
whereas the roma Macao, roma Eur and collegio romano sta-
tions recorded them for 36, 35 and 32 years, respectively.
to estimate the return period of the event of 31 Jan. 2014,
use was made of the rainfall probability curves pertaining to the
intensity of 1-24 hours (tab. 3).
the results show that, even on an hourly basis, the stations
roma Flaminio and tevere castel giubileo have the longest
return periods (especially for rainfall cumulated in 6 and more
hours) whereas, at the other stations, the recorded event lay
within the standard or quasi-standard range. additionally, it is
worth noting that the estimated return periods (in the range of 30
to about 100 years) are definitely longer for the hourly recorded
event than for the daily recorded one, disregarding the previ-
ously described uncertainties.
in brief, statistical analyses validate the assumption that the
Fig. 4 - Rainfall probability curves with return periods of 2, 4, 10, 20, 50 and 100 years for daily cumulated rainfall at the Collegio Romano (7), Tevere
Castel Giubileo (29), Roma Eur (16), Roma Flaminio (17) and Roma Macao (18) stations. The black dots denote the rainfall values recorded by
the stations during the event of 31 Jan. 2014
Tab. 2 - Estimated return periods (years) for rainfall cumulated in 2, 5, 10, 20, 30, 60, 90, 120 and 180 days prior to 30 Jan. and on 31 Jan. 2014
background image
d. alessi, f. bozzano, a. di lisa, c. esposito, a. fantini, a. loffredo, s. martino, f. mele, s. moretto,
a. noviello, a. prestininzi, p. sarandrea, G. scarascia muGnozza, l. schilirò & c. varone
22
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
sector most severely stricken by the event of 2014 was also the
one historically most prone to extreme rainfall (at least on the
basis of daily records). Moreover, unlike the three other stations
(at which the recorded event does not appear to be exceptional),
the two most representative stations of the north-western sector
are also those for which the longest return periods were calcu-
lated. this confirms that the event was extremely localised (as
evidenced by areal rainfall) and particularly intense (especially in
terms of rainfall cumulated in 6 and more hours).
analysis of the landslides
Inventory
a detailed engineering-geological survey began in the morn-
ing of 31 Jan. and went on in the following 3 weeks. the survey
collected qualitative and quantitative data on the landslides trig-
gered by the exceptional rainfall that rome experienced from 31
Jan. to 2 Feb. 2014.
the acquisition of field data (through systematic inspections)
encompassed the direct observation of morphological and strati-
graphic evidence, the collection of photographic material and the
identification of the landslide location via gps. the objective-
ness of the collected dataset was checked by a comparative pro-
cedure, i.e. taking into account the data inventoried by different
researchers for each landslide.
the survey was conducted immediately after the critical
rainfall event, to correlate it with the landslides identified and to
investigate their original geomorphological features (see plate i,
section 5, for some examples).
Based on morphological evidence, the landslides were classified
from the standpoint of their type of movement under the classifica-
tion of v
arnES
(1978). therefore, they were distinguished into falls,
translational slides, rotational slides, earth flows, debris flows and
complex slides. in places where human action had obliterated the
Fig. 5 - Rainfall probability curves with return periods of 2, 4, 10, 20, 50 and 100 years for hourly cumulated rainfall at the Collegio Romano (7), Tevere
Castel Giubileo (29), Roma Eur (16), Roma Flaminio (17) and Roma Macao (18) stations. The black dots denote the rainfall values recorded by
the stations during the event of 31 Jan. 2014
Tab. 3 - Estimated return periods (years) for rainfall cumulated in 1, 3,
6, 12 and 24 hours during the event of 31 Jan. 2014
background image
GeoloGical risks in larGe cities: the landslides triGGered in the city of rome (italy)
by the rainfall of 31 January-2 february 2014
23
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
kinematic records of the triggering mechanism, the landslide was
inventoried as “nd” (i.e. with an unclassifiable type of movement).
out of the 68 landslides surveyed (plate i, section 1), only 39
had such a size as to be mapped in detail (landslide body and esti-
mated surface area). among the landslides, 5 were classified as falls,
22 as rototranslational slides, 19 as translational slides, 5 as earth
flows, 2 as debris flows, 11 as complex slides (v
arnES
; 1978) and 4
as not classified based on the collected observational data (Fig. 6)
all the landslides surveyed were differentiated into multiple
lithological classes based on their dominant lithotypes (tab. 4).
the graphs of Fig. 7 exhibit the results of the lithological dis-
tribution for each type of landslide, while the graph in plate i,
section 3 shows the percentage distribution of the landslide by
geological formation.
a unique alphanumerical code was assigned to each of the
landslides (alphabetical characters for the type of movement fol-
lowed by a progressive number). For instance, the code rtr4
identifies rotational slide no. 4 (tab. 5).
From 31 Jan. to 18 Feb. 2014, 68 landslides (most of which
occurring from 31 Jan. to 7 Feb.) were inventoried.
the areal distribution of the landslides proved to be non-
homogeneous within the area under review, since the landslides
Fig. 6 - Percentage distribution of the 68 landslides inventoried in
Rome’s urban area after the exceptional rainfall of 31 Jan- - 2
Feb. 2014, by type of movement
Tab. 4 - Association of lithological classes and formational units from
CARG (C
arta
G
eoloGiCa
d
’i
talia
, 2008)
Fig. 7 - Percentage distribution of the 68 landslides inventoried in
Rome’s urban area after the exceptional rainfall of 31 Jan.- 2
Feb. 2014, by lithotypes involved and type of landslide
Tab. 5 - Relationship between landslide ID and landslide mechanism
background image
d. alessi, f. bozzano, a. di lisa, c. esposito, a. fantini, a. loffredo, s. martino, f. mele, s. moretto,
a. noviello, a. prestininzi, p. sarandrea, G. scarascia muGnozza, l. schilirò & c. varone
24
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
were concentrated in the north-western part of rome (Fig. 8).
the reliefs of Monte Mario and Monte ciocci (Fig. 9) had a
high concentration of landslides: 25 of the 68 landslides invento-
ried, i.e. over 30% of the total. in particular, 12 slides affected the
sE slope of Monte Mario, while 13 occurred on the sE and sW
slopes of Monte ciocci.
Most of these slides were translational and rotational and
involved relatively shallow portions of the slopes, consisting of
man-made fills and non-volcanoclastic lithotypes, mostly clayey,
from the Monte Vaticano Formation. to a lesser extent, they af-
fected dominantly sandy deposits belonging to the Monte Mario
Formation
and to the conglomeratic lithofacies of the Ponte
Galeria Formation (Monte Ciocci unit Auctt.
) (Fig. 10a).
By contrast, the landslides spared the hills located on the left
bank of the tiber; the only exception was the western slope of the
Fig. 8 - Location of landslides inventoried in Rome’s urban area after the exceptional rainfall of 31 Jan- - 2 Feb. 2014; each of them is associated with a
unique alphanumerical code. The landslides are colour-coded based on their mechanism of rupture
Fig. 9 - Statistical analysis of the 68 landslides inventoried. Distribution of
the number of landslides detected at Monte Ciocci (purple), Monte
Mario (yellow) and in the remaining urban area of Rome (green)
background image
GeoloGical risks in larGe cities: the landslides triGGered in the city of rome (italy)
by the rainfall of 31 January-2 february 2014
25
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
the remaining landslides were more unevenly distributed in
the urban area of rome and anyway ever in its northern quadrant,
as in the case of the landslide on the slope along the inner lane
of the grande raccordo anulare (gra) ring road, the one along
via del Foro italico and the falls observed along viale tiziano,
which involved cemented sands supposedly belonging to the
Valle Giulia Formation.
Most of the landslides were classified as dominantly rota-
tional and subordinately translational (tab. 6). rotational slides
also had the largest areal extent, even if the most extensive slide
“tr7” (occurred along via s. simoni on the south-eastern side
of Monte ciocci on 31 Jan., see plate i, section 5a) had a trans-
lational kinematics. it should be pointed out that the areal data
shown in tab. 6 and in the graphs do not refer to all of the land-
slides, but only to those extending over a surface area of more
than 35 m
2
(39 landslides).
the numerous slope instabilities identified may be classified
as shallow slides, in that their maximum depth (rotational slide
of via cavalieri di vittorio veneto “rtr15”, see plate i, section
5D) is about 5 m, whereas most of the landslides are less than 1
villa glori hill, which had a translational slide of limited extent in-
volving sandy travertine deposits from the Valle Giulia Formation.
another area with landslides induced by rainfall was the
western portion of rome’s urban area, along via della Magliana
and via della Maglianella (Fig. 10b). here, 10 landslides oc-
curred within dominantly sandy sedimentary deposits. among
them, a debris flow involved sandy silts presumably belonging to
the Ponte Galeria Formation.
numerous landslides were also observed inside “urban
parks”, including parco del pineto and riserva naturale della
Marcigliana (Marcigliana nature reserve). at parco del pineto,
the landslides were shallow and of limited extent and originated
inside sandy-silty deposits with thin intercalations of gray clay
ascribed to the Ponte Galeria Formation. some slides at the ris-
erva naturale della Marcigliana unusually affected the tuff cliffs
belonging to the Tufi Stratificati Varicolori di La Storta Forma-
tion
. in particular, the translational slide “tr10”, involving in
part the soil and in part the debris generated by tuff weathering,
invaded the roadway of via della Marcigliana, which was thus
closed to vehicle traffic (Fig. 11).
Fig. 10 - Examples of landslides triggered in Rome by the exceptional
rainfall event of 31 Jan.-2 Feb. 2014: a) complex side along
Via Simone Simoni (Monte Ciocci); b) rotational slide along
Via della Maglianella
Fig. 11 - Examples of landslides triggered in Rome by the exceptional
rainfall event of 31 Jan.-2 Feb. 2014: translational slide along
Via della Marcigliana
Tab. 6 - Number of landslides and their surface area (maximum - max;
average - av; total - tot) by type of mechanism of rupture. The
analysis was conducted on the 39 mappable landslides
background image
d. alessi, f. bozzano, a. di lisa, c. esposito, a. fantini, a. loffredo, s. martino, f. mele, s. moretto,
a. noviello, a. prestininzi, p. sarandrea, G. scarascia muGnozza, l. schilirò & c. varone
26
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
ume (astM D854, 2010) 26.22 kn/m
3
; natural water content
(astM D2216, 2010) 25.7%; plastic and liquid atterberg limits
(astM D4318, 2010) 22.9% and 40.1% respectively, and min-
eralogical activity 0.31. Direct shear tests (astM D3080, 2004)
yielded: a value of 27° for the internal friction angle correspond-
ing to the peak shear strength; a drained cohesion of about 20 kpa,
and a value of 25° for the internal friction angle corresponding
to the residual shear strength. Furthermore, based on in-situ tests
performed by the Municipality of rome in the same landslide
area, the undrained cohesion is of about 300 kpa.
in spite of their limited surface area, 40% of the landslides
m deep. Due to the limited thickness of the landslide masses, they
are generally constituted of weathered soils. to exemplify the
physical and mechanical properties of the weathered soil layers
involved in the landslides on the Monte Mario hill, some undis-
turbed samples were obtained from the main scarp of the afore-
mentioned landslide “rtr15”. these samples were tested at the
laboratory of “geologia applicata” of the Department of Earth
sciences, university of rome “sapienza”. as indicated by lab
tests, the sampled soil is a weathered silty-clayey mixture classi-
fied as cL under the uscs classification (astM D2487, 2010).
Lab-determined values are as follows: solid weight per unit vol-
Fig. 12 - Examples of damage caused in Rome by the landslide triggered by the exceptional rainfall event of 31 Jan.- 2 Feb. 2014: a) rototranslational slide
of Via del Foro Italico (Tangenziale nord), detail of damage to the low wall bordering the roadway; b) rototranslational slide along Via Cavalieri
di Vittorio Veneto, which was then closed to vehicle traffic; c) translational slide on the eastern side of Monte Ciocci along Via S. Simoni; d) detail
of damage caused by the landslide along Via S. Simoni to the roofed car parks
background image
GeoloGical risks in larGe cities: the landslides triGGered in the city of rome (italy)
by the rainfall of 31 January-2 february 2014
27
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
with the inspirE (infrastructure for spatial information in Eu-
rope) Directive.
the deliverables from the project of identification, charac-
terisation and classification of the landslides observed in rome’s
urban area consist of: detailed thematic maps (scale 1:1,000),
subsequently projected onto the topographic base ctr 1:5000 of
the Latium region, and of the on-line Webgis inventory posted
on cEri website (www.ceri.uniroma1.it).
at each stage of the research project (Fig. 14), the instability
data were validated by systematically checking whether the ele-
ments viewed in the gis environment matched the data collected
during field surveys.
the geodatabase is made up of different tables and feature-
classes, interconnected by key fields (e.g. the unique identifier of
the landslide, iD) through one-to-many and one-to-one relations,
which return the data associated with the landslide, e.g.: date of
observation, photographs, location and geographic coordinates
in Wgs84 utM 33n, mechanism of rupture and morphometric
features (namely, height of the main scarp, width and length of
the landslide body). Furthermore, for each landslide, the geodata-
base describes the lithotypes involved and the belonging “forma-
tion” reported in the official geological map, sheet roma 374,
scale 1:50,000 (F
unicELLo
& G
iorDano
, 2008). the database also
specifies whether the landslide caused or did not cause damage to
buildings and structures and the type of damage (tab. 7).
the distribution of the landslides generally confirmed the one
reported in official inventories (avi, iFFi, pai, prg) that iden-
tify the centres of landslides in rome’s urban area in the morpho-
logical highs and hillsides of the valleys, carved by the stream
network see plate i, section 1).
the “piano regolatore generale” (master land-use plan) of
caused major damage. Most of them posed a high risk to build-
ings and infrastructures of critical importance to urban transport.
in particular, at Monte Mario, the rotational slides of via trion-
fale “rtr16” (see plate i, section 5D) and of via cavalieri di vitto-
rio veneto “rtr15” (Fig. 12b) disrupted transport routes and made
it necessary to evacuate some buildings from 8 Feb. on. among the
landslides arising on the eastern side of Monte ciocci, two caused
damage to buildings: the complex landslide “coM1” (Fig. 12a)
and the translational slide “tr7” made some roofed car parks inac-
cessible owing to the accumulation of material (Fig. 12c-d).
along via cassia, in the section extending from piazza dei giuo-
chi Delfici to via pareto, the shallow translational landslide “tr16”
involved the weathered layer of the Valle Giulia Formation, invading
the roadway that was subsequently protected by jersey barriers.
the landslides that caused the most serious damage to trans-
port routes, especially to high-speed roads, took place near via del
Foro italico (“rtr12”) and along the gra ring road near casal del
Marmo (“rtr14”) on 31 Jan. the rotational landslide “rtr12”
(Fig. 12a) affected sands, sandy silts and man-made fills over an
area of approximately 320 m
2
, entailing the collapse of the re-
taining wall bordering the roadway and the pouring of part of the
landslide debris onto the roadway. also the landslide that occurred
along the gra had a rototranslational mechanism of rupture and
involved silty-sandy deposits. the detachment area extended over
about 160 m
2
, whereas the slide material, filling the inner lane of
the gra, disrupted road traffic but caused minor structural damage
to the retaining wall bordering the roadway. Furthermore, multiple
scarps with a height of about 50 cm were detected in the area uphill
of the landslide, pointing to a possible retrogressive mechanism.
a quantitative analysis of the damage caused by the differ-
ent slides, based on the inventory discussed here, evidences that
translational landslides were responsible for 41% of the damage
(11 slides), rototranslational landslides for 29% (8 slides), where-
as the remaining types of movement caused less than 15% of the
damage. it is worth emphasising that the falls did not induce in-
frastructural damage (Fig. 13).
all the collected data were fed to a geodatabase (see plate i,
section 4) and managed, queried and processed in a gis environ-
ment. Finally, they were implemented as metadata in accordance
Fig. 13 - Percentage of damage associated with each mechanism of rupture
Fig. 14 - Workflow for implementing the geodatabase and the related
queries via WebGIS
background image
d. alessi, f. bozzano, a. di lisa, c. esposito, a. fantini, a. loffredo, s. martino, f. mele, s. moretto,
a. noviello, a. prestininzi, p. sarandrea, G. scarascia muGnozza, l. schilirò & c. varone
28
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
also by anthropogenic action that changed the gradient of and the
stresses within the slope (a
Manti
et alii, 2008). the latest activa-
tion on the slope of viale tiziano took place in november 2007 and
was very similar to the reactivation of 1972 (a
Manti
et alii, 2012)
and to the landslide recorded in February 2014.
Landslides distribution analysis
a morphometric analysis of the 68 landslides was conducted
on the cartographic base onto which the field surveys had been
projected. the analysis made it possible to map the statistical dis-
tribution of the landslides vs. their surface area and the gradient
of the slope involved. the analysis took into consideration only
the 39 landslides whose areal extent could be mapped.
the diagrams of Fig. 15 show these distributions, divided by
size of the landslide: Fig. 18a (left) refers to all of the landslides,
while Fig. 15b (right) concerns only those extending for more
than 1,000 m
2
(about 25% of the total). though Fig. 15a shows
an apparently random distribution, Fig. 15b evidences a correla-
tion between the gradient of the slope and the size of the process.
then, a weighted landsliding index (iF%) was calculated.
this index is given by the percentage ratio of the landslide area
for each lithotype to the total landslide area (tab. 8). the in-
dex was calculated only for the 39 landslides whose areal extent
could be determined.
rome/2007 (carta di
vulnerabilità geologica del territorio comu-
nale - map of geological vulnerability of the municipal area)
shows areas historically affected by landslides. the plan indicates
that 56% of the landslides triggered by the rainfall event of 31
Jan. 2014 developed in areas previously classified as unstable.
in particular, in the map of geological vulnerability of the
municipal area, the sE slopes of Monte Mario and Monte ciocci
feature a wide belt that is prone to gravitational deformations.
consequently, the landslides observed along via cavalieri di vit-
torio veneto - via trionfale are located in an area already mapped
as “landslide-prone”.
as regards the earth flow recently reactivated on the slope of
Monte ciocci along via Labriola, the first documented activa-
tion leads back to 1960 and required the evacuation of 60 people,
whereas the latest one occurred in 1998, when the translational
slide evolved into an earth flow that damaged the retaining wall
bordering the road (a
Manti
et alii, 2008). also the complex slide,
detected on the sE side of Monte ciocci along via s. simoni, is
the reactivation of a historical landslide occurred in December
2008 (a
Manti
& F
abbri
, 2014).
the falls observed along viale tiziano occurred in an area his-
torically susceptible to this type of gravitational movement. Bib-
liographic sources of the 1930s report the evolutionary trend of the
travertine walls of viale tiziano, with falls and accumulation of
material downslope (D
E
a
nGELiS
D’o
SSat
, 1932). the main event
mentioned in the databases of avi and iFFi is the one of 1972,
which mobilised about 500 m
3
of soil. this event was induced not
only by geological and geomorphological predisposing factors, but
Tab. 7 - Database report
Tab. 8 - Explanatory table of the weighted landslide susceptibility index
(IF %)
background image
GeoloGical risks in larGe cities: the landslides triGGered in the city of rome (italy)
by the rainfall of 31 January-2 february 2014
29
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
tent. the use of different colours for geomorphological symbols
makes it possible to identify the type of mechanism of rupture,
described by an appropriate legend (Fig. 16).
the Webgis offers different query options: free text search,
by address or geographic coordinates (latitude, longitude), and
guided search, by type of movement or merely by delineating the
area of interest on the map (Fig. 17).
By pressing “stampa” (print), the data of the map may be
printed in .pdf format or as images in different formats.
Each level that the user may select, by a single click of the
WebGIS inventory
the Webgis for the inventory discussed in this paper has been
made publicly available on cEri’s website (www.ceri.uniroma1.
it), in the Webgis section, since 9 May 2014. users may view the
location of the landslides and the related geomorphological maps.
the nature of the queryable data depends on the viewing
scale: for scales of up to 1:10,000, the events are displayed
as dots with respect to the centroids of the landslide body. at
smaller scales, the geomorphological maps are divided into ar-
eal and linear forms of landslides with a minimum mappable ex-
Fig. 15 - Distribution of the landslides (areal extent vs. relative slope gradient): a) all landslides, b) only landslides having an areal extent of more than
1,000 m
.
. The analysis was made on the 39 mappable landslides
Fig. 16 - On-line WebGIS: example of simplified geomorphological map using different colours
background image
d. alessi, f. bozzano, a. di lisa, c. esposito, a. fantini, a. loffredo, s. martino, f. mele, s. moretto,
a. noviello, a. prestininzi, p. sarandrea, G. scarascia muGnozza, l. schilirò & c. varone
30
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
Fig. 17 - On-line WebGIS: query windows for guided search and related options
Fig. 18 - On-line WebGIS: pop-up window displaying a single landslide
background image
GeoloGical risks in larGe cities: the landslides triGGered in the city of rome (italy)
by the rainfall of 31 January-2 february 2014
31
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
along the eastern foot of the hills. this phenomenon might have
decreased the temperature of air moving uphill at Monte Mario
and Monte ciocci, making it water vapour-saturated and inducing
condensation and precipitation. nevertheless, considering an av-
erage adiabatic temperature decrease (vertical thermal gradient)
of 1°c/100 m of ascent along the hills (P
inna
, 1977) and the small
elevation of rome’s hills (Monte Mario 144 m and Monte ciocci
77 m above sea level), the orographic cause of the concentrated
rainfall that lashed the north-western portions of the city can be
ruled out. indeed, the small elevation of the hills does not permit
water vapour to reach saturation.
hence, it is more realistic to take into account weather condi-
tions at a wider scale, i.e. the weather conditions of italy on 31
Jan. 2014, the day with the maximum intensity of rainfall.
the high number of landslides that ravaged rome from 31
Jan. 2014 to 7 Feb. 2014 is thus related to its particular weather
conditions, which led to an exceptional concentration of the in-
tensity of hourly rainfall (return period of rainfall cumulated in
mouse or via the search option, is associated with a pop-up win-
dow; this window displays the basic data of the landslide, e.g. iD
of the event, mechanism of rupture, location, lithotype involved
and belonging geological formation (according to the legend of
sheet 374 of F
unicELLo
& G
iorDano
, 2006), as well as a photo-
graph of the landslide.
the pop-up window also comes with: a link to a detailed da-
tasheet of the landslide, with the morphometric data collected in
the field (Fig. 19); a wider description of the phenomenon; photo-
graphs; an except from the official geological map (F
unicELLo
&
G
iorDano
, 2006) to a scale of 1:10,000, and a geomorphological
map to a scale of 1:1,500.
discussion
an analysis was made of the spatial distribution of the rainfall
event that triggered landslides in rome between the end of Janu-
ary and the beginning of February. the analysis reported herein
highlights the non-homogeneous distribution of the event, since it
mostly affected the nW portions of the urban area, especially the
slopes of Monte Mario and Monte ciocci, and caused the highest
number of landslides just in that area (Fig. 20).
in view of the above, it might be assumed that the spatial dis-
tribution of the phenomenon was affected by local orography and
that the heavy rainfall originated from the ascent of air masses
Fig. 20 - Lithological map of the city of Rome, with the location of the in-
ventoried landslides and the distribution of rainfall on 31 Jan.
2014 - Thiessen polygon method
Fig. 19 - On-line WebGIS: detailed technical schedule of a single
landslide
background image
d. alessi, f. bozzano, a. di lisa, c. esposito, a. fantini, a. loffredo, s. martino, f. mele, s. moretto,
a. noviello, a. prestininzi, p. sarandrea, G. scarascia muGnozza, l. schilirò & c. varone
32
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
with a return period of more than 50 years; ii) the dominant in-
volvement of the nW sector, whose hills have more differences in
height than other sectors of the urban area, and iii) the occurrence
in such sector of dominantly non-volcanoclastic deposits.
such anomaly highlights the importance of reconstructing
triggering scenarios, with a view to predicting the spatial distri-
bution and types of triggered landslides (trigger-controlled sta-
tistics), rather than relying on general data provided by medium-
long term statistics. indeed, the latter data disregard the triggers,
placing greater emphasis on the role of landslide predisposing
factors (proneness-controlled statistics).
all this stresses the potential value for the city of rome of
modelling and quantifying specific hazard scenarios, so as to ad-
dress geological risks in urban areas and pursue city management
policies aimed at mitigating their effects and damage.
acknowledGments
the authors thank: roma capitale and, in particular, igna-
zio Marino (Mayor), paolo Masini (member of the municipal
government in charge of “sviluppo delle periferie, infrastrut-
ture e manutenzione urbana” - development of outskirts, infra-
structure and urban maintenance), Maurizio pucci from “Di-
rezione promozione, pianificazione strategica e coordinamento
attuativo di progetti speciali, per lo sviluppo e la valorizzazione
della città di roma e le sue risorse” - department of promo-
tion, strategic planning and coordination of the implementa-
tion of special projects for enhancing the value of rome and
its resources), for their logistic support; the technical special-
ists from the Municipality of rome, Maurizio allevi, angelo
canalini, Mariachiara galiano, Fabrizio Mazzenga, theo uber,
for their technical support; stefano casini, giuseppe De pisa,
Fabrizio Foschi from roma natura, for providing data and
technical support in the areas of the urban parks; andrea Luzzi
Franzoni from the “centro funzionale della protezione civile
della regione Lazio” (functional civil protection centre of the
Latium region), for providing data on and interpretations of the
rainfall event discussed in this paper.
6 hours: 100 years) and not of the intensity of daily rainfall (un-
exceptional return periods of roughly 20 years). in fact, given its
meridian direction, the storm system mostly affected the non-
volcanoclastic deposits that oucrop near the reliefs along the right
bank of the tiber (especially those of Monte Mario and Monte
ciocci). these deposits are particularly prone to instabilities ow-
ing to their geomorphological setting (steep slopes) and to the
strong anthropogenic pressure that they experienced in the sec-
ond half of last century (b
ozzano
et alii, 2006). Moreover, their
prevalently clayey-silty or sandy nature generally makes them
more susceptible to landslides induced by intense precipitation.
the analysis of the landslide distribution, carried out as part of
the research project and focused on the event occurred from 31
Jan. to 7 Feb. 2014, shows that 69% of the phenomena developed
inside dominantly sandy and sandy-silty deposits (Monte Mario,
Ponte Galeria, Valle Giulia Formations
) and only 6% inside the
“tuffs” from the sabatini Mts. volcanic district.
By contrast, the inventory of historically documented land-
slides (a
Manti
, 2008) demonstrates that most of them (dominantly
falls) involved the southern sector of rome, namely volcanic de-
posits from the alban hills volcanic district. in effect, the inven-
tory indicates that 41% of the landslides reported in rome involve
“tuffs” and “pozzolanas” and only 1.7% dominantly sandy depos-
its, e.g. those ascribable to the Monte Mario, Ponte Galeria and
Valle Giulia Formations. Based on the findings from this study, this
apparent statistical anomaly may be due to three main factors: 1)
spatial distribution of rainfall, confined to the nW zone of rome’s
urban area; 2) non-homogeneous geological setting of the hills lo-
cated on the right and left banks of the tiber; 3) higher topographic
elevation of reliefs on the right bank of the tiber.
conclusions
the analysis of the landslides triggered in rome by the excep-
tional rainfall of late January-early February 2014 showed an ab-
normal statistical distribution vs. the one of historical landslides.
the anomaly was correlated with: i) the distribution of rainfall
cumulated in 6 hours, which represented an exceptional episode
references
a
LvarEz
W., a
MMErMan
a.J., r
EnnE
P.r., K
arnEr
D.b., t
ErrEnato
n. & M
ontanari
a., (1996) - Quaternary fluvial-volcanic stratigraphy and geochronology
of the Capitoline hill in Rome. geology, 24: 751-754.
a
Manti
M., c
ESi
c. & v
itaLE
v. (2008) - Le frane nel territorio di Roma (ed) “La Geologia di Roma”. Memorie Descrittive della carta geologica d’italia,
80: 83-117.
a
Manti
M., c
hiESi
v. & G
uarino
P.M. (2012) - The 13 November 2007 rock-fall at Viale Tiziano in Rome (Italy). nat. hazards Earth syst. sci., 12: 1-12.
a
Manti
M. & F
abbri
M. (2014) - convenzione protezione civile roma capitale - ordine dei geologi del Lazio e progetto pilota comune di roma -
ispra - ogL. 11 febbraio 2014. http://geologilazio.it/public/file/2014/03/Amanti_Fabbri_prot_civile.pdf
a
nziDEi
M., c
araPEzza
M.L., E
SPoSito
a., G
iorDano
G., t
archini
L. & L
ELLi
M. (2008) - The Albano Maar Lake high resolution bathymetry and dissolved
CO
2
budget (Colli Albani District, Italy): constrains to hazard evaluation. J. volcanol. geotherm. res., 171: 258-268.
a
rchivio
P
rEviSioni
M
EtEoroLoGichE
cnr-isac. http://www.isac.cnr.it/dinamica/projects/forecasts/moloch/
a
Scani
F., b
ozzano
F., b
uccELLato
a., D
EL
M
ontE
M., M
attEucci
r. & v
ErGari
F. (2008) - Evoluzione del paesaggio e antiche vie di drenaggio nell’area
background image
GeoloGical risks in larGe cities: the landslides triGGered in the city of rome (italy)
by the rainfall of 31 January-2 february 2014
33
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
de “Il Castellaccio” (Roma) da indagini geologiche, geomorfologiche e archeologiche. geologica romana 41: 55-77.
astM D2487 (2010) - Standard Practice for Classification of Soils for Engineering Purposes (Unified Soil Classification System). 3-5.
astM D854 (2010) - Standard Test Methods for Specific Gravity of Soil Solids by Water Pycnometer, 2.
astM D2216 (2010) - Standard Test Methods for Laboratory Determination of Water (Moisture) Content of Soil and Rock by Mass. 1-5.
astM D4318 (2010) - Standard Test Methods for Liquid Limit, Plastic Limit, and Plasticity Index of Soils. 1.
astM D3080 (2004) - Standard Test Method for Direct Shear Test of Soils Under Consolidated Drained Condition. pp. 1-3.
v
ariouS
a
uthorS
(2011) - La pericolosità sismica nel Lazio. a cura di S
caraScia
M
uGnozza
G. Quaderni dell’ “italian Journal of Engineering geology and
Environment”, 4.
avi (1996) - Aree vulnerate italiane. a cura di coordinamento della protezione civile al gruppo nazionale per la Difesa dalle catastrofi idrogeologiche
(gnDci) del consiglio nazionale delle ricerche (cnr). http://avi.gndci.cnr.it
b
aSiLi
r., v
aLEnSiSE
G., v
annoLi
P., b
urrato
P., F
racaSSi
u., M
ariano
S., t
ibErti
M.M. & b
oSchi
E. (2008) - The Database of Individual Seismogenic
Sources (DISS), version 3: summarizing 20 years of research on Italy’s earthquake geology. tectonophysics, Doi:10.1016/J.tEcto.2007.04.014
b
EncivEnGa
M., D
i
L
orEto
E. & L
iPEri
L. (1995) - Il regime idrologico del Tevere, con particolare riguardo alle piene nella città di Roma. Memorie
descrittive della carta geologica d’italia, 50: 125-172.
b
ozzano
F., G
iacoMi
a.c., M
artino
S., c.P.vv.F. r
oMa
(2011) - Scenari di danneggiamento indotto nella città di Roma dalla sequenza sismica aquilana
del 2009. italian Journal of Engineering geology and Environment, 11(2): 5-22.
b
ozzano
F., M
artino
S. & P
riori
M (2006).- Natural and man-induced stress evolution of slopes: the Monte Mario hill in Rome. Environmental geology,
50(2): 505-524.
c
araPEzza
M.L., b
aDaLaMEnti
b., c
avarra
L. & S
caLzo
a. (2003) - Gas hazard assessment in a densely inhabited area of Colli Albani Volcano (Cava dei
Selci, Roma). J. volcanol. geotherm. res. 123: 81-94.
c
aStiGLionE
L., c
oPPoLa
G., D’a
rMEnto
E. & r
oSSo
F. (2014) - Gli eventi franosi ed alluvionali del 31 gennaio 2014 a Riano (Roma). professione geologo,
39: 10-14.
c
Entro
F
unzionaLE
r
EGionE
L
azio
(2014) - Rapporto di evento del 31 gennaio - 04 febbraio 2014: 41
c
iFELLi
F., D
onati
S., F
uniciELLo
F., G
aSParini
c., S
aracEni
a. & t
ErtuLLiani
a. (1997) - Il risentimento a Roma del terremoto umbro-marchigiano del 14
ottobre 1997. proceedings of Xvi cnr-gngts, 11-13 november 1997, roma
c
iotoLi
G., E
tioPE
G., F
LorinDo
F., M
arra
F., r
uGGiEro
L., S
auEr
P.E. (2013) - Sudden deep gas eruption nearby Rome’s airport of Fiumicino. geophysical
research Letters, Doi: 10.1002/2013gL058132
c
oLoSiMo
p. (1974) - Il dissesto di Viale Tiziano in Roma: Studio di geologia ambientale applicato alla tecnologia dell’Architettura. geologia tecnica, 3/1974
c
orazza
a., L
EonE
F. & M
azza
r. (2002) - Il quartiere di Monteverde a Roma: sviluppo urbanistico e dissesti in un’area urbana. geologia
dell’ambiente - anno X - 1/2002.
cpti04 (2004) - Catalogo parametrico dei Terremoti Italiani dal 217 a.C. al 2002. ingv. Website http://emidius.mi.ingv.it/cPti04/
DBMi04 (2004) - Database macrosismico Italiano. ingv. Website http://emidius.mi.ingv.it/Dbmi04
D
E
a
nGELiS
D’o
SSat
G. (1932) - La geologia e le catacombe romane/ memoria quinta di Gioacchino de Angelis D’Ossat. Memorie della pontificia
accademia delle scienze, nuovi Lincei, s.2, 16: 873-911-
Diss Working group (2010). Database of Individual Seismogenic Sources (DISS), Version 3.1.1: A compilation of potential sources for earthquakes
larger than M 5.5 in Italy and surrounding areas. http://diss.rm.ingv.it/diss/, © ingv 2010 - istituto nazionale di geofisica e vulcanologia - all rights
reserved; Doi:10.6092/ingv.it-Diss3.1.1
EuMEtsat - Monitoring weather and climate from the space. Copyright {2014} EuMEtsat. http://www.eumetsat.in
F
accEna
c., F
uniciELLo
r. & M
arra
F. (1995) - Inquadramento geologico strutturale dell’area romana. in La Geologia di Roma - Il centro storico. Memorie
Descrittive della carta geologica D’italia 50: 31-47.
F
roSini
p. (1977) - Il Tevere, le inondazioni di Roma e i provvedimenti presi dal governo italiano per evitarle. accademia nazionale dei Lincei. roma.
F
unicELLo
r. & G
iorDano
G. (2008
a
) - Foglio 374 - roma. carta geologica d’italia alla scala 1:50.000 - progetto carg.
F
unicELLo
r. & G
iorDano
G. (2008
b
) - La nuova Carta geologica di Roma: Litostratigrafia ed organizzazione stratigrafica. in: La geologia di Roma. Dal
centro storico alla periferia. Memorie Descrittive della carta geologica d’italia, 80: 39-85.
G
rEEnWooD
J.a., L
anDWEhr
J.M., M
ataLaS
n.c. & W
aLLiS
J.r. (1979) - Probability weighted moments: definition and relation to parameters of several
distribution expressable in inverse form. Water resources research, 15(5): 1049-1054.
G
iorDano
G. (2008) - I vulcani di Roma: storia eruttiva e pericolosità. in F
uniciELLo
r., P
raturLon
a. & G
iorDano
G. (
ED
) - La Geologia di Roma. Memorie
Descrittive della carta geologica d’italia, 80: 87-96.
h
EiKEn
G., F
uniciELLo
r., D
E
r
ita
D. & P
arotto
M. (2006) - I sette colli Guida geologica a una Roma mai vista. Milano, raffaello cortina.
h
oSKinG
J.r.M., W
aLLiS
J.r., W
ooD
E.F. (1984) - Estimation of the generalized extreme value distribution by the method of probability weighted moments.
background image
d. alessi, f. bozzano, a. di lisa, c. esposito, a. fantini, a. loffredo, s. martino, f. mele, s. moretto,
a. noviello, a. prestininzi, p. sarandrea, G. scarascia muGnozza, l. schilirò & c. varone
34
Italian Journal of Engineering Geology and Environment, 1 (2014)
© Sapienza Università Editrice
www.ijege.uniroma1.it
institute of hydrology, report no. 89, Wallingford, England, 27: 251-261.
iFFi (2007) - Inventario dei fenomeni franosi in italia realizzato dall’ISPRA e dalle Regioni e Province Autonome. http://www.isprambiente.gov.it/it/
progetti/iffi-inventario-dei-fenomeni-franosi-in-italia
ingv (2009) - Mappa sul risentimento macrosismico del terremoto del 6 aprile 2009. http//terremoto.rm.ingv.it
K
arnEr
D.b. & M
arra
F. (1998) - Correlation of fluviodeltaic aggradational sections with glacial climate history: a revision of the classical Pleistocene
stratigraphy of Rome. geol. soc. am. Bull., 110: 748-758.
L
E
G
aLL
J. (1953) - Le Tibre fleuve de Rome dans l’Antiquitè. publ. de l’institut d’art et d’archéologie de l’université de paris. paris.
L
EonE
F. (2014) - L’evento alluvionale di inzio anno che ha interessato l’area romana. professione geologo, 39: 15-16.
L
uciani
r. (1985) - Roma sotterranea. catalogo della mostra. comune di roma - ass.to alla cultura. Fratelli palombi Editori. roma.
M
arra
F. & r
oSa
c. (1995) - Stratigrafia e assetto geologico dell’area romana. in La Geologia di Roma. Il centro storico. Mem. Descr. carta geol. d’it.,
50: 49-118.
M
oLin
D., c
aStEnEtto
S., D
i
L
orEto
E., G
uiDoboni
E., L
iPEri
L., n
arciSi
b., P
aciELLo
a., r
iGuzzi
F., r
oSSi
a., t
ErtuLiani
a. & t
rania
G. (1995) - Caratteri
della sismicità. Mem. Descr. carta geol. d’it., 5: 391-403.
J
EnKinSon
a.F. (1955) - The frequency distribution of the annual maximum (or minimum) values of meteorological events. Quarterly Journal of the royal
Meteorological society, 87: 158-171.
K
arnEr
D.b., M
arra
F. & r
EnnE
P.r. (2001
b
) - The history of the Monti Sabatini and Alban Hills volcanoes: groundwork for assessing volcanic-tectonic
hazards for Rome. J. volc. and geoth. res., 107: 185-215.
pai (2012) - Piano di Assetto Idrogeologico a cura dell’autorità di bacino del tevere.
prg r
oMa
(2007) - Carta di pericolosità e vulnerabilità geologica del territorio comunale. piano regolatore generale del comune di roma a cura di
M
orDiaGLiani
D.
P
arotto
M. (2008) - Evoluzione paleogeografica dell’area romana: una breve sintesi. in: La geologia di Roma. Dal centro storico alla periferia. Mem.
Descr. carta geol. d’it., 80: 25-38.
P
inna
M. (1977) - Climatologia. utEt, torino. Xv, 442 pp.: ill.; 24 cm + 4 c. di tav.
P
rEStininzi
a. (2000) - La valutazione del rischio di frana: metodologie ed applicazioni al territorio della Regione Lazio. Dipartimento di scienze della
terra. volume unico, regione Lazio con cD allegato.
P
rEStininzi
a., P
uGLiESE
a. & r
oMEo
r.W. (2005) - Proposed seismic classification of Italy and related actions. italian Journal of Engineering geology and
Environment, 05/1. Doi:10.4408/iJEgE.2005-01.o-04.
r
EMEDia
G., a
LESSanDroni
M.G. & M
anGianti
F. (1998) - Le piene eccezionali del fiume Tevere a Roma Ripetta. università degli studi di L’aquila, Dip. di
ingegneria delle strutture, delle acque e del terreno (Disat n. 3).
S
ciotti
M. (1986) - Alcune osservazioni sulla situazione stratigrafica di un’area franosa a Roma. agi - Xvi convegno nazionale di geotecnica, 14-16
maggio, Bologna.
S
ELLa
P., b
iLLi
a., M
azzini
i., D
E
F
iLiPPiS
L., P
izzino
L., S
ciarra
a. & Q
uattrocchi
F. (2014) - A newly-emerged (August 2013) artificially-triggered fumarole
near the Fiumicino airport, Rome, Italy. Journal of volcanology and geothermal research (2014), Doi: 10.1016/J.JvoLgEorEs.2014.05.008 (in print).
v
arnES
D.J. (1978) - Slope movement types and processes. in: S
chuStEr
r.L. & K
rizEK
r.J. (E
DS
.), special report 176 Landslides: analysis and control.
transportation research Board, national research council, Washington D.c., 11-33.
Received April 2014 - Accepted May 2014
Statistics