Nicola CasagliGiovanni Gigli, Veronica Tofani

Warning systems: alert

thresholds

Department of Earth SciencesCENTRE OF COMPETENCE OF THE CIVIL PROTECTION DEPARTMENT

PRESIDENCY OF THE COUNCIL OF MINISTERS

Presentation structure

• Theory: predicting the time of failure using monitoring time series, from Apollonius of Perga to nowadays

• Example #1: Monte Beni rockslide, a case of successful prediction

• Example #2: Ruinon: on-going monitoring and early-warning of the most dangerous rockslide in Italy

• Example #3: Stromboli: real-time forecasting of slope instability and volcano eruptions

• Example #4: Urban landslides in Calabria Region: from prediction to governance

Apollonius of Perga(Perga, 262 a.C. – Murtina, 190 a.C.)

The Conics

The hyperbolas

Saito (1965)

• Method to predict failure based on creep rupture• Linear relationship between inverse velocity and

time during periods of acceleration to failure (i.e. hyperbolic acceleration)

1/ve

loci

ty

Time

Time of failure

0

Landslide behaviour

Time tVe

loci

ty v

Limitequilibrium

New limitequilibrium

collapse

tf = objective of prediction

(hyperbolic) acceleration

Fukuzono (1985)

( ) ( ) ( )( )α

αα

−

−=−

−−

1

11

Avv

tt ifif

α

⎟⎠⎞

⎜⎝⎛=

dtdxA

dtxd2

2

for α >1:

Power law relationship between acceleration and velocity before failure, tested on laboratory physical models

Inverse-velocity methodFukuzono (1985)

( )[ ] 11

11

1 )(1 −− −−=⇒∞= ααα ttAv fvf i

If α = 2 :

V-1 = A(tf-t)HYPERBOLA

Barry Voight (1988)A relation to describe rate-dependent material failureScience (1988)

A method for prediction of volcano eruptionsNature (1988)

July 1995, Montserrat'sSoufriere Hills volcano

May 1980, Mt. St. Helens volcano

Generalized Saito law

Ý Ω −α ⋅ Ý Ý Ω − A = 0Ω= any appropriate material quantity

A and α = experimental material constants

Dots refer to differentiation with respect to time

Voight (1991)

Vajont, 1963Voight (1991)

Voight (1991)Mt. St. Helens, 1980

Azimi graphical methodAzimi et al. (1988)

Some examples

Rose & Hungr (2006)

La Clapiere

Rose & Hungr (2006)

Landslide volume: 18 milion m3

Open pit mine applications(Rose & Hungr, 2006)

Stabilization progress monitoring

(Rose & Hungr, 2006)

Vajont 1963

Hendron & Patton (1985)

Vaiont landslideHutchinson (2001)

Vajont landslideHutchinson (2001)

Unexpected failures

• Measurement error

• Random noise• Local movements• Anthropic activity• Earthquakes• Meteorologic

factors• Structural

constraints• Brittle failure• Invalid

assumptions

Hutchinson (2001)

Non linearityPetley (2003)

Data from A: Angeli et al. (1989); B: Salt (1985); C: CNR IRPI (2002); D: USGS (2002)

Non linearityPetley (2003)

• Two styles of accelerating behaviour are seen in landslides– Linearity = rupture surface development / brittle failure /

first-time failures– Asymptotic = sliding on existing surfaces / ductile

deformation / continuous or reactivated failures

• These considerations are based on a new model for the development of progressive failure in brittle landslides tested in laboratory

My personal experience

• The linear model applies to the most part of slope failures (both first time and reactivations) in a “simple”geological environment (i.e. with a single main driving factor of instability)

• Non-linear (asympthotic) behaviour has been observed in first time failures during volcano eruptions

• In many cases reactivated failures exhibit a brittle behavior (related to the linear model) – Key reference: HUTCHINSON J.N. (1987) - Mechanisms producing large displacements in landslides on pre-existing shears. 1st Sino-British Geol. Conf., Tapei, Memoir of the Geological Survey of China, 9, 175-200.

Casagli (today)

Example #1: Monte Beni

Montebeni (Italy) April 19th 2002

Perimetral crack

debris cone debris conedebris cone

MASSIVE BASALTS

BASALT BRECCIAS

DISTURBED BRECCIAS

LIMESTONES

THRUST SURFACEWITH LAMINATED CHERTS

VEGETATED SLOPEDEPOSITS

talus

Tectonically disturbed breccias

Bedded limestones

Caothic clay shales

Basalt

Basalt breccias

Thrust fault

bedding

Monitoring with distometers

Continuous-wave stepped-frequency (CW-SF) radar based on a Network

Analyser (NWA) operating in the frequency band 17.0-17.1 GHz

Ground-based InSAR

The synthetic aperture is obtained sliding the antennas

along a linear rail

2.8m

target area

RxTx

sled

sourceNWA

computer

linear rail

European Commission Joint Research Centre

Target visibility

LiSA can operate in any condition of visibility(eg. during night, rainstorms and fog)

LISA Applications: Monte Beni landslide

Start: 8/5/2002 13:59

End: 13/5/2002 18:12

Interval: 124 h

Acquisition time: 40 min

Peak velocity: 0.48 mm/h

Mean Velocity: 0.16 mm/h

08/05/2002 13:5908/05/2002 17:1008/05/2002 20.2808/05/2002 23:3709/05/2002 00:0709/05/2002 04:0009/05/2002 06:3509/05/2002 08:4009/05/2002 11:2009/05/2002 14:0509/05/2002 16:5009/05/2002 19:2009/05/2002 23:5510/05/2002 03:1010/05/2002 05:5010/05/2002 09:4510/05/2002 11:5210/05/2002 13:5010/05/2002 16:4511/05/2002 19:0011/05/2002 21:2011/05/2002 23:4211/05/2002 19:0011/05/2002 21:2011/05/2002 23:4212/05/2002 02:4512/05/2002 06:0012/05/2002 08:3012/05/2002 11:2012/05/2002 14:3012/05/2002 17:3512/05/2002 20:1012/05/2002 22:4013/05/2002 02:1513/05/2002 05:4613/05/2002 08:5513/05/2002 12:4013/05/2002 12:4013/05/2002 15:4013/05/2002 18:12

LOS DISPLACEMENT (mm)

cum

ulat

ed d

ispl

acem

ent

(mm

)

point 1:displacement = 57 mmvelocity = 11 mm/day

point 5:displacement = 13 mmvelocity = 3 mm/day

point 4:displacement = 22 mmvelocity = 4 mm/day

point 3:displacement = 25 mmvelocity = 5 mm/day

point 2:displacement = 37 mmvelocity = 7 mm/day

point 1

point 2point 3

point 4

point 5D

ISP

LAC

EMEN

T (m

m)

m

m

elapsed time (min) elapsed time (min)

elapsed time (min) elapsed time (min)elapsed time (min)cu

mul

ated

dis

plac

emen

t (m

m)

cum

ulat

ed d

ispl

acem

ent

(mm

)

cum

ulat

ed d

ispl

acem

ent

(mm

)

cum

ulat

ed d

ispl

acem

ent

(mm

)Displacement histories

November 2002: early warning system

Traffic light 1

Traffic light 2 Network of wire extensometers, bar extensometers and clinometers

Displacement thresholds which activate two traffic lights

Early December: failure of sensorsarea 1

-50

0

50

100

150

200

250

300

350

400

30/11/02 0.00 2/12/02 0.00 4/12/02 0.00 6/12/02 0.00 8/12/02 0.00 10/12/02 0.00 12/12/02 0.00 14/12/02 0.00 16/12/02 0.00 18/12/02 0.00 20/12/02 0.00

data

mm

(cm

per

AL)

, °c DR1

DR2EF1FA1FA2AL1 TA1

area2

-10.00

40.00

90.00

140.00

190.00

240.00

290.00

340.00

30/11/020.00

2/12/020.00

4/12/020.00

6/12/020.00

8/12/020.00

10/12/020.00

12/12/020.00

14/12/020.00

16/12/020.00

18/12/020.00

20/12/020.00

data

mm

, cm

per

AL,

°c

DR3DR4 EF2EF3FA3AL3AL2TA2

area3

-20

30

80

130

180

230

280

330

380

430

27/11/0212.00

29/11/0212.00

1/12/0212.00

3/12/0212.00

5/12/0212.00

7/12/0212.00

9/12/0212.00

11/12/0212.00

13/12/0212.00

15/12/0212.00

17/12/0212.00

19/12/0212.00

data

mm

, cm

per

AL,

°c DR5

FA4FA5FA6EF4TA2DR6

area4

-50

-30

-10

10

30

50

70

90

27/11/0212.00

28/11/0212.00

29/11/0212.00

30/11/0212.00

1/12/0212.00

2/12/0212.00

3/12/0212.00

4/12/0212.00

5/12/0212.00

6/12/0212.00

7/12/0212.00

8/12/0212.00

data

mm

, °c

EF5EF6EF7TA 4

Monitoring time series

0

500

1000

1500

2000

2500

3000

4/4 24/4 14/5 3/6 23/6 13/7 2/8 22/8 11/9 1/10 21/10 10/11 30/11 20/12

InSAR monitoring

collapse

Cum

ulat

ed d

ispl

acem

ent

(mm

)

Extensometer monitoring

delta 1-2 delta 3-2 delta 12-8 delta a'-b' delta 34-35' delta 36-37 delta 45-47

Inverse velocity method

0.0000

0.1000

0.2000

0.3000

0.4000

0.5000

0.6000

0.7000

0.8000

0.9000

1.0000

1/9 9/9 17/9 25/9 3/10 11/10 19/10 27/10 4/11 12/11 20/11 28/11 6/12 14/12 22/12 30/12 7/1

1/v

(gio

rno/

mm

)

1/v1-2 1/v3-2 1/v12-8 1/va'-b' 1/v34-35' 1/v36-37 1/v45-47

collapse

0.0000

0.2000

0.4000

0.6000

0.8000

1.0000

1.2000

1.4000

1.6000

1.8000

13/7 23/7 2/8 12/8 22/8 1/9 11/9 21/9 1/10 11/10 21/10 31/10 10/11 20/11 30/11 10/12 20/12 30/12

1/v

(gio

rno/

mm

)

failure

Benchmarks 1-2

Inverse velocity method

01-set

21-set

11-ott

31-ott

20-nov

10-dic

30-dic

19-gen

01-set 11-set 21-set 01-ott 11-ott 21-ott 31-ott 10-nov 20-nov 30-nov 10-dic 20-dic 30-dic 09-gen

Benchmarks 12-8

Azimi graphical method

Dispatches to civil protection authorities

• The landslide, until now sensitive to rainfall, is now increasing its rate of displacement independently from precipitation (letter of Prof. Canuti to the Mayor of Firenzuola, November 26th, 2002)

• The main landslide is on the point of collapsing. It is necessary that the evacuation measures will be strictly respected during Christmas holidays (letter of Prof. Canuti to the Mayor of Firenzuola, December 23rd, 2002)

• The main landslide will collapse within the first days of January 2003 (report of Prof. Canuti to the Civil Protection Commission, December 27th, 2002)

Numerical model of failure

Regional road

Perimetral crack

Landslide volume:500 000 m3

Angle of friction:34°

Graphical elaboration:Conefall(www.quanterra.org)

Risk scenario

December 14th 2002

Rock fall of 20 000 m3

December 15th 2002

Intitiation of the major collapse

December 28th 2002

Collapse of 500 000 m3 of rock

Today