Folie 1

1
Martin Mergili Institute of Geography, Univ. of
Innsbruck
Katharina Schratz Inst. of Mathematics, Univ. of
Innsbruck Stella Maris Moreiras CRICYT-IANIGLA,
Mendoza, Argentina Wolfgang Fellin Inst. of
Infrastructure, Univ. of Innsbruck Alexander
Ostermann Inst. of Mathematics, Univ. of
Innsbruck Johann Stötter Institute of
Geography, Univ. of Innsbruck
Simulation of debris flows in the Central Andes
an integrated model approach based on Open Source
GIS
funded by the Tyrolean Science Funds
2
background
debris flows are ... ? mixtures of water and
mobilized soil moving down
slopes or channels ? often caused by heavy or
prolonged rainfall or extreme
snow melt ? initiating from failure of
(saturated) soil on steep
slopes, or from surface runoff with high
sediment concentration
?
?
modelling and prediction of debris flows ?
empirical approaches ? statistical
approaches ? physically-based models ?
combined methods
3
specific aims
?
model framework for spatially distributed
simulation of debris flows, based on Open Source
GIS (freely available) ? designed for small
catchments (few km²) ? precipitation and snow
melt as meteorological
input ? both slope failures and
sediment-laden runoff as initiation ?
including simulation of runout distance
and patterns of deposition
?
identification of potentially critical rainfall
(or snowmelt) events for the occurrence of debris
flows along the international road corridor
between Western Argentina and Central Chile.
4
GRASS GIS implementation
defined source areas
defined rainfall event
r.debrisflow
infiltration
soil water flow
surface runoff
slope failure
erosion
debris flow?
runout
area of impact
5
GRASS GIS implementation
Green-Ampt (1911) infiltration model
r.debrisflow
assumption sharp wetting front moving downwards
R
R
infiltration capacity f f(d, K, ??, R, ?)
R
d f?t/??
dR/??
saturated soil
bedrock
soil at initial moisture content
R surface water depth ?t length of time step
6
GRASS GIS implementation
infinite slope stability model
r.debrisflow
wht. of moist soil G ? ?x d
normal force N G cos S shear force T G sin S
seepage force Fs ?x d ?w sin S shear
resistance Tf N tan f c ?x / cos S
factor of safety FS Tf / (T Fs)
d
T
Fs
saturated soil
N
G
?, c , f
bedrock
?x
soil at initial moisture content
? spec. weight of moist soil ?w spec. weight
of water c combined cohesion (soil root) f
angle of internal friction
FS gt 1 cell is stable FS lt 1 cell is pot.
unstable
7
GRASS GIS implementation
Rickenmann (1990) model
r.debrisflow
fluid discharge qflow 1/n R1.67 (tan
S)0.5
critical fluid discharge qcrit f(D50, S)
qflow
potential qload,pot bedload discharge
f(qflow-qcrit, D90/D30, S)
qload, pot
erosion
qflow
erosion/deposition d l-qload,pot/v
sediment concentration c l / (lR)
qload
qflow
qload, pot
d
bedrock
qload
deposition
qflow
qload, pot
bedrock
qload
saturated soil
qflow
qload, pot
bedrock
qload
soil at initial moisture content
soil at initial moisture content
R hydraulic radius l depth of bedload n
Mannings n v flow velocity t length of time
step
8
GRASS GIS implementation
r.debrisflow
Corominas et al. (2003) (for
unobstucted flow path)
vol (m³)
26
vol lt 800 m³
vol 800 - 2000 m³
23
21
vol gt 2000 m³
vol (m³)
Rickenmann (1999)
hdist 1.9 vol0.16 vdist0.83
vdist (m)
hdist (m)
9
GRASS GIS implementation
Vandre (1985) distribution of debris
flow deposit can be estimated
mobilized volume does not influence
runout distance
r.debrisflow
vdist (m)
hdist lt 0.4 vdist
hdist (m)
10
4
runout with random walk, weighted for local slope
angle uniform or wedge-shaped distribution of
deposit
?
?
10
international road Western Argentina Central
Chile
11
international road Western Argentina Central
Chile
12
international road Western Argentina Central
Chile
13
international road Western Argentina Central
Chile
14
(No Transcript)
15
(No Transcript)
16
terrain
17
soil characteristics
18
soil characteristics
geotechnical laboratory
soil
triaxial experiment ? angle of internal friction
f ? soil cohesion c
?
pedotransfer tables
grain size distribution
soil hydrological parameters
?
19
further parameters
geomorphological and vegetation units ? mapping
in the field and from orthophotos
?
meteorological data ? daily values of
temperature, precipitation, and further
available data ? conversion into scenarios of
short rainfall events
?
?
historical data ? reports on debris flow
impacts on the international
road, and on quantities of material to be
removed from publications,
road maintenance agency, and
newspapers
20
GRASS GIS implementation
rainfall event 100 mm in 1.5 hours
corresponding to highest recorded daily
precipitation
r.debrisflow
depth of wetting front
areas of initiation
areas of scouring
areas of deposition
sediment budget from surface runoff
400 m
21
GRASS GIS implementation
observations
r.debrisflow
results are realistic, but much more validation
with a larger number of study areas is required
?
still to many parameters that have to be
calibrated (sediment transport, runout)
?
the tool provides a good base, but it still has
to be considered as a first try which should be
optimized with regard to some aspects
?
its public availability as an Open Source product
is therefore important anybody can look into the
details and come up with remarks for possible
improvements
?
www.uibk.ac.at/geographie/personal/mergili/scripts
22
The Savage-Hutter theory
for the mechanics of granular flows
first results of an implementation as module of
the OpenSource software GRASS GIS
r.avalanche
23
basics of the theory
continuum theory for description of motion
of finite mass avalanche over a rough inclined
slope
?
based on system of partial differential
equations of mass and momentum balances
?
24
basics of the theory
theory is only valid for smoothly changing
topographies, compared to the dimensions of the
avalanche
?
solutions do exist for several topographic
situations, but are rather complex for realistic
topographies
?
27.03.2007 presentation by Prof. Kolumban
Hutter, ETH Zürich about the Savage-Hutter
theory, its potential and its limitations
25
GRASS GIS implementation
r.avalanche
first try for simple topography
t 10 s
v 2.8 m s-1
500 m
depth of flow multiplied with 1000
26
GRASS GIS implementation
r.avalanche
first try for simple topography
t 20 s
v 5.5 m s-1
500 m
depth of flow multiplied with 1000
27
GRASS GIS implementation
r.avalanche
first try for simple topography
t 30 s
v 8.3 m s-1
500 m
depth of flow multiplied with 1000
28
GRASS GIS implementation
r.avalanche
first try for simple topography
t 40 s
v 11.0 m s-1
500 m
depth of flow multiplied with 1000
29
GRASS GIS implementation
r.avalanche
first try for simple topography
t 50 s
v 13.8 m s-1
500 m
depth of flow multiplied with 1000
30
GRASS GIS implementation
r.avalanche
first try for simple topography
t 60 s
v 16.5 m s-1
500 m
depth of flow multiplied with 1000
31
GRASS GIS implementation
r.avalanche
first try for simple topography
t 70 s
v 19.2 m s-1
500 m
depth of flow multiplied with 1000
32
GRASS GIS implementation
r.avalanche
first try for simple topography
t 80 s
v 21.6 m s-1
500 m
depth of flow multiplied with 1000
33
GRASS GIS implementation
r.avalanche
first try for simple topography
t 90 s
v 24.5 m s-1
500 m
depth of flow multiplied with 1000
34
GRASS GIS implementation
r.avalanche
first try for simple topography
t 92 s
v 25.0 m s-1
500 m
depth of flow multiplied with 1000
35
GRASS GIS implementation
r.avalanche
first try for simple topography
t 94 s
v 25.6 m s-1
500 m
depth of flow multiplied with 1000
36
GRASS GIS implementation
r.avalanche
first try for simple topography
t 96 s
v 26.2 m s-1
500 m
depth of flow multiplied with 1000
37
GRASS GIS implementation
r.avalanche
first try for simple topography
t 98 s
v 20.5 m s-1
500 m
depth of flow multiplied with 1000
38
GRASS GIS implementation
r.avalanche
first try for simple topography
t 100 s
v 17.8 m s-1
500 m
depth of flow multiplied with 1000
39
GRASS GIS implementation
r.avalanche
first try for simple topography
t 102 s
v 17.3 m s-1
500 m
depth of flow multiplied with 1000
40
GRASS GIS implementation
r.avalanche
first try for simple topography
t 104 s
v 15.7 m s-1
500 m
depth of flow multiplied with 1000
41
GRASS GIS implementation
r.avalanche
first try for simple topography
t 106 s
v 14.3 m s-1
500 m
depth of flow multiplied with 1000
42
GRASS GIS implementation
r.avalanche
first try for simple topography
t 108 s
v 13.1 m s-1
500 m
depth of flow multiplied with 1000
43
GRASS GIS implementation
r.avalanche
first try for simple topography
t 110 s
v 11.9 m s-1
500 m
depth of flow multiplied with 1000
44
GRASS GIS implementation
r.avalanche
first try for simple topography
t 112 s
v 9.9 m s-1
500 m
depth of flow multiplied with 1000
45
GRASS GIS implementation
r.avalanche
first try for simple topography
t 114 s
v 9.1 m s-1
500 m
depth of flow multiplied with 1000
46
GRASS GIS implementation
r.avalanche
first try for simple topography
t 116 s
v 7.5 m s-1
500 m
depth of flow multiplied with 1000
47
GRASS GIS implementation
r.avalanche
first try for simple topography
t 118 s
v 6.0 m s-1
500 m
depth of flow multiplied with 1000
48
GRASS GIS implementation
r.avalanche
first try for simple topography
t 120 s
v 4.8 m s-1
500 m
depth of flow multiplied with 1000
49
GRASS GIS implementation
r.avalanche
first try for simple topography
t 130 s
v 0.0 m s-1
500 m
depth of flow multiplied with 1000
50
GRASS GIS implementation
r.avalanche
first try for simple topography
t 140 s
v 0.0 m s-1
500 m
depth of flow multiplied with 1000
51
GRASS GIS implementation
r.avalanche
first try for simple topography
t 150 s
v 0.0 m s-1
500 m
depth of flow multiplied with 1000
52
GRASS GIS implementation
r.avalanche
first try for simple topography
t 160 s
v 0.0 m s-1
500 m
depth of flow multiplied with 1000
53
GRASS GIS implementation
r.avalanche
first try for simple topography
t 170 s
v 0.0 m s-1
500 m
depth of flow multiplied with 1000
54
GRASS GIS implementation
r.avalanche
first try for simple topography
t 180 s
v 0.0 m s-1
500 m
depth of flow multiplied with 1000
55
GRASS GIS implementation
r.avalanche
first try for simple topography
t 190 s
v 0.0 m s-1
500 m
depth of flow multiplied with 1000
56
GRASS GIS implementation
r.avalanche
first try for simple topography
t 200 s
v 0.0 m s-1
500 m
depth of flow multiplied with 1000
57
GRASS GIS implementation
shock
r.avalanche
58
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 4 s
v 2.0 m s-1
500 m
depth of flow multiplied with 100
59
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 8 s
v 4.0 m s-1
500 m
depth of flow multiplied with 100
60
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 12 s
v 5.8 m s-1
500 m
depth of flow multiplied with 100
61
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 16 s
v 7.5 m s-1
500 m
depth of flow multiplied with 100
62
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 20 s
v 9.1 m s-1
500 m
depth of flow multiplied with 100
63
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 24 s
v 10.7 m s-1
500 m
depth of flow multiplied with 100
64
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 28 s
v 11.7 m s-1
500 m
depth of flow multiplied with 100
65
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 32 s
v 12.0 m s-1
500 m
depth of flow multiplied with 100
66
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 36 s
v 11.9 m s-1
500 m
depth of flow multiplied with 100
67
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 40 s
v 11.4 m s-1
500 m
depth of flow multiplied with 100
68
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 44 s
v 11.4 m s-1
500 m
depth of flow multiplied with 100
69
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 48 s
v 10.8 m s-1
500 m
depth of flow multiplied with 100
70
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 52 s
v 10.7 m s-1
500 m
depth of flow multiplied with 100
71
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 56 s
v 10.4 m s-1
500 m
depth of flow multiplied with 100
72
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 60 s
v 9.9 m s-1
500 m
depth of flow multiplied with 100
73
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 64 s
v 9.6 m s-1
500 m
depth of flow multiplied with 100
74
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 68 s
v 9.6 m s-1
500 m
depth of flow multiplied with 100
75
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 72 s
v 11.0 m s-1
500 m
depth of flow multiplied with 100
76
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 76 s
v 11.0 m s-1
500 m
depth of flow multiplied with 100
77
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 80 s
v 10.2 m s-1
500 m
depth of flow multiplied with 100
78
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 84 s
v 8.5 m s-1
500 m
depth of flow multiplied with 100
79
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 88 s
v 7.3 m s-1
500 m
depth of flow multiplied with 100
80
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 92 s
v 7.0 m s-1
500 m
depth of flow multiplied with 100
81
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 96 s
v 6.0 m s-1
500 m
depth of flow multiplied with 100
82
GRASS GIS implementation
r.avalanche
first try for realistic topography
t 100 s
v 5.5 m s-1
500 m
depth of flow multiplied with 100
83
minor shock
GRASS GIS implementation
r.avalanche
84
GRASS GIS implementation
observations
r.avalanche
in general, flow patterns are simulated well,
including shocks connected to decrease of slope
angle
?
very aggressive behaviour of the simulated
avalanche regarding velocity and patterns of
spreading
?
rather inert behaviour of the avalanche when
changing flow direction on realistic topography
(moving up on the opposite slope)
?
inclusion of mechanical concepts for movement
on complex topography is required as well as a
careful choice of parameter values
?