mouline island
Transcription
mouline island
From calculated runout-zones to hazard zonation - Examples - PD Dr. Thomas Glade [email protected] 0 Lecture overview Application of the infinite slope model: Example of Bonn, Germany Empirical and physically-based modeling: Examples of Bíldudalur, Island Empirical modeling of Rock fall and its application in a GIS: Example of Bayern, Germany 3-D trajectory analysis for mitigation of rock fall: Example of La Désirade, French West Indies Examples for numerical simulations National scale analysis 1 1 Application of the Infinite Slope Model c´+ z ∗ cos 2 β × (γ − m × γ w ) × tan φ´ FS = = τ γ × z × sin β × cos β s FS s τ c´ z zw β γ m γw φ´ = = = = = = = = = = = Factor of Safety (<1 unstable; ≥1 stable) shear strength (resisting forces) [kN/m2] shear stress (driving forces) [kN/m2] effective cohesion [kN/m2] depth of shear plane [m] height of ground water table [m] slope [°] unit weight of soil [kN/m3] relation z / zw (0 < m < 1) [-] moist unit weight of soil [kN/m3] effective friction angle [°] 2 Application of the Infinite Slope Model: Bonn m=0 m = 0,5 m=1 N A B C 1 Mouline-Richard & Glade, 2003 2 3 km 3 2 Application of the Infinite Slope Model: Bonn Region m=0 m = 0,5 m=1 Mouline-Richard & Glade, 2003 4 Application of the Infinite Slope Model: Validation 100,00 active landslides = 0,61 % of the total area with a slope angle > 7° 90,00 80,00 70,00 60,00 fos >= 1,8 fos >= 1,3 - < 1,8 50,00 fos >= 1 - < 1,3 fos < 1 40,00 30,00 20,00 10,00 0,00 m=0 m = 0,1 m = 0,2 m = 0,3 m = 0,4 m = 0,5 m = 0,6 m = 0,7 m = 0,8 m = 0,9 m=1 Mouline-Richard & Glade, 2003 5 3 6 Mouline-Richard & Glade (2004) Application of the Infinite Slope Model: Asumptions Unlimited slope Within each pixel similar structure Constant depth of shear plane Geotechnical conditions do not change Hydrological changes are not included Vegetation is not considered => Shallow translational landslides 7 4 Empirical and physically-based modeling: Examples of Bíldudalur, Island Bíldudalur, Westfjords Glade & Jensen, 2004 8 Aerial photography of Bíldudalur, view to North Iceland - Photopraphs Bíldudalur, Westfjords 9 (Photo: Matz Wibelund) 5 Schematic profile of west fjord slopes near settlements Glade, 2005 10 Approach for debris flow modeling: Bíldudalur, Island Use of empirical and semi-empirical models Dividing area into units of similar settings Focus on correlation between rainstorm events, catchment size and respective run-out distance Empirical relationship between length of run-out and slope angle Ratio of horizontal and vertical distance & catchment size Use of back-analysis to adapt models to the conditions in the study area Scenario modeling: • Calculation of run-outs for different sized rain-storm events 11 6 Debris flow map: Bíldudalur, Island UNIT V UNIT IV UNIT III UNIT II UNIT I Glade & Jensen, 2004 12 Debris flow model Volume with VW k = P = A = Vw = k * P * A = Event magnitud of water [m3] in a particular period Discharge-coefficient (0,85) Rainfall P [m] in a particular period Catchment [m2] Transport distance L = 1,2 Vwd0,19 * H0,78 with L = Transport distance [m] Vwd = Debrid flow magnitude (70% sediment + 30% water) [m3] H = Hight between lowest deposition and source area [m] Rickenman, 1999 Scenario model is based on rainfall events (2yr / 10yr / 50yr) 13 Glade, 2005 7 Assumptions for empirical debris flow modeling: Bíldudalur, Island Coherent distribution of rainfall in catchment Comparable surface structures Minor water loss through infiltration, ground water recharge, and evaporation Minor delay between max. rainfall intensity and max. discharge Similar conditions of the catchment and the triggering event – both in time and space Unlimited sediment availability 14 Debris flows and calculated run-outs 15 Glade & Jensen 2004 8 Approach for rock fall modeling: Bíldudalur, Island Physically-based model Dividing area into units of similar settings Determination of characteristic profile in relation to rock fall Extrapolation of resulting values into respective unit with consideration of local features Use of Colorado Rock fall Simulation Program (CRSP) • 2-D rock fall model • Input variables: • Surface roughness • Tangential coefficient of frictional resistance • Normal coefficient of restitution 16 Approach for rock fall modeling: Bíldudalur, Island 2 dim. Model (CRSP4.0) Transport distance = f (rock size, shape, verticale profile, surface roughness) Scenario modelling: Based on MC simulation for rock sizes (1.9t/10.7t/38.7t) 17 9 Zones of transport distances for rock falls Glade & Jensen 2004 18 Assumptions for empirical debris flow modeling: Bíldudalur, Island Representativeness of slope profiles for total unit Rocks do not break during movement (worst case scenario) Rock form is round and does not change during movement Characteristics of catchments and the triggering event does not change neither in time nor in space Unlimited sediment availability 19 10 Classification: From Run-out to HAZARD Hazard Class Rock Fall: Rock weight [t] Debris flow: Triggering rainstorm event [mm / Ret. Period] < 1.9 68 / 2yr Moderate 1.9 – 11.3 92 / 10yr Low 11.3 – 38.7 117 / 50yr >38.7 > 117 / 50yr High Very Low 20 21 Glade 2002 Glade (2002) 11 Empirical modeling of Rock fall and its application in a GIS: Example of Bayern, Germany Documentation and information system for mass movements in Bavarian Alps: GEORISK GIS-based system Empirical approach: global angle model Implementation in GIS-environment (ArcGIS) Cazzaniga et al., 2005 22 Empirical modeling of Rock fall and its application in a GIS: Example of Bayern, Germany Maximum run-out zone is determined by: • Minimum global angles between the horizontal line and the line connecting the farthest blocks and different points within the detachment area or the top of the talus 23 12 Empirical modeling of Rock fall and its application in a GIS: Example of Bayern, Germany • Use of two angles • Shadow angle (angle between horizontal line & top of talus) • Geometrical slope angle (angle between the horizontal line and top of the detachment zone) Results are compared to a process-based trajectory model 24 3-D modeling for rock fall map – general approach: Bavaria, Germany (1/3) 1. Localisation of potential detachment zone/starting point • Use of DEM 25 13 3-D modeling for rock fall map – general approach: Bavaria, Germany (2/3) 2. Data preparation • • Generation of necessary attributes for “viewshed-function” Checking the angles between every point and starting points 26 3-D modeling for rock fall map – general approach: Bavaria, Germany (3/3) 3. Modeling Viewshed-function: Starting points of rock falls Limiting horizontal (lateral spread) and vertical angles (run out) • Checking for errors Cazzaniga et al., 2005 • • 27 14 3-D modeling for rock fall map – results: Bavaria, Germany (1/3) Test area: red areas are potential starting points of rock falls, extracted from GEORISK 28 3-D modeling for rock fall map – results: Bavaria, Germany (1/3) Calculated danger areas by using of viewshed function 29 15 3-D modeling for rock fall map – results: Bavaria, Germany (1/3) 30 Result of rock fall modeling applying global angle model; orange areas are accumulation and detachment areas 3-D modeling for rock fall map – results: Bavaria, Germany (1/3) 31 Result of rock fall modeling applying process based trajectory model; brown areas are accumulation and detachment areas 16 3-D modeling for rock fall map – results: Bavaria, Germany (1/3) in 80% the models produced the same output in 8% empirical model more pessimistic in 12% trajectory model more pressimistic Comparison of modeling outputs: Red & green: coincidence Orange: empirical model is more pessimistic Yellow: trajectory model is more pessimistic 32 Example of La Désirade, French West Indies Rock fall risk management project • Application of 3-D trajectory rock fall model • DEM • Starting points • Rebound conditions • Hazard and multi-risk map • Determination of solutions & risk prevention plan Leroi, 2005 33 17 Application of 3-D trajectory rock fall model a Main window of 3-D trajectory model b Computed 3-D trajectories c Trajectories with impact on existing buildings d Design of protecting fences and inclusion into DEM e Trajectories after included fences f Location of protecting fences across the island 34 Hazard and multi-risk map High Hazard Final risk map Medium to high hazard Location of protecting fences 35 18 Example of numerical simulations © Fausto Guzzetti 36 CURRY VILLAGE LEDGE TRAIL ROCK FALL Direkt Rock fall (Source area Ledge Trail) 15.8.2001, 14.9.2001 und 25.9.2001 1 100% 100% 39 39 Model vs. vs. Mapping Mapping Model 80% 80% 60% 60% 213 213 130 130 97 97 89 89 50 50 87 87 40% 40% 76 76 20% 20% 47 47 30 30 1-2 1-2 3-5 3-5 36 36 53 53 43 43 32 32 0% 0% 1-2 37 NUMBER OF BOULDERS 3-5 6-10 11-25 26-5051-100 >100 6-10 6-10 11-25 11-25 26-50 26-50 51-100 51-100 >100 >100 Number Number of of boulders boulders © Fausto Guzzetti 19 Presentation of a rock fall simulation (Yosemite National Park) 38 National distribution of floods and landslides in Italy Guzzetti (2000) 39 20 Landslide susceptibility map -USA Godt et al. (1999) 40 Example national Scale: Data availability • DEM (25m resolution) • Geology (1 : 1,250,000) • Landslide distributions for two regions (Bonn, Rheinhessen) 41 21 Methodology Combination of slope angle & lithology defines susceptibility class • Review of existing classifications (coasts) • Questionnaire asking for expert opinion • Transferred presedence (Prinz 1997) Susceptibility classes Analysis on 25m – Results upscaled to 150m 42 Expert judgment Region Lithology / geological formation Experts (Location) Bavarian Alps Calcatrous, dolomit, marls, sedimentary dep. Prof. Bunza (München) Lake Constanze Lower sweet water molasse PD Dr. Theilen-Willige (Stockach) Bonn Region Tertiary clays and sands / Devonian Series Dr. Schmidt (Bonn) MecklenburgVorpommern (coast) Glacial and fluvioglacial deposits Dr. Tiepolt & Dr. Gurwell (Rostock) Göttingen Muschelkalk Series Dipl.-Geogr. Scholte (Osterode) Rheinhessen Oligocene marls, clays / Miocene Limestone Dr. Jäger (Heidelberg) Schleswig-Holstein (Coast) Glacial and fluvioglacial deposits Dr. Schmidz & Dr. Ziegler (Flintbeck) Schwäbische ~ / Fränkische Alb Muschelkalk / Keuper / Jurassic Series Prof. Moser (Erlangen) Thüringer Becken Lower Muschelkalk / Upper Sandstone (Triassic) Dr. Beyer & Prof. Schmidt (Halle) 43 22 Susceptibility classes Class Description Very low At human discretion no danger Low Building damage and directly affected people unlikely Moderate Building damage possible, people probably endangered High Building destruction likely, people endangered 44 Weightning Options Lithology (n=219) Oligocene Marls Greywacke Loess ..... Slope angles [°] Susceptibility > 15 High 10-15 Moderate 5-10 Low <5 Very low > 60 High 10-60 Low <10 Very Low >33 High 22-33 Moderate 12-22 Low ..... ...... 45 23 Shaded relief of Germany 46 1 : 2,750,000 National landslide susceptibility map Suscep. = f (Lithology; Slope angle) 47 1 : 2,750,000 Dikau & Glade (2003) 24 Potentially affected slopes 0.2 0.6 5.4 Very low Low Moderate High 93.8 48 Percentage Susceptibility within slope classes 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 88.2% 8.3% 2.4% 0.78% 0.23% 0.09% % on total area High Moderate Low Very low 0°-<10° 10°-<20° 20°-<30° 30°-<40° 40°-<50° >=50° Slope angle classes 49 25 Discussion 1/2 High susceptible regions include: • Alpine regions • Steep cruestas • Deeply dissected valleys in the low mountain ranges (e.g. Mittelrhein; Mosel) • Coasts along North and East Sea • Failures along natural river banks National scale analysis require different approaches and methods Classified susceptibility classes Results have not been statistically validated with existing data Further regions need to be surveyed 50 Summary • Rock fall analysis for a slope profil • Debris flow modelling based on trigger • Estimated event magnitude & frequency => Hazard • Calculated run-out is based on field evidence • Risk Analysis is performed on single objectes • Social impacts of detailed results is crucial • Scenarios can be analysed 51 26 Change of disposition • General disposition Relief / Topography Geology & material properties Vegetation • Variable disposition • Triggering Event Rainfall (Extreme, long prolonged wet periods) Snowmelt Earthquakes Anthropogenic interference Climate fluctuations – seasons Geotechnical properties Material availability Based on Zimmermann et al. 1999 52 Advantages of spatial modelling • Abstraction to key-issues • Subjectivity by model development and choice • Objectivity: Repetition of similar analysis gives identical results • Unambiguous rules - Concepts and structures - Uniformity based on objective criteria - Transparency is inherent - Transferability is possible • Potential for scenarios 53 27 Disadvantages of spatial modelling • Reduction to single parameter indispensable • Commonly statistical relation (if - when) • Danger: Essential, important process-determining parameter will not be considered • Quality has to be ensured • Assumptions have to be reflected for interpretations • Transferability has to be critically questioned 54 Scientific challenges • Development of process-specific methods • Scale dependent choice of methods is important • Spatial models have to be improved, or further developed • Validation of results is essential for the judgement of the quality • Scenarios of events 55 28 References Cazzaniga, C., Sciesa, E., Thüring, M. and Zonta, M.F. (eds.) 2005: Mitigation of hydrogeological risk in alpine catchments – “CatchRisk”. Final report of the Program INTERREG II B – Alpine Space. pp. 189. Glade, T. 2005: Linking debris-flow hazard assessments with geomorphology. Geomorphology 66, 189-213. Glade, T. and Jensen, E.H. 2004: Landslide hazard assessments for Bolungarvík and Vesturbyggð, NW-Iceland. Reykjavik: Icelandic Meteorological Office. Leroi, E. 2005: Global rockfalls risk management process in ‘La Désirade‘ Island (French West Indies). Landslides 2, 358-365 Mouline-Richard, V. and Glade, T. 2003: Regional slope stability analysis for the Bonn region. Engineering Geology. 56 29