finding a representative friction angle for landslide areas in...

1
Finding a Representative Friction Angle for Landslide Areas in Palos Verdes Using Inϐinite Slope Analysis By Cameron Yu Presented: Dec 17, 2012 Introduction My project involves the occurrence of landslides in my home town of Palos Verdes, California. As part of my research, I found a landslide inventory map and an underlying geology map of the Palos Verdes area, and what was determined was that all of the landslides have occurred over PreQuaternary Bedrock. Since I have decided to use innite slope analysis in order to determine where landslides would occur, the fact that the landslides occur over bedrock dees logic about landslides. Innite slope analysis depends on the fricƟon angle of the underlying soil, and bedrock does not necessarily have a fricƟon angle and thus, makes innite slope analysis obsolete. What may have happened is that a ssure could have been created in the bedrock due to an earthquake, and the crack in the bedrock could have been lled with some kind of clay or other soil. Thus, the landslides are actually occurring because the bedrock is slipping on the clay lm. Working under this assumpƟon, the goal of my project is to nd a representaƟve fricƟon angle for the areas where landslides occur in the Palos Verdes area. I will do this by creaƟng a raster map that shows an area’s factor of safety against a landslide, and I will try to match up the landslide map and the underlying geology to match up by varying the fricƟon angle unƟl the areas with known landslides have the lowest factor of safety (<1.5 in this case). Under the assumpƟon that the landslides occur because of lowstrength soil lling fractures of bedrock, the knowledge gained from this project will shed light on a characterisƟc of the underlying soil. Methodology The main output of this project is a raster showing the factor of safety against landslide in the Palos Verdes area, and this was achieved using innite slope analysis. The equaƟon used was: Where φ is the fricƟon angle of the underlying soil, and β is the slope of the surface. The main data layers used are listed below, including their source: The .pdf les were converted to .jpeg les, georeferenced into ArcMap, and converted into .shp les. The ElevaƟon and Underlying Geology maps were then put through the following model, of which the output was the Factor of Safety map. The model is shown below: Data Title Source Format ElevaƟon Map USGS Raster Underlying Soils California Department of ConservaƟon .PDF Landslide Inventory California Department of ConservaƟon .PDF Area of Study I focused my area of study to the Palos Verdes Peninsula, as there are a number of landslides that have occurred there. This area includes the following USGS quadrangles: Redondo Beach, San Pedro, and Torrance. Below is the area of study in the context of Los Angeles County, and the map includes the landslide inventory of the area. Results Figures 3 and 4 are the two main input maps. Each geologic unit in Figure 3 is assigned a fricƟon angle, φ. It is this parameter and the slope angle that is derived from Figure 4 that are input to the model to create Figures 58. The fricƟon angle for the geologic unit Ls, which stands for landslides, is the angle represented in the Ɵtles of the maps. The red areas show a factor of safety of <1.5, which are the areas most suscepƟble to landslides. Figure 5 is the factor of safety of the Palos Verdes area if we are to assume that the landslides occur on bedrock, which is what the old geologic maps suggest. As the gure above suggests, landslide occurrence is highly unlikely in the inventoried areas assuming a fricƟon angle of 45° because the inventoried areas hardly contain any red. The fricƟon angle was varied in increments of 5° to produce maps similar to Figures 68. As the fricƟon angle is lowered, the amount of red area that appears in the landslide areas increases, which is what we can expect based on the innite slope analysis equaƟon. Conclusion Based on the factor of safety maps created by the model, the angles at which the landslide areas have the most red are when φ = 15° and φ = 10°. However, using φ = 10° in innite slope analysis is a more reasonable fricƟon angle to use for the landslide units. When the fricƟon angle changes from 15° to 10°, there is, visually, a signicant more amount of red in the landslide areas, thus not only using φ = 10° more conservaƟve when calculaƟng factor of safety, but it is also more accurate as well. Through my literature review, I found a geologic map from the CA Department of ConservaƟon of an area near Palos Verdes that includes landslides in the map. The fricƟon angle the map used was φ = 10°, which validates the ndings from my model. Figure 1: Area of Study Figure 2: Factor of Safety Model Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8

Upload: others

Post on 09-Aug-2021

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Finding a Representative Friction Angle for Landslide Areas in ...sites.tufts.edu/gis/files/2013/11/Yu_Cameron.pdfFinding a Representative Friction Angle for Landslide Areas in Palos

FindingaRepresentativeFrictionAngleforLandslideAreasinPalosVerdesUsingIn initeSlopeAnalysis

ByCameronYuPresented:Dec17,2012

IntroductionMy project involves the occurrence of landslides in my home town of Palos Verdes, California. As part of my research, I found a landslide inventory map and an underlying geology map of the Palos Verdes area, and what was determined was that all of the landslides have occurred over Pre‐Quaternary Bedrock. Since I have decided to use infinite slope analysis in order to determine where landslides would occur, the fact that the landslides occur over bedrock defies logic about landslides. Infinite slope analysis depends on the fric on angle of the un‐derlying soil, and bedrock does not necessarily have a fric on angle and thus, makes infinite slope analysis obsolete. What may have happened is that a fissure could have been created in the bedrock due to an earthquake, and the crack in the bed‐rock could have been filled with some kind of clay or other soil. Thus, the landslides are actually occurring because the bed‐rock is slipping on the clay film. Working under this assump on, the goal of my project is to find a representa ve fric on angle for the areas where land‐slides occur in the Palos Verdes area. I will do this by crea ng a raster map that shows an area’s factor of safety against a land‐slide, and I will try to match up the landslide map and the un‐derlying geology to match up by varying the fric on angle un l the areas with known landslides have the lowest factor of safe‐ty (<1.5 in this case). Under the assump on that the landslides occur because of low‐strength soil filling fractures of bedrock, the knowledge gained from this project will shed light on a characteris c of the underlying soil. 

Methodology The main output of this project is a raster showing the factor of safety against landslide in the Palos Verdes area, and this was achieved using infi‐nite slope analysis. The equa on used was: 

   Where φ is the fric on angle of the underlying soil, and β is the slope of the surface. The main data layers used are listed below, including their source: 

The .pdf files were converted to .jpeg files, georeferenced into ArcMap, and converted into .shp files. The Eleva on and Underlying Geology maps were then put through the following model, of which the output was the Factor of Safety map. The model is shown below:  

Data Title  Source  Format 

Eleva on Map  USGS  Raster 

Underlying Soils  California Department of Conserva on  .PDF 

Landslide Inventory  California Department of Conserva on  .PDF 

AreaofStudyI focused my area of study to the Palos Verdes Peninsula, as there are 

a number of landslides that have occurred there. This area includes 

the following USGS quadrangles: Redondo Beach, San Pedro, and Tor‐

rance. Below is the area of study in the context of Los Angeles Coun‐

ty, and the map includes the landslide inventory of the area. 

ResultsFigures 3 and 4 are the two main input maps. Each geologic unit in Figure 3 is assigned a fric on angle, φ. It is this parameter and the 

slope angle that is derived from Figure 4 that are input to the model to create Figures 5‐8. The fric on angle for the geologic unit Ls, 

which stands for landslides, is the angle represented in the  tles of the maps. The red areas show a factor of safety of <1.5, which are 

the areas most suscep ble to landslides. 

 

 

 

 

 

 

 

 

 

 

 

Figure 5 is the factor of safety of the Palos Verdes area if we are to assume that the landslides occur on bedrock, which is what the 

old geologic maps suggest. As the figure above suggests, landslide occurrence is highly unlikely in the inventoried areas assuming a 

fric on angle of 45° because the inventoried areas hardly contain any red. The fric on angle was varied in increments of 5° to pro‐

duce maps similar to Figures 6‐8. As the fric on angle is lowered, the amount of red area that appears in the landslide areas increas‐

es, which is what we can expect based on the infinite slope analysis equa on.

ConclusionBased on the factor of safety maps created by the model, the angles at which the landslide areas have the most red are 

when φ = 15° and φ = 10°. However, using φ = 10° in infinite slope analysis is a more reasonable fric on angle to use for 

the landslide units. When the fric on angle changes from 15° to 10°, there is, visually, a significant more amount of red in 

the landslide areas, thus not only using φ = 10° more conserva ve when calcula ng factor of safety, but it is also more ac‐

curate as well. Through my literature review, I found a geologic map from the CA Department of Conserva on of an area 

near Palos Verdes that includes landslides in the map. The fric on angle the map used was φ = 10°, which validates the 

findings from my model. 

Figure 1: Area of Study 

Figure 2: Factor of Safety Model 

Figure 3  Figure 4  Figure 5 

  Figure 6  Figure 7  Figure 8