Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the...

24
LANDSLIDE SUCCEPTABILITY MAPPING (Case study of SRILANKA)

Transcript of Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the...

Page 1: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

LANDSLIDE SUCCEPTABILITY MAPPING(Case study of SRILANKA)

Page 2: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

Bivariate statistical method

1. Landslide rupture

2. Relevant factors (parameters) for the prediction of landslide :

- Lithology - Slope- Landuse - Aspect- Soiltype - Curvature

Statistical Map

Page 3: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

Statistical Analysis Map

3. Weight value for each factor : Landslide Index

- Dens. clas : Landslide density within parameter class

- Density map : Landslide density within entire map

- N pixel (Si) : Number of pixels, which contain landslides per parameter class

- N pixel (Ni): Total number of pixels in a parameter class

Page 4: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

Statistical Analysis Map

Six weight value maps will be calculated:

1. lithology weight map

2. Soil type weight map

3. Landuse weight map

4. Slope weight map

5. Aspect weight map

6. Curvature weight map

Hazard succeptibility map

Page 5: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

Data Available

• Soil map

• Contour lines(10m intervals)

• Land use map

• Landslide rupture map

• Reference coordinate system for Srilanka: Central Meridian, False Northing Latitude of origin, Scale factor, false northings and false eastings are 200,000meters,

• Used Software : ArcGIS 9.3

Page 6: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

Work Flow

1. Created a file geodatabase (ArcGIS 9.3)2. Imported our features( all shapefiles)3. Rasterisation of our feature Landslide rupture (define the extent of

our raster by using mask)4. Generated DEM using contours: define cell size and mask

5. Reclassification : Aspect, slope and curvature

6. Rasterization : lithology, landuse and soil type features ( polygon to raster)

7. Zonal tables ( zonal statistics as table)

8. Join tables to corresponding classes

9. Calculate six weight value maps

10. Hazard susceptibility map (sum up all weight value maps) (Weighted Sum Operation)

Page 7: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

1.Rasterization:LandSide Rupture Feature

• Cell size : 20m • Assign the extent of our raster • by using given mask feature

• Total number of pixel (2752)

Page 8: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

2.Generation Aspect Slope and Curvature

1. DEM from Contour lines

(Spatial Analyst Tools TIN management)

Page 9: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

Continue….

Spatial analyst tools (Aspect, slope and curvature)

Aspect Slope Curvature

All these rasters do not have values ( no Attribute tables, floating rasters)

Page 10: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

3.Rasterization:Lithology Landuse and Soil Polygon Raster

Soiltype : 4 classes Lithology : 3 classes Landuse : 21 classes

Page 11: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

4.Reclassification

Spatial analyst tools Reclass Reclassify

Aspect 5 classes are made

Page 12: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

Continue…..

Slope 5 classes

Curvature 5 classes

Page 13: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

Continue…….

Slope CurvatureAspect

Page 14: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

Reclassification of LanduseRaster

Landuse 4 classes

Page 15: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

5.Zonal Tables

Spatial analyst tools Zonal Zonal statistics as table

• We want to calculate the number of pixels of landslide that fall in each class of our raster slope,

• Repeated the same process is done for 5 remaining Rasters

Page 16: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

6.Join Tables to Correspond Rasters

• Join tables of each raster to its corresponding zonal statistic table,• Total number of pixels of that raster in each zonal statistic table,• Use “Field calculator”, added a new field of weight in our table

Page 17: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

7.Field Calculator

Calculate the weight valuesby introducing the given formula in field calculator

Page 18: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

8.Calculate Six Weight Value Maps

e.g. Aspect-ve values mean Low Risk Area+ve values mean High Risk Area

Page 19: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

9.Weight Maps

Aspect weight map Slope weight map

Page 20: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

Continue…..

Curvature weight map Soiltype weight map

Page 21: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

Continue….

Lithology weight map Landuse weight map

Page 22: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

10.Final Landslide Susceptibility Map

Page 23: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

Thank you for your attention.

Page 24: Bivariate statistical method 1. Landslide rupture 2. Relevant factors (parameters) for the prediction of landslide : - Lithology - Slope - Landuse - Aspect.

24

QUESTIONS???