How to Use ArcGIS Pro to Assess Landslide Susceptibility

| |

A landslide is defined by the United States Geological Surveys (USGS) as the movement of a mass of a rock, debris or earth down a slope. Landslides are a significant geologic threat that can be found in virtually every part of the world

That said, this tutorial considers how landslide susceptibility can be assessed using ArcGIS Pro. Based on the methodology of the tutorial as shown in figure 1, the following factors will be considered; hydrology/rivers, roads, slope, and Normalized Difference Vegetation Index (NDVI)

Figure 1: Flow chart showing steps to calculate landslide susceptibility using ArcGISPro.
Figure 1: Flow chart showing steps to calculate landslide susceptibility using ArcGISPro.

Preparing the Road Data

  1. Open and create a new project in ArcGIS Pro. 
  2. Import the road data and boundary data using the “Add Data” icon from the menu bar. In case the road shapefile exceeds the boundary, use the clip tool as searched from the geoprocessing toolbox to limit the extent of the road data to the boundary of our area of interest. 
  3. In the “clip (Spatial Analyst)” window, set “input features” to road data and “clip features” to the boundary data. 
  4. Click “Run.”
Figure 2: Illustration of "Add Data" and "Clip" tool
Figure 2: Illustration of “Add Data” and “Clip” tool

“Euclidean distance” tool from Arctoolbox can be used to specify an allowable distance around the road feature.

  • In the search bar from the geoprocessing window, type and search “Euclidean distance.” 
  • From the “Euclidean distance” window, specify “input raster or feature source’’ as the road shapefile, and also “output distance raster’’ as desired output location and name. “Maximum distance” may be set to specify the allowable distance around the road feature. For the purposes of this tutorial, “Maximum distance” is left to default.
Figure 3: illustration of "Euclidean distance"
Figure 3: illustration of “Euclidean distance”

PREPARING SLOPE DATA

  1. Download slope data according to an area of interest from USGS.com or import GPS coordinates with elevation data which spans over the study area using the “Add data” icon.

For the purposes of this tutorial, Inverse Distance Weighted (IDW) would be used to spatially interpolate GPS coordinates using elevation data.

  • Type and search, “IDW” from the search bar in the geoprocessing toolbox. Select, “IDW (Spatial Analyst tools)” from the options that pops up.
  • From the IDW window, specify input point features as the GPS coordinates. Also, specify “Z value field” as the field of the elevation data.
  • In order to specify the processing extent to the boundary, select the environments tab closer to the parameters tab. 
  • From the environments tab, set coordinate system as “same as our boundary data” or “same as road data.” Also, set mask as “same as boundary data.” 
  • After, select the “Parameters tab,” and click on “Run.”
Figure 4: illustration of “IDW (Parameters and Environments tab)”
Figure 4: illustration of “IDW (Parameters and Environments tab)”

PREPARING NDVI DATA

For the purposes of this tutorial, the Landsat 7 imagery will be used for the calculation of NDVI, which can be downloaded freely from USGS.com. If there are lines in the imagery, it is expedient to first perform line correction using QGIS, ArcGIS or ArcGIS Pro. This video may be helpful (https://www.youtube.com/watch?v=Og9vQP3e9uk&t=184s). 

Once line correction has been performed over the Bands 4 and 5 of the landsat 7 imagery, calculation of NDVI can begin using the following steps;

  1. Use the “Add data” icon from the menu bar to import the corrected Band 4 and 5 onto the map canvas.
  2. Type and search “raster calculator” from the search bar.
  3. The formula, “float (‘Landsat 7 Band 4’-‘Landsat 7 Band 5’)/float (‘Landsat 7 Band 4’ + ‘Landsat 7 Band 5’) can be used for the calculation of NDVI.
  4. Specify “Output Raster” and select “Run.”
Figure 5: Illustration of NDVI
Figure 5: Illustration of NDVI

PREPARING HYDROLOGY/RIVER DATA

  1. Use the “Add data” icon to import Digital Elevation Model (DEM) data onto the map canvas.
  2. Type and search “Fill” from the search bar of the geoprocessing toolbox. 
  3. From the “Fill” window, set “Input Surface Raster (ISR)” as the DEM data. Also, set “Output Surface Raster (OSR)” as the desired output name and location and click “Run.”
Figure 6: Illustration of “Fill”
Figure 6: Illustration of “Fill”
  • Type and search “Flow Direction” from the search bar of the geoprocessing toolbox.
  • From the “Flow Direction” window, set “ISR” as results of “Fill,” Also, set “OSR” as the desired output name and location and click “Run.”
  • Type and search “Flow Accumulation” from the search bar of the geoprocessing toolbox.
  • From the “Flow Accumulation” window, set “Input Flow Direction Raster (IFDR)” as results of “Flow Direction.” Also, set “Output Flow Direction Raster” as the desired output name and location, and click “Run.”
  • Type and search “Raster Calculator” from the search bar of the geoprocessing toolbox. 

Per the Area of interest, calculation can be set to, “Flow Accumulation>3000.”  Depending on the number of detailed rivers desired as the output, the threshold value can be selected from between 500-5000.  After, click “Run.”

Figure 7: Illustration of “Flow accumulation” and “Raster calculator”
Figure 7: Illustration of “Flow accumulation” and “Raster calculator”
  • Type and search “Stream Link” from the search bar of the geoprocessing toolbox.
  • Set “Input Stream Raster (ISTR)” as results of “Raster Calculator” and set “IFDR” as the results of “Flow Direction.” Set Output name and location, and click “Run.”
  • Type and search “Stream Order” from the search bar of the geoprocessing toolbox.
  • Set “ISTR” as results of the “Stream Link,” and set “IFDR” as the results of “Flow Direction.” Set Output name and location, and click “Run.”
  • Type and search “Stream to Feature” from the search bar of the geoprecessing toolbox.
  • Set “ISTR” as results of the “Stream Order,” and set “IFDR” as the results of the “Flow Direction.” Set Output name and location, and click “Run.”

“Euclidean distance” tool from Arctoolbox can be used to specify an allowable distance    around the river features.

  1. In the search bar from the geoprocessing window, type and search “Euclidean distance.” 
  2. From the “Euclidean distance window,” specify ‘’input raster or feature source’’ as the results of the “Feature to stream,” and also ‘’output distance raster’’ as our desired output location and name. 
Figure 8: Illustration of “Stream to feature” and “Euclidean distance”
Figure 8: Illustration of “Stream to feature” and “Euclidean distance”

LANDSLIDE SUSCEPTIBILITY ASSESSMENT

  1. Type and search “Reclassify (Spatial Analyst)” from the geoprocessing search bar.
  2. Specify “input raster” as results of “Euclidean distance roads.” 
  3. Click “Classify” and set the number of classes to 9. Set “Output Raster” name and location and click “Run.”
Figure 9: Illustration of "Reclassify" tool.
Figure 9: Illustration of “Reclassify” tool
  • Repeat Steps for results of “Euclidean Distance Rivers,” “Slope,” and “NDVI.”
  • Type and search “Weighted Overlay” from the search bar of the geoprocessing toolbox.
  •  From the arrow down icon closer to the Raster tab, select the results of “Euclidean distance rivers,” “Slope,” “Euclidean distance roads,” and “NDVI.”

Influences can be shared among these factors percentage-wise, to sum up to 100. For the purposes of this tutorial, the following influence should be used; NDVI=35%, slope=15%, rivers=30%, and roads=20%. Click “Run.”      

Figure 10: Illustration of "Symbology".
Figure 10: Illustration of “Symbology”

A landslide susceptibility map will be generated. To change the labeling and coloring of the results of the landslide map (Weighted Overlay), double-click on the color palette of the resultant overlay map.

Figure 11: Landslide susceptibility map of Greater Accra, Ghana
Figure 11: Landslide susceptibility map of Greater Accra, Ghana

Related


Enter your email to receive the weekly GIS Lounge newsletter: