GEOMORPHOLOGY IN HAIDA GWAII

*Although I was inspired to conduct this project after reading literature analyzing the coupling relationship between stream hillslope and drainage basin area, I decided to modify my own study and use landslide area instead of the drainage basin area. It is crucial to distinguish this difference as the landslide area is a small component of the entire drainage basin area. As such, I have been careful to avoid using saying ‘channel area’ which may suggest I had used the area of the entire drainage basin in my calculations, which I have not.*

 

Aside from learning about hydrological modeling and watershed delineation, the other main focus of this research project was to attempt observing a relationship between hillslope and landslide area for both Riley Creek and Gregory Creek landslides. I decided to approach this goal by extracting and graphing two different datasets:
 

Dataset 1 – Plot the instantaneous slope at the point of landslide activation against the entire area of landslide flow

​

Dataset 2 – Plot the instantaneous slope at defined lengths along the landslide hillslope against just the landslide area below that point

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

*Before attempting to edit the geometry of the landslide data, I stored some identification data in the landslide features such as Landslide ID, Basin ID, Catchment ID, Contour, Contour ID, etc…*

​

1) Activation Points

First I made a new Activation Point Feature Class and created points where I estimated the origin of each landslide flow. This was done by manually editing the Landslide Polygon Feature Class and snapping points to the edges or vertices. I made use of the contour map to determine which end of the landslide features were at the top of the hillslope and assigned the activation points as close as possible to large streams flowing through this top region. Although most landslides originated from a distinguishable large stream, as expected, some did not and their activation points were defined as either the landslides highest elevation point or as the center of the polygon’s edge.       

 

 

 

 

 

 

 

 

​

​

​

​

​

2) Slope Model

I next created a Slope Raster Model in units of degrees with the Slope Tool, using the Canadian DEM Model as input.

​

​

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

​

3) Extract Data

The next step used a combination of the Extract Multi Values to Points Tool and Spatial Join Tool to store local slope values and landslide area within the attribute table of the Activation Point Feature Class.

​

4) Export Data

A CSV file of the attribute table was then exported into Excel for plotting using the Table to Excel Tool.

​

​

 

​

*Before attempting to edit the geometry of the landslide data, I stored some identification data in the landslide features such as Landslide ID, Basin ID, Catchment ID, Contour, Contour ID, etc…*

​

For this second dataset analysis, I aimed to use a similar method as described above, creating a set of points into which I could store slope and area data. This time, however I wanted to make a series of points along the length of the landslide, as well as calculate the downslope landslide area as opposed to each landslides’ entire area. Montgomery & Foufoula-Gorgiou (1993) describe a method of measuring basin contributing area per unit contour length, however I was unable to determine the technical procedure employed. Consequently, I decided to design my own rudimentary method of measuring points along the landslide using contour lines.

​

1) Overlay Spatial Analysis/Sub-Landslides

Employing a combination of overlay spatial analysis, I was able to create a new Sub-Landslide Polygon Feature Class where the Landslide Polygon Features were cut into smaller components based on the contour lines. This was accomplished with both the Intersect and Union Tools. I made sure to create a new Sub Landslide ID Field as well as Sub Landslide Area Field, which I later used to calculate the cumulative downslope landslide area.

 

 

 

 

 

 

 

 

 

2) Contour Points

In order to plot the coupling relationship I again had to create a Central Contour Point Feature Class in which to store the data fields. This was done using the Feature to Point Tool which generates an approximate centroid, a point at the geometric center of a polygon.

​

​

 

 

 

 

 

 

​

3) Sub-Landslide Centroid Points

Then I again used the Extract Multi Values to Points and Spatial Join Tools to store centroid slope values, however this time I stored the area of each sub-landslide components.

​

4) Downslope Landslide Area/Data Export

The last remaining data values I needed were the landslide areas downslope of each Contour Point Feature. I had planned was to store a variable in each polygon adding up the cumulative sub-landslide area moving downslope. This could then be subtracted from the known total landslide area to find the remaining landslide area below each sub-landslide component. However, the calculation posed to be quite a challenge as the method of sub-landslide polygon generation I used had flaws as many of the landslides from the raw shapefile data already had subdivides. Had the landslides been represented by a single whole polygon this would not have been an issue, but the raw data was messy. Additionally, the total number of sub-landslide components making up each individual landslide was unknown. I briefly thought about attempting to manually calculate and input the values, however this would not have been possible as I was working with about 1900 polygon features. Although I was unable to calculate this field in ArcMap, I exported the attribute tables to Excel with the Table to Excel Tool and was successful after making a custom function “=IF(A3<>A2, B3, C2-D2)”.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

​

​

​

​

​

​

​

​

​

​

​

​