GEOMORPHOLOGY IN HAIDA GWAII

In the end, both of the datasets collected showed a small positive correlation between hillslope and landslide area.

In Figure 1, the Gregory Creek slides have a steeper coupling relationship relative to those in Riley Creek. It also seems that the Gregory Creek plot has a stronger R² value (coefficient of determination), suggesting the Gregory Creek trendline more accurately replicates the observed data than the Riley Creek model.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

For both the Riley and Gregory Creek datasets, Figure 2 illustrates much shallower sloping regression trendlines. Although positive, this suggests that the inclusion of data points further down the hillslope causes the slope-area relationship to decrease and reach an inflection point. This is consistent with the findings of Montgomery et al [1993] seen in Figure 3, which characterized an inflection of the slope-area graph as being a transitional zone changing from debris-flow dominated channels to lower-gradient colluvial channels. Because I chose to use landslide area in this study instead of drainage basin area, my results only pertain to the hillslope regions of the watershed. Figure 3 shows this channel head region as having a positive relationship prior to reaching the inflection point. Both of the datasets I collected are predominantly points of channel head, and therefore I believe the reduction in slope between Figure 1 and Figure 2 is supported in this model, as the Figure 2 data includes points further down the hillslope yet still considered as channel head.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

In Figure 3 it is noted that hillslope terrain tends to be convex in nature, while surface elevation further down the drainage basin often flattens out or becomes concave. I attempted to simulate this as well, using the Curvature Tool to create a Second Derivative Surface Raster. This value is the longitudinal change in surface slope; the rate of change in slope. After overlaying the landslide features above this Curvature Model, I discovered that many of the landslide areas directly correlate to the concavity of the terrain. This is illustrated in Figure 4 and Figure 5 as the convex curvature of the surface indicated by the dark raster cells highly correlated with the outlines of the overlying landslide areas.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

​

 

 

 

​

In the future, I plan to pursue similar types of geomorphology-based GIS research. If I were to re-attempt this project I would suggest using drainage basin area in my analysis;  not just landslide area. I would additionally look into finding an improved method of data capture throughout the entire channel basin as well as taking linear cumulative measurements along the dominant stream in each basin.