Characterizing stream planform geometry using a novel application of spectral analysis

Abstract

Quantifying the curvature in stream planform geometries traditionally uses discrete analysis methods to calculate sinuosity or channel index. The challenge with discrete methods is in selecting the appropriate length of channel to use as a window size in the calculations. In this study, a new methodology was introduced to guide and optimize window size selection with comparisons to more traditional methods. Channel index was the metric used in this study, defined as the ratio of stream length to straight line distance. Three methods were used to calculate channel index: classical method using total stream length, modified classical method using infrastructure-defined segments for window size, and a novel critical point method using continuous channel index spectra to determine window size. As a case study, each method was applied to an agriculturally dominant watershed in southwestern Ontario at Parkhill Creek to compare planform geometry changes between 1954 and 2015. This creek is historically known to have been straightened artificially for agricultural purposes. The novel critical point method was the best of the three, providing the best spatial resolution while generating optimized window sizes to calculate channel index for each individual tributary as well as the main creek channel. Using the critical point method, the creek was determined to be, on average, 23% straighter in 2015 than in 1954. The results of this study demonstrate the advantages of the novel critical point method in analyzing stream planform geometries and can be adapted for research in other streams.