Structural comparison of multiple rivers across Canada: the importance of scale for characterizing the river complexity

English Abstract:

Ongoing developments in remote sensing and geographical information science are massively improving efficiencies in analyzing earth surface features. Various applications of these techniques are widely used to assess environmental and ecological processes. With the development of new satellite sensors and the lowered cost availability of high resolution multi spectral imagery, fluvial geomorphology is experiencing a revolution in mapping streams both at high resolution and over large spatial extents. Exploiting the power of aerial or satellite imagery is thus particularly useful in Riverscape framework. This study presents a multi-scale remote sensing method that helps describe the hydraulic structures of wetted habitats. Metrics of habitat description are calculated from i) depth maps obtained through spectral analysis on multiple satellite images and ii) complementary modelized velocity maps, in order to assess the spatial structure of the selected rivers. A principal component analysis on the set of metrics was used to cluster river reaches in regards to their “structural signature”. We tested our method in a set of 10 Canadian rivers and investigated the link between the local channel morphology and the regional geomorphic contexts. Results highlight the benefits of such a method for characterizing 1) the habitat heterogeneity and 2) the natural (alluvial) versus controlled (non-alluvial) river variability. Results also demonstrate the importance of using remote sensing tools in a multi-river analysis not only for constraint saving in time and onerous field work, but also for uniformly characterize river structures across the country. Applications of this approach can include 1) specific fish habitat detection for regional management, 2) understanding how river habitat heterogeneity affects fish distribution and 3) guidance for river restoration site location.