Understanding Riverine Habitat Inundation Patterns: Remote Sensing Tools and Techniques

Abstract

Regulated flow releases are key components in the ecological restoration of river ecosystems. Understanding relationships among hydrology, habitat inundation, and ecological processes at appropriate spatial scales is required to optimise ecological outcomes. This study used high resolution remotely sensed data to develop a three dimensional (3D) channel and floodplain riverine habitat inundation model. High-resolution LiDAR and digital imagery data were used to develop 3D digital-elevation models. Data segmentation techniques were used to map channel and non-channel areas, while in-channel habitats were identified through topographic characteristics. Object Oriented analysis was used to separate overbank vegetated and non-vegetated areas. Field surveys were undertaken to record the location and diameter of large woody debris (LWD), and canopy foliar density in addition to general landscape description. A qualitative evaluation of the accuracy of the habitat map showed that LiDAR-based discrimination of topographic features and the derivation of a canopy foliar density layer was successful. Mapping of LWD was unsuccessful due to size relative to the LiDAR point cloud density and vertical resolution. The high resolution remotely sensed data used in this study provide a useful tool for understanding river processes at reach scales by linking inundation patterns to geomorphic features, providing a framework for predicting ecological outcomes from river flow regimes.