Long-reach Biotope Mapping: Deriving Low Flow Hydraulic Habitat from Aerial Imagery

Understanding of the type and distribution of hydraulic habitat along watercourses is valuable from an ecological and a morphological perspective. The data quantify system state and may be used against benchmark criteria to define system status level and degradation. Current mapping techniques are subjective, time consuming and expensive when carried out over long reaches often requiring specialist field skills. This paper proposes a novel approach to hydraulic habitat mapping using readily available aerial imagery (GoogleEarth and Bing maps) to generate long-reach digital elevation models, which are subsequently used in a 2D modelling domain (JFlow+) to predict hydraulic habitat in the form of biotope types and distribution from Froude number classification. The approach is tested on a 1-km reach of the river Wharfe, England, a morphologically and hydraulically varied watercourse. Biotope mapping of the study reach recorded a distribution of 49% pools, 33% glides and 17% riffles, compared with an observed 54% pools, 32% glides, 13% riffles and 1% broken standing waves/chutes, suggesting that gross biotope distribution may be reliably mapped using the technique when compared with field mapping but that depth estimation error leads to classification issues around transition zones. The improved spatial detail and objective mapping achieved by the technique also provide valuable sub-feature detail on hydraulic habitat variation not picked up by conventional survey. The ease of digital elevation model construction allows for rapid assessment of extended reaches offering an efficient mechanism for whole river ecological assessment, flagging critical sites that would benefit from more detailed field assessment. Copyright (c) 2016 John Wiley & Sons, Ltd.