Automatic Delineation of Functional River Reach Boundaries for River Research and Applications

The river reach is a pervasive term within contemporary river research and applications. Yet, despite its prevalence, there is a notable lack of consistency in its definition. This paper identifies the prevalence of two broad types of reach definition within the academic literature: operational, and functional, and argues that a functional definition is more suitable for applications within river research and management. A range of statistical sequence zonation algorithms that were originally derived for geological well log analysis are compared for their ability to automatically identify functional reach boundaries. An analysis of variance -based global boundary hunting algorithm is identified as the most suitable. Two potential practical applications of automatic reach delineation are demonstrated: identification of functional reach boundaries first, in a sequence of predicted sediment transport capacity for reach-based sediment transport models; and second, in a sequence of RHS Habitat Quality Assessment scores for identification of areas in a catchment in need of habitat restoration efforts. Finally, the paper discusses how this type of functional reach identification procedure might be applied in many other areas of river research and applications, and how a multivariate version of a statistical zonation algorithm might prove useful in facilitating integrated catchment management by identifying reach boundaries common across all variables of interest in the system.