Characterizing fluvial systems at basin scale by fuzzy signatures of hydromorphological drivers in data scarce environments

Despite the relevance of river hydromorphology (HYMO) for integrated water resource management, consistent geomorphic information at the scale of whole river basin is still scarce, especially in emerging economies. In this paper, we propose a new, scalable and globally applicable framework to analyze and classify fluvial systems in data-scarce environments. The framework is based on a data-driven analysis of a multivariate data set of 6 key hydro-morphologic drivers derived using freely available remote-sensing information and several in situ hydrological time series. Core of the framework is a fuzzy classifier that assigns a characteristic signature of HYMO drivers to individual river reaches. We demonstrate the framework on the Red River Basin, a large, trans-boundary river basin in Vietnam and China, where human-induced morphological change, concretely endangering local livelihoods, is contrasted by very limited HYMO information. The derived HYMO information covers spatial scales from the entire basin to individual reaches. It conveys relevant information on subbasin hydro-morphologic characteristic as well as on local geomorphologic forms and processes. The fuzzy classifier successfully distinguishes abrupt from continuous downstream change and spatially dissects the river system in segments with homogeneous hydro-morphologic forcings. Successful numerical modelling of morphologic forms and process rates based on the HYMO signatures indicates that the multivariate, basin-scale classification captures relevant morphological drivers, outperforms an analysis based on local drivers only, and can support river management from diverse, morphology related perspectives over a wide range of scales.