Predicting urban channel morphology amidst multiple complexities

Urbanisation has resulted in rapid geomorphic changes and ecosystem degradation in many streams worldwide. The question of how channel response to urbanisation can be accurately predicted amid multiple driving factors remains a significant scientific challenge. Numerous efforts have been made to address this issue over the past 80 years. Nevertheless, channel response resulting from urbanisation differs across physiographic regions, making it difficult to predict. The main goal of this study is to investigate the most important factors influencing morphological changes, such as channel area, in streams with urbanised catchments. To accomplish this aim, we reviewed the literature on the application of hydraulic geometry models to urban-impacted rivers worldwide. We set out to investigate two specific questions: (i) to what extent and under what circumstances can urban stream bankfull cross-sectional area be predicted by catchment area and impervious cover metrics? and (ii) is stream bankfull width more susceptible than stream bankfull depth to urban-induced enlargement? However, over 90 % of the urban channels studied in the literature we reviewed showed irregular response patterns to urbanisation, owing to the complex interplay of local factors, such as geology, sediment, hardpoints, and riparian vegetation. Our most important conclusion is that local context is an important controller of stream channel response to urbanisation. While catchment area is a good predictor of channel area in some urban sites, it is a poor predictor of channel area in other urban sites. This contrasts with non-urban sites, where catchment area usually predicts channel area more strongly. In addition, impervious cover alone, which represents the direct hydrological impact of urbanisation, is insufficient to fully explain channel enlargement and incision. Therefore, models of channel response to urbanisation must consider the specific local factors that drive or limit these responses in order to be effective as predictive and explanatory tools.