Colin Phillips, Claire Masteller, Louise J. Slater , Kieran Dunne, Simona Francalanci, Stefano Lanzoni, Dorothy Jane Merritts, Eric Lajeunesse, Douglas J Jerolmack 

Many cities and settlements are organized around alluvial rivers, which are self-formed channels composed of gravel, sand and mud. Much of the time alluvial river channels are oversized, in that they could accommodate greater water flow; yet during extreme storms they are woefully undersized, and potentially catastrophic flooding can occur. Considering widely varying hydroclimates, sediment supply, geologic constraints and varying vegetation, it is not altogether obvious that rivers should achieve an average channel geometry that is pattern stable – let alone predictable from theory. Yet, natural rivers follow remarkably consistent hydraulic-geometry scaling relations, that are reproduced in laboratory experiments. Starting with the constraint that channel formation requires that fluid stress exceeds the threshold for sediment entrainment, we review the explanatory power of threshold channel models. Moreover we explore how deviations from threshold channel theory relate to higher-order dynamics of fluid and sediment transport — essentially perturbations to the threshold state — clarifying
misconceptions regarding model applicability. Finally, we demonstrate the utility of the threshold channel framework for understanding channel patterns and responses to variations in external forcing such as hydroclimate and land use. Accurate field determination of the entrainment threshold itself is a notorious problem, and emerges as a central challenge in further development and application of threshold channel theory.