Abstract
Traditional hydrology treats water volume (discharge, storage) as the fundamental descriptor of river systems, but this view neglects the mechanical action that flowing water exerts on landscapes—a dimension we term hydrodynamic action. We propose a dual-ontology framework characterizing rivers by two complementary dimensions: water volume ("how much water?") and hydrodynamic action ("what can the water do?"). We define hydrodynamic action ℋ = ρu2 [Pa], the streamwise momentum flux density, at the point scale; momentum flux F = ∫ρu2dA [N] at the cross-section scale; and a basin-scale hydrodynamic budget ∂𝐌/∂t = 𝐆 − 𝐃 + 𝐅in − 𝐅out, with 𝐆 and 𝐃 the momentum source and sink terms. The hydrodynamic characteristic time τℋ = M/D parallels water residence time τw = V/Q; the two are independent and can differ by 2–3 orders of magnitude. Using daily discharge, stage, and cross-sectional records from five Yangtze stations spanning ~2,000 km (Zhutuo to Datong), we provide the first quantification of this decoupling: (i) hydrodynamic action falls ~51% downstream while water volume increases 2.6-fold (−9% end-to-end for the ten-year-window stations); (ii) at matched discharge, hydrodynamic action at Yichang declined by 43% in the post-TGD period—far exceeding the 9% period-mean discharge reduction—through ℋ ∝ u2 nonlinearity and channel incision; and (iii) the hydrodynamic regime shifts from "intense and abrupt" (mountain reaches) to "mild and persistent" (alluvial plains). The framework subsumes established theories (Shields, Leopold–Maddock hydraulic geometry, ecological flow regimes) while exposing blind spots of volume-only metrics in erosion, channel stability, habitat, and engineering impact assessment.