Global Riverbank Slope Patterns Inferred from SWOT-Derived Hypsometry Curves

Riverbank slopes are a fundamental component of river geometry, governing channel–floodplain connectivity, cross-sectional shape, and overbank flow behavior. These slopes constrain ecohydrological processes such as hyporheic exchange, riparian inundation, floodplain residence times, and hydrodynamic models. Despite their importance, riverbank slopes remain poorly quantified at regional to global scales due to the limited ability of remote sensing to resolve near-bank topography and the impracticality of applying high-resolution topobathymetric models across extensive river networks.

The Surface Water and Ocean Topography (SWOT) satellite provides global, simultaneous observations of river width and water surface elevation (WSE), but does not directly measure channel depth, bank geometry, or cross-sectional area. A further limitation is that raw SWOT width measurements exhibit systematic bias as a function of cross-track distance from nadir. As a result, repeated observations of the same river node from different satellite passes can exhibit width differences that exceed expected natural variability, reflecting sensor- and geometry-related bias rather than true hydraulic change. To mitigate this effect, we apply automated width corrections during data processing using the river-spatial-scale repository (available via GitHub at https://github.com/SWOTAlgorithms/river-spatial-scale). This framework implements bias adjustment through spatial comparison of measurements across nodes and applies configurable windowed smoothing to reduce cross-track-dependent errors while preserving natural along-river variations; additional quality filtering ensures consistent and physically meaningful width estimates for hypsometric analysis.

We examine monotonic width–WSE hypsometric curves using these bias-corrected SWOT data, as a proxy for inferring effective riverbank slope. This approach builds on hydraulic geometry theory, which posits that observable hydraulic variables can constrain unobserved channel properties. We construct hypsometric curves using SWOT vector products at node scale, in which individual nodes are spaced at approximately 200 m, across a global set of rivers. We evaluate performance and robustness in comparison with in situ measurements of bathymetry along six rivers in the United States, Colombia, France, and Italy, spanning diverse climatic, geomorphic, hydraulic, and anthropogenic conditions, including systems with significant channel modification and hydraulic infrastructure. We analyze the shape, slope, and regression behavior of the hypsometric curves to estimate effective riverbank slopes and assess spatial variability along river corridors. Results demonstrate that SWOT-derived hypsometric curves provide a practical and scalable proxy for riverbank slope, enabling global-scale characterization of channel geometry using SWOT data. This approach offers new opportunities to investigate spatial patterns in hydraulic geometry, evaluate landscape controls on river form, and support hydrologic, hydraulic, and geomorphic applications in data-limited regions.