Global Bankfull River Width Estimates for Major Rivers

Global coverage of river reaches with an upstream drainage area greater than 25 km² and width exceeding 30 meters. This dataset provides machine learning-derived estimates of bankfull river width alongside environmental covariates for each reach. The data was generated by synthesizing satellite-derived width estimates from the GRWL database and MERIT Hydro hydrography.

Use Cases

• Modeling river discharge and flood inundation based on estimated channel geometry.
• Analyzing relationships between river width and environmental factors like slope, aridity, and bedrock properties.
• Validating and improving global hydrography and digital elevation models.
• Studying global patterns in river morphology and their controls.

Strengths

• Provides a global estimate of a key hydraulic variable (bankfull width) derived from a synthesis of major satellite and DEM sources.
• Includes multiple predictive environmental covariates (e.g., slope, sinuosity, permeability, aridity index) for each river reach.
• Dataset is associated with a peer-reviewed publication in Geophysical Research Letters, indicating scientific validation.

Limitations

• Metadata on exact row count, file size beyond '~1.4GB', and specific update dates is missing from all provided sources.
• Limited to rivers wider than 30 meters, excluding smaller streams and headwaters.
• The dataset's derivation relies on the accuracy and resolution of underlying DEM and satellite data sources.

Provenance

Source

Peirong Lin (Princeton University) and co-authors, as described in the associated Geophysical Research Letters article.

Collection Method

Machine learning predictions leveraging the GRWL database and MERIT Hydro, vectorized from DEMs under a 25 km² channelization threshold.

Geography

Global