New article in Nature Communications

Nyberg et al. (2023) introduce the Global Channel Belt (GCB) model in a new publication published in Nature Communications. The authors use advances in machine learning to identify patterns that define landforms characterizing the extent of river channel belts, including the river channel and their associated levees, bars, splays, and overbank features. The global surface area of channel belts is found to cover an area of approximately 30.5 x 105 km2, or ~7 times larger than the extent of active river channels alone. This new dataset has the potential to aid in a wide range of applications, including understanding river evolution over time, assessing flood risks, analyzing freshwater budgets, management of river ecosystems, and improved assessment of biogeochemical cycles of carbon on our climate. Moreover, the methods developed show the potential to create global landform maps to define morphology of features that describe processes of formation. These maps cannot be created using traditional pixel-based classification techniques and open a wide new avenue in Earth observation research. 
If you're interested in learning more, please check out our article in Nature Communications https://www.nature.com/articles/s41467-023-37852-8 or explore our interactive map https://bjornburrnyberg.users.earthengine.app/view/gcbm.