EGU26: Suspended Sediment Fluxes and Decadal Trends in the Humid Tropics: Machine Learning Reconstruction and Coupled Modelling in Upper Blue Nile Tributaries (06 May 2026)

Using machine learning methods — in particular Quantile Regression Forests (QRF) — continuous daily sedigraphs (1990–2020) are reconstructed for two catchments in the Upper Blue Nile Basin and used for the first time to calibrate the coupled hydro-sedimentological model WASA-SED. Decadal analyses reveal contrasting trajectories: whilst intensifying rainfall in the Gilgel Abay catchment strongly increased streamflow but sediment loads declined in recent years, both streamflow and sediment loads rose markedly in the Gumara catchment — driven primarily by wetland loss and rapid urban expansion.

👉Read abstract