Integrating multi-algorithm machine learning with zircon REE geochemistry to trace source-to-sink provenance in the Mekong River Basin

Fangbin Liu

doi:10.1016/j.geomorph.2026.110276

Integrating multi-algorithm machine learning with zircon REE geochemistry to trace source-to-sink provenance in the Mekong River Basin

2026-03-11

0

PRE

AI

F

Fangbin Liu *

Z

Zhaohan Guo

X

Xiaoming Shen

Z

Zhe Wang

M

Min Wang

王妍媖 (Yanying Wang)

X

Xin Ma

H

Honghua Liu

DOI：10.1016/j.geomorph.2026.110276

Original

Original request for help

Save

Abstract

En 中文

• Machine learning and zircon REE classify granitoid sources in the Mekong River. • S-type granitoids dominate sediment signatures, especially from the lower knickzone. • SHAP values reveal key geochemical features driving sediment provenance predictions.

Keywords:

Machine learning

Zircon REE

Sediment provenance

Granitoid sources

SHAP values

Integrating multi-algorithm machine learning with zircon REE geochemistry to trace source-to-sink provenance in the Mekong River Basin

Abstract

Journal

Organization