Sediment yield in the Lower Jinsha River Basin, China: Controlling factors and spatiotemporal patterns

The spatiotemporal pattern of sediment yield is critical for the management of mega cascade reservoirs in the Lower Jinsha River Basin (LJRB), despite that the scarcity of information has limited our knowledge. This study used a stepwise statistical approach to determine the synergistic factors controlling sediment yield, and relevant models were established to provide predictions for ungauged areas in the mountainous basin. Existing observations for 18 catchments (102-104 km2) displayed an imbalanced distribution pattern of specific sediment yield (SSY = 83-1269 t km-2 a � 1) in the LJRB, while 6 combinations of local characteristics were found to properly explain the variability (63-92%). The synthesis of temperature, slope proportion, and drainage area exhibited the most explanatory power, where the temperature factor appeared to exert the dominant effect on SSY. This study has effectively simplified the redundant factor dataset by multivariate and partial correlation methods, lowering the bar for modeling analysis. Predictions have been made to present full distribution patterns of sediment yield in the LJRB, characterized by high yields in the middle (cold and steep) and low yields at both ends (warm and flat). This enabled us to detect considerable ungauged areas with abundant sediment yields between cascade reservoirs Xiluodu and Baihetan. Regionalized SSY has decreased by 10-71% over the past 50 years, associating with the superior vegetation condition in response to rising temperatures. Moreover, 2050 projections via two extreme Shared Socioeconomic Pathways (SSP119 and SSP585) suggested that regionalized SSY would continue to decrease by 32-51% on average. The current assessment and future prediction of sediment yield in the LJRB should be useful for engineering purposes, and the proposed modeling approach can provide reference for upper lands and mountainous basins.