Improving sediment load estimations: The case of the Yarlung Zangbo River (the upper Brahmaputra, Tibet Plateau)

Abstract

Suspended sediment load of a river represents integrated results of soil erosion, landform change and ecosystem variation occurring within the river basin. Accurate estimation of suspended sediment load is helpful for distinguishing the impact of natural and anthropogenic factors on earth system processes within river basin under global climate change. Suspended sediment load, with long-term in-situ observation limited by harsh conditions, can be estimated by sediment rating curves and various subdivision methods with suspended sediment concentration and discharge data of low-frequency. New sediment rating curve subdivision methods based on flood ranks and suspended sediment concentration stages were proposed in this study. The flood ranks subdivision method, which defined a flood to begin when discharge exceeded the critical discharge value (rank1) or the previous peak flow (rank2 and the later), and end when fluctuating discharge reached the peak flow of current flood again (which is the beginning of next flood) or lowered to the critical discharge value (the last rank), was suitable for application in the basin where sediments were mainly transported and exhausted in early events. The suspended sediment concentration stages subdivision method, i.e. the rising and falling limbs of suspended sediment concentration separated by peak dates estimated based on accumulated precipitation, was suitable for application in the basin where soil erosion was closely related to precipitation. Compared to the traditional sediment rating curve, and seasonal, discharge classes and discharge stages subdivision methods, these newly proposed methods can improve suspended sediment concentration and subsequently suspended sediment load estimation in the middle reach of the Yarlung Zangbo River with higher coefficients of determination and Nash-Sutcliffe efficiency coefficients and lower bias and root-mean-square errors. Moreover, combination of the two separately developed methods presented further improved estimation. The newly proposed sediment rating curve subdivision methods could be helpful for suspended sediment load estimation and therefore are useful for water resource management within river basins.