Key points are not available for this paper at this time.
Estimation of sediment concentration is crucial for several theoretical as well as practical applications. The sediment rating curve method is widely adopted to estimate sediment concentration by relating sediment concentration (C) and river discharge (Q). Underestimation of sediment concentration by the rating curve method often happens as it ignores the dynamic relationship between sediment concentration (C) and river discharge (Q). In this study, we recognize the dynamic relationship between C and Q and focus on sediment recession events only. A power-law trend is observed for both the C–Q and (−dC/dt)−C relationship during recession: −dC/dt=fCe. The variation of exponent e is relatively less due to its dependence on static basin properties, whereas coefficient f varies to a larger extent since it is controlled by dynamic basin properties such as soil moisture and sediment availability condition. The dynamic factors responsible for coefficient f evolve slowly with time, which means there is an influence of past data on rating relation. Here, we propose the estimation of coefficient f using past discharge data only from which sediment concentration can be predicted at any river cross section. Unlike the normal sediment rating curve (SRC) technique, the proposed model can predict future suspended sediment concentration during a recession event at daily time scale. The efficacy of the model has been tested with 80 US Geological Survey basins with 75th, 50th, and 25th values of Nash–Sutcliffe efficiency as 0.78, 0.68, and 0.41, where SRC gives, 0.39, 0.2, and −0.04, respectively.
Mohanty et al. (Sat,) studied this question.