A brief review of recent global trends in suspended sediment estimation studies

抄録

Sediment has the potential to influence the landscape, economy and way of life. If the estimation is accurate and the distribution characteristics are known, sediment can be used as a resource. In this brief review, we evaluated recent advances in suspended sediment estimation techniques between 2011 and 2022. The three most popular techniques are Sediment Rating Curves, Machine Learning Models and Physical and/or Numerical Models. Their potential and limitations are briefly explored. Our findings indicate that Machine Learning Models perform better, whereas Sediment Rating Curves offer flexibility in field applications. On the other hand, the paucity of necessary high-resolution data for model calibration and validation hinders the performance of Physical and/or Numerical Models. The sampling frequency of input sediment data emerges as a pivotal factor influencing the performance of all methods. To overcome this issue, a semi-automatic surrogate method might be useful. Also, particle size could be included in the rating curves to obtain additional information. This is especially valuable for understanding sediment dynamics and the process of delta formation.