The study of using learning methods for the estimation of sediment transport

What is it about?

The phenomena of sediment transport is so complex, non-linear, and stochastic that one cannot rely on the conventional methods for a quick response regarding how much volume of sediment is transported to downstream for a given discharge. This study introduces a machine-learning method that uses a combination of laboratory and field data sets to fairly estimate the rate of sediment transport using important geometrical and hydraulic information. The results show the superiority of the proposed method in comparison to conventional methods of sediment estimation.

Featured Image

Why is it important?

The paper shows the ability of machine learning methods as an alternative approach for sediment transport with no need for knowing the complex relationships between dependent and independent variables.