Abstract
The porosity of gravel riverbed material often is an essential parameter to estimate the sediment transport rate, groundwater-river flow interaction, river ecosystem, and fluvial geomorphology. Current methods of porosity estimation are time-consuming in simulation. To evaluate the relation between porosity and grain size distribution (GSD), this study proposed a hybrid model of deep learning Long Short-Term Memory (LSTM) combined with the Discrete Element Method (DEM). The DEM is applied to model the packing pattern of gravel-bed structure and fine sediment infiltration processes in three-dimensional (3D) space. The combined approaches for porosity calculation enable the porosity to be determined through real time images, fast labeling to be applied, and validation to be done. DEM outputs based on the porosity dataset were utilized to develop the deep learning LSTM model for predicting bed porosity based on the GSD. The simulation results validated with the experimental data then segregated into 800 cross sections along the vertical direction of gravel pack. Two DEM packing cases, i.e., clogging and penetration are tested to predict the porosity. The LSTM model performance measures for porosity estimation along the z-direction are the coefficient of determination (R2), root mean squared error (RMSE), and mean absolute error (MAE) with values of 0.99, 0.01, and 0.01 respectively, which is better than the values obtained for the Clogging case which are 0.71, 0.14, and 0.03, respectively. The use of the LSTM in combination with the DEM model yields satisfactory results in a less complex gravel pack DEM setup, suggesting that it could be a viable alternative to minimize the simulation time and provide a robust tool for gravel riverbed porosity prediction. The simulated results showed that the hybrid model of the LSTM combined with the DEM is reliable and accurate in porosity prediction in gravel-bed river test samples.
Original language | English |
---|---|
Pages (from-to) | 128-140 |
Number of pages | 13 |
Journal | International Journal of Sediment Research |
Volume | 38 |
Issue number | 1 |
DOIs | |
State | Published - Feb 2023 |
Externally published | Yes |
Funding
In current study, a new proposed hybrid model of a deep learning LSTM combined with a DEM model is proposed for predicting porosity in gravel-bed rivers. At first, the high-resolution of 3D bed structure, formed by fine sediment infiltrating into gravel randomly packing the bed, was simulated using the DEM. Next, the outcomes of the DEM (i.e., simulated sediment distributions and simulated porosity) were compared with the experimental data to get relative improvement applying the LSTM model. To reduce the DEM computational time, the deep learning LSTM was used to learn the behavior of the grains in forming porosity characteristics of the gravel-bed structure. With the deep learning LSTM model, the numerical porosity values were generated with porosity as a label and the GSD as inputs along the cross sections at 800 different elevations in the gravel-bed structure. In addition, a structural cross-section image of the DEM simulated gravel-bed was generated to obtain the GSD for the porosity values. Finally, the output derived from the DEM model was used to support the deep learning LSTM model input for prediction of the gravel-bed porosity.
Keywords
- Bed porosity
- Discrete element method
- Gravel-bed river
- LSTM
- Numerical modeling