The Prediction of Fine Sediment Distribution in Gravel-Bed Rivers Using a Combination of DEM and FNN
Large amounts of fine sediment infiltration into void spaces of coarse bed material have the ability to alter the morphodynamics of rivers and their aquatic ecosystems. Modelling the mechanisms of fine sediment infiltration in gravel-bed is therefore of high significance. We proposed a framework for...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-05-01
|
Series: | Water |
Subjects: | |
Online Access: | https://www.mdpi.com/2073-4441/12/6/1515 |
id |
doaj-e0384ff5414c4e9a81b0b5a7e8a42d71 |
---|---|
record_format |
Article |
spelling |
doaj-e0384ff5414c4e9a81b0b5a7e8a42d712020-11-25T03:04:10ZengMDPI AGWater2073-44412020-05-01121515151510.3390/w12061515The Prediction of Fine Sediment Distribution in Gravel-Bed Rivers Using a Combination of DEM and FNNVan Hieu Bui0Minh Duc Bui1Peter Rutschmann2Institute of Hydraulic and Water Resources Engineering, Technische Universität München, Arcisstrasse 21, D-80333 München, GermanyInstitute of Hydraulic and Water Resources Engineering, Technische Universität München, Arcisstrasse 21, D-80333 München, GermanyInstitute of Hydraulic and Water Resources Engineering, Technische Universität München, Arcisstrasse 21, D-80333 München, GermanyLarge amounts of fine sediment infiltration into void spaces of coarse bed material have the ability to alter the morphodynamics of rivers and their aquatic ecosystems. Modelling the mechanisms of fine sediment infiltration in gravel-bed is therefore of high significance. We proposed a framework for calculating the sediment exchange in two layers. On the basis of the conventional approaches, we derived a two-layer fine sediment sorting, which considers the transportation of fine sediment in the form of infiltration into the void spaces of the gravel-bed. The relationship between the fine sediment exchange and the affected factors was obtained by using the discrete element method (DEM) in combination with feedforward neural networks (FNN). The DEM model was validated and applied for gravel-bed flumes with different sizes of fine sediments. Further, we developed algorithms for extracting information in terms of gravel-bed packing, grain size distribution, and porosity variation. On the basis of the DEM results with this extracted information, we developed an FNN model for fine sediment sorting. Analyzing the calculated results and comparing them with the available measurements showed that our framework can successfully simulate the exchange of fine sediment in gravel-bed rivers.https://www.mdpi.com/2073-4441/12/6/1515fine sediment distributionDEMFNNalgorithmsgravel-bed |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Van Hieu Bui Minh Duc Bui Peter Rutschmann |
spellingShingle |
Van Hieu Bui Minh Duc Bui Peter Rutschmann The Prediction of Fine Sediment Distribution in Gravel-Bed Rivers Using a Combination of DEM and FNN Water fine sediment distribution DEM FNN algorithms gravel-bed |
author_facet |
Van Hieu Bui Minh Duc Bui Peter Rutschmann |
author_sort |
Van Hieu Bui |
title |
The Prediction of Fine Sediment Distribution in Gravel-Bed Rivers Using a Combination of DEM and FNN |
title_short |
The Prediction of Fine Sediment Distribution in Gravel-Bed Rivers Using a Combination of DEM and FNN |
title_full |
The Prediction of Fine Sediment Distribution in Gravel-Bed Rivers Using a Combination of DEM and FNN |
title_fullStr |
The Prediction of Fine Sediment Distribution in Gravel-Bed Rivers Using a Combination of DEM and FNN |
title_full_unstemmed |
The Prediction of Fine Sediment Distribution in Gravel-Bed Rivers Using a Combination of DEM and FNN |
title_sort |
prediction of fine sediment distribution in gravel-bed rivers using a combination of dem and fnn |
publisher |
MDPI AG |
series |
Water |
issn |
2073-4441 |
publishDate |
2020-05-01 |
description |
Large amounts of fine sediment infiltration into void spaces of coarse bed material have the ability to alter the morphodynamics of rivers and their aquatic ecosystems. Modelling the mechanisms of fine sediment infiltration in gravel-bed is therefore of high significance. We proposed a framework for calculating the sediment exchange in two layers. On the basis of the conventional approaches, we derived a two-layer fine sediment sorting, which considers the transportation of fine sediment in the form of infiltration into the void spaces of the gravel-bed. The relationship between the fine sediment exchange and the affected factors was obtained by using the discrete element method (DEM) in combination with feedforward neural networks (FNN). The DEM model was validated and applied for gravel-bed flumes with different sizes of fine sediments. Further, we developed algorithms for extracting information in terms of gravel-bed packing, grain size distribution, and porosity variation. On the basis of the DEM results with this extracted information, we developed an FNN model for fine sediment sorting. Analyzing the calculated results and comparing them with the available measurements showed that our framework can successfully simulate the exchange of fine sediment in gravel-bed rivers. |
topic |
fine sediment distribution DEM FNN algorithms gravel-bed |
url |
https://www.mdpi.com/2073-4441/12/6/1515 |
work_keys_str_mv |
AT vanhieubui thepredictionoffinesedimentdistributioningravelbedriversusingacombinationofdemandfnn AT minhducbui thepredictionoffinesedimentdistributioningravelbedriversusingacombinationofdemandfnn AT peterrutschmann thepredictionoffinesedimentdistributioningravelbedriversusingacombinationofdemandfnn AT vanhieubui predictionoffinesedimentdistributioningravelbedriversusingacombinationofdemandfnn AT minhducbui predictionoffinesedimentdistributioningravelbedriversusingacombinationofdemandfnn AT peterrutschmann predictionoffinesedimentdistributioningravelbedriversusingacombinationofdemandfnn |
_version_ |
1724682507184504832 |