Sign in Subscribe
Artificial neural networks

Advanced Neural Network Approaches for Optimal Check Dam Site Selection in Sub-tropical Climates

Strategic Insights

— 2 min read

Researchers at American University have developed a new method for identifying optimal check dam sites in the Kangsabati River Basin using advanced neural network models. The study, which was supported by Fundacao para a Ciencia e a Tecnologia (FCT), used Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and Recurrent Neural Networks (RNN) to assess the suitability of check dam sites for water conservation, sediment trapping, soil erosion control, and groundwater recharge. The research found that the CNN model demonstrated the best performance, achieving a sensitivity of 0.93, specificity of 0.85, and AUC scores of 0.91 during training and 0.82 during validation.

Key Takeaways:

  • The study used three advanced neural network models: Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and Recurrent Neural Networks (RNN) to identify optimal check dam sites in the Kangsabati River Basin.
  • The CNN model demonstrated the best performance, achieving a sensitivity of 0.93, specificity of 0.85, and AUC scores of 0.91 during training and 0.82 during validation.
  • Spatial analysis indicated that 14.33% of the study area was classified as highly suitable and 17.21% as very highly suitable in the CNN model.
  • The research found that the optimal check dam sites could contribute significantly to sustainable water resource management in the region.
  • The study concluded that future studies should focus on land use, climate change impacts, and design modifications to improve long-term check dam performance.
  • The research has been peer-reviewed and published in the journal Advances in Space Research.
  • The study included authors from American University Sharjah, Department of Civil Engineering: Rabin Chakrabortty, Tarig Ali, Malay Pramanik, Abu Reza Md. Towfiqul Islam, Chaitanya Baliram Pande, Romulus Costache, and Mohamed Abioui.

Statistics:

  • 14.33% of the study area was classified as highly suitable in the CNN model.
  • 17.21% of the study area was classified as very highly suitable in the CNN model.
  • The CNN model achieved a sensitivity of 0.93 and specificity of 0.85.
  • The CNN model had an AUC score of 0.91 during training and 0.82 during validation.

Sources:

  • Fundacao para a Ciencia e a Tecnologia (FCT) - provided financial support for the research
  • American University - performed the research and published the findings
  • Advances in Space Research - published the research in the journal
  • Elsevier Sci Ltd - publisher of Advances in Space Research
  • Rabin Chakrabortty, American University Sharjah, Department of Civil Engineering - contacted for additional information
  • Tarig Ali, Malay Pramanik, Abu Reza Md. Towfiqul Islam, Chaitanya Baliram Pande, Romulus Costache, and Mohamed Abioui - additional authors of the research.

Read more