Presenters-0707: Difference between revisions
From CSDMS
No edit summary |
No edit summary |
||
| Line 15: | Line 15: | ||
|CSDMS meeting youtube code=0 | |CSDMS meeting youtube code=0 | ||
|CSDMS meeting youtube views={{Youtube_0}} | |CSDMS meeting youtube views={{Youtube_0}} | ||
|CSDMS meeting participants= | |CSDMS meeting participants=18 | ||
}} | }} | ||
{{Presenters additional material | {{Presenters additional material | ||
|Working group member=Terrestrial Working Group, Coastal Working Group, Marine Working Group, Education and Knowledge Transfer (EKT) Working Group, Cyberinformatics and Numerics Working Group, Hydrology Focus Research Group, Chesapeake Focus Research Group, Critical Zone Focus Research Group, Human Dimensions Focus Research Group, Geodynamics Focus Research Group, Ecosystem Dynamics Focus Research Group, Coastal Vulnerability Initiative, Continental Margin Initiative, Artificial Intelligence & Machine Learning Initiative, Modeling Platform Interoperability Initiative, River Network Modeling Initiative | |Working group member=Terrestrial Working Group, Coastal Working Group, Marine Working Group, Education and Knowledge Transfer (EKT) Working Group, Cyberinformatics and Numerics Working Group, Hydrology Focus Research Group, Chesapeake Focus Research Group, Critical Zone Focus Research Group, Human Dimensions Focus Research Group, Geodynamics Focus Research Group, Ecosystem Dynamics Focus Research Group, Coastal Vulnerability Initiative, Continental Margin Initiative, Artificial Intelligence & Machine Learning Initiative, Modeling Platform Interoperability Initiative, River Network Modeling Initiative | ||
}} | }} | ||
Latest revision as of 12:17, 24 January 2026
CSDMS 2026: Modeling Landscapes in Motion
Advancing Spatiotemporal Modeling with Deep Learning: CNN–LSTM Integration
Abstract
This workshop introduces participants to a hybrid Deep Learning (DL) framework that integrates Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks to capture complex spatiotemporal dynamics. Such architectures are increasingly important for modeling and forecasting in Earth and Environmental Sciences, where processes evolve across both space and time. While flooding will be used as a motivating example, through predicting inundation maps and water depth in two-dimensional domains, the framework applies broadly to other domains, including geomorphological evolution, groundwater dynamics, coastal wave processes, and climate-driven hydrological responses. Participants will begin with a conceptual overview of CNNs, LSTMs, and their integration into a unified pipeline, followed by hands-on implementation using PyTorch. A small-scale two-dimensional flume dataset will serve as the training example, chosen for its simplicity, yet the same approach scales to complex, real-world systems and large spatial domains. The workshop will also emphasize the advantages of hybrid deep learning approaches compared to conventional physics-based models (e.g., hydrodynamic simulators such as HEC-RAS), particularly in terms of computational efficiency and scalability for large-scale or long-duration simulations. Finally, participants will learn best practices, recognize common challenges, and explore strategies for adapting CNN–LSTM architectures to a wide range of spatiotemporal applications. For this workshop, a basic understanding of machine learning in Python is recommended to maximize the benefits of this session.
Please acknowledge the original contributors when you are using this material. If there are any copyright issues, please let us know (CSDMSweb@colorado.edu) and we will respond as soon as possible.
Of interest for:
