Wednesday, 13 January 2021: 2:00 PM-3:30 PM
Description:
This session focuses on the machine learning, deep learning, and other data driven methods for hydrological science, simulation, and forecasting.
We encourage contributions on model diagnostics, identification, and benchmarking, and are particularly interested in the development of benchmarks based on data-driven (e.g., machine learning) applications that can be used to evaluate the performance of models using a priori metrics and expectations of performance. The use of novel techniques to assess distributed data or models that focus on impacts to, and understanding of, coupled land-atmosphere and hydrometeorological processes and prediction is also encouraged. In addition, this session also solicits studies of data-driven modeling advances that improve our understanding of physical processes and the representation of observational information in hydrological applications.
Topics include, but are not limited to:
1) Novel development and applications of data-driven modeling, including spatiotemporal aspects
2) Novel machine learning theory for hydrology and geoscience, including integrating physical constraints
3) Assessing quality of model/data products or the fidelity of models of complex terrestrial hydrologic systems
4) Integration of process-based and data-driven approaches
5) Improved understanding of the reliability and extrapolatability of machine learning and other data-driven approaches
6) Deriving scaling relationships or process insights from data
7) Benchmarking with data-driven models
Abstract Submissions Closed
See more of: 35th Conference on Hydrology