-
- H13D-05: Comparing Multi-Model Approaches to Simulate Streamflow Across the Contiguous USA
-
NOLA CC
Author(s):Generic 'disconnected' Message
Cyril Thebault, University of Calgary (First Author, Presenting Author)
Wouter Knoben, University of Calgary
Nans Addor, Fathom
Martyn Clark, University of Calgary
Predicting flow in rivers is crucial for water resources management, protecting populations against floods, and supporting industries, farming, and power generation. To this end, research and operational communities have developed hydrological models to provide a simplified representation of the real-system. The complexity of processes within catchments makes hydrological modelling particularly challenging and a single model can not accurately represent streamflow everywhere at all time. To overcome this limitation, multi-model approaches have been developed to leverage an ensemble of models by capitalizing on their individual strengths and their complementarity.
This study compares various multi-model approaches to improve streamflow. We analyzed data from 559 rivers across the U.S.A. and used a large ensemble of 78 models that are configured to focus on either high-flow or low-flow conditions. The results show small difference between the different multi-model approaches. This work also highlighted very promising results for a potential single model solution, selected from a large ensemble of models.
Scientific DisciplineNeighborhoodType
Enter Note
Go to previous page in this tab
Session
