Target groups are both scientists and practitioners in the field of operational reservoir management. Water Authorities interested in improving their early warning system by means of model updating techniques are also welcome.
This course addresses different data assimilation techniques within the operational forecasting context, particularly for hydrological forecasting application such as reservoir management. Data assimilation is an essential part of operational forecasting. It allows to improve the estimate of the initial model state variables and to increase the forecast accuracy considering the most recent observations of the system. Data assimilation methods applied in forecasting systems range from sequential to batch processing types and include a variety of techniques like Ensemble Kalman Filter, Particle Filtering, Asynchronous Filtering, Variational Assimilation, among others.
This course addresses different data assimilation techniques within the operational forecasting context, particularly for hydrological forecasting application such as reservoir management. We provide an overview of operational forecasting systems and then present a detailed description of multiple data assimilation techniques with a practical application using conceptual lumped models to facilitate hands-on exercises to the participants. Theory and practical exercises correspond to 60% and 40% of the course time respectively.
Upon completion, the participant should be able to:
- Recognise the main challenges in operational hydrological forecasting in case of reservoir management
- Explain the concept of model updating using data assimilation techniques
- Identify the advantages and disadvantages of each updating method to improve operational forecasting
- To be able to apply different data assimilation approaches to operational reservoir management applications and analyses their usefulness
- Discuss the concept of forecast uncertainty