Short Course: The Science and Practice of Operational Ensemble Hydrological Prediction

May 21 - 23, 2019

Location: NCAR | Boulder, CO


Andy Wood
Martyn Clark


Over the last decade, hydrologic ensemble prediction has progressed from a research application to a central thrust of operational streamflow and hydrologic forecasting system development around the world. The overarching objective of ensemble forecasting is to quantify and include uncertainty in predictions and thereby to support risk-based decision-making related to hydrologic resources and hazards. Such forecasts are now generated for major river basins across Europe, North America, Australia and other continents, at lead times from days to seasons, supporting needs from early flood warning and water management to the direct fine-grained operation of hydropower systems.

The design of ensemble prediction systems to support decision-making is a multi-disciplinary challenge. It requires an understanding not only of earth system predictability, meteorological and climate prediction, hydrologic and river system modeling, but also of a range of statistical techniques including data assimilation, model output post-processing, verification, and economic and risk analysis concepts (e.g., cost-loss assessment).

This 3-day short course draws together experts involved in developing, using and promoting ensemble hydrologic forecasting systems, and will teach attendees about the history, status and concepts of ensemble hydrologic forecasting systems. The course will also demonstrate hands-on examples of key techniques for a number of case study basins.

This short course leverages the experience of the international Hydrologic Ensemble Prediction Experiment, a community of practice that has advocated and developed such systems since 2004. Participants will gain an understanding of different strategies for designing an ensemble prediction system, and of the tradeoffs associated with different choices in system development.

Eligibility and Requirements

Applicants must apply to attend the workshop. Applications will be accepted from current and incoming graduate students, post-docs, and professionals working in hydrology and/or the atmospheric sciences. Instructors will review applications to verify in advance that the applicant has the necessary skills and experience to successfully participate in the course and get the most out of the course topics.

  • Prior hydrologic and/or atmospheric modeling experience is advised.
  • Must be comfortable working in a UNIX environment, running command line operations and text editors.
  • Familiarity with the scripting languages such as R or Python is recommended.