Several water sectors are involved with hydrological forecasting at different time scales (from hours to months and seasons ahead) and spatial scales (from local catchments to regional, national and global scales). Forecasts of river flows or water levels are essential for different activities, such as managing floods and the risk of inundation, anticipating droughts, operating hydropower plants, planning navigation schedules and agriculture irrigation, or assessing water resources for any future needs.

Hydro-meteorological forecasting systems are increasingly being developed to provide skilful predictions of possible future scenarios and estimate uncertainties to support decision-making. Probabilistic forecasting and uncertainty estimation can be done using a wide range of techniques, including ensemble techniques, statistical post-processing or a combination of these. The science of hydrological forecasting has progressed fast in the past decade and has shown that advanced forecasting systems can improve preparedness for water-related risks and add value to optimized decisions. However, the large variety of existing techniques and the huge amount of data to be handled can be seen as too complex for operational users. A better understanding of the existing techniques, their strengths and limitations, is needed to enable a confident use of probabilistic hydrological forecasts in hydrology and an efficient communication of risks. Within the framework of a forecast – decision – re sponse system, effective forecasting requires that attention is given to issues such as forecast post-processing, real-time data assimilation, visualization, communication, decision-making and training.

This short course comprises an introduction to these topics. It includes lecturers, game-play exercises and real-case examples of hydrological forecasting. Special attention is paid to current challenges to research scientists, practitioners and operational users. The main contents of the course are:

1.) Terminology and basic concepts: uncertainty, risk and probability, real-time data assimilation, differences between hydrologic simulation and forecasting.
2.) Techniques for estimating the predictive uncertainty: ensembles and statistical post-processing.
3.) Forecast verification/evaluation: how good is my (probabilistic) forecast? How valuable is it for decision-making?
4.) Using probabilistic forecasts in operational practice.

The course will be given by Prof. Marie-Amélie Boucher (UQAC, Québec, Canada) and Dr. Jan Verkade (Deltares, Delft, The Netherlands).

This course is supported by HEPEX, a community of research and practice to advance hydrologic ensemble prediction: www.hepex.org and the Young Hydrologic Society (YHS): https://younghs.com/.