Real-time watershed-scale hydrologic probabilistic forecasting with an integrated groundwater - surface water model

Authors: Xiaoyong Xu*, , Steven Frey, Aquanty Inc., Waterloo, ON, Canada, Graham Stonebridge, Aquanty Inc., Waterloo, ON, Canada, Omar Khader, Aquanty Inc., Waterloo, ON, Canada, Edward Sudicky, Aquanty Inc., Waterloo, ON, Canada
Topics: Water Resources and Hydrology
Keywords: Hydrological forecasting, Ensemble prediction, HydroGeoSphere, NAEFS
Session Type: Paper
Day: 4/5/2019
Start / End Time: 8:00 AM / 9:40 AM
Room: Taylor, Marriott, Mezzanine Level
Presentation File: No File Uploaded


There is growing recognition that real-time hydrologic modeling would benefit assessment of water-related risks and water resources management across various scales. Until recently, water resources modeling and forecasting typically relied upon surface water models in which the 3D subsurface variably-saturated flow is either ignored or only loosely connected to surface water and deterministic meteorological prediction products. This work presents the implementation and evaluation of real-time ensemble forecasting for the South Nation watershed, Canada with the HydroGeoSphere model, a high-resolution 3D fully integrated groundwater-surface water/variably-saturated subsurface flow simulator. The predication system is able to automatically ingest real-time probabilistic and deterministic weather forecast fields from multiple sources, such as the North American Ensemble Forecast System (NAEFS), ECCC’s Global Deterministic Prediction System (GDPS), The Weather Company’s gridded forecasts, and NCEP’s Global Forecast System (GFS). A methodology based on a combination of the variational assimilation and an evolving initial condition (IC) library is used to provide the IC for issuing hydrological forecasts. Preliminary evaluation showed that the probabilistic (ensemble) hydrological forecasts outperformed the deterministic forecasts and a simple extrapolation, and could help capture water resources modeling and forecasting uncertainties and improve our capability to assess and predict hydrologic risks.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login