EGU General Assembly 2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Ensemble Water Level Prediction System: Improving the Representation of Model Uncertainty

Natacha Bernier1, Oleksandr Huziy2, Keith Thompson3, Pengcheng Wang1, Benoit Pouliot2, and Syd Pell1
Natacha Bernier et al.
  • 1Environment and Climate change Canada, Meteorological Research Division, Canada (
  • 2Environment and Climate change Canada, Meteorological Service of Canada, Canada
  • 3Dalhousie University, Department of Oceanography, Canada

Concern over increased flooding and the need for earlier and more reliable risk forecasts motivate the continued development of operational forecasts of coastal water level. We report here on results from a year long ensemble of total water level forecasts calculated using a dynamical ocean model forced with ensemble atmospheric forcing and tidal boundary conditions. We focus on the east coast of Canada. The domain includes the Gulf of St. Lawrence, the Labrador Shelf, the Scotian Shelf, and the Gulf of Maine. The water level ensemble is made of a control and 20 perturbed members. Individual forecasts are produced twice daily for 16 days.


The novelty of the present study is in the exploration of perturbations of the ocean contributions. In addition to examining how uncertainty in atmospheric forcing maps into flood risk, we also explore the feasibility, and impact, of perturbing the ocean tides. We use a recent case study to demonstrate our findings.


How to cite: Bernier, N., Huziy, O., Thompson, K., Wang, P., Pouliot, B., and Pell, S.: Ensemble Water Level Prediction System: Improving the Representation of Model Uncertainty , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-22142,, 2020

Comments on the presentation

AC: Author Comment | CC: Community Comment | Report abuse

Presentation version 1 – uploaded on 04 May 2020 , no comments