**4. Conclusions**

The unstructured-grid, Finite-Volume Coastal Ocean Model (FVCOM) was used to simulate the flows in Discovery Passage including the adjoining Lower Campbell River, British Columbia, Canada. Discovery Passage located between the mainland of British Columbia and Vancouver Island along with many smaller islands. Two applications of high resolution 3D modeling in this area were conducted and presented in this paper. One is for the Lower Campbell River from the upstream as far as the John Hart Hydroelectric dam. The horizontal resolution varies from 0.27 m to 32 m in the unstructured triangular mesh. The triangular grid for this FVCOM application has 16,625 nodes with 32,845 elements for existing dam tailrace runs. Model output was verified with observations and compared with the previous 2D MIKE21 model results with the same model grid and bottom roughness distribution. The model demonstrated a good performance in simulating the flow structures and capturing the volume transport split ratio at the left and right channels of First Island, located some 150 m downstream of the powerhouse as areas of concern.

The second application is to compute ocean currents immediately above the seabed along the present and potential underwater electrical cable crossing routes between Campbell River and Quadra Island across Discovery Passage. The horizontal grid resolution varies from approximately 10 m for the study area, incorporating a multibeam high resolution bathymetric data set. In the vertical, a sigma-coordinate system was applied with 15 levels. Higher resolution was used near the bottom with inter-layer spacing ranging from 0.125 to 0.0005 of total water depth. The high vertical resolution at near bottom levels allows simulation of currents at 0.02, 0.07, 0.17, 0.38, 0.76, and 1.51 m above the seabed given the maximum water depths of 100 m. Model results were verified using available tidal gauge and current meter data throughout the Discovery Passage. The model behaves very well so as to simulate the strong tidal currents at very high resolution in the horizontal and vertical. The one year maximum velocity distribution crossing the channel along the routes was examined, which is useful for extreme value analysis as related to engineering design. Differences were found in the distribution of the maximum near surface velocities and the near bottom velocities along the crossing routes.

#### **Acknowledgments**

We wish to express our appreciation to SNC-Lavalin Inc. and Terra Remote Sensing Inc. for the opportunity to undertake this study. We would also like to thank Northwest Hydraulic Consultants for observational data and early 2D MIKE21 model inputs and results.

#### **Author Contributions**

Y. Lin wrote the manuscript and led the modeling study. D. B. Fissel reviewed/edited the manuscript and supervised the design and planning of the model study.

#### **Conflicts of Interest**

The authors declare no conflict of interest.
