Numerical Modelling

The power of our computer model "COCIRM-ASL" lies in its ability to predict conditions in regions where data is sparse or when extensive data collection is expensive or impractical.

Technical Papers
Numerical Modeling Brochure
Numerical Modeling Projects

ASL's Super Computer for Numerical Modeling

ASL has developed super computing capabilities using MPICH (Message Passing Interface). This is a leading edge parallel processing system. The system has multiple nodes, each with multiple CPUs. It is being used to improve ASL's existing current, sediment transport, ice tracking, and plume models to work faster and in more detail. The system distributes large groups of calculations to a number of computers to generate results faster. It greatly reduces the run-time of our high resolution 3D circulation models.

Numerical modeling is a powerful method of visualizing the dynamic behaviour of physical systems. We have developed a three-dimensional computer model (COCIRM-ASL) capable of accurately simulating water circulation in:
  • Rivers
  • Estuaries
  • Coastal Waters
  • Continental Shelf and Deeper Waters
Our model is founded solidly on the science of fluid dynamics for circulation including such natural forces as:
  • Tides
  • Density stratification and buoyancy
  • Wind stress
  • Drag arising from the shoreline and bottom
  • Coriolis

Wave Model Study of a new Marina Development in Victoria Harbour, BC, Canada

The variable discharge from such engineered works as dams, power stations and sewage treatment stations can readily be included in the model. Our model has been fully calibrated and validated through comparison with extensive data sets in a variety of project environments.

Features and Benefits

Successful calibration and validation of a numerical model against field measurements is an affirmation of our understanding of the natural environment being studied. The power of our computer model "COCIRM-ASL" lies in its ability to predict currents, temperature, salinity and sediment in regions where data is sparse or when extensive data collection is expensive or impractical. COCIRM-ASL can undertake "what if" studies to investigate the impact on river, estuarine or coastal circulation patterns of the placement, for instance, of:
  • A new dam or plant
  • The effect of changing discharge levels or operational configurations.
  • Coastal engineering structure.
For application of COCIRM-ASL to a new project area, one need only input the geometry of the modeled domain (shoreline, bathymetry, engineering structures), open boundary conditions, and physical properties that are known. The distribution and behaviour of key properties can be readily simulated within the ASL model, including:
  • Tidal currents and circulations
  • Temperature, salinity, and suspended sediment concentration (SSC)
  • Bottom siltation and scour
  • Biological or chemical distributions: plankton abundance, coliform concentrations, oil spills

Simulated surface current during maximum flood without Burrard Generating Station (BGS) in operation.

Model Description

COCIRM-ASL uses hydrodynamic pressure, sigma-transform, and variety of turbulence parameters. It solves for the time-dependent, three-dimensional velocities (u, v, w), temperature(T), salinity(s), SSC(c) as well as water surface elevation (Jiang, 1999). It also includes wetting/drying and nested sub-grid schemes, capable of incorporating tidal flats, buoyant jets and relatively small interested areas.

The model boundary conditions consist of the momentum flux (wind stress) at the water surface and the shear stress at the bottom in terms of quadratic or linear law.
At open boundaries, the water levels, velocities or radiation conditions are specified. In the case of discharge from a dam, the resulted currents are oriented to the same direction as the spillway.

A semi-implicit finite difference method is applied in COCIRM-ASL. The numerical solution method has the advantages of a minimum degree of implicitness, good stability (unconditionally stable when one neglects horizontal diffusion) and consistency, and high computational efficiency at a low computational cost. Grid sizes can range from 10 m to kilometers in size.