Keywords: nested model, hydrodynamics, interpolation, one-way nesting, DIVAST.

The present research deals with the development of a one-way nested version of the combined hydrodynamic and water quality model: Depth Integrated Velocity and Solute Transport (DIVAST). DIVAST is a depth integrated, time-variant model capable of simulating two-dimensional distributions of currents, water surface elevations and various water quality parameters within the modelling domain. The test-case for the development process was an artificial rectangular harbour. It is hoped to apply the nested model to a validated primary production model of Cork Harbour, Ireland. For accurate water quality predictions, the Cork Harbour model uses a spatial resolution of 30m. By using a nested model, significant savings on computational cost can be achieved.

The nested model approach allows one to increase spatial resolution in a sub-region of the model domain without incurring the computational expense of fine resolution over the entire domain. The model developed during this research is a one-way (or passive) nested model. For this class of model the coarse resolution flow field affects the fine resolution region by providing boundary conditions for the fine grid domain, however, there is no mechanism by which the evolution in the fine resolution region can affect the flow field in the coarse grid (and hence its own boundary conditions). A number of versions of the nested model were developed to investigate the effect of boundary interpolation techniques on model performance. The interpolation techniques examined were the zeroth-order, linear and quadratic interpolation schemes.

The accuracy of the nested model was determined by comparing the computed current velocities with those calculated by a fine-grid model with the same spatial resolution as the nested domain. Using the zero-order interpolation technique, it was found that perturbations in current velocity occurred in the region of the open nested boundary. These perturbations were significantly reduced by the use of the linear interpolation technique and were eliminated when the quadratic interpolation technique was employed. For the vast majority of the model domain the error in velocity predictions between the nested and fine-grid models was less than 2%. The error was largest in the region close to the back wall of the harbour. However, the actual magnitude of the error in this region was very small, with the maximum less than 0.004 m/s. It is thought that this slight error is due to a small reduction in mass flux across the nested domain boundary which is a common problem with nested models. Overall, the nested model was found to reproduce the results from the fine-grid model with a high degree of accuracy.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £120 +P&P)