Keywords: adaptive mesh, one-way nested model, solute transport, DIVAST.

An adaptive mesh solute transport model is introduced. The model is a one-way, multiply nested version of the hydrobiological model DIVAST. An inherent problem with the static structured grids used in many hydrobiological models is that the computational cost is not optimal. In order to resolve detailed features within the fixed grid geometry it is necessary to increase the resolution across the whole domain, thereby significantly increasing computational effort. Classical zoom nested models provide high resolutions only in the areas of interest, thus significantly reducing the computational cost; however grid structure is still fixed and static. The adaptive mesh model provides for both static and dynamic grids and changing grid structure. The adaptive mesh grids can therefore track features of interest within the model domain while minimising the area over which high resolution is required giving further computational savings.

A nested simulation involves one outer grid which contains one or more inner nested grids. Each nested region (or child) is entirely contained within a single coarser grid (the parent). The model allows multiple levels of nesting, in which case children are also parents. The fine grids may be telescoped to any depth and several fine grids may share the same parent at the same level of nesting. Nested grid open boundary data is interpolated (both in time and space) from parent grid data; a linear technique is used for temporal interpolation while an inverse distance weighted technique is used for spatial interpolation. The adaptive mesh scheme allows for any nested domain to move anywhere within the boundaries of its parent domain. All nests are eligible to be moving nests and the user must specify whether a nested grid is to be static or dynamic. The model provides two methods of facilitating dynamically moving nests during the model simulation: specified and automatic. For a specified move, the timing of a nest move and the extent of the lateral move is defined entirely by the user. For the automatically moving nest, grid movement is controlled by the model itself according to some predetermined rule.

The model was extensively tested by simulating solute transport in Galway Bay. In order to determine model behaviour and accuracy results were compared against those from a fine resolution model of Galway Bay. A very high level of correlation was observed between both sets of results indicating that the adaptive mesh model behaves well and is capable of computing solute transport to a high degree of accuracy. Analysis of the computation times showed that the adaptive mesh model offers a 20% computational saving over the classical zoom nested model and a 77% saving over the non-nested fine resolution model.

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 £140 +P&P)