Revision as of 17:15, 7 January 2015 view source Mitch3007 (talk \| contribs) Bureaucrats, Administrators 1,585 edits →Grid Setup and Cell Size ← Older edit		Revision as of 17:22, 7 January 2015 view source Mitch3007 (talk \| contribs) Bureaucrats, Administrators 1,585 edits →Grid Setup and Cell Size Newer edit →
Line 24: ==Grid Setup and Cell Size== North-south oriented 10, 15 and 30m fixed grid domains were developed. A series of fixed, north-south oriented grid domains were developed. The coarsest fixed grid resolutions were sufficiently detailed to represent the river bathymetry, so no 1D/2D coupling was incorporated. Where possible, a fully 2D solution is preferable over a 1D/2D linked representation provided the 2D resolution of the primary flowpaths is sufficiently fine. A minimum recommended 2D resolution across the river for TUFLOW for this model is a 30m grid (it is recommended that a minimum of 4 fixed grid cells be used for modelling a primary flowpath). For this model the upper sections near the two inflows are not represented well, especially the northern Tributary Inflow. However, this is unlikely to have an adverse effect on the objectives of the modelling, but would affect the comparison to the high water marks near these inflows. When conceptualising the model domain, a fully 2D solution is preferable over a 1D/2D linked representation provided the 2D resolution of the primary flowpath/s is sufficiently fine. 'Sufficiently fine' is representing the channel with at minimum 4 fixed grid cells perpendicular to the direction of flow. Here, the 30m fixed grid resolution was determined sufficiently detailed to represent the river bathymetry. For this model the upper sections near the two inflows are not represented well, especially the northern Tributary Inflow. However, this is unlikely to have an adverse effect on the objectives of the modelling, but would affect the comparison to the high water marks near these inflows. As there is no 1D/2D linking, Challenge 2 provided an opportunity to provide some comparisons between our two different 2D solvers. Results from TUFLOW (finite difference implicit solution) and TUFLOW’s GPU Module (explicit finite volume GPU solver over a grid) are presented.

FMA Challenge 2: Difference between revisions