TINs have the advantage compared to grid-based elevation models in that they require less memory than grids. In addition, linear features are more accurately represented with TINs than with DEMs. When using grids to model channels and other linear features, edges must be always oriented along the horizontal, vertical, or diagonal directions. TINs eliminate this data redundancy and are thus better suited for modeling streams and other linear features. TINs can be constructed so that triangle edges conform to features and are not restricted to lie in the horizontal, vertical, and diagonal directions. The TIN data structure is also often more efficient because the terrain model can be adapted readily to the surface being modeled. In areas where the terrain is flat, only few a points need to be utilized (Nelson et al., 1999).
Grid-based watershed modeling is advantageous over TIN-based watershed delineation in that grids have a simpler data structure than TINS, grid- based data is very abundant, and grid-based models are reproducible. Other disadvantages of TINs result when inserting breaklines. Inserting breaklines may result in small or long thin triangles which, in turn, will cause difficulties in
numerical round off or tolerance problems. TINs have the major that large TINs are difficult to work with, and editing pits and flat a very time consuming process, especially when areas are
disadvantage in triangles can be
determining al., 1999a).
TIN-based watershed delineation is based on the process of tracing a flow across triangle surfaces. Because each triangle has a flat surface, the mathematics