| This thesis is a comparison and evaluation of the flood severity (Depth, Velocity, and D*V or D∼V combination relationship etc.) of two representative hydrodynamic models: one-dimensional DAMBRK, and two-dimensional MIKE21. In addition, this thesis also includes a literature review on several proposed dam-break life loss estimation models, on some hydrodynamic models, and some discussion on the model development for life loss estimation.;This thesis takes Logan River First Dam, Utah, as the study case, and finds that if topographic data is properly defined in 1D models, and sinuosity is properly taken into account, their results will be quite close to 2D models' (refers to depth; the velocity distribution of 1D models is mostly unreasonable), but this is just the problem (topographic definition) 1D models usually cannot outreach themselves. Actually 2D hydraulic models are also influenced by topographic data: the finer the topography, the better the outcomes. The resulting differences between the MIKE21 run with the 10 m DEM (the derived one) and the MIKE21 run with the 8 ft DEM (the original one) are much less than those between the MIKE21 run with the 10 m DEM and the DAMBRK new run based on 8 ft DEM. (8 ft = 2.4384 m);The downstream boundary condition just impacts the flow states of its vicinity in both 1D and 2D models.;Flood severity comparisons indicate that, for the cases examined in this thesis, Velocity had a much stronger role in determining the shape of the Depth*Velocity curves than Depth. This finding suggests that in performing flood routing for life loss and property loss estimation, particular attention should be paid to velocity. This pattern is more marked with MIKE21 than DAMBRK. |
