ABSTRACT
Despite the availability of excellent forecasting and numerical flood models, there remain gaps in our understanding of flood systems, and hence in the applications of these models. However, the size, complexity, and diversity of the flood systems pose considerable challenges in gaining a comprehensive understanding of flood system elements and their relationships. A model is expected to provide complex information about the system in a robust and effective manner. It provides the means of making key explicit assumptions and considerations about the system. Recent realizations in the flood control industry of the need to make the integrated decisions based on 1D as well as 2D hydraulic models to tackle spatially distributed input hydrographs’ routing time, and on-line and off-line storage issues is notable. The bathymetry of a complex landscape does not always permit the identification of flow split and storage locations from field reconnaissance trips. The complex physical characteristics of watersheds with a split flow, tributary flow, in-channel sand and gravel mine pits, critical flow breakout locations, braids, perched channels, roadway crossings, and dam pool area can exacerbate the ability to extract flow inundation information solely from a 1D hydraulic model. A 2D model that utilizes the hydrographs from a rainfall-runoff model as inputs does a better job at capturing the resolution necessary to identify the flood risk zones in such complex terrains. While the power of hydrologic and hydraulic modeling is appreciated worldwide, the outputs from these models need to be carefully assessed to make sure they represent the true watershed flow dynamics. The selection of the right class of models and knowledge-based integration of multiple models are the watershed-specific problems in flood risk assessment.
