ABSTRACT

The wise selection of modeling approaches with an appropriate level of complexity for the study objectives is critical for robust inference. In this paper, the structure of a cost–performance grid designed for flood modeling is presented. The grid is developed to compare different flood modeling approaches of variable complexity and to provide guidance on the proper selection of the data-model couple. The methodology involves defining metrics to quantify the three variables: data costs, model costs, and performance. Preliminarily, 18 flood modeling applications in the literature were arbitrarily selected and analyzed to guide the implementation of the grid. The cost–performance diagram allows us to trace a cost–performance curve and group applications in four zones corresponding to four modeling approaches (empirical and geomorphic, hydrological, hydraulic, and coupling). The grid is a tool to support the comparison, classification, and future selection of cost-effective modeling approaches. It is flexible and can be extrapolated to other modeling objectives.

摘要

对于研究目标而言，明智地选择具有适当复杂度的建模方法对于进行可靠的推理至关重要。在本文中，提出了一种为洪水建模而设计的成本－性能网格的结构。开发网格的目的是比较可变复杂性的不同洪水建模方法，并为正确选择数据模型对提供指导。该方法涉及定义度量以量化三个变量：数据成本，模型成本和性能。初步地，任意选择和分析了文献中的18个洪水建模应用程序，以指导网格的实施。成本绩效图允许我们在四个区域（经验和地貌，水文，液压和联轴器）。网格是支持比较，分类和将来选择具有成本效益的建模方法的工具。它很灵活，可以外推到其他建模目标。