Abstract

Floods are among natural disasters that increasingly threaten society, especially with current and future climate change trends. Several tools have been developed to help planners manage the risks associated to flooding, including the mapping of flood-prone areas, but one of the major challenges is still the availability of detailed data, particularly bathymetry. This manuscript compares two modeling approaches to produce flood maps. An innovative large-scale approach that, without bathymetric data, estimates by inverse modeling the bed section for a given flow and a given roughness coefficient through 1 D/2D hydraulic modeling (LISFLOOD-FP). And a local approach, with a detailed coupled 1 D/2D hydraulic model (HEC-RAS) that uses all available information at the bed and floodplain (LiDAR and bathymetry). Both implementations revealed good performance values for flood peak levels as well as excellent fit indices in describing the areal extent of flooding. As expected, the local approach is more accurate, but the results of the large-scale approach are very promising especially for areas lacking bathymetric data and for large-scale governmental programs.

摘要

洪水是日益威胁社会的自然灾害之一，尤其是在当前和未来的气候变化趋势下。已经开发了多种工具来帮助规划者管理与洪水相关的风险，包括绘制洪水易发地区的地图，但主要挑战之一仍然是详细数据的可用性，尤其是水深测量。这份手稿比较了两种生成洪水地图的建模方法。一种创新的大规模方法，在没有测深数据的情况下，通过一维/二维水力建模 (LISFLOOD-FP) 对给定流量和给定粗糙度系数的床截面进行逆向建模。还有一种本地方法，具有详细的耦合一维/二维水力模型 (HEC-RAS)，该模型使用河床和洪泛区的所有可用信息（激光雷达和水深测量）。两种实施都显示了洪水峰值水平的良好性能值以及描述洪水区域范围的出色拟合指数。正如预期的那样，局部方法更准确，但大规模方法的结果非常有希望，特别是对于缺乏测深数据的地区和大规模的政府项目。