Optimising Flood Risk Management Interventions in Catchments and Cities

详细的洪水模型用于评估城市和集水区未来增加的洪水风险，并设计洪水衰减特征，包括绿色基础设施（GI）或自然洪水管理（NFM）。重要的是，明智地使用有限的洪水管理资金，最佳地定位GI或NFM功能，从而最大限度地减少流量和雨水下水道泄漏。该项目的目的是开发一个自动化框架，在未来气候情景下，利用一系列NFM或GI在集水区和城市中进行洪水风险管理方案的优化设计。该框架将提供优化NFM特征的位置、类型、大小和数量的能力，通过机器学习优化算法实现，以最大限度地降低特定投资在一系列气候情景下对生命和财产的风险。将对不同NFM特征“种群”的影响进行多次模拟，由优化算法控制，并进行迭代，直到获得最优解。这一过程为了解不同洪水情景下的集水区动态提供了一个新的层次，从而为洪水主管部门提供了一个可靠的设计解决方案。

Detailed flood models are used to assess increased future flood risk in cities and catchments, and to design flood attenuation features including green infrastructure (GI) or natural flood management (NFM). It is important that the limited funding available for flood management is spent wisely, locating GI or NFM features optimally so they maximally reduce flows and storm sewer spills. The aim of this project is to develop an automated framework for the optimal design of flood risk management options using a range of NFM or GI in catchments and cities under future climate scenarios. The framework will provide the capability to optimise the location, type, size and number of NFM features, achieved by a machine learning optimisation algorithm to minimise risk to life and property for a given investment for a range of climate scenarios. Multiple simulations of the impact of different "populations" of NFM features will be made, controlled by an optimisation algorithm, and iterated until an optimal solution is obtained. This process provides a new level of understanding of catchment dynamics under different flood scenarios, resulting in a robust design solution for use by the lead flood authority.