Abstract
The mitigation of societal damage from urban floods requires fast hydraulic models for emergency and planning purposes. The simplified mathematical model Cellular Automata is combined with Motion Cost fields, which score the difficulty to traverse an area, to the urban inundation model CAMC. It is implemented with simple matrix and logic operations to achieve high computational efficiency. The development concentrated on an application in dense urban built-up areas with numerous buildings. CAMC is efficient and flexible enough to be used in a “live” urban flood warning system with current weather conditions. A case study is conducted in the German city of Wuppertal with about 12,000 buildings. The water depth estimation of every time step are visualized in a web-interface on the basis of the virtual globe NASA WorldWind. CAMC is compared with the shallow water equations-based model ANUGA. CAMC is approximatively 5 times faster than ANUGA at high spatial resolution and able to maintain numerical stability. The Nash-Sutcliffe coefficient (0.61), Root Mean Square Error (0.39 m) and Index of Agreement (0.65) indicate acceptable agreement for water depth estimation but identify different areas where important deviations occur. The estimation of velocity performs considerably less well (0.34 for Nash-Sutcliffe coefficient, 0.13 ms − 1 for Root Mean Square Error, and 0.39 for Index of Agreement) because CA ignores momentum conservation.
Highlights
Climate change and urbanisation are driving forces in the increasing occurrence and severity of urban floods [1,2,3,4]
This study describes the development of a cellular automata (CA) model that relies on simple matrix and logic operations, while avoiding the relatively computationally expansive Manning’s equation [33,35] and iterative sorting algorithms [34] as an integral component of its CA formulation
It shows that the majority of the area, especially the elevated cells, were very quickly dried because the limited amount of runoff was drained into the depression
Summary
Climate change and urbanisation are driving forces in the increasing occurrence and severity of urban floods [1,2,3,4]. A famous example of the devastation which an urban flood can cause is the flash flood of 21 July 2012 in Beijing [5,6]. The total value of property damage was approximately 10 billion RMB (1.6 billion USD). Such tragedies underline the importance of continuing efforts in the development of urban inundation models [7]. Urban floods can be mitigated by smoothing peaks and slowing flows with appropriate stormwater management. Fast and accurate inundation models would benefit the corresponding planning processes by facilitating the application of spatial optimization [10,11]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
