2D Urban Flood Modelling – The Better Way?

Fiona Macdonald and James Jorgensen

ABSTRACT

Recent advances in hydraulic modelling software and hardware, as well as the increased availability of detailed ground models, have led to an increase in the use of two-dimensional (2D) modeling to assess urban flooding. The purpose of this paper is to examine the potential applications of this modelling technique, present advantages and limitations, and compare it to traditional 1D modelling.

Case studies from recent projects in the United Kingdom and New Zealand will be used to illustrate the concepts discussed.

This paper will focus on the use of 2D modelling to assess urban flooding. The below are some of the potential applications:

  • Rapid Flood Hazard Mapping. A 2D model can be used, in isolation, to quickly provide an idea of potential flooding locations and extent without having to model the underground network or spend time on representing the above- ground network in 1D.
  • Flood Prone Area Analysis. In a similar fashion, a 2D model can be used in isolation to identify areas where flooding may occur should the primary drainage system fail.
  • Detailed Flood Hazard Mapping. A coupled 1D-2D model can be created to accurately understand the capacity of the primary stormwater drainage system, and to model the effect of surcharge and flooding under larger events. These coupled models allow flow to be interchanged between the 1D and 2D models, with the 2D only used where required, to save on computational effort and runtime.
  • Integrated Catchment Planning. In many urban areas, under a large rainfall event the primary stormwater and waste water systems, secondary drainage, and watercourses effectively operate as a single co-joined system. For example piped networks may surcharge and overflow from manhole lids to produce overland flow and flooding, while com bined sewer overflows may interact with watercourses through their outlets, with flows potentially in either direction. While historically these systems have been modelled independently, recent software developments allow the integra tion of 1D, 2D and open channel models in a single model. These allow 1D models to be used in areas with pipe and uni-directional channel flows, 2D models to be used where required e.g. floodplains, and for flow to be exchanged dy -namically between the models. This means that the whole network can be looked at in an integrated fashion, and the effect of changes to one system on the others determined.

Permanent link: