全球变化知识资源中心

Urban flooding is increasingly pervasive, with dreadful impacts on people and development assets. Whilst the frequency of occurrence of this hazard, mainly from pluvial events, is of fundamental importance in climate change and earth sciences research, the severity of its impacts motivates major debates within the context of flood risk management and sustainable urban development. The present study focuses on the development of a novel flood model, as a contribution to meeting the challenges of flood risk assessment within data poor urban areas such as in Nigeria. The new model combines the full functionality of cellular automata (CA) framework with a semi-implicit finite difference numerical scheme (SIFDS), whilst the resulting algorithms were programmed within MATLAB (TM) programming platform. In this study, computation complexity and distributed topographic data requirement, both which are associated with flood modelling, and which tend to present a major limitation to flood modelling in the developing countries (DCs) are being addressed. A highly urbanized area within the Lagos metropolis of Nigeria was chosen as a case study to validate the model and to simulate the July 10th 2011 flooding event. A 2-m horizontal resolution LiDAR DEM, published Manning's friction coefficients and rainfall intensity, were used as data inputs into the new flood model. Simulated results compared well with actual flooding inundations, reported by urban residents, and detailed in some literature and by the media. The Pearson correlation coefficient (r) between predicted flood depth and estimated values is 0.968. It is expected that the challenges of urban flooding in Lagos particularly and in the DCs generally will be better addressed if robust, but low-cost flood models are developed and utilized in the assessment of flood damage.