A coarse-grid approach to representing building blockage effects in 2D urban flood modelling

Abstract

The latest information and communications technology has enabled flood modelling in urban areas using high quality terrain data to simulate the detailed flow dynamics in local areas. However, the computational cost rises exponentially as the resolution goes finer. The advance of computing hardware is still a limiting factor for large-scale area or risk/uncertainty analysis modelling with fine resolution that describes the details of building features. Grid coarsening is the straightforward way to reduce the computing efforts for 2D flood modelling. The traditional approach to grid coarsening usually takes the average elevation of a fine grid as the new terrain model for the coarse grid. This approach often results in loss of information that introduces errors to modelling. In this study, the building features in coarse grids were abstracted using the building coverage ratio (BCR) and the conveyance reduction factor (CRF) parameters in a 2D model to simulate flooding in urban areas. The outcome of 2D case studies showed the proposed model can minimise the errors due to terrain averaging and provide a much better accuracy of modelling results at a marginally increased computing cost.

Point Cloud Data Fusion for Enhancing 2D Urban Flood Modelling

“…For some less complex areas, quasi 2D approaches could be capable of handling such situations. However, a fully two-dimensional (2D) numerical models are more appropriate for simulating flow dynamics in the more complex flood situations (Neal et al, 2009;Hai et al, 2010;Chen et al, 2012a;Chen et al, 2012b;Rychkov et al, 2012;Smith et al, 2012). Moreover, coupled 1D-2D models (Section 2.5) can be further applied to achieve benefits from both 1D and 2D models (Hsu et al, 2000b;Chen et al, 2006).…”

“…When a simple 1D model is coupled with a fully 2D model ( Fig. 2-23), their coupled 1D-2D model simulation results are typically capable of replicating exceedance floodwater flows between 1D-channels and 2D floodplains (Neal et al, 2009;Hai et al, 2010;Chen et al, 2012a;Chen et al, 2012b;Rychkov et al, 2012;Seyoum et al, 2012;Smith et al, 2012). When these coupled 1D-2D models have emerged, they are become more widely used to simulate urban floods, which commonly found in pluvial and fluvial flood simulations, especially for complex cities (Vojinović & Tutulic, 2009).…”

“…Efficiency and accuracy are two major indicators for evaluating 2D urban-flood model performances. Typically, efficiency and accuracy are commonly in conflict with one to another (Chen et al, 2012a). While accuracies of adopting finer scales of spatial and temporal resolutions and enhancing 2D model schematics with more key components of local urban features should improve higher accuracy for 2D model simulation, efficiencies of reducing resolutions in space and time scales should also be taken into account for simplifying complexities, speeding up processes, and minimising computational cost.…”

“…These two major indicators, i.e. accuracy (i to iii) and efficiency (iv to vii) are commonly in conflict with one to another (Chen et al, 2012a). Almost all chapters mentioned earlier are mainly focused on improving accuracy of 2D urban flood model by multi-source view (MSV) method for enhancing 2D model schematics, other accuracy improvements should be found elsewhere.…”

“…However, averaging elevations could also introduce loss of information to the final 2D model schematics. In some recent research, new grid adjustment methods could minimise such losses and improve accuracy of simulation results with reducing computational cost (Chen et al, 2012a). Either traditional or new grid adjustment method still need a proper quality of topographic data for representing key component of urban features.…”