ABSTRACT
Urban flooding at the block- or district-level is characterized by comparatively larger horizontal length scales than vertical ones. Therefore, a number of laboratory experiments of urban flooding were based on geometrically distorted scale models. Here, we investigate the influence of model geometric distortion on the flow processes in two simple settings representative of urban flooding: a single prismatic street and a street junction. We specifically compare the upstream flow depths predicted from laboratory observations using different geometric distortions. The predicted upstream flow depths consistently decrease as the model geometric distortion is increased. The bias induced by model geometric distortion, estimated as the relative difference between predictions from distorted and undistorted models is of order of 15% and 7% for the prismatic street and the street junction, respectively. This difference may be attributed to a difference in the relative importance of frictional losses and local head losses in the two settings. The outcomes of the study provide valuable guidance for the design of laboratory models of urban flooding.
