Geometric Dependence of the Scalar Transfer Efficiency over Rough Surfaces

We performed a series of wind-tunnel experiments under neutral conditions in order to create a comprehensive database of scalar transfer coefficients for street surfaces using regular block arrays representing an urban environment. The objective is to clarify the geometric dependence of scalar transfer phenomena on rough surfaces. In addition, the datasets we have obtained are necessary to improve the modelling of scalar transfer used for computational simulations of urban environments; further, we can validate the results obtained by numerical simulations. We estimated the scalar transfer coefficients using the salinity method. The various configurations of the block arrays were designed to be similar to those used in a previous experiment to determine the total drag force acting on arrays. Our results are summarized as follows: first, the results for cubical arrays showed that the transfer coefficients for staggered and square layouts varied with the roughness packing density. The results for the staggered layout showed the possibility that the mixing effect of air can be enhanced for the mid-range values of the packing density. Secondly, the transfer coefficients for arrays with blocks of non-uniform heights were smaller than those for arrays with blocks of uniform height under conditions of low packing density; however, as the packing density increased, the opposite tendency was observed. Thirdly, the randomness of rotation angles of the blocks in the array led to increasing values of the transfer coefficients under sparse packing density conditions when compared with those for cubical arrays.