Effects of building-height heterogeneity on area-averaged transfer velocity in the street surface—Wind tunnel experiments using salinity change technique

As a parameterization of turbulent fluxes in urban canopy layer, the resistance network in the canyon has been studied with our originally devised water evaporation technique using filter paper. This method is useful to examine the effect of canyon geometry on transfer velocity in two-dimensional models and also to compare the local difference of transfer velocity between the kinds of surfaces in homogeneous cubic array. About heterogeneous building arrangements, however, it is difficult to investigate the spatial-averaged transfer velocity using this technique because of a problem of representativeness. In this study, an experimental method to estimate the area-averaged transfer velocity was newly devised and tried to clarify the effect of heterogeneity of building height in wind tunnel experiments.

In the new method, the water evaporation rate is measured not by weight loss but by salinity change. A square vessel (60cm*60cm) with 5cm depth was buried in wind tunnel working section as the rim-top matched to surrounding surface. The building models are arranged within it, then fill the vessel with saltwater to bathe the foot of them. After half and hour, stop wind tunnel fan, and remove the building models from the vessel. Then, stir the salt-water in the vessel sufficiently, and take it into four sample bottles. The salinity of these samples were measured with salinometer (accuracy:0.003 per mill) as well as samples of first poured saltwater. From the salinity-change, evaporation rate was calculated. Though this method is, of course, available for only about street surface not including wall and roof and is inferior to filter paper method in the accuracy, it is excellent for area-averaged investigations especially for the heterogeneous morphology.

As a result, it was confirmed that the vertical heterogeneity of buildings has much greater effects on area-averaged transfer velocity than that of horizontal variation under building-height constant.