Search

Spectral proper orthogonal decomposition (SPOD) is applied as a post-processing technique to elucidate the relationship between turbulent motion and pollutant removal in a two-dimensional street canyon with an aspect ratio of one and a uniform roof height. A pollutant is continuously emitted from a line source set on the ground of the target canyon. First, the SPOD technique is used to decompose the velocity field obtained from a large-eddy simulation. The external large-scale coherent structures and waves caused by the Kelvin–Helmholtz instabilities are extracted and visualized using the SPOD modes. The SPOD cospectra are defined to further examine the phase relationship between the streamwise and vertical velocity components and concentrations. Based on the SPOD cospectra, the contribution of the turbulent structures at various scales to pollutant removal are quantitatively estimated. The results reveal that both the large-scale coherent structures and Kelvin–Helmholtz instabilities could cause ejection events at the canyon roof level and thus contribute to pollutant removal. Although the former occupy a large proportion of the turbulence kinetic energy, a smaller vertical turbulent mass flux is also observed. Conversely, the latter contribute to stronger ejection and sweep events with stronger vertical components. However, the shapes of the modes indicate that the external large-scale coherent structures also play a role in triggering and transporting the small-scale turbulence at the roof level, which may indirectly contribute to pollutant removal.

Coherent flow structures and pollutant dispersion in a spanwise-long street canyon are investigated using a parallelized large-eddy-simulation model. Low- and high-concentration branches, starting from the downwind top corner and upwind bottom corner, respectively, are detected in the time-averaged field of pollutant concentration, and detailed structures of in-canyon flow and pollutant dispersion following the two branches are demonstrated. When turbulent eddies impinge on the upper downwind wall, low- and high-concentration blobs with U-shaped flow structures appear and move downward. The downdrafts tilt away from the downwind bottom corner and impinge on the canyon bottom, driving horizontally diverging flows. Cellular structures of low-concentration centres and high-concentration edges are induced by the downdrafts and diverging flows. The diverging flows push low-concentration air toward the downwind and upwind building walls, resulting in local divergence and convergence of pollutants on both walls. Time series of pollutant concentration at multiple points illustrate that pollutant concentration at the pedestrian level is highly sensitive to the diverging flows. The multiresolution spectra show that time scales of variations of pollutant concentration and vertical velocity component increase from the canyon top to the pedestrian-level centre, indicating longer time-scale flow structures are dominant inside the street canyon. The multiresolution cospectra also show that the time scale of vertical turbulent transport of pollutants increases from the canyon top to the pedestrian-level centre. At the two bottom corners, however, short and long time-scale transports occur together, confirming that the low-concentration diverging flows transport pollutants downward while short time-scale turbulence transports pollutants upward.

Techniques for improving the removal of pollution from urban canyons are crucial for air quality control in cities. The removal mainly occurs at the building roof level, where it is supported by turbulent mixing and hampered by roof shear, which tends to isolate the internal canyon region from the atmospheric flow. Here, a modification of roof infrastructures is proposed with the aim of increasing the former and reducing the latter, overall enhancing the removal mechanisms. The topic is investigated by numerical experiment, using large-eddy simulation to study the paradigmatic case of a periodic square urban canyon at $$ Re=2 \times 10^4$$ R e = 2 × 10 4 . Two geometries are analyzed: one with a smooth building roof, the other having a series of solid obstacles atop the upwind building roof. The pollutant is released at the street level. The simulations are successfully validated against laboratory and numerical datasets, and the primary vortex displacement detected in some laboratory experiments is discussed. The turbulence triggered by the obstacles destroys the sharp shear layer that separates the canyon and the surrounding flow, increasing the mixing. Greater vertical turbulent mass fluxes and more frequent ejection events near the upwind building (where pollution accumulates) are detected. Overall, the obstacles lead to a reduction in the pollution concentration within the canyon of about $$34\%$$ 34 % .

An experimental field campaign is designed to unveil mechanisms responsible for turbulent exchange processes when mechanical and thermal effects are entwined. The focus is an urban street canyon with a mean aspect ratio H/W of 1.65 in the business centre of a mid-size Italian city (H is the mean building height and W is the mean canyon width). The exchange processes can be characterized by time scales and time-scale ratios specific to either mechanical or thermal process. Time scales describe the mixing caused by momentum and heat exchange within different canyon layers, while their rates are surrogates of their efficacy. Given that homogeneous mixing does not always occur within the canyon, several time scales are estimated at different levels, showing that mechanical and thermal processes may both contribute to enhance mixing. By computing mechanical time scales, it is found that the fastest mixing occurs at the canyon rooftop level for perpendicular or oblique wind directions, while slow mixing occurs for parallel directions. Thermal processes are faster than the mechanical ones and are particularly efficient for perpendicular wind directions. By calculating the time-scale ratios, exchange processes are found to facilitate mixing for most wind directions and to regulate the pollutant-concentration variability in the canyon. This variability can be associated with the local-circulation regime, demarcated as thermally driven or inertially driven using a buoyancy parameter, i.e., the ratio between thermal and inertial forcings. Using this approach, a generalization of the results is proposed, enabling the extension of the current investigation to different street-canyon aspect ratios.

We investigate the flow characteristics around step-up street canyons with various building aspect ratios (ratio of along-canyon building length to street-canyon width, and upwind building height to downwind building height) using a computational fluid dynamics (CFD) model. Simulated results are validated against experimental wind-tunnel results, with the CFD simulations conducted under the same building configurations as those in the wind-tunnel experiments. The CFD model reproduces the measured in-canyon vortex, rooftop recirculation zone above the downwind building, and stagnation point position reasonably well. We analyze the flow characteristics, focusing on the structural change of the in-canyon flows and the interaction between the in- and around-canyon flows with the increase of building-length ratio. The in-canyon flows undergo development and mature stages as the building-length ratio increases. In the development stage (i.e., small building-length ratios), the position of the primary vortex wanders, and the incoming flow closely follows both the upstream and downstream building sidewalls. As a result, increasing momentum transfer from the upper layer contributes to a momentum increase in the in-canyon region, and the vorticity in the in-canyon region also increases. In the mature stage (i.e., large building-length ratios), the primary vortex stabilizes in position, and the incoming flow no longer follows the building sidewalls. This causes momentum loss through the street-canyon lateral boundaries. As the building-length ratio increases, momentum transfer from the upper layer slightly decreases, and the reverse flow, updraft, and streamwise flow in the in-canyon region also slightly decrease, resulting in vorticity reduction.

The effect of upstream roughness and canyon width on turbulent street-canyon flow is investigated, using wind-tunnel measurements made in a horizontal plane at near roof level of a street canyon and stereoscopic particle image velocimetry. Three upstream roughness arrays and two canyon width (W) to height (h) aspect ratios (AR = W/h = 1 and 3) are used; the arrays consist of three-dimensional cubes (plan area density, λp = 25%), 1h-spaced two-dimensional bars (skimming flow, λp = 50%) and 3h-spaced two-dimensional bars (wake-interference flow, λp = 25%). Understanding the spanwise structure of the flow and how it interacts with large-scale structures is necessary to reliably predict the mean pollutant transport in the lateral direction along the canyon and to further investigate the three-dimensional behaviour of turbulent street-canyon flows. The mean turbulent statistics are presented, whilst two-point correlations and integral length scales are computed for the different configurations. The results show a significant effect of upstream roughness on these quantities. The total turbulent kinetic energy and shear stress are found to be highest for the wake-interference flow regimes and lowest for the skimming-flow regimes. It is found that the three-dimensional upstream roughness configurations result in a significantly weaker correlation in the spanwise direction at canyon roof level, with a similar trend observed in the spanwise integral length scales. The shear-layer thickness is found to be related to the magnitude of the correlations near roof level of the street canyon.

Built terrains, with their complexity in morphology, high heterogeneity, and anthropogenic impact, impose substantial challenges in Earth-system modelling. In particular, estimation of the source areas and footprints of atmospheric measurements in cities requires realistic representation of the landscape characteristics and flow physics in urban areas, but has hitherto been heavily reliant on large-eddy simulations. In this study, we developed physical parametrization schemes for estimating urban footprints based on the backward-Lagrangian-stochastic algorithm, with the built environment represented by street canyons. The vertical profile of mean streamwise velocity is parametrized for the urban canopy and boundary layer. Flux footprints estimated by the proposed model show reasonable agreement with analytical predictions over flat surfaces without roughness elements, and with experimental observations over sparse plant canopies. Furthermore, comparisons of canyon flow and turbulence profiles and the subsequent footprints were made between the proposed model and large-eddy simulation data. The results suggest that the parametrized canyon wind and turbulence statistics, based on the simple similarity theory used, need to be further improved to yield more realistic urban footprint modelling.

We experimentally investigate the effect of a typical building covering: the gable roof, on the flow and air exchange in urban canyons. In general, the morphology of the urban canopy is very varied and complex, depending on a large number of factors, such as building arrangement, or the morphology of the terrain. Therefore we focus on a simple, prototypal shape, the two-dimensional canyon, with the aim of elucidating some fundamental phenomena driving the street-canyon ventilation. Experiments are performed in a water channel, over an array of identical prismatic obstacles representing an idealized urban canopy. The aspect ratio, i.e. canyon-width to building-height ratio, ranges from 1 to 6. Gable roof buildings with 1:1 pitch are compared with flat roofed buildings. Velocity is measured using a particle-image-velocimetry technique with flow dynamics discussed in terms of mean flow and second- and third-order statistical moments of the velocity. The ventilation is interpreted by means of a simple well-mixed box model and the outflow rate and mean residence time are computed. Results show that gable roofs tend to delay the transition from the skimming-flow to the wake-interference regime and promote the development of a deeper and more turbulent roughness layer. The presence of a gable roof significantly increases the momentum flux, especially for high packing density. The air exchange is improved compared to the flat roof buildings, and the beneficial effect is more significant for narrow canyons. Accordingly, for unit aspect ratio gable roofs reduce the mean residence time by a factor of 0.37 compared to flat roofs, whereas the decrease is only by a factor of 0.9 at the largest aspect ratio. Data analysis indicates that, for flat roof buildings, the mean residence time increases by 30% when the aspect ratio is decreased from 6 to 2, whereas this parameter is only weakly dependent on aspect ratio in the case of gable roofs.

As urbanization progresses, more realistic methods are required to analyze the urban microclimate. However, given the complexity and computational cost of numerical models, the effects of realistic representations should be evaluated to identify the level of detail required for an accurate analysis. We consider the realistic representation of surface heating in an idealized three-dimensional urban configuration, and evaluate the spatial variability of flow statistics (mean flow and turbulent fluxes) in urban streets. Large-eddy simulations coupled with an urban energy balance model are employed, and the heating distribution of urban surfaces is parametrized using sets of horizontal and vertical Richardson numbers, characterizing thermal stratification and heating orientation with respect to the wind direction. For all studied conditions, the thermal field is strongly affected by the orientation of heating with respect to the airflow. The modification of airflow by the horizontal heating is also pronounced for strongly unstable conditions. The formation of the canyon vortices is affected by the three-dimensional heating distribution in both spanwise and streamwise street canyons, such that the secondary vortex is seen adjacent to the windward wall. For the dispersion field, however, the overall heating of urban surfaces, and more importantly, the vertical temperature gradient, dominate the distribution of concentration and the removal of pollutants from the building canyon. Accordingly, the spatial variability of concentration is not significantly affected by the detailed heating distribution. The analysis is extended to assess the effects of three-dimensional surface heating on turbulent transfer. Quadrant analysis reveals that the differential heating also affects the dominance of ejection and sweep events and the efficiency of turbulent transfer (exuberance) within the street canyon and at the roof level, while the vertical variation of these parameters is less dependent on the detailed heating of urban facets.

The influence of roof-edge roughness elements on airflow, heat transfer, and street-level pollutant transport inside and above a two-dimensional urban canyon is analyzed using an urban energy balance model coupled to a large-eddy simulation model. Simulations are performed for cold (early morning) and hot (mid afternoon) periods during the hottest month of the year (August) for the climate of Abu Dhabi, United Arab Emirates. The analysis suggests that early in the morning, and when the tallest roughness elements are implemented, the temperature above the street level increases on average by 0.5 K, while the pollutant concentration decreases by 2% of the street-level concentration. For the same conditions in mid afternoon, the temperature decreases conservatively by 1 K, while the pollutant concentration increases by 7% of the street-level concentration. As a passive or active architectural solution, the roof roughness element shows promise for improving thermal comfort and air quality in the canyon for specific times, but this should be further verified experimentally. The results also warrant a closer look at the effects of mid-range roughness elements in the urban morphology on atmospheric dynamics so as to improve parametrizations in mesoscale modelling.

We demonstrate how application of the stochastic estimation method can be employed to combine spatially well-resolved wind-tunnel particle image velocimetry measurements with instantaneous velocity signals from a limited number of sensors (six sonic anemometers located within the canyon in the present case) to predict full-scale flow dynamics in an entire street-canyon cross-section. The investigated configuration corresponds to a street-canyon flow in a neutrally stratified atmospheric boundary layer with the oncoming flow being perpendicular to the main canyon axis. Data were obtained during both full-scale and 1:200-scale wind-tunnel experiments. The performance of the proposed method is investigated using both wind-tunnel data and signals from five sonic anemometers to predict the velocity from the sixth one. In particular, based on analysis of the influence of the high-frequency velocity fluctuations on the quality of the reconstruction, it is shown that stochastic estimation is able to correctly reproduce the large-scale temporal features of the flow with the present set-up. The full dataset is then used to spatially extrapolate the instantaneous flow measured by the six sonic anemometers and perform detailed analysis of instantaneous flow features. The main features of the flow, such as the presence of the shear layer that develops over the canyon and the intermittent ejection and penetration events across the canyon opening, are well predicted by stochastic estimation. In addition, thanks to the high spatial resolution made possible by the technique, the intermittency of the main vortical structure existing within the canyon is demonstrated, as well as its meandering motion in the canyon cross-section. It is also shown that the canyon flow, particularly its spanwise component, is affected by large-scale fluctuations of low temporal frequency along the canyon axis. Finally, the proposed techniques based on wind-tunnel data can prove useful for a priori design of field experiments to determine the optimum location of sensors beforehand.

Large-eddy simulation is conducted to investigate the effect of fetch on the pollutant-removal mechanism from a two-dimensional street canyon with a building-height to street-width (aspect) ratio of 1. The line sources were placed within the first, second, third, fifth, seventh and tenth canyons, and the six tracer gases are simultaneously released by a ground-level continuous pollutant line source placed parallel to the spanwise axis at the canyons. The mean concentration and the deviation of the concentration fluctuation within the canyon roughly reach a near-constant value downwind of the seventh canyon, which is similar to the behaviour of the turbulent intensities. In the first canyon, pollutant removal is affected by both advective flow and turbulent flow; however, the turbulent motions mainly affect pollutant removal from the top of the canyon as the fetch increases. In the first and third canyons, the low-momentum fluid does not always affect pollutant removal, but does so gradually as the fetch increases.

The ventilation processes in three different street canyons of variable roof geometry were investigated in a wind tunnel using a ground-level line source. All three street canyons were part of an urban-type array formed by courtyard-type buildings with pitched roofs. A constant roof height was used in the first case, while a variable roof height along the leeward or windward walls was simulated in the two other cases. All street-canyon models were exposed to a neutrally stratified flow with two approaching wind directions, perpendicular and oblique. The complexity of the flow and dispersion within the canyons of variable roof height was demonstrated for both wind directions. The relative pollutant removals and spatially-averaged concentrations within the canyons revealed that the model with constant roof height has higher re-emissions than models with variable roof heights. The nomenclature for the ventilation processes according to quadrant analysis of the pollutant flux was introduced. The venting of polluted air (positive fluctuations of both concentration and velocity) from the canyon increased when the wind direction changed from perpendicular to oblique, irrespective of the studied canyon model. Strong correlations ( $$>$$ 0.5) between coherent structures and ventilation processes were found at roof level, irrespective of the canyon model and wind direction. This supports the idea that sweep and ejection events of momentum bring clean air in and detrain the polluted air from the street canyon, respectively.

A single-layer urban canopy model that captures sub-facet heterogeneity and various hydrological processes is further developed to explicitly incorporate trees within the urban canyon. The physical processes associated with trees are shortwave/longwave radiation exchange, including mutual interception and shading by trees and buildings and multiple reflections, sensible heat and latent heat (through transpiration) exchange, and root water uptake. A computationally-efficient geometric approach is applied to the radiation exchanges, requiring a priori knowledge of view factors. These view factors are first obtained from independent Monte Carlo ray-tracing simulations, and subsequently simple relations, which are functions of canyon aspect ratio and tree-crown ratio, are proposed to estimate them. The developed model is evaluated against field observations at two urban sites and one suburban site, showing improved performance for latent heat flux compared to the previous version that only includes ground vegetation. The trees in the urban canopy act to considerably decrease sensible heat flux and increase latent heat flux, and these effects are found to be more significant in the more dense urban site. Sensitivity tests are then performed to examine the effects of tree geometry relative to canyon geometry. The results indicate that the tree-crown size relative to canyon width is the most influential parameter to decrease sensible heat flux and increase latent heat flux, resulting in cooling of the urban area.

Urban flow and turbulence are driven by atmospheric flows with larger horizontal scales. Since building-resolving computational fluid dynamics models typically employ steady Dirichlet boundary conditions or forcing, the accuracy of numerical simulations may be limited by the neglect of perturbations. We investigate the sensitivity of flow within a unit-aspect-ratio street canyon to time-dependent perturbations near the inflow boundary. Using large-eddy simulation, time-periodic perturbations to the streamwise velocity component are incorporated via the nudging technique. Spatial averages of pointwise differences between unperturbed and perturbed velocity fields (i.e., the error kinetic energy) show a clear dependence on the perturbation period, though spatial structures are largely insensitive to the time-dependent forcing. The response of the error kinetic energy is maximized for perturbation periods comparable to the time scale of the mean canyon circulation. Frequency spectra indicate that this behaviour arises from a resonance between the inflow forcing and the mean motion around closed streamlines. The robustness of the results is confirmed using perturbations derived from measurements of roof-level wind speed.

Although predictability is a subject of great importance in atmospheric modelling, there has been little research on urban boundary-layer flows. Here the predictability of street-canyon flow is examined numerically via large-eddy simulation of a unit-aspect-ratio canyon and neutrally stratified atmosphere. In spectral space there is indication of cascade-like behaviour away from the canyon at early times, but the error growth is essentially independent of scale inside the canyon; in physical space the error field is rather inhomogeneous and shows clear differences among the canyon, shear layer and inertial sublayer. The error growth is largely driven by the shear layer: errors generated above roof level are advected into the canyon while contributions from intermittent bursting and in situ development within the canyon play a relatively minor role. This work highlights differences between the predictability of urban flows and canonical turbulent flows and should be useful in developing modelling strategies for more realistic time-dependent urban flows.

The surface albedo of two large cities in Japan was measured using a pyranometer mounted on a helicopter to avoid the bidirectional reflectance distribution. The daytime albedo was 0.12 in the cities, which was less than that of a nearby forest (0.16). The albedo was dependent on building structure in the cities; the albedo was lower in areas with more buildings, and decreased as the aspect ratio of street canyons increased. There are two reasons for this dependency: the multiple reflection of radiation in the building canopy, as has been shown in many previous studies, and the sparse vegetation in urban areas. These two factors concurrently determine the albedo in a real city, where the vegetation amount decreases as the plan roof ratio increases.

The impact of diurnal variations of the heat fluxes from building and ground surfaces on the fluid flow and air temperature distribution in street canyons is numerically investigated using the PArallelized Large-eddy Simulation Model (PALM). Simulations are performed for a 3 by 5 array of buildings with canyon aspect ratio of one for two clear summer days that differ in atmospheric instability. A detailed building energy model with a three-dimensional raster-type geometry—Temperature of Urban Facets Indoor-Outdoor Building Energy Simulator (TUF-IOBES)—provides urban surface heat fluxes as thermal boundary conditions for PALM. In vertical cross-sections at the centre of the spanwise canyon the mechanical forcing and the horizontal streamwise thermal forcing at roof level outweigh the thermal forces from the heated surfaces inside the canyon in defining the general flow pattern throughout the day. This results in a dominant canyon vortex with a persistent speed, centered at a constant height. Compared to neutral simulations, non-uniform heating of the urban canyon surfaces significantly modifies the pressure field and turbulence statistics in street canyons. Strong horizontal pressure gradients were detected in streamwise and spanwise canyons throughout the day, and which motivate larger turbulent velocity fluctuations in the horizontal directions rather than in the vertical direction. Canyon-averaged turbulent kinetic energy in all non-neutral simulations exhibits a diurnal cycle following the insolation on the ground in both spanwise and streamwise canyons, and it is larger when the canopy bottom surface is paved with darker materials and the ground surface temperature is higher as a result. Compared to uniformly distributed thermal forcing on urban surfaces, the present analysis shows that realistic non-uniform thermal forcing can result in complex local airflow patterns, as evident, for example, from the location of the vortices in horizontal planes in the spanwise canyon. This study shows the importance of three-dimensional simulations with detailed thermal boundary conditions to explore the heat and mass transport in an urban area.

Large-eddy simulations were conducted to investigate the mechanism of pollutant removal from a three-dimensional street canyon. Five block configurations with aspect ratios (building height to length) of 1, 2, 4, 8 and \(\infty \) were used to create an urban-like array. A pollutant was released from a ground-level line source at the centre of the target canyon floor. For smaller aspect ratios, the relative contribution of the turbulent mass flux to net mass flux at the roof level, which was spatially averaged along the roof-level ventilation area, was closer to unity, indicating that turbulent motions mainly affected pollutant removal from the top of the canyon. As aspect ratio increased, the relative contribution became smaller, owing to strong upwind motions. However, the relative contribution again reached near unity for the infinite aspect ratio (i.e. a two-dimensional street canyon) because of lowered lateral flow convergence. At least 75 % of total emissions from the three-dimensional street canyon were attributable to turbulent motions. Pollutant removal by turbulent motions was related to the coherent structures of low-momentum fluid above the canyons. Though the coherent structure size of the low-momentum fluid differed, the positions of low-momentum fluid largely corresponded to instantaneous high concentrations of pollutant above the target canyon, irrespective of canyon geometry.

Flow resistance, ventilation, and pollutant removal for idealized two-dimensional (2D) street canyons of different building-height to street-width (aspect) ratios $$AR$$ are examined using the friction factor $$f$$ , air exchange rate (ACH), and pollutant exchange rate (PCH), respectively, calculated by large-eddy simulation (LES). The flows are basically classified into three characteristic regimes, namely isolated roughness, wake interference, and skimming flow, as functions of the aspect ratios. The LES results are validated by various experimental and numerical datasets available in the literature. The friction factor increases with decreasing aspect ratio and reaches a peak at $$AR = 0.1$$ in the isolated roughness regime and decreases thereafter. As with the friction factor, the ACH increases with decreasing aspect ratio in the wake interference and skimming flow regimes, signifying the improved aged air removal for a wider street canyon. The PCH exhibits a behaviour different from its ACH counterpart in the range of aspect ratios tested. Pollutants are most effectively removed from the street canyon with $$AR = 0.5$$ . However, a minimum of PCH is found nearby at $$AR = 0.3$$ , at which the pollutant removal is sharply weakened. Besides, the ACH and PCH are partitioned into the mean and turbulent components to compare their relative contributions. In line with our earlier Reynolds-averaged Navier–Stokes calculations (Liu et al., Atmos Environ 45:4763–4769, 2011), the current LES shows that the turbulent components contribute more to both ACH and PCH, consistently demonstrating the importance of atmospheric turbulence in the ventilation and pollutant removal for urban areas.

An analysis of the dynamics of the flow over a street canyon immersed in an atmospheric boundary layer is presented, using particle image velocimetry measurements in a wind tunnel. Care was taken to generate a 1:200 model scale urban type boundary layer that is correctly scaled to the size of the canyon buildings. Using proper orthogonal decomposition (POD) of the velocity field and conditional averaging techniques, it is first shown that the flow above the opening of the canyon consists of a shear layer separating from the upstream obstacle, animated by a coherent flapping motion and generating large-scale vortical structures. These structures are alternately injected into the canyon or shed off the obstacle into the outer flow. It is shown that unsteady fluid exchanges between the canyon and the outer flow are mainly driven by the shear layer. Finally, using POD, the non-linear interaction between the large-scale structures of the oncoming atmospheric boundary layer and the flow over the canyon is demonstrated.

Large-eddy simulations are conducted to investigate the effects of the incoming turbulent structure of the flow on pollutant removal from an ideal canyon. The target canyon is a two-dimensional street canyon with an aspect ratio of 1.0 (building height to street width). Three turbulent flows upwind of the street canyon are generated by using different block configurations, and a tracer gas is released as a ground-level line source at the centre of the canyon floor. Mean velocity profiles for the three flows are similar, except near the roof. However, the root-mean-square values of the velocity fluctuations and the Reynolds shear stress increase with the friction velocity of the incoming turbulent flow. The spatially-averaged concentration within the canyon decreases with increasing friction velocity. Coherent structures of low-momentum fluid, generated above the upwind block configurations, contribute to pollutant removal, and the amount of pollutant removal is directly related to the size of the coherent structure.

We develop a double-canyon radiation scheme (DCEP) for urban canopy models embedded in mesoscale numerical models based on the Building Effect Parametrization (BEP). The new scheme calculates the incoming and outgoing longwave and shortwave radiation for roof, wall and ground surfaces for an urban street canyon characterized by its street and building width, canyon length, and the building height distribution. The scheme introduces the radiative interaction of two neighbouring urban canyons allowing the full inclusion of roofs into the radiation exchange both inside the canyon and with the sky. In contrast to BEP, we also treat direct and diffuse shortwave radiation from the sky independently, thus allowing calculation of the effective parameters representing the urban diffuse and direct shortwave radiation budget inside the mesoscale model. Furthermore, we close the energy balance of incoming longwave and diffuse shortwave radiation from the sky, so that the new scheme is physically more consistent than the BEP scheme. Sensitivity tests show that these modifications are important for urban regions with a large variety of building heights. The evaluation against data from the Basel Urban Boundary Layer Experiment indicates a good performance of the DCEP when coupled with the regional weather and climate model COSMO-CLM.

This study summarizes the effects of avenues of trees in urban street canyons on traffic pollutant dispersion. We describe various wind-tunnel experiments with different tree-avenue models in combination with variations in street-canyon aspect ratio W/H (with W the street-canyon width and H the building height) and approaching wind direction. Compared to tree-free street canyons, in general, higher pollutant concentrations are found. Avenues of trees do not suppress canyon vortices, although the air ventilation in canyons is hindered significantly. For a perpendicular wind direction, increases in wall-average and wall-maximum concentrations at the leeward canyon wall and decreases in wall-average concentrations at the windward wall are found. For oblique and perpendicular wind directions, increases at both canyon walls are obtained. The strongest effects of avenues of trees on traffic pollutant dispersion are observed for oblique wind directions for which also the largest concentrations at the canyon walls are found. Thus, the prevailing assumption that attributes the most harmful dispersion conditions to a perpendicular wind direction does not hold for street canyons with avenues of trees. Furthermore, following dimensional analysis, an estimate of the normalized wall-maximum traffic pollutant concentration in street canyons with avenues of trees is derived.

A validated large-eddy simulation model was employed to study the effect of the aspect ratio and ground heating on the flow and pollutant dispersion in urban street canyons. Three ground-heating intensities (neutral, weak and strong) were imposed in street canyons of aspect ratio 1, 2, and 0.5. The detailed patterns of flow, turbulence, temperature and pollutant transport were analyzed and compared. Significant changes of flow and scalar patterns were caused by ground heating in the street canyon of aspect ratio 2 and 0.5, while only the street canyon of aspect ratio 0.5 showed a change in flow regime (from wake interference flow to skimming flow). The street canyon of aspect ratio 1 does not show any significant change in the flow field. Ground heating generated strong mixing of heat and pollutant; the normalized temperature inside street canyons was approximately spatially uniform and somewhat insensitive to the aspect ratio and heating intensity. This study helps elucidate the combined effects of urban geometry and thermal stratification on the urban canyon flow and pollutant dispersion., Singapore National Research Foundation (Singapore-MIT Alliance for Research and Technology (SMART))

We develop an urban canopy scheme coupled to a mesoscale atmospheric numerical model and evaluate the simulated climate of an Australian city. The urban canopy scheme is based on the Town Energy Budget approach, but is modified to efficiently represent the predominately suburban component of Australian cities in regional climate simulations. Energy conservation is improved by adding a simple model of air-conditioning to prevent the urban parametrization acting as an energy sink during the Australian summer. In-canyon vegetation for suburban areas is represented by a big-leaf model, but with a largely reduced set of prognostic variables compared to previous approaches. Although we have used a recirculation/venting based parametrization of in-canyon turbulent heat fluxes that employs two canyon wall energy budgets, we avoid using a fixed canyon orientation by averaging the canyon fluxes after integrating over 180° of possible canyon orientations. The urban canopy scheme is evaluated by simulating the climate for Melbourne, Australia after coupling it to The Air Pollution Model. The combined system was found to predict a realistic climatology of air temperatures and winds when compared with observations from Environmental Protection Authority monitoring stations. The model also produced a plausible partitioning of the urban energy budget when compared to urban flux-tower studies. Overall, the urban canyon parametrization appears to have reasonable potential for studying present and predicting changes in future Australian urban climates in regional climate simulations.

Urban surface and radiation processes are incorporated into a computational fluid dynamics (CFD) model to investigate the diurnal variation of flow in a street canyon with an aspect ratio of 1. The developed CFD model predicts surface and substrate temperatures of the roof, walls, and road. One-day simulations are performed with various ambient wind speeds of 2, 3, 4, 5, and 6 ms−1, with the ambient wind perpendicular to the north–south oriented canyon. During the day, the largest maximum surface temperature for all surfaces is found at the road surface for an ambient wind speed of 3 ms−1 (56.0°C). Two flow regimes are identified by the vortex configuration in the street canyon. Flow regime I is characterized by a primary vortex. Flow regime II is characterized by two counter-rotating vortices, which appears in the presence of strong downwind building-wall heating. Air temperature is relatively low near the downwind building wall in flow regime I and inside the upper vortex in flow regime II. In flow regime II, the upper vortex expands with increasing ambient wind speed, thus enlarging the extent of cool air within the canyon. The canyon wind speed in flow regime II is proportional to the ambient wind speed, but that in flow regime I is not. For weak ambient winds, the dependency of surface sensible heat flux on the ambient wind speed is found to play an essential role in determining the relationship between canyon wind speed and ambient wind speed.

Very few attempts have so far been made to quantify the momentum and turbulent kinetic energy (TKE) budgets within real urban canopies. In this study, sonic anemometer data obtained during the Joint Urban 2003 field campaign in Oklahoma City, U.S.A. were used for calculating the momentum and TKE budgets within a real-world urban street canyon. Sonic anemometers were deployed on multiple towers in the lower half of the canyon. Gradients in all three principal directions were included in the analyses. The storage and buoyancy terms were found to have negligible contributions to both the momentum and TKE budgets. The momentum budgets were generally found to be more complex than a simple balance of two physical processes. The horizontal terms were found to have significant and sometimes dominant contributions to the momentum and TKE budgets.

Large-eddy simulation (LES) is conducted to investigate the mechanism of pollutant removal from a two-dimensional street canyon with a building-height to street-width (aspect) ratio of 1. A pollutant is released as a ground-level line source at the centre of the canyon floor. The mean velocities, turbulent fluctuations, and mean pollutant concentration estimated by LES are in good agreement with those obtained by wind-tunnel experiments. Pollutant removal from the canyon is mainly determined by turbulent motions, except in the adjacent area to the windward wall. The turbulent motions are composed of small vortices and small-scale coherent structures of low-momentum fluid generated close to the plane of the roof. Although both small vortices and small-scale coherent structures affect pollutant removal, the pollutant is largely emitted from the canyon by ejection of low-momentum fluid when the small-scale coherent structures appear just above the canyon where the pollutant is retained. Large-scale coherent structures also develop above the canyon, but they do not always affect pollutant removal.

We present results from fast-response wind measurements within and above a busy intersection between two street canyons (Marylebone Road and Gloucester Place) in Westminster, London taken as part of the DAPPLE (Dispersion of Air Pollution and Penetration into the Local Environment; www.dapple.org.uk ) 2007 field campaign. The data reported here were collected using ultrasonic anemometers on the roof-top of a building adjacent to the intersection and at two heights on a pair of lamp-posts on opposite sides of the intersection. Site characteristics, data analysis and the variation of intersection flow with the above-roof wind direction (θ ref ) are discussed. Evidence of both flow channelling and recirculation was identified within the canyon, only a few metres from the intersection for along-street and across-street roof-top winds respectively. Results also indicate that for oblique roof-top flows, the intersection flow is a complex combination of bifurcated channelled flows, recirculation and corner vortices. Asymmetries in local building geometry around the intersection and small changes in the background wind direction (changes in 15- min mean θ ref of 5°–10°) were also observed to have profound influences on the behaviour of intersection flow patterns. Consequently, short time-scale variability in the background flow direction can lead to highly scattered in-street mean flow angles masking the true multi-modal features of the flow and thus further complicating modelling challenges.

In this experimental work both qualitative (flow visualisation) and quantitative (laser Doppler anemometry) methods were applied in a wind tunnel in order to describe the complex three-dimensional flow field in a real environment (a street canyon intersection). The main aim was an examination of the mean flow, turbulence and flow pathlines characterising a complex three-dimensional urban location. The experiments highlighted the complexity of the observed flows, particularly in the upwind region of the intersection. In this complex and realistic situation some details of the upwind flow, such as the presence of two tall towers, play an important role in defining the flow field within the intersection, particularly at roof level. This effect is likely to have a strong influence on the mass exchange mechanism between the canopy flow and the air aloft, and therefore the distribution of pollutants. This strong interaction between the flows inside and outside the urban canopy is currently neglected in most state-of-the-art local scale dispersion models.

Wind-tunnel studies of dispersion processes of traffic exhaust in urban street canyons with tree planting were performed and tracer gas concentrations using electron capture detection (ECD) and flow fields using laser Doppler velocimetry (LDV) were measured. It was found that tree planting reduces the air exchange between street canyons and the ambience. In comparison to treeless street canyons, higher overall pollutant concentrations and lower flow velocities were measured. In particular, for perpendicular approaching wind, markedly higher concentrations at the leeward canyon wall and slightly lower concentrations at the windward canyon wall were observed. Furthermore, a new approach is suggested to model porous vegetative structures such as tree crowns for small-scale wind-tunnel applications. The approach is based on creating different model tree crown porosities by incorporating a certain amount of wadding material into a specified volume. A significant influence of the crown porosity on pollutant concentrations was found for high degrees of porosity, however, when it falls below a certain threshold, no further changes in pollutant concentrations were observed.

An urban canopy model is developed for use in mesoscale meteorological and environmental modelling. The urban geometry is composed of simple homogeneous buildings characterized by the canyon aspect ratio (h/w) as well as the canyon vegetation characterized by the leaf aspect ratio (σ l ) and leaf area density profile. Five energy exchanging surfaces (roof, wall, road, leaf, soil) are considered in the model, and energy conservation relations are applied to each component. In addition, the temperature and specific humidity of canopy air are predicted without the assumption of thermal equilibrium. For radiative transfer within the canyon, multiple reflections for shortwave radiation and one reflection for longwave radiation are considered, while the shadowing and absorption of radiation due to the canyon vegetation are computed by using the transmissivity and the leaf area density profile function. The model is evaluated using field measurements in Vancouver, British Columbia and Marseille, France. Results show that the model quite well simulates the observations of surface temperatures, canopy air temperature and specific humidity, momentum flux, net radiation, and energy partitioning into turbulent fluxes and storage heat flux. Sensitivity tests show that the canyon vegetation has a large influence not only on surface temperatures but also on the partitioning of sensible and latent heat fluxes. In addition, the surface energy balance can be affected by soil moisture content and leaf area index as well as the fraction of vegetation. These results suggest that a proper parameterization of the canyon vegetation is prerequisite for urban modelling.

The transfer processes within and above a simulated urban street canyon were investigated in a generic manner. Computational fluid dynamics (CFD) was used to aid understanding and to produce some simple operational parameterisations. In this study we addressed specifically the commonly met situation where buoyancy effects arising from elevated surface temperatures are not important, i.e. when mechanical forces outweigh buoyancy forces. In a geophysical context this requires that some suitably defined Richardson number is small. From an engineering perspective this is interpreted as the important case when heat transfer within and above urban street canyons is by forced convection. Surprisingly, this particular scenario (for which the heat transfer coefficient between buildings and the flow is largest), has been less well studied than the situation where buoyancy effects are important. The CFD technique was compared against wind-tunnel experiments to provide model evaluation. The height-to-width ratio of the canyon was varied through the range 0.5–5 and the flow was normal to the canyon axis. By setting the canyon’s facets to have the same or different temperatures or to have a partial temperature distribution, simulations were carried out to investigate: (a) the influence of geometry on the flow and mixing within the canyon and (b) the exchange processes within the canyon and across the canyon top interface. Results showed that the vortex-type circulation and turbulence developed within the canyon produced a temperature distribution that was, essentially, spatially uniform (apart from a relatively thin near-wall thermal boundary layer) This allowed the temperatures within the street canyon to be specified by just one value Tcan, the canyon temperature. The variation of Tcan with wind speed, surface temperatures and geometry was extensively studied. Finally, the exchange velocity uE across the interface between the canyon and the flow above was calculated based on a heat flux balance within the canyon and between the canyon and the flow above. Results showed that uE was approximately 1% of a characteristic wind velocity above the street canyon. The problem of radiative exchange is not addressed but it can, of course, be introduced analytically, or computationally, when necessary.

Air flow inside an array of cubes is simulated. Cubes (edge length 0.15 m) are arranged in a regular array, separated by 0.15 m in the streamwise and spanwise directions. Numerical simulations are performed based on Reynolds-averaged Navier–Stokes equations (RANS), solved in a computational fluid dynamics model (CFD), with standard k–e turbulent closure (two prognostic equations are solved for the turbulent kinetic energy k and its dissipation e, respectively). Simulations are validated against wind-tunnel data using a technique based on hit-rate calculations, and calculated statistical parameters. The results show that the horizontal velocity is very well modelled, and despite some discrepancies, the model that fulfils the hit-rate test criteria gives useful results that are used to investigate three-dimensional (3-D) flow structures. The 3-D analysis of the flow shows interesting patterns: the centre of the canyon vortex is at 3/4 of the canyon height, and stronger downward than upward motions are present within the canyon. Such behaviour is explained by the presence of a compensation flow through the side of the canyon, which enters the canyon from the upper part and exits from the lower part. This complex 3-D structure affects the tracer dispersion, and is responsible for pollutant transport and diffusion.

A major problem in urban climate modelling is determining how the heat fluxes from various canyon surfaces are affected by canyon flow. To address this problem, we developed a water evaporation method involving filter paper to study the distribution of the convective transfer velocity in urban street canyons. In this method, filter paper is pasted onto a building model and the evaporation rate from the paper is measured with an electric balance. The method was tested on 2D (two-dimensional) street canyon models and 3D model arrangements. Moreover, in this technique, it is easy to restrict the flux within an arbitrary surface in question. That is, the evaporation distribution on a surface can be studied by using several small pieces of filter paper. In the 2D case, the wall transfer velocity was strongly dependent on the canyon aspect ratio for perpendicular wind directions and it varied widely with height within both windward and leeward wall surfaces. For 3D cubic arrays, the relation to canyon aspect ratio was largely different from that of the 2D canyon. And, as a case study, the variation of wind direction was investigated for a city-like setting. The area-averaged transfer velocity was insensitive to wind direction but its local deviation was significant. Finally, we measured the transfer velocity for a clustered block array surrounded by relatively wide streets. The effect of spatial heterogeneity on the transfer velocity was significant. Moreover, for a fixed total building volume, the transfer velocity was considerably larger when the building height varied than when it was uniform. Therefore, the water evaporation method with filter paper is expected to be useful for studying the transfer velocity and ventilation rates in urban areas with various canyon shapes.

The Town Energy Budget (TEB) model, a detailed urban parameterisa- tion using a generalised canyon geometry, coupled with the Regional Atmospheric Modelling System (RAMS) is used to simulate the wintertime local circulation in the megacityenvironmentofthemetropolitanareaofSaoPaulo(MASP)inBrazil.Model simulations are performed using actual topography and land-use fields. Comparison with a simple urban parameterisation based on the LEAF-2 scheme is also shown. Validation is based on comparison between model simulations and observations. Sen- sitivity tests with TEB reveal an important interaction between the sea breeze and the MASP heat island circulation. Even though topography is known to play an impor- tant role in the MASP region's weather, in these tests the simulations were performed without topography in order to unambiguously identify the interaction between the two local circulations. The urban heat island (UHI) forms a strong convergence zone in the centre of the city and thereby accelerates the sea-breeze front toward the centre of the city. The presence of the urban region increases the sea-breeze front propa- gation mean speed by about 0.32ms −1 when compared with the situation of no city. After its arrival in the urban region, the sea-breeze front stalls over the centre of the city for about 2h. Subsequently, the sea breeze progresses beyond the city when the heat island dissipates. Thereafter, the sea breeze propagates beyond the urban area at a decelerated rate compared to a simulation without an UHI.

Wind and temperature measurements from within and above a deep urban canyon (height/width = 2.1) were used to examine the thermal structure of air within the canyon, exchange of heat with the overlying atmosphere, and the possible impacts of surface heating on within-canyon air flow. Measurements were made over a range of seasons and primarily analysed for sunny days. This allowed the study of temperature differences between opposing canyon walls and between wall and air of more than 15°C in summer. The wall temperature patterns follow those of incoming solar radiation loading with a secondary daytime effect from the longwave exchange between the walls. In winter, the canyon walls receive little direct solar radiation, and temperature differences are largely due to anthropogenic heating of the building interiors. Cool air from aloft and heated air from canyon walls is shown to circulate within the canyon under cross-canyon flow. Roofs and some portions of walls heat up rapidly on clear days and have a large influence on heat fluxes and the temperature field. The magnitude and direction of the measured turbulent heat flux also depend strongly on the direction of flow relative to surface heating. However, these spatial differences are smoothed by the shear layer at the canyon top. Buoyancy effects from the heated walls were not seen to have as large an impact on the measured flow field as has been shown in numerical experiments. At night canyon walls are shown to be the source of positive sensible heat fluxes. The measurements show that materials and their location, as well as geometry, play a role in regulating the heat exchange between the urban surface and atmosphere.

In order to investigate the microclimatic conditions in a street canyon, a physical model was used to conduct the Joint ATREUS-PICADA Experiment (JAPEX) in situ experimental campaign. Four lines of buildings simulated by steel containers were installed to form three parallel street canyons at 1:5 scale, with width/height aspect ratio approximately 0.40. The reference wind and atmospheric conditions were measured, as well as the flow velocity and direction in the street. Preliminary results concern street canyon ventilation and thermal effects on in-canyon airflow, and show that vortical motions appear for reference wind directions perpendicular to the street axis. The presence of adjacent rows of buildings did not appear to significantly influence the flow character within the canyon for the case of a low aspect ratio corresponding to a skimming flow regime. The flow structure was not significantly affected by the thermal effects although some slight interference occurred in the lower part of the canyon. An analysis of horizontal temperature gradients indicated that a thin boundary layer develops near the heated facade. These facts imply that the thermal effects are considerable only very close to the wall.

We present a scheme for parameterising scalar transfer in the urban boundary layer, which is divided into an inertial layer and a roughness layer. The latter is further divided into a shear layer and a canyon layer. In the inertial layer, scalar transfer is determined by turbulence related to canyon macroscopic features, while in the roughness layer, it is determined by shear-generated turbulence, canyon vortex and vortex-generated turbulence. We first describe a conceptual model for the canyon flow and the aerodynamic resistance network, and then estimate the resistances from the point of view of drag partition and vortex advection. The results are compared with the measurements from wind-tunnel experiments. It is found that for small canyon aspect ratio, σc, the transfer velocity increases with σc, reaching a maximum at around σc=0.5 and then decreases with σc. We also show that the scheme is not sensitive to adjustable parameters

Flow over urban surfaces depends on surface morphology and interaction with the boundary layer above. However, the effect of the flow on scalar fluxes is hard to quantify. The naphthalene sublimation technique was used to quantify scalar vertical fluxes out of a street canyon under neutral conditions. For an array of eight canyons with aspect ratio H/W=0.75 (here, H is building height and W is the street width), increased flux was observed in the first two or three canyons for moderate and low roughness upstream. This is consistent with predictions of the length scale for initial adjustment of flow to an urban canopy. The flux was constant after the initial adjustment region and thus dependent only on local geometry. For a street canyon in the ‘equilibrium’ part of the array, each facet of the street canyon was coated with naphthalene to simulate scalar release from street, walls and roof, to evaluate the effect of street canyon geometry on fluxes for H/W=0.25, 0.6, 1 and 2. Fluxes from the roof and downstream wall were considerably larger than fluxes from the street and upstream wall, and only the flux from the downstream wall exhibited a simple decrease with H/W. For each H/W there was a monotonic decrease between downstream wall, street and upstream wall transfer. This suggests that flow decelerates around the recirculation region in the lee of the upstream building, i.e. a recirculating jet rather than a symmetrical vortex. The addition of a second source within the street canyon resulted in reduced fluxes from each facet for H/W>0.25, due to increased concentration of naphthalene in the canyon air.

A practical model is developed for the vertical flux of a scalar, such as heat, from an urban street canyon that accounts for variations of the flow and turbulence with canyon geometry. The model gives the magnitude and geometric dependence of the flux from each facet of the urban street canyon, and is shown to agree well with wind-tunnel measurements described in Part I. The geometric dependence of the flux from an urban street canyon is shown to be determined by two physical processes. Firstly, as the height-to-width ratio of the street canyon increases, so does the roughness length and displacement height of the surface. This increase leads to a reduction in the wind speed in the inertial sublayer above the street canyons. Since the speed of the circulations in the street are proportional to this inertial sublayer wind speed, the flux then reduces with the inertial sublayer wind speed. This process is dominant at low height-to-width ratios. Secondly, the character of the circulations within the street canyon also varies as the height-to-width ratio increases. The flow in the street is partitioned into a recirculation region and a ventilated region. When the street canyon has high height-to-width ratios the recirculation region occupies the whole street canyon and the wind speeds within the street are low. This tendency decreases the flux at high height-to-width ratios. These processes tend to reduce the flux density from the individual facets of the street canyon, when compared to the flux density from a horizontal surface of the same material. But the street canyon has an increased total surface area, which means that the total flux from the street canyon is larger than from a horizontal surface. The variations in scalar flux from an urban street canyon with geometry is over a factor of two, which means that the physical mechanisms responsible should be incorporated into energy balance models for urban areas.

The influence of building geometry on the radiation terms ofthe surface energy balance is a principal reason for surfacetemperature differences between rural and urban areas.Methods exist to calculate the radiation balance in an urban area,but their validity across the range of urban geometries andmaterials has not been carefully considered.Here the exchange of diffuse radiation in an urban street canyon isinvestigated using a method incorporating all reflections of radiation.This exact solution is compared to two commonly used approximationsthat retain either no reflections, or just one reflection of radiation.The area-averaged net radiative flux density from the facets of the canyondecreases in magnitude monotonically as the canyon aspect ratio increases.The two approximate solutions possess unphysical differences from thismonotonic decrease for high canyon aspect ratios or low materialemissivities/high material albedos.The errors of the two approximate solutions are small for near blackbodymaterials and small canyon aspect ratios but can be an order ofmagnitude for intermediate material properties and deep street canyons.Urban street canyon models need to consider at least one reflectionof radiation and multiple reflections are desirable for full applicability.

Transport of pollution and heatout of streets into the boundary layer above is not currently understood and so fluxes cannot be quantified. Scalar concentration within the street is determined by the flux out of it and so quantifying fluxes for turbulent flow over a rough urban surface is essential. We have developed a naphthalene sublimation technique to measure transfer from a two-dimensional street canyon in a wind tunnel for the case of flow perpendicular to the street. The street was coated with naphthalene, which sublimes at room temperature, so that the vapour represented the scalar source. The transfer velocity wT relates the flux out of the canyon to the concentration within it and is shown to be linearly related to windspeed above the street. The dimensionless transfer coefficient wT/Uδ represents the ventilation efficiency of the canyon (here, wT is a transfer velocity,Uδ is the wind speed at the boundary-layer top). Observed values are between 1.5 and 2.7 ×10-3 and, for the case where H/W→0 (ratio of buildingheight to street width), values are in the same range as estimates of transfer from a flat plate, giving confidence that the technique yields accurate values for street canyon scalar transfer. wT/Uδ varies with aspect ratio (H/W), reaching a maximum in the wake interference regime (0.3 < H/W < 0.65). However, when upstream roughness is increased, the maximum in wT/Uδ reduces, suggesting that street ventilation is less sensitive to H/W when the flow is in equilibrium with the urban surface. The results suggest that using naphthalene sublimation with wind-tunnel models of urban surfaces can provide a direct measure of area-averaged scalar fluxes.

A model for the urban canyon is formulated for meteorologicalconditions of weak winds at night time. Thermal radiation, conductivity and convection are simulated by means of the Monte Carlo method. These are the main physical processesof energy transfer that give rise to the characteristic temperaturedistribution in these systems. The model has been satisfactory tested under ideal conditions for which analytical solutions exist.The predictions of the model under morerealistic conditions accurately reproduce the observationalresults. A strong temperature gradient across streets, with the canyon corners up to 4 °C warmer than the canyon centre, is found for the deepest canyons. This theoretical predictionhas been successfully verified with measurementstaken in a number of streets of the city of Granada in Spain.

The atmospheric katabatic flow in the foothills of the Front Range of the Rocky Mountains has been monitored by a network of towers and sodars for several years as part of the Atmospheric Studies in COmplex Terrain (ASCOT) program. We used three years of data from the network to explore the dependence on surface cooling and channeling by winds above the canyon of (1) profiles of the mean and variance of the vertical (perpendicular to the geopotential) component of motion and (2) the mean component of the wind perpendicular to the local terrain of Coal Creek Canyon. Previously we found that the magnitude of the near-surface temperature difference decreases with increasing surface cooling in light winds, apparently because of increasing turbulence caused when increasing drainage winds interact with surface topography. The variance of vertical velocity exhibits three types of vertical profiles, corresponding to different cooling rates and external wind speeds. The mean variance was found to depend strongly on a locally derived Richardson number.

A numerical model which simulates the dispersion of passive scalars within an urban canyon is proposed and some initial verification of the model undertaken. It is then used to investigate three-dimensional characteristics of the concentration field established for flow perpendicular to canyons with emissions released near the floor of the canyon by motor vehicles. In particular, the influence of canyon geometry on the concentration field is studied and some general observations made concerning the location of maximum concentrations.

A model for the calculation of the turbulence flow field and air pollutant concentrations in urban canyons is developed. A two-dimensional set of hydrodynamical equations and a threedimensional diffusion equation are solved numerically with a personal computer. Different boundary conditions were investigated. Three flow regimes were found: without vortex, with one vortex, and with two vortexes, within an urban canyon. The influence of building density and wind speed components along the street was also investigated.

A two-dimensional, time-dependent flow model coupled with a radiative transfer module has been applied to examine the characteristics of nocturnal flow in a steep canyon in the Rocky Mountains in Colorado. The effect of nighttime surface cooling on drainage flow is examined and compared with observations. In a complementary study, tracer data have been analyzed to estimate the mass flux from a tributary canyon and to examine processes of transport and diffusion. Simulations indicate that the strength and structure of the drainage wind are controlled mainly by terrain features, ambient wind conditions, and effective radiative cooling rates. The transport of tracer from a lower secondary vortex to an upper primary vortex is largely controlled by diffusional processes; removal of tracer from the canyon is controlled by the primary vortex and its interaction with the ambient wind. Differences between mass fluxes from model simulations and those calculated from experiments involve uncertainties in both the structure of the model and the analysis of data.