Evaluation of Urban Local-Scale Aerodynamic Parameters: Implications for the Vertical Profile of Wind Speed and for Source Areas

Nine methods to determine local-scale aerodynamic roughness length \((z_{0})\) and zero-plane displacement \((z_{d})\) are compared at three sites (within 60 m of each other) in London, UK. Methods include three anemometric (single-level high frequency observations), six morphometric (surface geometry) and one reference-based approach (look-up tables). A footprint model is used with the morphometric methods in an iterative procedure. The results are insensitive to the initial \(z_{d}\) and \(z_{0}\) estimates. Across the three sites, \(z_{d}\) varies between 5 and 45 m depending upon the method used. Morphometric methods that incorporate roughness-element height variability agree better with anemometric methods, indicating \(z_{d}\) is consistently greater than the local mean building height. Depending upon method and wind direction, \(z_{0}\) varies between 0.1 and 5 m with morphometric \(z_{0}\) consistently being 2–3 m larger than the anemometric \(z_{0}\). No morphometric method consistently resembles the anemometric methods. Wind-speed profiles observed with Doppler lidar provide additional data with which to assess the methods. Locally determined roughness parameters are used to extrapolate wind-speed profiles to a height roughly 200 m above the canopy. Wind-speed profiles extrapolated based on morphometric methods that account for roughness-element height variability are most similar to observations. The extent of the modelled source area for measurements varies by up to a factor of three, depending upon the morphometric method used to determine \(z_{d}\) and \(z_{0}\).     

Katabatic Flow: A Closed-Form Solution with Spatially-Varying Eddy Diffusivities

The Nieuwstadt closed-form solution for the stationary Ekman layer is generalized for katabatic flows within the conceptual framework of the Prandtl model. The proposed solution is valid for spatially-varying eddy viscosity and diffusivity (O’Brien type) and constant Prandtl number (Pr). Variations in the velocity and buoyancy profiles are discussed as a function of the dimensionless model parameters \(z_0 \equiv \hat{z}_0 \hat{N}^2 Pr \sin {(\alpha )} |\hat{b}_\mathrm{s} |^{-1}\) and \(\lambda \equiv \hat{u}_{\mathrm{ref}}\hat{N} \sqrt{Pr} |\hat{b}_\mathrm{s} |^{-1}\), where \(\hat{z}_0\) is the hydrodynamic roughness length, \(\hat{N}\) is the Brunt-Väisälä frequency, \(\alpha \) is the surface sloping angle, \(\hat{b}_\mathrm{s}\) is the imposed surface buoyancy, and \(\hat{u}_{\mathrm{ref}}\) is a reference velocity scale used to define eddy diffusivities. Velocity and buoyancy profiles show significant variations in both phase and amplitude of extrema with respect to the classic constant \(\textit{K}\) model and with respect to a recent approximate analytic solution based on the Wentzel-Kramers-Brillouin theory. Near-wall regions are characterized by relatively stronger surface momentum and buoyancy gradients, whose magnitude is proportional to \(z_0\) and to \(\lambda \). In addition, slope-parallel momentum and buoyancy fluxes are reduced, the low-level jet is further displaced toward the wall, and its peak velocity depends on both \(z_0\) and \(\lambda \).     

Aerodynamic Properties of Rough Surfaces with High Aspect-Ratio Roughness Elements: Effect of Aspect Ratio and Arrangements

We examine the effect of varying roughness-element aspect ratio on the mean velocity distributions of turbulent flow over arrays of rectangular-prism-shaped elements. Large-eddy simulations (LES) in conjunction with a sharp-interface immersed boundary method are used to simulate spatially-growing turbulent boundary layers over these rough surfaces. Arrays of aligned and staggered rectangular roughness elements with aspect ratio >1 are considered. First the temporally- and spatially-averaged velocity profiles are used to illustrate the aspect-ratio effects. For aligned prisms, the roughness length (\(z_\mathrm{o}\)) and the friction velocity (\(u_*\)) increase initially with an increase in the roughness-element aspect ratio, until the values reach a plateau at a particular aspect ratio. The exact value of this aspect ratio depends on the coverage density. Further increase in the aspect ratio changes neither \(z_\mathrm{o}\), \(u_*\) nor the bulk flow above the roughness elements. For the staggered cases, \(z_\mathrm{o}\) and \(u_*\) continue to increase for the surface coverage density and the aspect ratios investigated. To model the flow response to variations in roughness aspect ratio, we turn to a previously developed phenomenological volumetric sheltering model (Yang et al., in J Fluid Mech 789:127–165, 2016), which was intended for low to moderate aspect-ratio roughness elements. Here, we extend this model to account for high aspect-ratio roughness elements. We find that for aligned cases, the model predicts strong mutual sheltering among the roughness elements, while the effect is much weaker for staggered cases. The model-predicted \(z_\mathrm{o}\) and \(u_*\) agree well with the LES results. Results show that the model, which takes explicit account of the mutual sheltering effects, provides a rapid and reliable prediction method of roughness effects in turbulent boundary-layer flows over arrays of rectangular-prism roughness elements.     

On the Relevance of $$\ln (z_{0} /z_{0T})= kB^{-1}$$

The assumption that the roughness Reynolds number \(( Re_{*})\) can be used as a basis for quantifying the boundary-layer property \({ kB}^{-1} (= \ln (z_{0}/z_{0T}))\) as in some modern numerical models is questioned. While \({ Re}_{*}\) is a useful property in studies of pipe flow, it appears to have only marginal applicability in the case of treeless terrain, as studied in the two experimental situations presented here. For both the daytime and night-time cases there appears to be little correlation between \({ kB}^{-1}\) and \({ Re}_{*}\). For daytime, the present studies indicate that the assumption \({ kB}^{-1} \approx 2\) is acceptable, while for night-time, the scatter involved in relating \({ kB}^{-1}\) to \({ Re}_{*}\) suggests there is little reason to assume a direct relationship. However, while the scatter affecting all of the night-time results is large, there remains a significant correlation between the heat and momentum fluxes upon which an alternative methodology for describing bulk air–surface exchange at night could be constructed. The friction coefficient (\(C_{f}\)) and the turbulent Stanton number \(({ St}_{*})\) are discussed as possible alternatives for describing bulk properties of the air layer adjacent to the surface. While describing the surface roughness in terms of the friction coefficient provides an attractive simplification relative to the conventional methodologies based on roughness length and stability considerations, use of the Stanton number shares many of uncertainties that affect \({ kB}^{-1}\). The transitions at dawn and dusk remain demanding situations to address.     

Influence of Mean Rooftop-Level Estimation Method on Sensible Heat Flux Retrieved from a Large-Aperture Scintillometer Over a City Centre

The sensible heat flux (H) is determined using large-aperture scintillometer (LAS) measurements over a city centre for eight different computation scenarios. The scenarios are based on different approaches of the mean rooftop-level \((z_{H})\) estimation for the LAS path. Here, \(z_{H}\) is determined separately for wind directions perpendicular (two zones) and parallel (one zone) to the optical beam to reflect the variation in topography and building height on both sides of the LAS path. Two methods of \(z_{H}\) estimation are analyzed: (1) average building profiles; (2) weighted-average building height within a 250 m radius from points located every 50 m along the optical beam, or the centre of a certain zone (in the case of a wind direction perpendicular to the path). The sensible heat flux is computed separately using the friction velocity determined with the eddy-covariance method and the iterative procedure. The sensitivity of the sensible heat flux and the extent of the scintillometer source area to different computation scenarios are analyzed. Differences reaching up to 7% between heat fluxes computed with different scenarios were found. The mean rooftop-level estimation method has a smaller influence on the sensible heat flux (?4 to 5%) than the area used for the \(z_{H}\) computation (?5 to 7%). For the source-area extent, the discrepancies between respective scenarios reached a similar magnitude. The results demonstrate the value of the approach in which \(z_{H}\) is estimated separately for wind directions parallel and perpendicular to the LAS optical beam.     

Optimizing the Determination of Roughness Parameters for Model Urban Canopies

We present an objective optimization procedure to determine the roughness parameters for very rough boundary-layer flow over model urban canopies. For neutral stratification the mean velocity profile above a model urban canopy is described by the logarithmic law together with the set of roughness parameters of displacement height d, roughness length \(z_0\), and friction velocity \(u_*\). Traditionally, values of these roughness parameters are obtained by fitting the logarithmic law through (all) the data points comprising the velocity profile. The new procedure generates unique velocity profiles from subsets or combinations of the data points of the original velocity profile, after which all possible profiles are examined. Each of the generated profiles is fitted to the logarithmic law for a sequence of values of d, with the representative value of d obtained from the minima of the summed least-squares errors for all the generated profiles. The representative values of \(z_0\) and \(u_*\) are identified by the peak in the bivariate histogram of \(z_0\) and \(u_*\). The methodology has been verified against laboratory datasets of flow above model urban canopies.     

Edge-to-Stem Variability in Wet-Canopy Evaporation From an Urban Tree Row

Evaporation from wet-canopy (\(E_\mathrm{C}\)) and stem (\(E_\mathrm{S}\)) surfaces during rainfall represents a significant portion of municipal-to-global scale hydrologic cycles. For urban ecosystems, \(E_\mathrm{C}\) and \(E_\mathrm{S}\) dynamics play valuable roles in stormwater management. Despite this, canopy-interception loss studies typically ignore crown-scale variability in \(E_\mathrm{C}\) and assume (with few indirect data) that \(E_\mathrm{S}\) is generally \({<}2\%\) of total wet-canopy evaporation. We test these common assumptions for the first time with a spatially-distributed network of in-canopy meteorological monitoring and 45 surface temperature sensors in an urban Pinus elliottii tree row to estimate \(E_\mathrm{C}\) and \(E_\mathrm{S}\) under the assumption that crown surfaces behave as “wet bulbs”. From December 2015 through July 2016, 33 saturated crown periods (195 h of 5-min observations) were isolated from storms for determination of 5-min evaporation rates ranging from negligible to 0.67 \(\hbox {mm h}^{-1}\). Mean \(E_\mathrm{S}\) (0.10 \(\hbox {mm h}^{-1}\)) was significantly lower (\(p < 0.01\)) than mean \(E_\mathrm{C}\) (0.16 \(\hbox {mm h}^{-1}\)). But, \(E_\mathrm{S}\) values often equalled \(E_\mathrm{C}\) and, when scaled to trunk area using terrestrial lidar, accounted for 8–13% (inter-quartile range) of total wet-crown evaporation (\(E_\mathrm{S}+E_\mathrm{C}\) scaled to surface area). \(E_\mathrm{S}\) contributions to total wet-crown evaporation maximized at 33%, showing a general underestimate (by 2–17 times) of this quantity in the literature. Moreover, results suggest wet-crown evaporation from urban tree rows can be adequately estimated by simply assuming saturated tree surfaces behave as wet bulbs, avoiding problematic assumptions associated with other physically-based methods.     

Parametric Study of Urban-Like Topographic Statistical Moments Relevant to a Priori Modelling of Bulk Aerodynamic Parameters

For a horizontally homogeneous, neutrally stratified atmospheric boundary layer (ABL), aerodynamic roughness length, \(z_0\), is the effective elevation at which the streamwise component of mean velocity is zero. A priori prediction of \(z_0\) based on topographic attributes remains an open line of inquiry in planetary boundary-layer research. Urban topographies – the topic of this study – exhibit spatial heterogeneities associated with variability of building height, width, and proximity with adjacent buildings; such variability renders a priori, prognostic \(z_0\) models appealing. Here, large-eddy simulation (LES) has been used in an extensive parametric study to characterize the ABL response (and \(z_0\)) to a range of synthetic, urban-like topographies wherein statistical moments of the topography have been systematically varied. Using LES results, we determined the hierarchical influence of topographic moments relevant to setting \(z_0\). We demonstrate that standard deviation and skewness are important, while kurtosis is negligible. This finding is reconciled with a model recently proposed by Flack and Schultz (J Fluids Eng 132:041203-1–041203-10, 2010), who demonstrate that \(z_0\) can be modelled with standard deviation and skewness, and two empirical coefficients (one for each moment). We find that the empirical coefficient related to skewness is not constant, but exhibits a dependence on standard deviation over certain ranges. For idealized, quasi-uniform cubic topographies and for complex, fully random urban-like topographies, we demonstrate strong performance of the generalized Flack and Schultz model against contemporary roughness correlations.     

Seasonal Variations in Drag Coefficient over a Sastrugi-Covered Snowfield in Coastal East Antarctica

The surface of windy Antarctic snowfields is subject to drifting snow, which leads to the formation of sastrugi. In turn, sastrugi contribute to the drag exerted by the snow surface on the atmosphere and hence influence drifting snow. Although the surface drag over rough sastrugi fields has been estimated for individual locations in Antarctica, its variation over time and with respect to drifting snow has received little attention. Using year-round data from a meteorological mast, seasonal variations in the neutral drag coefficient at a height of 10 m \((C_{{ DN}10})\) in coastal Adelie Land are presented and discussed in light of the formation and behaviour of sastrugi based on observed aeolian erosion patterns. The measurements revealed high \(C_{{ DN}10} \) values \((\ge \) 2 \(\times \) 10\(^{-3})\) and limited drifting snow (35% of the time) in summer (December–February) versus lower \(C_{{ DN}10} \) values \((\approx \) 1.5 \(\times \) \(10^{-3})\) associated with more frequent drifting snow (70% of the time) in winter (March–November). Without the seasonal distinction, there was no clear dependence of \(C_{{ DN}10} \) on friction velocity or wind direction, but observations revealed a general increase in \(C_{{ DN}10} \) with rising air temperature. The main hypothesis defended here is that higher temperatures increase snow cohesion and the development of sastrugi just after snow deposition while inhibiting the sastrugi streamlining process by raising the erosion threshold. This increases the contribution of the sastrugi form drag to the total surface drag in summer when winds are lighter and more variable. The analysis also showed that, in the absence of erosion, single snowfall events can reduce \(C_{{ DN}10} \) to \(1\,\times \,10^{-3}\) due to the burying of pre-existing microrelief under newly deposited snow. The results suggest that polar atmospheric models should account for spatial and temporal variations in snow surface roughness through a dynamic representation of the sastrugi form drag.     

Understanding Diurnality and Inter-Seasonality of a Sub-tropical Urban Heat Island

We quantify the spatial and temporal aspects of the urban heat-island (UHI) effect for Kanpur, a major city in the humid sub-tropical monsoon climate of the Gangetic basin. Fixed station measurements are used to investigate the diurnality and inter-seasonality in the urban–rural differences in surface temperature (\({\Delta } T_\mathrm{s}\)) and air temperature (\({\Delta } T_\mathrm{c}\)) separately. The extent of the spatial variations of the nighttime \({\Delta } T_\mathrm{c}\) and \({\Delta } T_\mathrm{s}\) is investigated through mobile campaigns and satellite remote sensing respectively. Nighttime \({\Delta } T_\mathrm{c}\) values dominate during both the pre-monsoon (maximum of 3.6 \(^\circ \hbox {C}\)) and the monsoon (maximum of 2.0 \(^\circ \hbox {C}\)). However, the diurnality in \({\Delta } T_\mathrm{s}\) is different, with higher daytime values during the pre-monsoon, but very little diurnality during the monsoon. The nighttime \({\Delta } T_\mathrm{s}\) value is mainly associated with differences in the urban–rural incoming longwave radiative flux (\(r^{2}=0.33\) during the pre-monsoon; 0.65 during the monsoon), which, in turn, causes a difference in the outgoing longwave radiative flux. This difference may modulate the nighttime \({\Delta } T_\mathrm{c}\) value as suggested by significant correlations (\(r^{2}=0.68\) for the pre-monsoon; 0.50 for the monsoon). The magnitude of \({\Delta } T_\mathrm{c}\) may also be modulated by advection, as it is inversely related with the urban wind speed. A combination of in situ, remotely sensed, and model simulation data were used to show that the inter-seasonality in \({\Delta } T_\mathrm{s}\), and, to a lesser extent, in \({\Delta } T_\mathrm{c}\), may be related to the change in the land use of the rural site between the pre-monsoon and the monsoon periods. Results suggest that the degree of coupling of \({\Delta } T_\mathrm{s}\) and \({\Delta } T_\mathrm{c}\) may be a strong function of land use and land cover.     

Scale Properties of Anisotropic and Isotropic Turbulence in the Urban Surface Layer

The scale properties of anisotropic and isotropic turbulence in the urban surface layer are investigated. A dimensionless anisotropic tensor is introduced and the turbulent tensor anisotropic coefficient, defined as C, where \(C = 3d_{3}\,+\,1 (d_{3}\) is the minimum eigenvalue of the tensor) is used to characterize the turbulence anisotropy or isotropy. Turbulence is isotropic when \(C \approx 1\), and anisotropic when \(C \ll 1\). Three-dimensional velocity data collected using a sonic anemometer are analyzed to obtain the anisotropic characteristics of atmospheric turbulence in the urban surface layer, and the tensor anisotropic coefficient of turbulent eddies at different spatial scales calculated. The analysis shows that C is strongly dependent on atmospheric stability \(\xi = (z-z_{\mathrm{d}})/L_{{\textit{MO}}}\), where z is the measurement height, \(z_{\mathrm{d}}\) is the displacement height, and \(L_{{\textit{MO}}}\) is the Obukhov length. The turbulence at a specific scale in unstable conditions (i.e., \(\xi < 0\)) is closer to isotropic than that at the same scale under stable conditions. The maximum isotropic scale of turbulence is determined based on the characteristics of the power spectrum in three directions. Turbulence does not behave isotropically when the eddy scale is greater than the maximum isotropic scale, whereas it is horizontally isotropic at relatively large scales. The maximum isotropic scale of turbulence is compared to the outer scale of temperature, which is obtained by fitting the temperature fluctuation spectrum using the von Karman turbulent model. The results show that the outer scale of temperature is greater than the maximum isotropic scale of turbulence.     

An Efficient Non-iterative Bulk Parametrization of Surface Fluxes for Stable Atmospheric Conditions Over Polar Sea-Ice

In climate and weather prediction models the near-surface turbulent fluxes of heat and momentum and related transfer coefficients are usually parametrized on the basis of Monin–Obukhov similarity theory (MOST). To avoid iteration, required for the numerical solution of the MOST equations, many models apply parametrizations of the transfer coefficients based on an approach relating these coefficients to the bulk Richardson number \(Ri_{b}\). However, the parametrizations that are presently used in most climate models are valid only for weaker stability and larger surface roughnesses than those documented during the Surface Heat Budget of the Arctic Ocean campaign (SHEBA). The latter delivered a well-accepted set of turbulence data in the stable surface layer over polar sea-ice. Using stability functions based on the SHEBA data, we solve the MOST equations applying a new semi-analytic approach that results in transfer coefficients as a function of \(Ri_{b}\) and roughness lengths for momentum and heat. It is shown that the new coefficients reproduce the coefficients obtained by the numerical iterative method with a good accuracy in the most relevant range of stability and roughness lengths. For small \(Ri_{b}\), the new bulk transfer coefficients are similar to the traditional coefficients, but for large \(Ri_{b}\) they are much smaller than currently used coefficients. Finally, a possible adjustment of the latter and the implementation of the new proposed parametrizations in models are discussed.     

The Effect of Breaking Waves on $$\hbox {CO}_2$$ Air–Sea Fluxes in the Coastal Zone

The influence of wave-associated parameters controlling turbulent \(\hbox {CO}_2\) fluxes through the air–sea interface is investigated in a coastal region. A full year of high-quality data of direct estimates of air–sea \(\hbox {CO}_2\) fluxes based on eddy-covariance measurements is presented. The study area located in Todos Santos Bay, Baja California, Mexico, is a net sink of \(\hbox {CO}_2\) with a mean flux of \(-1.3\, \upmu \hbox {mol m}^{-2}\hbox {s}^{-1}\) (\(-41.6\hbox { mol m}^{-2}\hbox {yr}^{-1}\)). The results of a quantile-regression analysis computed between the \(\hbox {CO}_2\) flux and, (1) wind speed, (2) significant wave height, (3) wave steepness, and (4) water temperature, suggest that the significant wave height is the most correlated parameter with the magnitude of the flux but the behaviour of the relation varies along the probability distribution function, with the slopes of the regression lines presenting both positive and negative values. These results imply that the presence of surface waves in coastal areas is the key factor that promotes the increase of the flux from and into the ocean. Further analysis suggests that the local characteristics of the aqueous and atmospheric layers might determine the direction of the flux.     

Estimating Sediment Mass Fluxes on Surfaces Sheltered by Live Vegetation

We present a simple model based on already existing and widely used equations for estimating particle mass fluxes on surfaces sheltered by live vegetation. Wind-tunnel measurements of vertical profiles of mass flux in three different dense live plant canopies, and as a function of the spatially averaged skin friction velocity \({u_{\tau }}'\), provide the baseline set of data. For the bare-sand surface, the total mass flux Q shows the typical \(b({u_\tau }' - {u_{\tau t}}')^{3 }\) increase with increasing skin friction velocity \({u_{\tau }}'\), where b is a constant and \({u_{\tau t}}'\) is the threshold at the onset of particle erosion. Similar relations, however, with different values for b and \({u_{\tau t}}'\) compared to the bare-sand surface were found for experiments with 5.25 and 24.5 plants \(\hbox {m}^{-2}\) and can be explained by the spatial variations of \(u_{\tau }\) for the canopy cases. Based on the resulting parameters b and \({u_{\tau t}}'\), which are found to be functions of the roughness density \(\lambda \), we present a final simple relation \(Q(\lambda ,\, {u_{\tau }}')\) used for estimating the total mass flux for surfaces sheltered by live vegetation.     

Water-Channel Estimation of Eulerian and Lagrangian Time Scales of the Turbulence in Idealized Two-Dimensional Urban Canopies

Lagrangian and Eulerian statistics are obtained from a water-channel experiment of an idealized two-dimensional urban canopy flow in neutral conditions. The objective is to quantify the Eulerian \((T^{\mathrm{E}})\) and Lagrangian \((T^{\mathrm{L}})\) time scales of the turbulence above the canopy layer as well as to investigate their dependence on the aspect ratio of the canopy, AR, as the latter is the ratio of the width (W) to the height (H) of the canyon. Experiments are also conducted for the case of flat terrain, which can be thought of as equivalent to a classical one-directional shear flow. The values found for the Eulerian time scales on flat terrain are in agreement with previous numerical results found in the literature. It is found that both the streamwise and vertical components of the Lagrangian time scale, \(T_\mathrm{u}^\mathrm{L} \) and \(T_\mathrm{w}^\mathrm{L} \), follow Raupach’s linear law within the constant-flux layer. The same holds true for \(T_\mathrm{w}^\mathrm{L} \) in both the canopies analyzed \((AR= 1\) and \(AR= 2\)) and also for \(T_\mathrm{u}^\mathrm{L} \) when \(AR = 1\). In contrast, for \(AR = 2\), \(T_\mathrm{u}^\mathrm{L} \) follows Raupach’s law only above \(z=2H\). Below that level, \(T_\mathrm{u}^\mathrm{L} \) is nearly constant with height, showing at \(z=H\) a value approximately one order of magnitude greater than that found for \(AR = 1\). It is shown that the assumption usually adopted for flat terrain, that \(\beta =T^{\mathrm{L}}/T^{\mathrm{E}}\) is proportional to the inverse of the turbulence intensity, also holds true even for the canopy flow in the constant-flux layer. In particular, \(\gamma /i_\mathrm{u} \) fits well \(\beta _\mathrm{u} =T_\mathrm{u}^\mathrm{L} /T_\mathrm{u}^\mathrm{E} \) in both the configurations by choosing \(\gamma \) to be 0.35 (here, \(i_\mathrm{u} =\sigma _\mathrm{u} / \bar{u} \), where \(\bar{u} \) and \(\sigma _\mathrm{u} \) are the mean and the root-mean-square of the streamwise velocity component, respectively). On the other hand, \(\beta _\mathrm{w} =T_\mathrm{w}^\mathrm{L} /T_\mathrm{w}^\mathrm{E} \) follows approximately \(\gamma /i_\mathrm{w} =0.65/\left( {\sigma _\mathrm{w} /\bar{u} } \right) \) for \(z > 2H\), irrespective of the AR value. The second main objective is to estimate other parameters of interest in dispersion studies, such as the eddy diffusivity of momentum \((K_\mathrm{{T}})\) and the Kolmogorov constant \((C_0)\). It is found that \(C_0\) depends appreciably on the velocity component both for the flat terrain and canopy flow, even though for the latter case it is insensitive to AR values. In all the three experimental configurations analyzed here, \(K_\mathrm{{T}}\) shows an overall linear growth with height in agreement with the linear trend predicted by Prandtl’s theory.     

Large-Eddy-Simulation Study of the Effects of Building-Height Variability on Turbulent Flows over an Actual Urban Area

Large-eddy simulation (LES) is used to investigate the effects of building-height variability on turbulent flows over an actual urban area, the city of Kyoto, which is reproduced using a 2-m resolution digital surface dataset. Comparison of the morphological characteristics of Kyoto with those of European, North American, and other Japanese cities indicates a similarity to European cities but with more variable building heights. The performance of the LES model is validated and found to be consistent with turbulence observations obtained from a meteorological tower and from Doppler lidar. We conducted the following two numerical experiments: a control experiment using Kyoto buildings, and a sensitivity experiment in which all the building heights are set to the average height over the computational region \(h_{all}\). The difference of Reynolds stress at height \(z=2.5h_{all}\) between the control and sensitivity experiments is found to increase with the increase in the plan-area index (\(\lambda _p\)) for \(\lambda _p > 0.32\). Thus, values of \(\lambda _p\approx 0.3\) can be regarded as a threshold for distinguishing the effects of building-height variability. The quadrant analysis reveals that sweeps contribute to the increase in the Reynolds stress in the control experiment at a height \(z= 2.5h_{all}\). The exuberance in the control experiment at height \(z=0.5h_{all}\) is found to decrease with increase in the building-height variability. Although the extreme momentum flux at height \(z=2.5h_{all}\) in the control experiment appears around buildings, it contributes little to the total Reynolds stress and is not associated with coherent motions.     

Turbulent Flow Over Large Roughness Elements: Effect of Frontal and Plan Solidity on Turbulence Statistics and Structure

Wind-tunnel experiments were carried out on fully-rough boundary layers with large roughness (\(\delta /h \approx 10\), where h is the height of the roughness elements and \(\delta \) is the boundary-layer thickness). Twelve different surface conditions were created by using LEGO? bricks of uniform height. Six cases are tested for a fixed plan solidity (\(\lambda _\mathrm{P}\)) with variations in frontal density (\(\lambda _\mathrm{F}\)), while the other six cases have varying \(\lambda _\mathrm{P}\) for fixed \(\lambda _\mathrm{F}\). Particle image velocimetry and floating-element drag-balance measurements were performed. The current results complement those contained in Placidi and Ganapathisubramani (J Fluid Mech 782:541–566, 2015), extending the previous analysis to the turbulence statistics and spatial structure. Results indicate that mean velocity profiles in defect form agree with Townsend’s similarity hypothesis with varying \(\lambda _\mathrm{F}\), however, the agreement is worse for cases with varying \(\lambda _\mathrm{P}\). The streamwise and wall-normal turbulent stresses, as well as the Reynolds shear stresses, show a lack of similarity across most examined cases. This suggests that the critical height of the roughness for which outer-layer similarity holds depends not only on the height of the roughness, but also on the local wall morphology. A new criterion based on shelter solidity, defined as the sheltered plan area per unit wall-parallel area, which is similar to the ‘effective shelter area’ in Raupach and Shaw (Boundary-Layer Meteorol 22:79–90, 1982), is found to capture the departure of the turbulence statistics from outer-layer similarity. Despite this lack of similarity reported in the turbulence statistics, proper orthogonal decomposition analysis, as well as two-point spatial correlations, show that some form of universal flow structure is present, as all cases exhibit virtually identical proper orthogonal decomposition mode shapes and correlation fields. Finally, reduced models based on proper orthogonal decomposition reveal that the small scales of the turbulence play a significant role in assessing outer-layer similarity.     

Mechanisms Responsible for the Observed Thermodynamic Structure in a Convective Boundary Layer Over the Hudson Valley of New York State

We examine cases of a regional elevated mixed layer (EML) observed during the Hudson Valley Ambient Meteorology Study (HVAMS) conducted in New York State, USA in 2003. Previously observed EMLs referred to topographic domains on scales of 10\(^{5}\)–10\(^{6}\) km\(^{2}\). Here, we present observational evidence of the mechanisms responsible for the development and maintenance of regional EMLs overlying a valley-based convective boundary layer (CBL) on much smaller spatial scales (<5000 km\(^{2})\). Using observations from aircraft-based, balloon-based, and surface-based platforms deployed during the HVAMS, we show that cross-valley horizontal advection, along-valley channelling, and fog-induced cold-air pooling are responsible for the formation and maintenance of the EML and valley-CBL coupling over New York State’s Hudson Valley. The upper layer stability of the overlying EML constrains growth of the valley CBL, and this has important implications for air dispersion, aviation interests, and fog forecasting.     

Effect of Surface Heterogeneity on the Boundary-Layer Height: A Case Study at a Semi-Arid Forest

We investigate the effects of an isolated meso-\(\gamma \)-scale surface heterogeneity for roughness and albedo on the atmospheric boundary-layer (ABL) height, with a case study at a semi-arid forest surrounded by sparse shrubland (forest area: \(28~\text{ km }^2\), forest length in the main wind direction: 7 km). Doppler lidar and ceilometer measurements at this semi-arid forest show an increase in the ABL height over the forest compared with the shrubland on four out of eight days. The differences in the ABL height between shrubland and forest are explained for all days with a model that assumes a linear growth of the internal boundary layer of the forest through the convective ABL upwind of the forest followed by a square-root growth into the stable free atmosphere. For the environmental conditions that existed during our measurements, the increase in ABL height due to large sensible heat fluxes from the forest (\(600~\text {W~m}^{-2}\) in summer) is subdued by stable stratification in the free atmosphere above the ABL, or reduced by high wind speeds in the mixed layer.     

Observations of Atmospheric Methane and Carbon Dioxide Mixing Ratios: Tall-Tower or Mountain-Top Stations?

Mountain-top observations of greenhouse gas mixing ratios may be an alternative to tall-tower measurements for regional scale source and sink estimation. To investigate the equivalence or limitations of a mountain-top site as compared to a tall-tower site, we used the unique opportunity of comparing in situ measurements of methane (\(\hbox {CH}_{4}\)) and carbon dioxide (\(\hbox {CO}_{2}\)) mixing ratios at a mountain top (986 m above sea level, a.s.l.) with measurements from a nearby (distance 28.4 km) tall tower, sampled at almost the same elevation (1009 m a.s.l.). Special attention was given to, (i) how local wind statistics and greenhouse gas sources and sinks at the mountain top influence the observations, and (ii) whether mountain-top observations can be used as for those from a tall tower for constraining regional greenhouse gas emissions. Wind statistics at the mountain-top site are clearly more influenced by local flow systems than those at the tall-tower site. Differences in temporal patterns of the greenhouse gas mixing ratios observed at the two sites are mostly related to the influence of local sources and sinks at the mountain-top site. Major influences of local sources can be removed by applying a statistical filter (\(5{\mathrm{th}}\) percentile) or a filter that removes periods with unfavourable flow conditions. In the best case, the bias in mixing ratios between the mountain-top and the tall-tower sites after the application of the wind filter was \({-}0.0005\pm 0.0010\) ppm for methane (September, 0000–0400 UTC) and \(0.11\pm 0.18\) ppm for \(\hbox {CO}_{2}\) (February, 1200–1600 UTC). Temporal fluctuations of atmospheric \(\hbox {CH}_{4}\) and \(\hbox {CO}_{2}\) mixing ratios at both stations also showed good agreement (apart from \(\hbox {CO}_{2}\) during summertime) as determined by moving bi-weekly Pearson correlation coefficients (up to 0.96 for \(\hbox {CO}_{2}\) and 0.97 for \(\hbox {CH}_{4}\)). When only comparing mixing ratios minimally influenced by local sources (low bias and high correlation coefficients), our measurements indicate that mountain-top observations are comparable to tall-tower observations.