Atmospheric turbulence and surface energy exchange in urban environments : results from the Basel Urban Boundary Layer Experiment (BUBBLE)

The present experimental study addresses turbulence
and exchange processes in the urban roughness
sublayer, namely the region from street
canyon floor up to 2.5 times the mean building
height. Measurements with ultrasonic anemometerthermometers
from three urban full-scale towers
provided new insights into vertical profiles of
mean flow, Reynolds stress, turbulent kinetic energy
(TKE), dissipation rate, as well as exchange
processes of heat, and partially water vapor and
CO2. With the help of ensemble profiles, which are
a surrogate for a real horizontal average, results are
discussed in the frame of an ‘urban family portrait’.
For the majority of realizations, the plane mixing
layer analogy matches processes in the urban roughness
sublayer much better than the classical boundary
layer theory. The observed patterns suggest a
conceptual division of the urban roughness sublayer
into three parts, namely the canyon layer, the roof
layer, and an above-roof layer.
In the canyon layer, local mechanical and thermal
turbulence production are of minor importance. Turbulence
is dominated by large coherent structures,
it is very intermittent and highly uncorrelated. The
majority of TKE is imported by turbulent and pressure
transport from the roof layer. The well known
street canyon vortex is only found on average and
only for selected configurations. Upwind roof shape
was determined as an important factor affecting its
dynamics.
In the roof layer, profiles are characterized by
strongest gradients and exchange is more efficient.
Here, local shear production is a strong source of
TKE. The skimming flow over the street canyons
creates an inflected mean wind profile, from which
instabilities evolve. Notable amounts of TKE and
temperature variance are exported from the roof
layer by sweeps into the upper street canyon and
by ejections into the above-roof layer. As a consequence,
dissipation rate is lower than locally produced
turbulence and neutral limits of velocity variances
are slightly lower than predicted with classical
(local) approaches.
In the above-roof layer, the mean wind profile approximates
the well known logarithmic form valid
in the inertial sublayer. And, integral statistics approach
surface layer values. Turbulent transport
processes of momentum and heat are dominated by
ejections. While shear production is the main source
of TKE in the roof layer below, here both, buoyancy
and shear production are important.
Finally, a network of spatially distributed energy
balance measurements allowed a quantitative estimation
of the urban energy balance modification.
For this purpose, the surface energy balance was simultaneously
measured over different land uses (urban,
suburban, rural).
The impact of a lower urban albedo is roughly counterbalanced
by a stronger long-wave emission, resulting
in a nearly equivalent net radiation over urban
and rural surfaces. Urban surfaces are characterized
by a strong storage term and a high Bowen ratio.
At night, turbulent flux densities remain upward
directed in dense urban environments. This is explained
by a strong nocturnal release of stored heat.
As a consequence, the urban inertial sublayer and
the roughness sublayer are thermally unstable most
of the time.