The Morning Development of the Urban Boundary Layer over London, UK, during the ACTUAL Project

The typical evolution of the Boundary Layer can be described as a succession of stages: I Night-time stable layer; II Formation of a shallow convective boundary layer (CBL) near the ground, known as Morning Transition (MT); III Rapid growth (RG) up to the level of the capping inversion; IV Consolidation of the well-mixed CBL, where the mixing height (MH) grows only slowly; and V Decay of thermally-driven turbulence and vertical mixing followed by the formation of a shallow stable layer at the ground. This evolution has been mostly studied for homogeneous, equilibrium and stationary flows over relatively smooth and homogeneous surfaces. Over urban areas, the modified properties of the urban surface tend to increase energy partitioned into sensible heat flux, and increase surface roughness. Hence, the urban boundary layer (UBL), its stability, thermodynamic properties, and MH are affected. The aim of this study was to test whether the evolution of the UBL was distinctly different to previous findings for rural and idealised boundary layers, using a novel combination of ground-based remote sensing and in situ instrumentation.

Instrumentation was deployed in London, UK, as part of the ACTUAL project (Advanced Climate Technology Urban Atmospheric Laboratory). Two identical instrumentation platforms were placed at a roof-top site (18 m) and the BT Tower (191 m), separated by 1.6 km. Platforms were equipped with (1) eddy covariance systems consisting of sonic anemometers (R3-50, Gill Instruments Ltd), hygrometers (LI-COR 7500 infra-red gas analyser) and net radiometers (Kipp and Zonen CNR4); and (2) weather stations (Vaisala WXT520). Stability of the layer between the roof-top station and BT Tower was estimated using a bulk Richardson number.

A heterodyne scanning Doppler lidar (Halo Photonics, Stream Line) was used for observing UBL structure. The Doppler lidar operated in two modes: Doppler Beam Swinging (DBS) and vertical stare mode. In DBS mode, the lidar beam was tilted in three orthogonal directions every 21 s and a vertical profile of wind velocity was calculated. In vertical stare mode, the lidar beam pointed vertically for 99 sec and measurements were taken every 3.6 s (sampling rate 0.286 Hz). Thus, within each hour, 30 DBS profiles were averaged to give an hourly mean wind speed profile; stare mode data were used to calculate hourly mean attenuated backscatter, and vertical velocity variance profiles. MH was defined as the height up to which vertical velocity variance was greater than 0.1 m2s-2. Lidar measurements from three urban sites within a radius of ~3 km were used. Measurements were conducted during the period 19 May 2011 to 11 January 2012; from this long database, only cases with small amounts of boundary layer cloud were selected. Days were categorised according to wind speed (low, moderate or high), and season (summer, winter).

In contrast with stable nocturnal boundary layers observed over rural areas, the nocturnal UBL was sometimes weakly convective and relatively deep. A new approach for defining the MT was introduced, by redefining the start of the MT as sunrise and the end as a sharp, quasi-linear increase in MH. The urban MT varied in length between 0.5 and 4 h and the growth rate during the RG stage had a strong positive relationship with the convective velocity scale, and a weaker, negative relationship with wind speed. The growth rate also had a positive correlation with MT stage duration, particularly in summer. This result highlighted the opposite effects of wind speed, namely that increased wind speed shortened the MT stage, yet reduced growth rate.

Given the complex dependence of UBL evolution on wind speed, wind shear was more closely examined. Wind shear was higher in the night-time and MT phases than the rapid growth phase, with a greater difference in summer. Shear production of turbulent kinetic energy near the MH top was around 6 times larger than at the surface in summer, and around 1.5 times larger in winter. These results suggest that elevated shear is important in governing mixing within the UBL. In summer-time under low winds, shear production was greater than buoyant production at night-time and in the MT stage near the top of the MH. Within the RG stage, buoyant production dominated at the surface, but shear production dominated from around half-way up the boundary layer. The results suggest that shear processes cannot be neglected during UBL evolution, even in cases of low wind speeds.