SUMMARYUsing numerical model analyses, it is shown that there is a coherent diurnal cycle of the West African monsoon winds. As has been observed in previous studies of arid and semi-arid areas, the winds are at their weakest in the afternoon when the convective boundary layer (CBL) is deep, and intensify overnight when the boundary-layer turbulence is much weaker. This diurnal cycle is maximized in the northern part of the monsoon layer, where the meridional pressure gradient and the diurnal cycle of the CBL are both strong.The diurnal cycle can also be resolved in surface and upper-air data, which show how the nocturnal meridional circulation acts to stratify the lower part of the monsoon layer. In contrast, mixing in the daytime CBL acts to maintain the baroclinicity, as has been observed in laboratory flows. This pattern has implications for the efficiency of the monsoon circulation in the continental water budget, as well as in mixing of trace gases and aerosols between the surface layer and the free troposphere. Vertical mixing occurs by day, while meridional advection, with isentropic upgliding and downgliding, is most efficient at night.Finally, high-resolution observations from the JET2000 experiment are used to show that there is mesoscale structure in the diurnally varying monsoon circulation. In the nocturnal flows, local circulations have been observed and appear to represent a response to recent deep convective events. In contrast, the daytime CBL properties at these scales have been shown in a previous study to map closely onto patterns of soil moisture, with horizontal advection playing a weaker role.

SUMMARYAnalyses of the African easterly jet (AEJ) are presented which are based on meridional transects of highresolution dropsonde observations made during JET2000, an aircraft campaign conducted in the last week of August 2000. The observations have confirmed that the AEJ is closely defined by geostrophic balance. The baroclinicity between the extreme northern and southern profiles accurately determines the altitude of the jet core, while the location and morphology of the jet core correspond to a locally-defined geostrophic wind measure. The potential-vorticity (PV) structure has also been found to accord with theoretical expectations, with distinctive positive-and negative-PV anomalies equatorward and poleward of the jet core respectively.The thermodynamic structure of the AEJ environment can be categorized into coherent layers. The monsoon layer is a humid zone connected to the land surface, extending northwards into the Sahel and increasing in depth towards the south. This layer is affected by the land surface on diurnal time-scales, through the growing convective mixed layer and through shallow cumulus clouds. Above the monsoon layer is the Saharan air layer (SAL), which can be identified as a layer of low static-stability and low PV. The SAL is deep where it merges with the Saharan boundary-layer in the north, and becomes thinner toward the south. It has been shown that the boundaries of the SAL can be approximated to good accuracy as adiabatic surfaces, meaning that the SAL comprises air which is adiabatically connected to the land surface via the Saharan boundary-layer. The upper region of the SAL is identified as a layer of high relative-humidity where altocumulus and stratocumulus layers are observed. Finally, the troposphere above the SAL is again almost pseudoadiabatic, with small baroclinicity which determines the closure of the AEJ core aloft. Through inspection of thermodynamic tracers, evidence of convective and lateral transport and exchange between these layers is also presented.

No abstract

SUMMARYA field campaign aimed at observing the near-surface flow field across and downwind of a mountain range on the Falkland Islands, South Atlantic, is described. The objective was to understand and eventually predict orographically generated turbulence. The instrumentation was based primarily on an array of automatic weather stations (AWSs), which recorded 30 s mean surface pressure, wind speed and direction (at 2 m), temperature and relative humidity for approximately one year. These measurements were supported by twice-daily radiosonde releases. The densest part of the AWS array was located to the south of the Wickham mountain range, across Mount Pleasant Airfield (MPA). In northerly flow the array provides a detailed study of the flow downwind of the mountain range. The dataset contains several episodes in which the flow downwind of the mountains is accelerated relative to the upwind flow. During some of these episodes short-lived (typically ∼1 hour) periods of unsteady flow separation are observed and these are associated with the formation of rotors aloft. Such events present a significant hazard to aviation at MPA. Examination of radiosonde profiles suggests that the presence of a strong temperature inversion at a height similar to the mountain height is a necessary condition for both downwind acceleration and the formation of rotors. The data are used to show that the downwind fractional speed-up is proportional to the non-dimensional mountain height (based on upstream near-surface winds and a depth-averaged Brunt-Väisälä frequency diagnosed from radiosonde data). Similarly, a relationship is established between a quantity that describes the spatial variability of the flow downwind of the mountains and the upstream wind and depth-averaged Brunt-Väisälä frequency. The dependence of the flow behaviour on the Froude number (defined in the usual way for two-layer shallow-water flow) and ratio of mountain height to inversion height is presented in terms of a flow regime diagram.