10th Conference on Mountain Meteorology and MAP Meeting 2002

Thursday, 20 June 2002
Rotor streaming, hydraulic jumps and flow separation in the Falkland Islands
Stephen D. Mobbs, University of Leeds, Leeds, West Yorkshire, United Kingdom; and S. J. Arnold, M. K. Hill, S. B. Vosper, and A. M. Gadian
The term "rotor streaming" is used by forecasters to describe recirculations associated with lee waves. This is an idealisation of a local wave-induced adverse pressure gradient. These gradients are observed to cause rapid near-surface deceleration; they are associated with strong turbulence and a high degree of wind variability.

Rotor streaming conditions are frequently encountered at the airport on the Falkland Islands. In northerly winds the airport is situated downwind of two ridges (each 600 m in altitude, separated by 15 km). Extremely unsteady reversals of the wind are observed downwind of the ridges. These are often associated with dramatic roll and overturning clouds.

A field programme gathering comprehensive observations of rotor streaming was conducted from October 2000 to October 2001. Twenty automatic weather stations with microbarographs were located in a north-south line across the ridges, extending for over 40 km. The data was supplemented by turbulence measurements from two 15 m masts, radiosonde ascents, sodar and video camera observations at the airport.

Results from the field programme show strong evidence that the rotor streaming is associated with a hydraulic jump-like structure downwind of the ridges, close to where the wave cloud is observed. This jump has a pressure drop of up to 5 hPa over a distance of a few km. Radiosonde profiles show a correspondence between rotor streaming and sharp boundary-layer capping inversions with temperature increases of over 10 degC in a vertical distance of 200 m. The role of temperature inversions in producing large amplitude responses to gravity-waves has been investigated using a linear numerical model. The model is initialised using radiosonde or NWP analysis profiles. Agreement between model and observations is good when using radiosonde profiles but less good when using NWP profiles. The differences between the results can be explained by limited resolution of temperature inversions in the NWP analyses.

Supplementary URL: