The Mesoscale Alpine Project Special Observing Period took place during the Autumn of 1999. One of the scientific objectives was to investigate the dynamics of gravity waves and gravity wave breaking in complex mountainous terrain. Three research aircraft, the UK Met Office C130, the NCAR Electra and the DLR Falcon, flew combined sorties stacked in the vertical to take measurements of wave activity. Flight level data and remote sensing instrumentation were used to achieve good sampling in both space and time. Results will be presented for IOP13 on 2 November 1999. On this day, strong south-westerly flow and increasing wind-speed with height up to the jet stream level provided conditions favourable for trapped lee-wave formation.

Time-series of potential temperature, vertical velocity and along- and across-wind components of horizontal velocity for this day were detrended and filtered to remove very fine scale detail. A coherent and steady wave-train is documented from the vertical velocity and potential temperature measurements which varies little in wavelength and amplitude with time. A short wavelength wave is observed at lower levels flown by the Electra, with a longer wavelength wave reaching higher altitudes. However, the waves in the along-wind component of the horizontal wind are absent at the C130 flight level (7.5km). Idealised numerical model simulations have shown that this is probably due to the wide maximum in the vertical velocity wave amplitude for the larger wave.

The lack of wave signature in the along-wind component leads to large scatter in the average momentum fluxes calculated from different runs at each level. In addition, sampling issues, instrument accuracy, turbulence and the effect of clouds may all compound the difficulty in obtaining accurate momentum fluxes from aircraft measurements. The new dynamical core of the Met Office Unified Model, which uses a non-hydrostatic, semi-Lagrangian advection scheme, has been used to simulate the 2 November case study at high resolution. Results from a 1km run will be shown and compared to the observations on that day, indicating how best to make use of the available data for purposes of improving NWP.