During winter 2002/2003, three balloon-borne frost point hygrometers measured high-resolution profiles of stratospheric water vapour above Ny-Ålesund, Spitsbergen. All measurements reveal a high H₂O mixing ratio of about 7 ppmv above 24 km, thus differing significantly from the 5 ppmv that are commonly assumed for the calculation of polar stratospheric cloud existence temperatures. The profiles obtained on 12 December 2002 and on 17 January 2003 provide an insight into the vertical distribution of water vapour in the core of the polar vortex.

Unlike the earlier profiles, the water vapour sounding on 11 February 2003 detected the vortex edge region in the lower part of the stratosphere. Here, a striking diminuition in H₂O mixing ratio stands out between 16 and 19 km. The according stratospheric temperatures clarify that this dehydration can not be caused by the presence of polar stratospheric clouds or earlier PSC particle sedimentation.

On the same day, ozone observations by lidar indicate a large scale movement of the polar vortex, while an ozone sonde measurement even shows laminae in the same altitude range as in the water vapour profile. Tracer lamination in the vortex edge region is caused by filamentation of the vortex. The link between the observed water vapour diminuition and filaments in the vortex edge region is highlighted by results of the MIMOSA contour advection model. In the altitude of interest, adjoined filaments of polar and mid-latitudinal air can be identified above the Spitsbergen region. A vertical cross-section reveals that the water vapour sonde has flown through polar air in the lowest part of the stratosphere. Where the low water vapour mixing ratio was detected, the balloon passed through air from a mid-latitudinal filament from about 425 to 445 K, before it finally entered the polar vortex above 450 K. The MIMOSA model results elucidate the correlation that on 11 February 2003 the frost point hygrometer measured strongly variable water vapour concentrations as the sonde detected air with different origins, respectively.

Instead of being linked to dehydration due to PSC particle sedimentation, the local diminuition in the stratospheric water vapour profile of 11 February 2003 has been found to be caused by dynamical processes in the polar stratosphere.