Some Relations to Upper Air Data

Abstract

Although, in this Atlas, we refer to surface data only (see also summarizing paper by Schönwiese et al., 1994) a few relations to upper air data (Bayer et al., 1994; Meyhöfer et al., 1996) should be mentioned. These data are available by radiosonde measurements performed in recent decades so that only the 1961–1990 reference period can be considered. Also, the upper air trend analysis is performed only for the area 30° W – 50° E / 60° – 72° N and 15° W – 50° E / 35° – 60° N. This is due to representativeness problems, especially in oceanic areas, although, in general, the representativeness of upper air data is considerably better than that of surface data. Note that Angell (1988) has used no more than 63 stations for a world-wide analysis of radiosonde data. For example, the 500 hPa (c. 5.5 km) geopotential height correlations drop to 0.9 at a distance of as much as c. 1500 km (Bayer et al., 1994). In the case of temperature, the situation is similar. Near the tropopause the station-to-station correlations are somewhat smaller which may be due to increased variability in this level (c. 200–300 hPa).