Verfahren zur Berechnung des Horizontalwindes für die freie Atmosphäre aus hemisphärisch gegebenen Geopotentialfeldern und Vergleich der Ergebnisse mit Messwerten aus Radiosondenmeldungen

Two sets of wind data were compared in this paper for two days with different circulation forms: computed winds from geopotential fields of the German weather Service (DWD) with measured winds at radiosonde stations. For the area between the North Pole and 20°N the windcomponents were computed with different methods in the six standard levels at grid-points and then interpolated onto the stations, where these computed components were compared with the measured components. The vector differences between the computed and the measured winds at the stations were used as a measure for the worth of the different kinds of wind-computation. Since the addition of acceleration and friction to the wind-computation yielded mainly worse results than the geostrophic wind, the reasons for this circumstance were sought after. At first, the geopotential heights of the 500-mb-level, which were measured at the stations, were interpolated onto the grid-points with a special triangular interpolation. But the results from these new fields were worse than the results from the geopotential fields of the DWD. Then the amplitudes of the given geopotential fields were enhanced and it was discovered, that the geostrophic winds estimated from these new fields did correspond better with the measured winds than the geostrophic winds obtained from the normal DWD-fields. That is, the analysis program of the DWD smoothes the geopotential fields too strongly. This too streng smoothing was conferred by a direct comparison of the geopotentials at the grid-points and at the stations. For the 500-mb-level the windcomponents were also computed from existent coefficients of orthogonal functions and compared with the measured values