Time and Space Variability of Spectral Estimates of Atmospheric Pressure

Abstract

The purpose of this paper is to analyze the temporal and spatial behavior of atmospheric pressure spectra. The literature shows many examples of pressure, wind and temperature spectra whose shapes display a remarkable degree of universality. Theories relying either on turbulence or internal waves have been suggested to account for such spectra. While the former accounts for features at the observed synoptic scales and the latter for the local scales, several difficulties remain especially at the intermediate or the so-called mesoscale range. As a preliminary step for our understanding of the physical mechanisms underlying the spectral behavior, a detailed analysis of the surface pressure was carried out having in mind to test the temporal and spatial variability of these spectra. The source of the data were two microbarograph stations in the Po Valley, 276 km apart; one in the plains 50 km south of the Alps, the other in the foothills of the Dolomites. The 60-day record was part of the Alpine Experiment (ALPEX) which took place in 1982. The study reveals new important phenomena concerning the behavior of spectra of atmospheric pressure: 1) Pressure records are intrinsically nonstationary and thus the analysis should more properly be conducted in terms of evolutionary spectra which account for the time variations of the frequency components 2) A relationship exists between a given frequency range and a corresponding nonstationarity scale, i.e., the minimum time interval needed by nonstationarity to manifest itself fully. 3) A quantitative discussion of the time variability of spectra demonstrates that their shape and energy content depend on different time segments. 4) Important differences of the spectra exist between the two stations, indicating a substantial effect of topography, particularly for periods below 40 min. The results of this work suggest that nonstationarity of the pressure spectra and their temporal and spatial behavior are important elements for our understanding of atmospheric and oceanic dynamics. It is still possible that universality remains a valid concept, but it must be recognized that averages over large temporal or spatial datasets may produce results that mask the underlying physical processes.