Statistics of Antarctic surface meteorology based on hourly data in 1957e2007 at Syowa Station

• Main Authors: Kaoru Sato A, Naohiko Hirasawa B
• Other Authors: The Pennsylvania State University CiteSeerX Archives
• Format: Text
• Language: English
• Published: 2007
• Subjects: Antarctic; Syowa Station; The Antarctic; Antarc*
• Online Access: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.513.9744; http://www-aos.eps.s.u-tokyo.ac.jp/~kaoru/CV/PDFs/Sato%26Hirasawa_PS_2007.pdf

Statistical characteristics of the surface meteorology are examined at Syowa Station (69000S, 39350E), which is located on an island near the coastal region of the Antarctic continent, based on operational observations made over the 50-year period from February 1957 through January 2007, which includes missing periods equivalent to 5 years. Statistics are obtained for the surface temperature, sea level pressure (SLP), and horizontal winds in terms of frequency distribution, frequency power spectra, seasonal variation, diurnal variation, inter-annual variation, and trends, using hourly observation data, and several interesting characteristics are elucidated. The mean temperature, SLP, and wind speed over the 50-year period are 10.5 C, 986 hPa, and 6.6 m s1, respec-tively. The frequency distribution of temperature is far from the normal one, because less variation exists in summer at higher temperatures. The predominant wind direction is northeasterly (southwestward), and a weak secondary peak is observed in the southerly (northward) direction in the frequency distribution. The directional constancy of winds is 0.78. The frequency spectra over a wide range of 2 h to 20 years exhibit clearly isolated peaks corresponding to annual and diurnal frequencies and their higher harmonics. An important finding is that the spectral shape is proportional to a power of the frequency with a transition frequency for all physical parameters. The transition frequencies correspond to about 5 days for temperature and winds and 3 days for SLP, most likely due to cyclonic activity. A peak in the 11-year solar cycle is not identified in any spectrum. Another interesting feature is the dominance of semi-annual and semi-diurnal variations in SLP, while annual and diurnal variations are dominant for temperature