Solar apparent radius variability: a new statistical approach to astrolabe multisite observations

• Published in: Monthly notices of the Royal Astronomical Society Vol. 369; no. 1; pp. 83 - 88
• Main Authors: Badache-Damiani, C., Rozelot, J. P.
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 11-06-2006; Blackwell Science; Oxford University Press
• Subjects: Astronomy; Comparative analysis; Earth, ocean, space; Exact sciences and technology; instrumentation: miscellaneous; Measurement; methods: data analysis; methods: statistical; Stars & galaxies; Sun: oscillations; Sun: oscillations; methods: statistical; instrumentation: miscellaneous; methods: data analysis; Correlation coefficient; Random noise; Variability; Astrolabes; Periodicity; Time series; Irradiance; Time variation; Cross correlations; Persistence; Diameter
• ISSN: 0035-8711; 1365-2966
• DOI: 10.1111/j.1365-2966.2006.10248.x

The results of the analysis of solar diameter measurements obtained by means of the astrolabe have been inconclusive so far. Temporal variations are found, but they differ in amplitude and phase. This paper presents the results of a new statistical approach to compare and analyse the astrolabe data obtained during the period 1998–2003 in Brazil, Chile, France and Turkey. We show that different time series can be considered as extracted from the same statistical population following the Gaussian law. As following on a variographic analysis, the series do not seem to be autocorrelated at small scale (40 d) and present the characteristics of a highly disorganized phenomenon. At medium scale (6 yr), a 1-yr periodicity can be seen. It could be the residual trace of a temporal latitudinal variability. At large scale (up to 30 yr, using only French and Chilean data available, respectively, from 1975 and 1990), a faint 11-yr component in the signal can be pointed out. As an 11-yr periodicity has been found in the stratospheric data, we have explored the possibility of a relationship between the measured radii and stratospheric phenomena. This would explain the discrepancy in amplitudes, the sign of the crossed correlation coefficients, and the apparent disorganized behaviour of the data. Therefore, a complementary non-determinist analysis has been done in order to distinguish a chaotic signal from a random noise. The Hurst exponent establishes the persistence of the series. It can be concluded that the solar measurements from astrolabes are not purely randomly distributed, but show the character of a deterministic series for which (i) the observations need to be far more accurate, as it is only possible if obtained from space or in real-time together with a seeing monitor; (ii) supposing that it is indeed possible to extract a solar component, data must be studied considering their heliographic latitude and (iii) data show the variability of the upper atmospheric conditions (maybe Quasi-Biennial Oscillations) and the stratospheric 11-yr periodicity, for which the modulation is caused by the solar irradiance.