| Abstract
| - Regular measurements of the Sun's diameter have been obtained by means of solar astrolabes for several decades. However, the variations found from those measurements are still very controversial, and the results implying changes in the solar radius are not conclusive. Since the quality of the measurements cannot be doubted, it is necessary to wonder what they are actually observing. We show here that some refraction effects, neglected so far, must be taken into account. Then, if the lower atmospheric layers merely add high-frequency noise (validated by the measured Fried parameter), we show here that the mid- to long-term variations take root in the upper troposphere - lower stratosphere region. Hence, the annual means of the Sun's apparent radius, measured at Calern (France) and Santiago (Chile), are linearly correlated with the geopotential height at 50-100 mb with correlation coefficients of −0.66 and 0.62, respectively, the opposite signs being the signature of the symmetry of the two hemispheres, North and South. In addition, the measurement dispersion is correlated with the variance of the wind speed above each observatory. Thus, it appears that the solar signal is amplified by the mechanisms dwelling in this interface zone between the lower stratosphere and the upper troposphere. Because these amplification mechanisms are unknown, it is difficult to extract pure solar information from the astrolabe signals, especially in sites where the seeing is less than 14-16 cm.
|
