| 其他摘要 | The Sun is the nearest star to our earth, the only one high-resolution imaging observed and studied in the universe. The Sun does not have a clear boundary,like those of rocky planets. Usually, the photosphere solar radius is de?ned as the distance from the center of the Sun to the outer boundary of the photo- sphere. China was the ?rst country in the world to measure it, which can date back to West Han Dynasty in about 1st century B. C. The solar radius has been measured systematically since 19th century. The standard solar radius value of 959.63” adopted by the IAU was ?rst obtained by Auwers. Since then, a variety of instruments and di?erent methods have been used to measure it: (1)Solar eclipses and Mercury transit, (2) Meridian circle measurements, (3) Driftscan technology, (4) Solar astrolabes, (5) Satellite direct angular measurements.Either from the ground or from orbit. However, the measured results are inconsistent. Some researchers have reported that the solar radius had shown a secular decrease. Other researchers have reported that the solar radius is stationary. Some researchers have found that the radius ?uctuates over periods of a decade days to several years. The thesis, ?rstly, introduces the history measurement of the solar radius and progress study of the solar radius, then several nonlinear analysis approaches are used to study the periodicities of solar radius data ob- tained in the Calern Observatory and Rio de Janeiro Observatory. The purpose of these analyses is to gain some physical understanding of the observation and the main results are listed as follows: 1. In order to study the long period variations of the solar radius, daily solar radius data from 1978 February to 2000 September obtained at the Calern Observatory are used. Continuous measurements of the solar radius are di?cult due to the weather, seasonal e?ects, and instrument characteristics. Thus, to analyze these data, we ?rst use the Dixon criterion to reject outliers, then we measure the cross-correlation between the solar radius and sunspot numbers. The result shows that the solar radius is antiphase with the sunspot numbers and shows lead times of 74 months relative to the sunspot numbers. Two nonlinear analysis methods including Lomb–Scargle and date compensated discrete Fourier transform are also applied to investigate the periodicity of the solar radius. Both methods yield similar signi?cance periodicities around ～ 1 yr, ～ 2.6 yr, and ～ 11yr. We discussed the possible mechanisms for these periods. The possible physical cause of the ～ 11 yr period is the cyclic variation of the magnetic pressure of the concentrated ?ux tubes at the bottom of the solar convection zone. 2. Combining the data of daily solar radius measured at Calern Observatory (R c ) and Rio de Janeiro Observatory (R r ), di?erent nonlinear analysis methods including EMD, DCDFT, CLEANest and Lomb–Scargle are used to study the periodicityies of these two data. The results are consistent with several former peridodicies obtained in other solar radius data. It indicates that di?erent periodicityies do exist in solar radius. On the other hand, spectra analysis for both raw data for no correction and corrected data for seeing e?ect shows that spectra peak are signi?cant. If atmospheric turbulence drives the ground-based solar radius measurements, both spectra can not get similar peaks. Thereby, ground-based solar radius measurements are valuable for probing solar radius variations. 3. We consider the solar radius in conjunction with sunspot areas to discuss the variation of the solar radius, and how magnetic ?elds a?ect the variation of the radius. Both R c and R r demonstrate the rotation period signal with statistical signi?cance by means of the auto-correlation function. In order to determine whether the yearly R ?uctuations within the considered time interval have a periodic component equal to the rotation period, a periodical function, y=p1+p2*sin(2*pi*x/p3+p4), is used to ?t the data of solar radius every year. The correlation coe?cient(cc) between the ?tting data and the original yearly data show that cc are statistically signi?cant in most of considered years. It means that the rotation period signal does clearly exist in R. The result of the correlation analysis between each year’s solar radius and sunspot areas show that all the statistically signi?cant correlation coe?cients are negative, which indicates that the solar radius is anti-phase with sunspot areas. Generally, the correlation coe?cients have obvious statistical signi?cance in the maxima of solar cycles, and their absolute values are larger than these correlation coe?cients of the minima of solar cycles. In the minima of solar cycles, the correlation coe?cients are not statistically signi?cant. This indicates that strong magnetic ?elds have a greater inhibitive e?ect on the radius change than that of weak magnetic ?elds. Surface magnetic ?elds are considered to be channels by which for interior energy can be transferred to the surface, as are convective processes. Magnetic ?elds inhibit convection, causing a decrease of heat transport e?ciency. The strong magnetic ?elds may lead to an increase in magnetic tension force, which would result in compression of the surrounding regions, which, in turn, will shrink the size of the Sun. In the last chapter, a brief summary of this thesis is provided and some unsolved questions about the solar radius are mentioned. And a proposal for future study is made. |
修改评论