| 其他摘要 | The variation of the total solar irradiance (TSI) is a basic characteristic of the Sun itself; to a great extent, it shaped the solar-terrestrial space environment and the structure of the Earth's upper atmosphere. TSI varies on all timescales on which it has been observed, and there is increasing evidence that it is one of the key factor that influence the climate. Early measurements of TSI on the ground are all subject to the obstruction of the Earth's atmosphere, and the real value of the TSI can not be obtained; therefore, it is impossible to investigate its variations. Since late 1978, observations of TSI with electrically calibrated radiometers from space have provided overlapping time series from different missions up to the present. This allows the construction of a continuous record for nearly 40 years. The measurements show that TSI varies on all timescales on which it has been observed, and the main drivers of TSI variability are thought to be magnetic features at the solar surface. In this thesis, data of sunspot numbers, surface magnetic flux, and TSI observations are used to study the variation of the photospheric magnetic field, as well as its influence on TSI. The main findings are summarized as below:1. Phase relations of the sunspot numbers (SN) and two TSI composite data are investigated. It is found that TSI and SN are positively correlated, and the former lags the latter by about 29 days which is about a solar rotation period; analyses of the four individual cycles show that, in solar cycle 21, 23, and 24, TSI lags SN by 28.9?30.3 days, while in solar cycle 22, the lag is only 21.8?22.3 days. The abnormality in solar cycle 22 is probably caused by its stronger magnetic field in sunspots compared with its adjacent cycles. The nonlinearity between TSI and SN is confirmed and interpreted here as the combination of the short-term effect of sunspots and faculae, and the long-term effect of magnetic network. The non-linear relation is further studied with the cross-wavelet transform and wavelet coherence methods. The common periodicity of TSI and SN at timescale of the solar cycle is clearly revealed, and high common power at timescale of the rotational period only intermittently appears around solar maxima. The variations of TSI and SN indicates significant coherence at timescales larger than about 4 years. Both the cross-wavelet transform and wavelet coherence analyses shown a high level of phase synchronization between TSI and SN with coherent phase angles in the low- frequency components corresponding to period scales around the dominating 11-year cycle, and their high-frequency components show a noisy behavior with strong phase mixing.2. The rotational characteristics of the solar photospheric magnetic field at four flux ranges are investigated together with the total flux of active regions (MF$_{ar}$) and quiet regions (MF$_{qr}$). The first four ranges (MF$_{1-4}$) are $(1.5-2.9)\times10^{18}$, $(2.9-32.0)\times10^{18}$, $(3.20-4.27)\times10^{19}$, and $(4.27-38.01)\times10^{19}$, respectively (the unit is Mx per element). Daily values of the flux data are extracted from magnetograms of the Michelson Doppler Imager on board the \textit{Solar and Heliospheric Observatory}. Lomb-Scargle periodograms shown that only MF$_2$, MF$_4$, MF$_{qr}$, and MF$_{ar}$ exhibit rotational periods. The periods of the first three types of flux are much very similar, i.e., 26.20, 26.23, and 26.24 days, respectively, while that of MF$_{ar}$ is larger, 26.66 days. This indicates that active regions rotate more slowly than quiet regions on average, and strong magnetic fields tend to repress the surface rotation. Sinusoidal function fittings and cross-correlation analyses reveal that MF$_{ar}$ leads MF$_2$ and MF$_4$ by 5 and 1 days, respectively. This is speculated to be related with the decaying of active regions. MF$_2$ and MF$_{ar}$ are negatively correlated, while both MF$_4$ and MF$_{qr}$ are positively correlated with MF$_{ar}$. At the timescale of the solar activity cycle, MF$_{ar}$ leads (negatively) MF$_2$ by around one year (350 days), and leads MF$_4$ by about 3 rotation periods (82 days). The relation between MF$_2$ and MF$_{ar}$ may be explained by the possibility that the former mainly comes form a higher latitude, or emerges from the subsurface shear layer. We conjecture that MF$_4$ may partly come from the magnetic flux of active regions; this verifies previous results that were obtained with indirect solar magnetic indices.3. Solar photospheric magnetic field plays a dominant role in the variability of total solar irradiance (TSI). The modulation of six categories of magnetic flux at specific ranges on TSI is characterized for the first time. Daily flux values of magnetic field at four ranges (MF$_{1-4}$) are extracted from MDI/SOHO, together with daily flux of active regions (MF$_\textrm{ar}$) and quiet regions (MF$_\textrm{qr}$). Cross-correlograms show that MF$_4$, MF$_\textrm{qr}$, and MF$_\textrm{ar}$ are positively correlated with TSI, while MF$_2$ is negatively correlated with TSI; the correlations between MF$_1$, MF$_3$ and TSI are insignificant. The correlation coefficient between MF$_4$ and TSI (0.76) is larger than that of MF$_\textrm{ar}$ (0.70) and MF$_\textrm{qr}$ (0.64), and it is further confirmed with the bootstrapping test, which implies that the impact of MF$_4$ on TSI is dominant. MF$_\textrm{ar}$ leads TSI by a solar rotational cycle, which is also confirmed with the bootstrapping test. The $\sim$27-day rotational variations in the flux data sets and TSI are clearly identified with the mode decomposition method, and the modulation of the magnetic flux on TSI is characterized at the rotational timescale. During solar maximum time, the relative amplitudes of the flux and TSI are much larger than that during solar minimum time. The negative correlation between MF$_\textrm{ar}$ and TSI at solar maximum time implies the instant darkening effect of sunspots. During solar minimum time, both MF$_\textrm{ar}$ and MF$_4$ are positively correlated with TSI. The impact of MF$_2$ on TSI is found to be similarly negative during either maximum or minimum at different timescales. The impact of a unique active region on TSI is depicted with the 2008 outburst. The amplitudes of the long-term variations of MF$_4$ and MF$_\textrm{ar}$ are in good agreement with TSI, while that of MF$_2$ lags TSI by about 1.5 years. These phenomena are explained with the evolution and property of magnetic features on the solar surface. These results deepen our understanding of the relation between solar surface magnetic field and TSI. |
修改评论