| 摘要: | 在這項研究中，我們通過將隨機森林演算法應用於麥哲倫星雲中的米拉變星的 I 波 段光變曲線，來調查和分類米拉變星的性質。米拉變星是已知的漸近巨星分支星，以 其長的脈衝週期和在光學波段中有著較大振幅的特徵。我們的分析將米拉變星分為兩 個不同的組別:常規米拉變星和非常規米拉變星。非常規米拉變星除了其主要脈衝週 期外，其光變曲線還表現出長周期的變化。我們的結果證實了在周光關係中常規氧富 (O-rich)米拉變星中在最大亮度時的周光關係散布較小。這個結果我們認為是未來測 量距離尺度工作中的寶貴工具，並且建議相應地應用。此外，我們收集了這些米拉變 星的多波段光度資料，以進行光譜能量分布 (SED) 擬合。擬合的結果表明，非常規米 拉扁星周圍有相當大的灰塵含量。根據這些結果，我們提出非常規米拉變星中觀察到 的周期性長期變化可能是由於灰塵引起的。 綜上所述，這項研究對麥哲倫星雲中的米 拉星變星的特性提供了新的見解。
此外，我們對於三角座星系M33中1637個已知的米拉變星進行系統性的分析。我 們收集了近18年的數個巡天觀測計畫中的 g,r,i 波段的觀測資料,其中包含泛星計畫 (Pan-STARRs)、帕洛馬瞬變工廠 (PTF)、帕洛馬瞬變工廠中期(iPTF)還有史維基瞬變 探測器 (ZTF). 結果顯示，完整的光變曲線對於計算並確定較遠距離米拉變星的週期 相當重要。此外，還使用了機器學習技術，並根據紅外線波段的週期-顏色的分布將 三角座星系中未分類米拉星進行分類，將其分類為為氧富(O-rich)米拉變星和碳富 (C-rich)米拉變星。最後藉由氧富(O-rich)米拉變星的最大光度時的周光關係去計 算出三角座星系的距離模數為24.67 ± 0.06星等，這個結果符合許多過往文獻所計算出 來的距離模數，因此我們認為最大光度時的周光關係在未來的距離測定研究中有很好 的發展性。;In this study, we employ the random forest algorithm to classify Mira variables and investigate their properties in the Magellanic Clouds using the I-band light curves of these stars. Mira variables, as asymptotic giant branch pulsating stars, are well-known for their characteristic long pulsation periods and large amplitudes in optical bands
Our analysis resulted in the classification of the Miras into two distinct groups: regular Miras and non-regular Miras. The non-regular Miras, unlike the regular Miras, exhibited a more complex variation in their light curves, which included a long-term modulation in addition to their primary pulsation periods. Our results confirm the presence of a period-luminosity relation for maximum light, with a smaller dispersion found in the regular oxygen-rich (O-rich) Miras. This relation is deemed to be a valuable tool in future distance scale work and is recommended to be applied accordingly.
In addition to analyzing the multi-band photometry data was also collected for the Miras to enable spectral-energy-distribution (SED) fitting. The results of the SED fitting indi- cated that a significant fraction of dust is present around the non-regular Miras. Given the results of the SED fitting, it is suggested that the presence of dust may account for non-regular Miras.
Furthermore, we performed a systematic analysis to determine and improve the pulsation periods of 1637 known Mira variables in M33. This was accomplished by analyzing gri- band light curves spanning approximately 18 years from several surveys, including the M33 variability survey, Panoramic Survey Telescope and Rapid Response System (Pan- STARRs), Palomar Transient Factory (PTF), intermediate Palomar Transient Factory (iPTF), and Zwicky Transient Facility (ZTF). The results showed that complete optical band light curves are essential for determining the periods of distant Miras. In addition, machine learning techniques were used to classify the Miras into O-rich and C-rich based
on the (J − Ks) period–color plane.
Finally, we derived the distance modulus to M33 using O-rich Miras at maximum light and our improved periods. The distance modulus was found to be 24.67±0.06 mag, which agrees with the recommended value in the literature. This study provides a comprehensive understanding of the Mira variables in M33, and the results of this analysis will be helpful for future research in this field.understanding of the Mira variables in M33, and the results of this analysis will be helpful for future research in this field. |
