基於機器學習LightGBM架構預測AE地磁指數

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：11

、訪客IP：3.133.141.1

姓名

蔡孟融(Meng-Jung Tsai) 查詢紙本館藏

畢業系所

太空科學與工程學系

論文名稱

基於機器學習LightGBM架構預測AE地磁指數
(Forecasting the AE Geomagnetic Indices with the Machine Learning LightGBM Framework)

相關論文

★ 磁暴與磁副暴的關係：檢視跨磁尾電流對 SYM-H 的貢獻	★ 磁尾的磁場延伸和偶極化現象與磁副暴發生位置的距離關係之探討
★ 二胞型極光與行星際磁場間的關係	★ 磁層頂位置之不對稱性研究
★ 兩類快速電漿流事件與夜側極光活動關係之研究	★ 太陽風對地球磁層頂內側磁場之影響
★ 磁層頂日下點對峙距離和行星際磁場錐角值關係的研究	★ 運用西蜜斯衛星資料研究低頻帶升調合唱波的重複發生週期之分布
★ 太空環境中的兩個觀測難題：前艏震波區域波擾動斜向傳播現象與接觸不連續面的存在證據	★ 徑向行星際磁場事件之特性及其對磁層之影響
★ 太空天氣對Formosat-2及Formosat-3異常事件影響之分析	★ 多能量通道之極區沉降粒子研究
★ 徑向行星際磁場下日側極光與電離層對流型態	★ 水星磁層對行星際磁場與太陽風動壓的反應
★ 應用長短期記憶遞迴神經網路預測Kp地磁指數

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

至系統瀏覽論文 (2025-12-30以後開放)

摘要(中)

極光電噴流(AE)指數通常用於監測極區磁場的擾動程度，可以藉助其發
展程度來判斷磁副暴發生過程，也可以透過AE指數的數值來了解太陽風、磁層以及電離層耦合狀況。本論文使用了1996年至2019年期間，解析度為一分鐘和一小時的太陽風和地磁指數資料來預測可能的地磁變化為提升預測準確度，本研究引入了多種太陽風磁層耦合函數作為輸入參數，包含太陽風能量參數、磁層頂參數、磁重聯率(Rcs)以及極冠指數(PC)，這些參數提供了磁層受到太陽風擾動後的額外資訊，比起只使用太陽風基本參數的多數先前研究，更能有效地預測AE指數?然而加入時間參數，包含年積日還有當地時間更能示AE年日變化。為解決先前研究時間延遲問題，本研究利用LightGBM，一種基於決策樹的梯度提升架構，建立了兩種AE預測模型：ModelL利用滑動窗口建立延遲特徵作為輸入；ModelN 則直接使用耦合與時間參數作為輸入。並透過ModelL 的延遲特徵重要性分析來改進ModelN 參數的輸入時間。結果顯示，在小時尺度下，ModelN 表現最佳(C.C.=0.93)；在分鐘尺度下，ModelL表現最佳(C.C.=0.89)。證實LightGBM模型能有效預測AE指數。此外，本研究發現PC和Rcs在AE指數預測中扮演重要角色，而Rcs具有替代PC的潛力。

摘要(英)

The Auroral Electrojet (AE) indices were created for monitoring geomagnetic field disturbances in the auroral zone. Their variations can be used to determine the development stages of substorms. We can use AE to understand the solar wind-magnetosphere-ionosphere (S-M-I) coupling driven by solar wind forcing. This study uses 1-min and 1-hr solar wind parameters and geomagnetic indices from 1996 to 2019 to construct two AE-predicting models. To improve their prediction accuracy, extra solar wind coupling functions introducedinto the models used as input parameters, including the solar wind energy in put functions, the magnetopause shape parameters, magnetic reconnection rate
(Rcs), and the polar cap index (PC). These functions consider the effects of solar wind in the magnetosphere, enabling them more efficient in predicting AE than the previous models that used only the solar wind parameters as inputs. In addition, incorporating time parameters such as day of year (DOY) and hour allows us to adapt the annual and diurnal variations of AE. Two models were developed in this study: ModelL, which incorporates ”Lagged features”? and ModelN, whichdirectly inputs the coupling and time parameters without includeing Lagged features. Meanwhile, the importance analysis of ModelL’s lagged features is used to determine the lag time for ModelN’s parameters. The results show that ModelN performs best (with a correlation coefficient of 0.93) at 1-hr resolution, while ModelL performs best (with a correlation coefficient of 0.89) at the 1-min resolution, showing that LightGBM is suitable for predicting AE. Furthermore, PC and Rcs are found to play an important roles in predicting AE, and Rcs has the potential to replace PC.

關鍵字(中)

★ AE指數
★ 太陽風-磁層-電離層耦合
★ 機器學習
★ 磁層頂
★ 磁重聯

關鍵字(英)

★ AE indices
★ solar wind-magnetosphere-ionosphere coupling
★ Machine Learning
★ Magnetopause
★ Magnetic Reconnection

論文目次

摘要 xi
Abstract xiii
誌謝 xv
目錄 xvii
圖目錄 xxi
表目錄 xxiii
使用符號與定義 xxv
一、緒論 1
1.1 研究背景 1
1.1.1 磁副暴 3
1.1.2 極光電噴流 6
1.2 AE 研究文獻回顧 8
1.2.1 線性模型與統計方法 8
1.2.2 非線性模型與數值模擬 9
1.2.3 機器學習與自回歸模型 10
1.3 研究動機與目的 13
二、資料與使用參數 15
2.1 資料來源 15
2.1.1 缺失值與特徵縮放 17
2.2 模型輸入參數介紹 18
2.2.1 參數介紹-太陽風參數 vBs 18
2.2.2 參數介紹-磁層頂參數 r0、α 20
2.2.3 參數介紹-極冠指數 PC index 23
2.2.4 參數介紹-行星際參數 ε、EKL 25
2.2.5 參數介紹-磁重聯率 Rcs 27
三、研究方法 31
3.1 機器學習模型介紹 31
3.1.1 決策樹 31
3.1.2 梯度提升決策樹, GBDT 33
3.1.3 LightGBM 35
四、 AE預測結果與分析 39
4.1 延遲特徵模型 (ModelL) 44
4.1.1 ModelhrL 12hr 延遲特徵模型 (小時) 44
4.1.2 Modelmin L 60min 延遲特徵模型（分鐘） 47
4.2 無延遲特徵模型 (ModelN ) 51
4.2.1 ModelhrN - 無延遲特徵模型 (小時) 51
4.2.2 Modelmin N - 無延遲特徵模型 (分鐘) 54
4.3 ModelN 和 ModelL 比較 57
4.4 輸入參數對模型影響 60
4.4.1 基本參數與耦合參數比較 60
4.4.2 耦合參數之間比較-去除磁層頂參數與 ε 61
4.4.3 增加時間參數的物理意義 62
五、討論與結論 67
5.1 討論 67
5.2 結論 68
參考文獻 71
附錄 A 預測指標 77
附錄 B 程式碼 79
B.1 特徵工程-Create Lagged Features 79

參考文獻

Ahn, B.-H. et al. (2000). “Seasonal and solar cycle variations of the auroral electrojet indices.” In: Journal of Atmospheric and Solar-Terrestrial Physics 62.14. ISSN: 1364-6826. URL: https://doi.org/10.1016/S1364-6826(00)00073-0. Akasofu, S.-I. (1964). “The development of the auroral substorm.” In: Planetary and Space Science 12.4, pp. 273–282. ISSN: 0032-0633. URL: https://doi.org/10.1016/0032-0633(64)90151-5. Akasofu, S.-I. (1968). “Polar and magnetospheric substorms.” In: Astrophysics and Space Science 2.3, pp. 91–100. ISSN: 0004-640X. URL: https://doi.org/10.1007/BF00651006. Akasofu, S.-I. (1979). “Relationships between the AE and Dst indices during geomagnetic storms.” In: Journal of Geophysical Research: Space Physics 84.A11, pp. 6660–6662. ISSN: 0148-0227. URL: https://doi.org/10.1029/JA084iA11p06660. Allen, J. H. and H. W. Kroehl (1975). “Spatial and temporal distributions of magnetic effects of auroral electrojets as derived from AE indices.” In: Journal of Geophysical Research 80.25, pp. 3667–3671. ISSN: 0148-0227. URL: https://doi.org/10.1029/JA080i025p03667. Amariutei, O. A. and N. Y. Ganushkina (2012). “Forecasting auroral electrojet AL index using combined autoregressive models and neural networks.” In: Space Weather 10.4, n/a–n/a. ISSN: 1542-7390. URL: https://doi.org/10.1029/2011SW000737. Baker, D. N. et al. (1981). “A high time resolution study of interplanetary parameter correlations with AE.” In: Geophysical Research Letters 8.2, pp. 179–182. ISSN: 0094-8276. URL: https://doi.org/10.1029/gl008i002p00179. Bala, R. and P. H. Reiff (2012). “Improvements in short-term forecasting of geomagnetic activity.” In: Space Weather 10.6, n/a–n/a. ISSN: 1542-7390. URL: https://doi.org/10.1029/2012sw000779. Bala, R. and P. H. Reiff (2014). “Validating the Rice neural network and the Wing Kp real-time models.” In: Space Weather 12.6, pp. 417–425. ISSN: 1542-7390. URL: https://doi.org/10.1002/2014sw001075. Basu, S. (1975). “Universal time variations in auroral zone magnetic activity.” In: Journal of Geophysical Research 80.4, pp. 471–474. ISSN: 0148-0227. URL: https://doi.org/10.1029/JA080i004p00471. Baumjohann, W. and R. Nakamura (2007). “Magnetotail Flows and Substorms.” In: Auroral Dynamics and Space Weather, pp. 79–109. Berthelier, A. (1976). “Seasonal effect on auroral sub storms.” In: Planetary and Space Science 24.9, pp. 881–885. ISSN: 0032-0633. URL: https://doi.org/10.1016/0032-0633(76)90142-7. Borovsky, J. E. (2008). “Solar wind control of the magnetospheric dynamics.” In: Journal of Geophysical Research: Space Physics 113.A8, n/a–n/a. ISSN: 0148-0227. URL: https://doi.org/10.1029/2007ja012684. Burton, R. K., R. L. McPherron, and C. T. Russell (1975). “An empirical relationship between interplanetary conditions and Dst.” In: Journal of Geophysical Research 80.31, pp. 4204–4214. ISSN: 0148-0227. URL: https://doi.org/10.1029/JA080i031p04204. Clauer, C. R. et al. (1981). “A technique for one to five minute prediction of geomagnetic activity.” In: Solar Terrestrial Predictions Proceedings 4, pp. 488–497. Daglis, I. A., R. M. Thorne, and W. J. Hughes (1999). “The global impact of magnetic storms.” In: Space Science Reviews 91.1-2, pp. 289–353. ISSN: 0038-6308. URL: https://doi.org/10.1023/A:1005110128555. Davis, T. N. and M. Sugiura (1966). “Auroral electrojet activity index AE and its relation to polar magnetic disturbances.” In: Journal of Geophysical Research 71.3, pp. 785–801. ISSN: 0148-0227. URL: https://doi.org/10.1029/JZ071i003p00785. Dorelli, J. C. (2019). “A Statistical Study of Dayside Magnetic Reconnection Rates Using the OMNI Solar Wind Database.” In: Journal of Geophysical Research: Space Physics 124.6, pp. 4600–4612. ISSN: 2169-9380. URL: https://doi.org/10.1029/2019JA026741. Frey, H. U. et al. (2003). “Properties of localized high-latitude dayside aurora.” In: Journal of Geophysical Research: Space Physics 108.A4, n/a–n/a. ISSN: 0148-0227. URL: https://doi.org/10.1029/2002ja009345. Gopinath, S. and N. Prince (2019). “Prediction of geomagnetic auroral electrojet index using machine learning models.” In: Advances in Space Research 64.1, pp. 1–10. ISSN: 0273-1177. URL: https://doi.org/10.1016/j.asr.2019.03.013. Grandin, M. (2017). “Magnetospheric Dynamics During Northward IMF: A Global Perspective.” PhD thesis. KTH Royal Institute of Technology. Gu, Y. et al. (2019). “Cloud-NARX model for AE index forecasting.” In: Space Weather 17.1, pp. 154–166. ISSN: 1542-7390. URL: https://doi.org/10.1029/2018sw002088. Haiducek, J. D. et al. (2017). “Physics-based modeling of the auroral electrojet indices using the Space Weather Modeling Framework.” In: Space Weather 15.2, pp. 288–304. ISSN: 1542-7390. URL: https://doi.org/10.1002/2016sw001555. Horton, W. and I. Doxas (1998). “Analysis and prediction of the auroral electrojet index.” In: Journal of Geophysical Research: Space Physics 103.A7, pp. 14619–14638. ISSN: 0148-0227. URL: https://doi.org/10.1029/98JA00826. Janzhura, A. et al. (2007). “On the relationship between the polar cap index PC and auroral electrojet index AE.” In: Journal of Atmospheric and Solar-Terrestrial Physics 69.9, pp. 1028–1037. ISSN: 1364-6826. URL: https://doi.org/10.1016/j.jastp.2007.02.013. Kan, J. R. et al. (1988). “A theory of substorms: Onset and subsidence.” In: Journal of Geophysical Research: Space Physics 93.A6, pp. 5624–5640. ISSN: 0148-0227. URL: https://doi.org/10.1029/ja093ia06p05624. Ke, G. and A. Ferreira (2020). What happens with missing values during prediction? · issue #2921 · Microsoft/LIGHTGBM. GitHub. URL: https://github.com/microsoft/LightGBM/issues/2921. Ke, G. et al. (2017). “Lightgbm: A highly efficient gradient boosting decision tree.” In: Advances in neural information processing systems 30. URL: https://proceedings.neurips.cc/paper_files/paper/2017/file/6449f44a102fde848669bdd9eb6b76fa-Paper.pdf. King, J. H. and V. O. Papitashvili (2006). “Solar wind spatial and temporal scales at 1 AU.” In: Journal of Geophysical Research: Space Physics 111.A8, n/a–n/a. ISSN: 0148-0227. URL: https://doi.org/10.1029/2006JA011739. Liemohn, M. W. et al. (2018). “Geomagnetic indices: A review of their utility and limitations.” In: Space Weather 16.1, pp. 4–24. ISSN: 1542-7390. URL: https://doi.org/10.1002/2017SW001759. Luo, Y. et al. (2013). “Statistical study of the solar wind drivers of the auroral electrojet indices.” In: Journal of Geophysical Research: Space Physics 118.10, pp. 6461–6471. ISSN: 2169-9380. URL: https://doi.org/10.1002/jgra.50549. McPherron, R. L. (1970). “Magnetospheric substorms.” In: Reviews of Geophysics 8.4, pp. 841–880. ISSN: 0034-6853. URL: https://doi.org/10.1029/RG008i004p00841. Mcpherron, R. L. et al. (1973). “Satellite studies of magnetospheric substorms on August 15, 1968, 8. Ogo 5 magnetic field observations.” In: Journal of Geophysical Research 78.19, pp. 3068–3081. ISSN: 0148-0227. URL: https://doi.org/10.1029/JA078i019p03068. Milan, S. E. et al. (2017). “Magnetospheric and ionospheric current systems during substorms.” In: Space Science Reviews 212.3-4, pp. 1533–1608. ISSN: 0038-6308. URL: https://doi.org/10.1007/s11214-017-0420-5. Meng, C.-I. and K. Liou (2004). “Dayside auroral dynamics.” In: Auroral Physics, pp. 125–152. Perreault, P. and S.-I. Akasofu (1978). “A study of geomagnetic storms.” In: Geophysical Journal International 54.3, pp. 547–573. ISSN: 0956-540X. URL: https://doi.org/10.1111/j.1365-246X.1978.tb04783.x. Raschka, S. and V. Mirjalili (2019). Python Machine Learning: Machine Learning and Deep Learning with Python, scikit-learn, and TensorFlow. 2nd ed. Packt Publishing. Rostoker, G. (1996). “Substorms: Their nature and role in space weather.” In: Geophysical Monograph 91, pp. 59–70. Shue, J.-H. et al. (1997). “A new functional form to study the solar wind control of the magnetopause size and shape.” In: Journal of Geophysical Research: Space Physics 102.A5, pp. 9497–9511. ISSN: 0148-0227. URL: https://doi.org/10.1029/97JA00196. Shue, J.-H. et al. (1998). “Magnetopause location under extreme solar wind conditions.” In: Journal of Geophysical Research: Space Physics 103.A8, pp. 17691–17700. ISSN: 0148-0227. URL: https://doi.org/10.1029/98JA01103. Stauning, P. (2012). “The Polar Cap (PC) Index and its relations to other geomagnetic indices.” In: Space Weather 10.1, n/a–n/a. ISSN: 1542-7390. URL: https://doi.org/10.1029/2011SW000697. Takalo, J. and J. Timonen (1997). “Neural network prediction of the auroral electrojet index AE.” In: Geophysical Research Letters 24.1, pp. 35–38. ISSN: 0094-8276. URL: https://doi.org/10.1029/96GL03711. Takalo, J. and J. Timonen (1998). “Prediction of the auroral electrojet index AE using the solar wind speed and the polar cap index PC as input.” In: Geophysical Research Letters 25.18, pp. 3589–3592. ISSN: 0094-8276. URL: https://doi.org/10.1029/98GL02470. Takalo, J. and J. Timonen (1999). “A neural network model for the auroral electrojet index AE.” In: Journal of Geophysical Research: Space Physics 104.A4, pp. 6959–6968. ISSN: 0148-0227. URL: https://doi.org/10.1029/1998JA900103. Troshichev, O. A. and L. V. Andrezen (1985). “The relationship between the polar cap magnetic activity index PC and the auroral electrojet activity indices AE, AL, AU.” In: Planet. Space Sci. 33.4, pp. 415–419. ISSN: 0032-0633. URL: https://doi.org/10.1016/0032-0633(85)90031-2. Troshichev, O. A., N. P. Dmitrieva, and B. M. Kuznetsov (1988). “Polar cap magnetic activity as a signature of substorm development.” In: Space Sci. Rev. 48.3, pp. 277–298. ISSN: 0038-6308. URL: https://doi.org/10.1007/BF00177166. Vennerstrom, S. et al. (1991). “On the use of the polar cap index, PC, in studies of the solar wind-magnetosphere interaction.” In: Journal of Geophysical Research: Space Physics 96.A1, pp. 35–47. ISSN: 0148-0227. URL: https://doi.org/10.1029/90JA01845. Yin, Lu et al. (2021). “JLGBMLoc-A Novel High-Precision Indoor Localization Method Based on LightGBM.” In: Sensors 21.8. ISSN: 1424-8220. URL: https://doi.org/10.3390/s21082722. Zhang, X.-Y. and M. B. Moldwin (2015). “Probabilistic forecasting analysis of geomagnetic indices for southward IMF events.” In: Space Weather 13.3, pp. 130–140. ISSN: 1542-7390. URL: https://doi.org/10.1002/2014sw001113.

指導教授

許志浤(Jih-Hong Shue)

審核日期

2025-1-13

推文