為了從電離層探測儀觀測到的電離圖中找出電離層參數與其對應的電子密度分布，本文將提出一套新的自動判讀演算法。在電離圖資料處理部份，將背景雜訊移除，並利用影像處理方式擷取正、異常波訊號軌跡以及得到電離層參數。本文將使用類拋物線分層模型(Quasi-Parabolic Segment, QPS)作為電子密度分布反演的理論，其初始參數將由國際電離層參考模型(International Reference Ionosphere, IRI)的電離層參數以及電離圖資料判讀的參數所提供。以初始參數作為參考並進行參數調整設定，即可得到上千組電離層參數，利用QPS model模擬出的電子密度分布進行實高分析，得到正常波訊號的合成軌跡。利用正常波合成軌跡與實際電離圖正常波軌跡的相互比較，找出所有電離層參數組中相關係數最大值和均方根誤差(RMSE)最小值的參數組當作最後的電離圖判讀結果。本文使用甘肅地區的原始電離圖資料進行自動判讀演算法分析，其判讀電離層參數之結果將對精準度和誤差部分進行討論。

In order to precisely estimate ionospheric parameters with corresponding electron density profiles form the ionogram observed by this ionosonde system, a new algorithm is developed to scale the ionogram traces. We first remove background noise from the ionogram. We use an image processing technique to identify the observed O- and X-wave traces, then obtain preliminary ionospheric parameters based on separated O- and X-wave traces. In this study, the quasi-parabolic segment (QPS) model is exploited to obtain the electron density profile, and the IRI model is also used to generate the initial inputs. According to the scaled ionospheric parameters from the O-wave traces combined with the inputs of the IRI model, it is easy to retrieve the QPS model-based electron density profiles. From the retrieved electron density profiles, the corresponding O-wave traces can be simulated and compared with the observed one. The root mean square errors (RMSEs) between observed and simulated O-wave traces are validated and the ionospheric parameters of the constructed trace with the minimum RMSE value and the best correlation-coefficient are then selected as the final result of the ionogram scaling. The accuracy and precision of the estimated ionospheric parameters by using the proposed algorithm are also analyzed and discussed.

(1) Croft, T. A., and H. Hoogasain (1968). Exact Ray Calculation in a Quasi-Parabolic Ionosphere With no Magnetic Filed. Radio Science, Vol. 3, No. 1.
(2) Davies, K. (1989). Ionospheric radio wave propagation. London, U. K.: P. Peregrinus on behalf of the Institution of Electrical Engineers.
(3) Dungey, J. W. (1956). The influence of the geomagnetic field on turbulence in the ionosphere. Journal of Atmospheric and Terrestrial Physics, 8, p. 39.
(4) Gonzalez, R. C., R. E. Woods and S. L. Eddins (2010). Digital Image Processing Using Matlab. Tata McGraw Hill Education Private Limited, pp.440-488
(5) Galkin, I. A. and B. W. Reinisch (2008). The New ARTIST 5 for all Digisondes. University of Massachusetts Lowell Center for Atmospheric Research, 600 Suffolk Street, Lowell, MA 01854
(6) Jiang, C. H., G. B. Yang, Z. Y. Zhao, Y. N. Zhang, P. Zhu and H. Q. Sun (2013). An automatic scaling technique for obtaining F2 parameters and F1 critical frequency from vertical incidence ionograms. Radio Sci., 48, pp. 739-751. doi:10.1002/2013RS005223
(7) Jiang, C. H., G. B. Yang, Z. Y. Zhao, Y. N. Zhang, P. Zhu, H. Q. Sun and C. Zhou (2014). A method for the automatic calculation of electron density profiles from vertical incidence ionograms. Journal of Atmospheric and Solar-Terrestrial Physics, 107, pp. 20-29.
(8) Kelley, M. C. (2009). The Earth's ionosphere. Academic Press.
(9) Lan, T., C. H. Jiang, G. B. Yang, Y.N. Zhang and Z. Y. Zhao (2017). Investigation of automatic scaling of the F2 layer for film ionograms. Progress in Geophysics (in Chinese), 32(1), pp. 56-65. doi:10.6038/pg20170107
(10) Ma, L. Y., Y. Sun, N. Z. Feng and Z. Liu (2000). Image Fast Template Matching Algorithm Based on Projection and Sequential Similarity Detecting. IEEE. doi:10.1109/IIH-MSP.2009.94
(11) Norman, R. J. (2003). An Inversion Technique for obtaining Quasi-Parabolic layer parameters from VI Ionogram. Australia: Radar Conference. Proceedings of the International. IEEE Press.
(12) Pezzopane, M. and C. Scotto (2005). The INGV software for the automatic scaling of foF2 and MUF(3000)F2 from ionograms: a comparison with ARTIST 4.01 from Rome data. Journal of Atmospheric and Solar-Terrestrial Physics, 67(12), pp. 1063-1073.
(13) Pezzopane, M.and C. Scotto (2008). A methodfor automatic scaling of F1 critical frequencies from ionograms. Radio Sci., 43. doi:10.1029/2007RS003723
(14) Reinisch, B. W. and X. Huang (1983). Automatic calculation of electron density profiles from digital ionograms: 3. Processing of bottomside ionograms. Radio Sci., 18(3), pp. 477-492.
(15) Scotto, C. and M. Pezzopane (2002). A software for automatic scaling of foF2 and MUF(3000) F2 from ionograms. In Proceedings of URSI 2002 (pp. 17-24). Maastricht.
(16) Scotto, C., M. Pezzopane and B. Zolesi (2012). Estimating the vertical electron density profile from an ionogram: On the passage from true to virtual heights via the target function method. Radio Sci., 47. doi:10.1029/2011RS004833
(17) Sun, H. Q., G. B. Yang, X. Cui, P. Zhu and C. H. Jiang (2015). Design of an Ultrawideband Ionosonde. IEEE Geoscience and Remote Sensing Letters, Vol. 12, No. 5
(18) Titheridge, J. E. (1985). Ionogram analysis with the generalised program Polan. Boulder: Report UAG-93. World Data Center A for Solar-Terrestrial Physics
(19) Tsai, L. C. and F. T. Barkey (2000). Ionogram analysis of ionograms: A generalized formulation. Radio Sci., 35(5), pp. 1173-1186. doi:10.1029/1999RS002170
(20) UAG-23A. (1972). URSI Handbook of Ionogram Interpretation and Reduction.
(21) Zabotin, N. A., J. W. Wright and G. A. Zhbankov (2006). NeXtYZ: Three-dimensional electron density inversion for dynasonde ionograms. Radio Sci., 41. doi:10.1029/2005RS003352
(22) 吳剛宏，改進GPS電波掩星法反演電離層E層電子濃度之誤差:COSMIC觀測與IRI模型模擬，國立中央大學太空科學研究所博士論文，2015
Digital ionogram Database http://umlcar.uml.edu/DIDBase/
Space Weather Service http://www.sws.bom.gov.au
Mathworks https://www.mathworks.com/help/images/ref/medfilt2.html
International Reference Ionosphere http://irimodel.org/