電離層F2 層臨界頻率（foF2）之自動判讀改進與正確性判別

Automatic Scaling of Critical Frequency of F2 Layer and the Accuracy Testing of Results

張雅莉(Ya-li Chang)

指導教授 ： 蔡龍治

國立中央大學/地球科學學院/太空科學研究所/碩士(2012年)

摘要

動態式電離層觀測儀所觀測到的電離圖，經過判讀系統的處理後可以得到一系列電離層的相關參數，透過分析這些參數可以了解電離層的概況。以模糊理論為概念發展的自動化電離圖判讀程式處理動態式電離層觀測儀花蓮站的觀測資料時，電離圖被正確判讀的比率在六成以下；經改良的foF2判讀程序是透過將電離層F層回波訊號單獨分離出來，再取其頻率最大值的方式決定foF2之值，此方法可將foF2之判讀結果提高到八成以上，亦能從自動化判讀結果看出foF2的日變化。

關鍵字

電離圖 ； 動態式電離層觀測儀 ； foF2 ； 自動判讀

並列摘要

The topic of this thesis is to improve the ionograms scaling results of the critical frequency of F2 layer (foF2).
Comparing the foF2 scaling results between the automatic scaling program by using a fuzzy segmentation and a connectedness algorithm and by human eyes, the absolute values of differences between two scaling results less than 1 MHz are less than 60% for the chosen 2304 ionograms.
The improved automatic scaling program filters the echo not belonging to the F layer, and then using the F layer echoes to determine the foF2. The absolute values of differences between two scaling results less than 1 MHz are more than 85%, better than the previous program. This thesis also discusses the accuracy of the scaling results.
The 2304 ionograms were obtained by the Dynasonde that was established at Dong-Hwa University in Hualien, Taiwan since 2010.

並列關鍵字

foF2 ； Autoscaling ； Dynasonde ； Ionograms

參考文獻

[2] Ding, Z.,Ning, B.,Wan, W.,Liu, L., Automatic scaling of F2-layer parameters from ionograms based on the empirical orthogonal function (EOF) analysis of ionospheric electron density, Earth Planets Space, 59, 51–58, 2007.

[7] Kelley, M. C. , The earth''s ionosphere: plasma physics and electrodynamics, Academic Press, Inc., 1989

[8] M. Hagenbuchner, J. Fulcher, Noise removal in ionograms by neural network, Neural Comput & Applic, 6:165-172, 1997.

[9] Pezzopane, M., Scotto, C., Highlighting the F2 trace on an ionograms to improve Autoscala performance, Computers & Geosciences 36, 1168–1177, 2010.

[11] R. Fisher, J. Fulcher, Improving the inversion of ionograms by combining neural network and data fusion techniques, Neural Comput & Applic, 7:3-16, 1998.