An automatic scaling method for obtaining the trace and parameters from oblique ionogram based on hybrid genetic algorithm. Issue 12 (9th December 2016)

• Record Type: Journal Article
• Authors: Song, Huan; Hu, Yaogai; Jiang, Chunhua; Zhou, Chen; Zhao, Zhengyu; Zou, Xianjian
• Is Part Of: Radio science. Volume 51:Issue 12(2016:Dec.)
• Page Start: 1838
• Page End: 1854
• Publication Date: 2016-12-09
• Subjects: oblique ionogram -- automatic scaling -- hybrid genetic algorithm; Radio meteorology -- Periodicals; Radio wave propagation -- Periodicals; 621.38405
• DOI: 10.1002/2016RS005987
• Languages: English
• ISSNs: 0048-6604

Abstract

Abstract: Scaling oblique ionogram plays an important role in obtaining ionospheric structure at the midpoint of oblique sounding path. The paper proposed an automatic scaling method to extract the trace and parameters of oblique ionogram based on hybrid genetic algorithm (HGA). The extracted 10 parameters come from F 2 layer and Es layer, such as maximum observation frequency, critical frequency, and virtual height. The method adopts quasi‐parabolic (QP) model to describe F 2 layer's electron density profile that is used to synthesize trace. And it utilizes secant theorem, Martyn's equivalent path theorem, image processing technology, and echoes' characteristics to determine seven parameters' best fit values, and three parameter's initial values in QP model to set up their searching spaces which are the needed input data of HGA. Then HGA searches the three parameters' best fit values from their searching spaces based on the fitness between the synthesized trace and the real trace. In order to verify the performance of the method, 240 oblique ionograms are scaled and their results are compared with manual scaling results and the inversion results of the corresponding vertical ionograms. The comparison results show that the scaling results are accurate or at least adequate 60–90% of the time. Key Points: Trace and 10 parameters from oblique ionograms are scaled automatically QP model, secant theorem, Martyn's equivalent path theorem, and hybrid genetic algorithm were used Scaling results of 240 oblique ionogram were presented