The dominant natural electromagnetic (EM) radiation in the extremely low frequency (ELF) range is due to global lightning activity. Radio waves of ELF band traveling along the surface of the ground are able to circle the globe and return to the starting point. Schumann Resonances (SR) are the EM phenomena which occur in the cavity formed by the conducting Earth and the ionosphere, with peak frequencies close to 8, 14, 20, 26 Hz, etc. The spectral characteristics of the SR modes are defined by their resonant mode amplitudes, center frequencies and half-widths. The characteristics of the SR became important in aerospace, marine applications, atmospheric studies, in addition to their relevance to global lightning studies due to their frequency band. The objective of this work is two fold: (i) to investigate the characteristics of SR parameters obtained at Sarkoy in Turkey / (ii) to model the nonlinear characteristics of the Near Earth Space Processes by forecasting the 1st SR mode intensities different time steps in advance using neural network modeling approach. The results show that the SR amplitudes exhibit the characteristics of Tropical African lightning activity and have maxima around 1400 UT. The neural network results show that the proposed model is able to forecast SR amplitudes from 0,5 to 36 hours in advance within reasonable error limits. Furthermore, a fuzzy neural network model with a non&ndash / linear optimization algorithm for the training phase is proposed and tested for the future work.
Identifer | oai:union.ndltd.org:METU/oai:etd.lib.metu.edu.tr:http://etd.lib.metu.edu.tr/upload/12608781/index.pdf |
Date | 01 September 2007 |
Creators | Altuntas, Emre |
Contributors | Tulunay, Yurdanur |
Publisher | METU |
Source Sets | Middle East Technical Univ. |
Language | English |
Detected Language | English |
Type | M.S. Thesis |
Format | text/pdf |
Rights | To liberate the content for public access |
Page generated in 0.0018 seconds