Trading on Schumann resonances with a neural network. chapter and author info
From the many transients that were observed at the Astronomical and Geophysical Observatory AGO of Comenius University near Modra, western Slovakia, in the vertical electric field component mainly during May and June ofa peculiar group of events could be recognized. According to the waveform analysis, these peculiar events in most cases consist of two overlapping transients with a characteristic time difference of 0.
The natural origin of the peculiar events was verified from the NCK data and the source location was determined from the second transient.
The results suggest that the two consecutive transients originated in the same thunderstorm. Furthermore, the phase spectrum analysis indicates that the sources have coherently excited the Earth-ionosphere cavity. These findings seem to support the idea that electromagnetic waves orbiting the Earth might trigger lightning discharges.
The possibility that electromagnetic waves may trigger discharges was first considered by Nikola Tesla. The modest attenuation at ELF 0. In this study, nine ELF receiving sites in Antarctica 3 sitesHungary, India, Japan, Poland, Spitsbergen and USA are used to provide power spectra at minute intervals in two absolutely calibrated magnetic fields and occasionally, one electric field, with up to five resonance modes each.
The observables are the extracted modal parameters peak intensity, peak frequency and Q-factor for each spectrum. The unknowns are calculated from the observables by the iterative inversion of an evolving 'sensitivity matrix' whose elements are the partial derivatives of each observable for all receiving sites with respect to each unknown quantity.
The propagation model includes the important day-night asymmetry of the natural waveguide.
To overcome the problem of multiple minima common in inversion problems of this kindlocation information from the World Wide Lightning Location Network has been used to make initial guess solutions based on centroids of stroke locations in each chimney.
Results for five consecutive days in Jan show UT variations with the African chimney dominating on four of five days, and America dominating on the fifth day.
Analysis of Financial Time Series in Frequency Domain Using Neural Networks
The amplitude variations in absolute source strength exceed that of the 'Carnegie curve' of the DC global circuit by roughly twofold.
Day-to-day variations in chimney source strength are of the order of tens of percent.
Examination of forward calculations performed with the global inversion solution often show good agreement with the observed diurnal variations at.