Neural networks for several ionogram features
Article first published online: 7 DEC 2012
Copyright 1995 by the American Geophysical Union.
Volume 30, Issue 6, pages 1831–1840, November-December 1995
How to Cite
1995), Neural networks for several ionogram features, Radio Sci., 30(6), 1831–1840, doi:10.1029/95RS02189., and (
- Issue published online: 7 DEC 2012
- Article first published online: 7 DEC 2012
- Manuscript Accepted: 20 JUL 1995
- Manuscript Received: 7 APR 1995
Several demonstration neural networks will be described that identify certain characteristics of ionograms. The characteristics identified are ordinary and extraordinary F region critical frequency, F1 cusp frequency, sporadic E type C and type H cusp frequencies, and spread F. Accuracy for the neural networks varies, ranging from 50% to 90%. The ionograms come from a midlatitude ionosonde station.