Use of long-term prediction in automatic vertical-incidence ionogram processing

Of decisive importance in diagnosing the ionospheric channel is the use of vertical-incidence sounding (VS) ionograms. The ionogram processing problem involves identifying characteristic points corresponding to actual signal modes, followed by approximating them in the form of traces. Within the framework of this problem, it is necessary to take the following steps: (1) to carry out a preprocessing of ionograms; (2) a compression of the data resulting in a substantial reduction in their amount and the identification of the times of arrival of the signal; and (3) a combination of the points, thus identified, into traces and referencing them to particular propagation modes. The solution of the first problem is achieved through use of statistical techniques for processing images. The cellular automaton is proposed in order to solve the second problem. The third problem is tackled by implementing Hopfield's method of artificial neural networks. As a consequence of the complicated character of the VS ionograms and due to the presence of scattered signals, errors are possible both when determining the times of arrival of signals and, especially, when combining them into traces. To solve this problem we suggest that use should be made of predicted values of critical values (foE, foF2) and heights h'F of regular layers obtained from a real-time version of the ionospheric model. An analysis of results made on the basis of the data from the ISTP chirp-sonde (Irkutsk) showed that this approach yields good results.© (2000) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.