Novel data stream techniques for real time HF radio weather statistics and forecasting

The HF radio environment often exhibits dramatic changes that span over many different time scales, ranging from milliseconds (burst radio transmissions) to about eleven years (the solar cycle). It is desirable to obtain statistics of this HF "radio weather" in order to make reliable forecasts. We have used real time data from a vector antenna in Uppsala, Sweden, to measure intensity, polarization, and direction of arrival of sources in the 100 kHz to 20 MHz frequency range. The digitized antenna signals are distributed over the internet and the real time processing is performed using a novel streaming data base manager, System S from IBM. The initial results of this research project are presented and its future applications are discussed.