Detecting Tornadoes by Radio

Abstract

The tornado season is now at its peak, and Plains and Midwestern States are being swept by dozens of the destructive funnels. Each year, more than 600 tornadoes materialize th United States, taking more lives than February's San Fernando earthquake. Last year, a total of 650 tornadoes caused 73 deaths. This toll is comparatively mild: On March 18, 1925, one of the great killers of all time ripped a 200-mile path through Missouri, Illinois and Indiana, killing 700 people. Several years ago, a number of observers-both amateur and professional -reported that by tuning their television sets to a channel for which there was no local station and turning the screen dark, they had picked up distinctive signals from tornadoes the vicinity. The most obvious interpretation of this ability of television sets to tune in a tornado, point out three scientists at Iowa State University, is that the storms emitted radio noise pulses the very-high-frequency (VHF) range of the electromagnetic spectrum. Dr. John L. Stanford, Michael A. Lind and Gene S. Takle therefore embarked upon a study, supported by the university's physics department, of the electromagnetic pulses generated by severe storms. It has been known for years that thunderstorms create electromagnetic noise; this is the cause of radio static during a storm. But the Iowa State scientists hoped to find identifiable differences between the signals emitted by different types of severe weather. The chief goals of the study are to find out if some fraction of tornadoes (or other severe weather such as hailstorms and windstorms) does indeed produce electromagnetic radiation different from that of thunderstorms, and if so, how this tornado radiation may be distinguished. The Iowa scientists also want to discover what fraction of tornadoes do not emit characteristic radiation, and, conversely, what fraction of tornado-like signals would be false indicators of a tornado. Finally, the three hope to uncover the physics of the phenomenon. Thunderstorms producing tornadoes often display unusually intense lightning, and a number of authors have discussed the electrical phenomena associated with tornadoes. There are many reports on electromagnetic noise generation at very low frequencies, and it has been shown that storm noise energy this region of the spectrum is due primarily to large-scale electrical strokes and is not as good an indicator of tornadic activity as higher frequencies. The researchers therefore set up an antenna system to monitor higher electromagnetic frequencies-670 kilohertz, 53 megahertz and 144 megahertz. The researchers recorded the electromagnetic signals received by the antenna on magnetic tape and then counted the pulse rates and charted them against time to give pulse rate per second. The results of their first year of observations, from six storms of the 1970 tornado season, were summarized recently the JOURNAL OF THE ATMOSPHERIC SCIENCES. The data reveal several possibly different types of electromagnetic radiation phenomena from storms: Radiation from ordinary lightning strokes creates isolated, high-amplitude peaks pulse rate. Another type is a gradual build-up pulse rate over a 15to 30-minute period. This is probably connected with convective processes occurring while the storm is building. In a third type, pulse-rate peaks occur almost periodically at intervals of about 10 minutes. These patterns appeared to be associated time with observed severe weather events and