IT has been predicted that the supersonic movement of the Moon's shadow during a solar eclipse will generate internal gravity waves in the atmosphere1. The bow wave so produced by the Moon's shadow will be manifested by travelling ionospheric disturbances (TIDs). During the March 7 eclipse last year some of the observations at Fort Monmouth, NJ, were devoted to TID observations using measurements of the integrated electron content of the atmosphere, obtained from the polarization of the beacon signal from the ATS-3 satellite at 137.35 MHz. Detection of the predicted motions, by variations in the electron content, depends on achieving sufficient precision in relation to an unknown effect. We therefore carried out an error analysis from more than 1,100 single measurements taken in the afternoon of March 7 (after the eclipse had finished. We found a maximum error of ±5.5 polarization degrees inherent in our new automatic system for the data points in Fig. 1, each of which is averaged over twenty-eight single digitized data points taken within four seconds at intervals of three minutes. This averaging is necessary and compensates for the spin effect on polarization. The same maximum error was observed experimentally when the fluctuations of electron content were studied during ionospherically quiet days. The data of Fig. 1 were obtained using an effective geomagnetic component of constant value, neglecting the unknown variations of this factor. The data were analysed for undulations by taking the differences between actual data and their running mean taken over nine data points, over twenty-seven minutes, in other words. The maximum error for these differences is then ±11 polarization degrees. For reasons of improved reliability, we assumed a maximum error of ±15 degrees or ±3.5×1011 electrons cm−2 for the variations in content deduced during the eclipse (Fig. 2).
Read full abstract