NEW (Sept. 21) — more detailed results of the Eclipse Experiment have been published by the sponsors — see this PDF document.
Both professional and citizen scientists conducted formal and informal investigations into the effect of the August 21 solar eclipse on HF radio propagation. Nathaniel Frissell, W2NAF, of HamSCI, has said it will take some time to get a more scientific analysis of data that was compiled during the Solar Eclipse QSO Party. He and others are investigating whether the sudden absence of sunlight during the eclipse — and especially of solar ultra-violet and x-rays — would briefly change the properties of the upper atmosphere.
Professional ionospheric researcher Dr. Phil Erickson, W1PJE, head of the Atmospheric Sciences Group at MIT’s Haystack Observatory, said he can say categorically that there was a definite, large, and measurable effect in the ionosphere from the eclipse.
“We saw a 2X reduction in electron density during the eclipse for at least 45 minutes to 1 hour,” Erickson told ARRL. “This reduction had direct impacts on HF propagation along the bottom side.” Erickson said many models and observations exist from previous eclipses that demonstrate these effects. Erickson said MIT researchers used a “megawatt-class Thomson scatter radar,” which can directly measure the plasma state of the ionosphere, including electron density, across a huge area in the eastern US.
“Scientists in the worldwide space physics community will be using these and many other eclipse observations to learn more about our ionosphere, space weather, and its effects on navigation and communication signals including Amateur Radio,” Erickson said. He has shared his data with the HamSCI team.
Bob Reif, W1XP, was in southeastern North Carolina where the eclipse totality was about 97% with two radios running multiple bands of WSPR. “Lots of data to look at,” he said, “but what jumped out was that at almost the exact time of maximum coverage of the sun at this location, 160 meters opened for about 30 minutes and then closed down again until the normal gray line. So, the D layer responded to the shadow of the Moon to some extent.”
Gene Greneker, K4MOG, in Georgia told ARRL that his own eclipse experiment “worked out rather well.” He set up an RFSpace NetSDR receiver at his location, locked to a 10 MHz GPS standard and tuned to WWV on 15 MHz. “The NetSDR provides in-phase and quadrature components of the WWV signal, which allows relative signal phase to be reconstructed from recorded data,” Greneker explained in a more-detailed account. “Signal phase-vector rotation change was chosen to indicate totality arrival, because phase is very sensitive to any change in propagation path length, possibly caused by ionospheric movement, up or down. Solar presence or absence can cause vertical ionospheric movement.”
Greneker recording the phase of the 15-MHz WWV signal from 1800 UTC until 1900 UTC on August 21. “Each time the path length changes by 1 wavelength, there is a 360° change in the phase of the signal,” he said. Greneker offset the path of totality to the south, running parallel to the path between WWV in Fort Collins, Colorado, and his location in Atlanta.
Greneker assumed that the reflection point off the ionosphere was south of Kansas City, and, he said, the minimum dip in the phase record occurred very close to totality at that location. “At 1809 hours UTC, when the totality point was parallel to the midpoint of the propagation path, the path length increased from zero wavelengths to 157 wavelengths, or 3,140 meters, during the intervening 9 minutes,” he reported. As totality moved southeast, solar radiation began to increase and the path length decreases as the ionospheric reflecting point moves downward.
Bob Skaggs, KB5RX, told ARRL he spent about 4 hours in the central part of Mission Valley, Montana, listening to conversations on various 20-meter frequencies with a low antenna. “At maximum of the eclipse, the propagation went almost to nothing for maybe about 15 or 20 minutes,” he said. “As the eclipse receded, signals came back up.” Skaggs tried 17 meters for 5 minutes at 1800 UTC and heard “no signals at all.”
He also said the local animal population responded to the eclipse as if evening were approaching.