Page 34 January February 2018 TCA

Figure 7: VA3ROM WSPR HUGE number – the digital equivalent of
Mid-eclipse 80m SHOUTING! Ask any Morse code operator
Band Opening. if a 6 dB SNR change up/down makes any
Courtesy: WSPRnet.org
difference to them.
Results from many other As for commercial broadcast AM radio
independent sources are
needed to prove/disprove band reception – quoting from my notes –
any hypothesis. “Four stations received between 900 and
1000 kHz after mid‑eclipse, but identified
Figure 8A is a graphical only one across Lake Superior (100 miles
analysis of the Reverse southeast) – WMPL 920 Hancock, Michigan.
Beacon network (RBN) The others weren’t readable enough to
Morse code data, as make out any station details because they
compiled by the Ham were riding the atmospheric noise, fading
Radio Science Citizen in and out. Most likely Great Lakes region
Investigation (HamSCI) stations to my south southeast.”
group, during the eclipse.
One of the HamSCI EXPLANATION
founders, is newly minted So why did this particular solar eclipse
PhD Nathaniel Frissell, affect propagation right across North
W2NAF, who wrote his America? First, it was a transcontinental
dissertation on using eclipse; and second, propagation is
Amateur Radio data affected by something called the total
networks to assist electron count (TEC) high above your
scientists, like himself. head in the ionosphere. It’s created by
In this case, thousands of solar UV (ionizing) radiation literally
real‑time reports (spots) blasting electrons free from atoms in the
clearly showed what upper atmosphere – turning atoms into
happened during the were removed, but all beacons, regardless positive ions (hence the name). The freed
eclipse: the 20 metre band behaved like it of power levels, were kept. The beacons’ electron particles interact with the Earth’s
would during late night, and signals were received signal‑to‑noise ratios (SNR or magnetic field (magnetosphere) creating
no longer being refracted back to earth S/N), as determined by K9AN, were negatively charged clouds of varying
because they now travelled up and out to analyzed with a scatter plot, which is used densities, “lumpiness”, energy levels and
the stars, while the low bands displayed to identify the type of relationship (if any) tilt angles that refract (or not) radio waves
typical night‑time propagation by travelling between two quantitative variables. The (photons) of specific wavelengths
farther up into the ionosphere before results produced less spectacular looking travelling though them.
being refracted. But there’s a slight graphs (see Figure 8B), but now they can The TEC changed by an amazing 40%,
problem with this graph – dramatic as it is give us comparative dB values. Their creating low band openings and high
– you can’t get any empirical values from (blue horizontal) trend lines, showing the band closings following along the path of
it. There appears to be some kind of general direction a group of points seem totality. The propagation effects expanded
reciprocal relationship going on with the to be heading, indicate the 20 metre band outwards, but weakened with increasing
data set, but how do we quantify it as a had about a 6 dB decrease in SNR, while distance, to well over 1000 kilometres.
number, preferably in dB, for any time the 80 metre band had about a 6 db
during the eclipse? increase (a power factor of four). TEC is affected by the Earth’s diurnal
cycle (higher on daylight side), solar cycle
I decided to analyze the corresponding Voice only operators would (probably) (higher with more sunspots), and
WSPR beacon data because these say this is only a 1 S‑unit increase/ geomagnetic storms (solar “wind”
automatons generate reliable, regular decrease and barely noticeable interaction with the magnetosphere).
repeating signals with embedded digital (perhaps), but in the data world it’s a
data (call, grid and power). Steve Franke,
K9AN’s impressive 24/7 WSPR streaming
station was used as the data source
because he’s almost due south of me
(same solar time zone) and was just north
of the path of totality.
K9AN’s eclipse spots were extracted from
the WSPRnet (archives go back to 2008),
and stations beyond 2000 kilometres (km)

Figure 8A: HamSCI North American Eclipse
Effects. RBN graph depicting percentage of
maximum spots per band received during the
Solar Eclipse QSO Party (SEQP). Shaded areas
represent the eclipse level: dark – no eclipse;
light gray – partial eclipse; white – total eclipse.
Courtesy: ARRL (QST, December 2017, p. 39)
32