Proxy Measurements and Propagation

Why do we count sunspots?

It's an interesting question since many hams seem to pay close attention to it. However, as we'll see, it's of greater importance to science historians and solar physicists, but misleading to HF propagation expectation. 

In the beginning, sunspots were counted because they were there! That was a monumental discovery since in many cultures and religions the Sun was declared to be perfect, without blemish. At first only the largest spots could be observed because, without instruments, the Sun is too bright to look at. Eventually it was a simple pinhole projector that allowed scientific observation. I did this myself a few times when I was young and pursuing my interest in astronomy.

With these simple instruments it was possible to discover the rotation of the Sun, and its differential rotation with latitude. Sunspot birth and death, sunspot groups and, with longer observation, solar cycles were recorded and studied. Progress was slow at first. The relationship to the ionosphere and radio propagation were wholly unknown because both were undiscovered. Over a century ago that changed.

Every ham knows that there is a correlation between sunspots and ionospheric radio propagation. The correlation is loose and of little use for answering whether there is, say, propagation from here to Egypt at 21 MHz right now. Another way of putting it is that sunspots are a poor proxy for ionospheric propagation. They're just easy to see and count.

Although the correlation is poor it is better than nothing. We would listen to WWV (before the internet!) for the latest data: sunspot number, solar flux and A and K indices. For a long time we had little else. It was often easier and more reliable to turn on the radio and listen or transmit and see what would happen. That's a bit like stepping outdoors to check the weather rather than rely on a low accuracy weather forecast, or none at all. As with the weather, we now have better data, proxies and prediction methods to rely on, and the internet to access them in real time.

A chart like the one above shows how wildly the sunspot number swings. To get meaningful information from the raw data it is necessary to smooth the data. That involves weighted averaging over a moving window of data. When you do that the correlation to propagation is greatly improved. The correlation is backward looking since the "prediction" is for the past. That doesn't help us today. Sunspot number and SSN (smoothed sunspot number) are informative to solar physicists, but not to hams. We need current information.

Ionosondes can tell us with good accuracy the ionization above us at a single location and time. The data, good as it is, tells us only a little about ionospheric propagation. That's why ionosondes, like weather stations are placed in many locations. The network is full of gaps since the interest these days is scientific and not commercial. There's little money to be made in 2023 from observing and predicting HF propagation.

Why can't we directly measure the Sun's affect on ionospheric propagation? Why do we rely on a proxy like sunspots? The major factor is EUV (extreme ultraviolet radiation). EUV radiation is correlated to sunspot activity but poorly, yet EUV is largely responsible for F-layer ionization. We need to directly observe EUV, not sunspots, for reliable propagation predictions. The attention paid by many hams to sunspots more often leads to confusion than enlightenment.

The NASA sourced diagram found on Wikipedia shows that electromagnetic wavelengths shorter than near UV radiation are blocked by the atmosphere. The ozone layer is largely responsible, but so is the ionosphere. Well, that must be true since if the ionosphere were transparent to EUV the ionosphere wouldn't ionize! This is not only good for propagation: EUV and more energetic radiation is hazardous to life. 

The downside is that EUV can't be directly monitored from the ground. It can only be observed directly by satellites, and that comes at a price. Ideally we want a measurement we can do at the Earth's surface that correlates well with EUV. That is, a reliable proxy!

EUV is closely correlated to solar radiation in a portion of the microwave spectrum which the atmosphere is largely transparent to. The Sun's noise across a 100 MHz spectrum near a wavelength of 10.7 cm is measured from ground observatories. With smoothing on the order of hours it is a good predictor of F-layer ionizing EUV radiation. That is, it's a good proxy for a major influence on radio propagation. This is the solar flux index (F10.7) that hams are familiar with.

Unfortunately, the current EUV level is insufficient to determine the current propagation. Propagation is far more complex than that. I would be remiss to focus only on EUV and its proxies, despite its outsized contribution.

Ions form during daylight in response to EUV, then gradually recombine during nighttime. The cycle repeats every day. The rates of ionization and de-ionization are not equal and, indeed, vary with the seasons: by latitude and hours of daylight.

This is not unlike what happens with the cycle of the seasons. The coldest temperatures in most of the northern hemisphere occur in late January, 5 weeks after the solstice when solar irradiation is minimum. For most of January the average rate of cooling is greater than that of heating. Only after the rates cross in late January does the temperature begin to climb. The same process in reverse occurs in July.

Sustained high SFI improves F-layer propagation with a net positive ionization rate over many days. Since we can measure the SFI each day, consecutive series (daily, hourly) of SFI values are needed to predict propagation via the F-layer. There is more than one observatory so measurements are possible around the clock. However, a high SFI where the Sun is shining does not contribute to ionization on the nighttime half of the globe.

Of course there's more to predicting propagation than EUV. X-rays from solar flares temporarily cause the D-layer to absorb short wave signals. Density, composition and speed of the solar wind, especially as enhanced by CME (corona mass ejections) can make the E-layer opaque or reflective and also disturb the F-layer at high latitudes. Visible and radio auroras are a useful but imperfect proxy for the relevant geomagnetic activity. There are many factors at play. Even the stratosphere plays a role, such as its likely influence on sporadic E propagation. 

As I write these words the SFI has been above 200 for several days. With the X-ray flux staying high there is also attenuation of signals being propagated by the F-layer. There are lots of signals on 10 meters but the signal levels are not what they could be. 

Now that the X-ray flux is dropping, conditions should become more favourable, at least until the next flare! This is good because, ideally, we want lots of sunspots that are middle aged, not new, to have sustained high EUV flux and without the constant flaring of young spots. Sunspots are a two-edged sword: the good of increased EUV and the bad of flaring and CME.

When the SFI is high I leave WSJT-X running with the 6 meter yagi pointing east then south in the afternoon when I'm busy doing other things, in hopes of decoding a signal from the tropics or South America. No luck yet but I did hear a US station working the TN8K DXpedition. Our chances improve as the solar cycle progresses and a sustained high EUV flux becomes routine.

In the meantime, it is wonderful to hear and work the world on 10 meters. It will only get better and 6 meters is next. At this northerly latitude I watch as daylight, and therefore EUV exposure, increases as the calendar progress towards spring.

Credit: xkcd (Jan 16, 2023)

In closing, watch the sunspot number as a harbinger of improved high bands performance, but pay closer attention to the SFI since it is the far better proxy for what to expect on the bands, right now when you turn on the rig. SFI and SSN track well but only when the latter is smoothed over many many months. Don't be misled by the sunspot number.