comments on the irrelevance of "WWV" numbers to 6m F2 propagation
by Shel Remington NI6E/KH6
(many thanks to KA3B newsletter)

I enjoyed the item by GJ4ICD in the December 1990 KA3B 6 Metre Report, and I certainly agree that the panic to get the latest WWV solar figures is unwarranted, and in fact it causes needless QRM and clutter on 28.885. Regular participants on that frequency have already heard me state some of the following, but it seems worthwhile to get it into print once and for all. Geoff is correct in that anyone who takes the time to graph the daily 2800 MHz solar flux as a function of any aspect of F2 propagation on 50MHz soon realizes that these two variables are generally not correlated. it has become a source of some amusement when naive 6-metre operators are overheard to say, "Gee, the flux is over 200, so why don't we have any openings?" or the converse, "But the flux is only 140, how come the band is red hot today?".

Being one of those privileged to monitor 28.885 for many hours every day, I can't help but notice that the number of people reporting new countries being worked on 6m often appears to be inversely proportional to the daily solar flux. quite the opposite of what we have been led to expect. The Cult has become so pervasive that, in a triumph of lily-gilding (where the lily is made out of hole cloth), one manufacturer of amateur equipment is now selling an expensive transceiver that incorporates a digital clock which chimes at :16 past each hour to warn the user to check the upcoming WWV solar report!!

The whole concept is reminiscent of numerology or astrology or any number of other pseudo-sciences, in which the adherents are so boggled by the true complexity of nature that they embrace anything which appears to offer simplicity. To see how this situation arose, we need to step back a couple of decades. In several landmark articles in QST starting with the DXers crystal ball series, Ed Tilton, W1HDQ, explored the possibility of predicting F-region propagation by means of methods accessible to amateurs. Initially, this concentrated on visual sunspot observation. While sunspot activity is very interesting from an astronomical viewpoint, it soon became apparent that radio amateurs, by and large, were not motivated to invest in the needed optics and take the time to make daily sunspot observations.

This is unfortunate for astronomy because the AAVSO American Sunspot Number Program is one of the few ongoing projects in which amateurs are making a significant contribution to the field. Anyway, Ed soon decided that since the 2800 MHz Ottawa solar flux is broadcast daily on WWV and thus accessible by anyone with a short-wave receiver, it might make a good substitute for the SSN. After all, anyone who examines a long-term smoothed graph of SSN verses flux can't help but notice that they track closely, and furthermore they both peak around the years when the F-region MUFs are at a maximum (e.g. 1958). Graphs and formulae were then published showing the mathematical relationship between SSN and smoothed flux, and these are reasonably accurate although not exact. DXers who had copies of the old CRPL books showing smoothed global F2 MUF distributions as a function of smoothed SSN could now use flux information. it appeared, for predictions. With the advent of personal computers, software was written based on the CRPL database and this software was welcomed with open arms because it was so convenient. Some writers proceeded to publish tables showing that for each historical sunspot maximum, the beginning and end of widespread F2 openings at 50MHz coincided with a particular level of smoothed SSN and flux, and this led to the widespread belief in the magic flux level of 200 for 6m DX!

All of these developments were logical and properly rooted in scientific fact But obviously something is wrong because, as noted above the obsessive checking of the daily WWV flux number has failed to produce any improvement in our prediction ability. What happened is that somewhere along the way (actually, many writers seem to have independently made the same fatal error), the critical adjective "smoothed" was omitted. This concept of smoothing means that the data are averaged over a 13-month period centered on the data in question, which unfortunately means that smoothed numbers are not available until 6.5 months later and close inspection of that data archives reveals that, while the smoothed variables do track closely, the (unsmoothed) daily variables do not.

When looked at on a daily or other short term basis, the flux diverges wildly from the SSN and both of them diverge from the levels of F-region ionization. The CRPL database was never intended to be used with daily SSN, and its accuracy when used that way is so poor as to render it almost useless. The same applies to all PC prediction software, and to the 200-flux "magic number". The reason why daily 2800 MHz flux and SSN fail to predict daily ionization is that those two wavelengths of solar radiation are spectrally far away from the wavelengths that actually excite the ionosphere. The flux wavelength is about 0.1 metre; the sunspot number is observed in the visual spectrum around 0.7 micrometre; while the F-region is primarily ionised by ultraviolet radiation around 0.03 micrometre. Note that these wavelengths differ by orders of magnitude. Since the spectral distribution of solar output varies tremendously (and unpredictable) with time, it is quite possible for the radio flux or SSN to be constant while the UV fluctuates, or vice versa. Unfortunately for amateur observers, the Earth's atmosphere is totally opaque in the short UV range, so measurements of such radiation can only be made from above the atmosphere. Satellite UV data are available at NOAA SEL, so perhaps they could be petitioned to make such data available via WWV, and while they 're at it, they could start giving us more data on the intensity of flares. the disappearance of filaments, and coronal hole activity. Meanwhile. the NOAA SEL BBS at 303-497-5000 (see QST Nov 1990 p.41) does include some items that should be more useful, at least, than the WWV flux: The X-ray background level and the proton fluence are both listed under option A, sub item E "daily indices". Anyone with a modem who connects to that BBS could do us all a favour by relaying those items on 28.885. It would be wonderful if someone in the Denver Boulder area could make such calls on a daily basis, since it would be a local call from that area. The WWV flux may not be totally useless.

Several 6m DXers have observed that MUFs tend to be elevated, beginning roughly 3 days after the flux peaks, and ending roughly when the flux bottoms out, in its 27-day periodicity. So it is the trend of the flux, rather than its actual value, that seems al least a little bit useful. Another item mentioned on WWV is the "solar activity for the last 24 hours", which is based on X-ray influence, although it's given on a very coarse 5 point scale from very low to very high. I believe that the x-ray activity has more in common with the short UV than the radio flux does, so that little item may actually be the most useful in the whole WWV report. Certainly the WWV predictions "for the next 24 hours" for both solar and geomagnetic activity are nothing but crude guesses and rarely prove accurate if those variables do anything but hold still. As for the magnetic indices, again there seems to be an ill-founded belief that "the quieter, the better" as Geoff points out. This appears to have begun with articles in CQ Magazine by a writer who shall remain nameless, in which formulae and graphs were published proclaiming such a relationship. And again, the idea has been perpetuated in recent propagation software. It is clear that, at 50MHz at least, prolonged periods of geomagnetic quiet actually appear to suppress the F2 MUF. A fact which I believe was at least partly responsible for the very poor conditions in November 1990. The one exception seems to be transpolar paths, where anything but extreme quiet appears preclude F2 propagation. On the other hand, a really major geomagnetic storm with A-indices exceeding say 80 or 100, also suppresses normal F2 (for example, the storm of March 1989), but when such a storm is subsiding, there can be spectacular F2 opening worldwide. So it appears that intermediate A and K indices may be the best. But even so, two dates with identical indices can differ dramatically; perhaps the afore mentioned proton fluence data on the SEL- BBS can help sort this out.

Finally, those attempting analysis of past events should be aware that the WWV numbers are preliminary. The final data is published in the monthly Solar-Geophysical data, and reprinted in the Journal of Geophysical Research, section A, available in many university libraries, and also, of course, they are archived in Boulder and at other World Data Centers. Certainly the 1800 UTC Boulder A-index so favoured by the cultists should never be used for serious analysis, because, as stated by WWV, that is a preliminary number encompassing only an 18-hour period, and it is usually changed at 2400 UTC to a semi-final value. Likewise, all the magnetic indices on WWV are local Boulder numbers, and should be supplanted by the final Planetary indices as given in SGD and JGR. Anyone using the formulae for conversion of flux to SSN or vice versa should keep in mind that such formulae are crude empirical approximations of a relationship which is not a simple function, so please leave off the decimals. For use with prediction software, the best numbers are the predicted smoothed monthly sunspot numbers issued by A. Koeckelenburgh of the sunspot index centre in Brussels; these are published in Sky & Telescope every month and tend to be quite accurate. This number for January 1991 is 134.

The most powerful predictor at this point seems to be that actual 50MHz F2 propagation events are most likely to recur at 27-day intervals. This has, of course, been known for several decades, I have found it to be surprisingly useful during Cycle 22.

Geoff raised several other points. On the matter of backscatter, or more accurately side scatter, I get lots of it at my geomagnetic dip latitude, and it always fits a certain pattern. Signals are weak and diffuse, whether the distances are short or long (such as ZC4-KH6 and 6W1-KH6), and the poorly-equipped stations along with those who insist on using SSB are never heard. The stations at both ends generally do have strong reception to the intermediate point to which the antennas are pointed if there is any activity there (South America in the two cases cited above). These are hallmarks of scatter. By great contrast, the true long path openings I've witnessed (and have described) show strong, clear signals, even from the small stations, and the antenna bearings never deviate noticeably from the great-circle long path azimuth. Added to all this, I cannot imagine any mechanism which might account for strong signal bent paths, so it's not surprising that they don't seem to exist, at least not at these frequencies.

On the matter of widespread sporadic-E, I'm not going to touch that one! But it reminds me that many times in this sunspot maximum, there have been sudden almost global onsets of 6m F2, simultaneously (e.g.; within a 5-minute period), and almost as suddenly, the band dies, everywhere simultaneously. Apparently the cause must be some burst of UV from the sun which briefly enhances the MUF over the entire sunward hemisphere. And does 28.885 ever get clogged up on those occasions.

UKSMG Six News issue 29, April 1991