Friday, September 25, 2015

The Mirror in the Sky - part 1

The daily activity of the Sun routinely tosses electrons right out of the atoms in the upper atmosphere.  The free electrons (and remaining ions) make those layers conduct electricity.  For radio waves, this is like having a mirror in the sky, that can be used to bounce high frequency (HF, othewise known as short wave) waves over the horizon, to anywhere on the planet, if conditions are right.  These waves literally bounce off the sky & are known as skywaves


I had not been on the 80-meter CW traffic nets for very long, when it became obvious that sometimes I could hear the other stations and sometimes I couldn’t.  Unfortunately for me, these early months after my General license were during the sunspot minimum.  Some days I would bring up for a look-see & I’d find myself dusting off my computer screen in order to find out if what I saw on the solar image was a sunspot or a piece of dirt!  At that time, the proportion of “dead nets” was, to say the least, higher that usual.

The local 80-nets are NVIS nets … Near-Vertical Incident Skywave.  The outgoing wave from the transmitter hits the ionosphere from a near-vertical direction & is reflected straight back down, illuminating the area within a few hundred kilometers, which is just the area that a local or Region traffic net is looking to cover.

It should, in my naïve & innocent mind, then be easy to see if the net was going to work or not by lookup up (on the internet) the daily predicted value of a number called foF2.  foF2 is the critical frequency (f) for the “ordinary” (o) wave reflected off the F2 layer of the ionosphere, for a vertical ray path (i.e. NVIS).  There is also an fxF2, relating to the “extraordinary” wave, but it seems that FoF2 is the standard number used to indicate the health of the ions in the upper atmosphere.  (The ordinary & extrordinary are circularly polarized in the opposite direction, but otherwise … well, the same?  A dipole antenna, such as used by hams, only gets half the energy contained in either, but cannot tell the difference.)

Below the critical frequency, the wave is reflected back.  Above it, the vertically traveling wave flies off into space & is gone.

There are web sites out there that give the current predicted foF2, based on detailed ionospheric models (including current values of sunspot number, solar flux, etc.).  Before each net, I note down the predicted foF2, then turn on the radio.  If the net is on 3.537 MHz & foF2 is higher than that, then (assuming the D layer isn’t still too strong … the D layer goes to bed about sundown), I should be able to hear the other stations on the net.  Unfortunately, the source web sites do not all agree with each other, any more than any two weather forecasters would agree.  This muddies the situation.

When the ionosphere has gone south for the winter (leaving us only a few ions) & the Sun is not very active, the 80 meter ham band is right on that hairy edge with the foF2.  Often, as you would expect with critical reflection, the nearby stations drop out, because the ionosphere isn’t strong enough to support truly vertical NVIS, while the stations a few hundred miles away will still come in.  Sometimes, we all may as well pack it in & go to bed.  And, sometimes all is hunky dory.

So picture this … a relatively new ham, who keys & writes with the same hand, struggling with CW (Morse code) in the first place, still having to write down every single character to follow what is going on, now also trying to keep notes on everyone else’s signal strength!  It got better, but it did take quite a awhile!

Naively, I expected 80 meters to behave in a similar way to the higher-frequency ham bands that are normally used for contacting distant (DX) stations.  For any given great-circle path, there is a LUF, or lowest usable frequency, determined by the D layer, and a MUF, or maximum usable frequency, determinde by how good the ionospheric mirror is between here & there.  These are the principles of radio propagation that are taught in the amateur radio license study manuals. In the case of the “low bands” (80 meters & 160 meters), the pattern is sort of followed, but not all that well.  Immediately, I noticed some wide discrepancies.  Two different web sites listed foF2 values that differed by up to 1 MHz.  Since the CW end of the 80 meter band is at 3.5 MHz, that is a 29% variation, which seemed like a lot to me.  Often the nets worked when they should not have & vice versa.

One of our nets, the Sixth Region Net, called RN6, has two sessions every night, one at 7:45 pm Pacific Time & one at 9:30 pm.  The second one should be tougher copy than the first one, right?  Because the Sun isn’t shining on our part of the upper atmosphere, most of the electrons would have found their way back into an atom.  Maybe, maybe not.  It depends, apparently.  Occasionally, the second net was the better one.

Even crazier, on rare occasions, I ran into a net called NYS on the RN6 frequency, 45 minutes before ours.  It was weak, but I could copy it well enough.  I also checked on the internet for the time & frequency.  NYS is the New York State traffic net!  80 meters is not supposed to have that much range.

So there is weirdness going on over our heads, that only short wave operators (hams or otherwise) know & care about.  That’s about where my understanding stayed until I found this (Eric Nichols clued me in to it) …

There various scientific installations around the world that probe the ionosphere, every 15 minutes!  Many of them post their results on the web.  The closest ionosonde station to me & the rest of SCN & RN6 is the one at Point Arguello CA (Vandenburg Air Force Base, actually).   The ionosonde sends pulses upward, while sweeping the frequency through the entire relevant range, normally 1 MHz to 15 MHz.  It is just like any other radar … the reflections that come back tell is what going on & the time delay tells how far up the action is.  It use circularly polarized antennas, so can tell the difference between an ordinary & an extraordinary wave, which the output indicates by color.

In fairly normal circumstances, the results look like this:

The horizontal axis is frequency in MHz & the vertical axis is reflection height in kilometers, as determined from the time delay between the pulse & its return.  Pink/red = ordinary wave.  Green = extraordinary wave.  Note that above about 7.8 MHz on this ionogram, there is no ordinary return.  This frequency is foF2.  The computer pins it down to 7.838 MHz (see upper left).  You can also see a reflection from the F1 layer here (below 4 MHz) & some E layer reflection (below 2.9 MHz), but that would constitute a “plot complication” for this post.  The black solid & dotted line is the computer model’s estimate of the ion density as a function of height.
Now, this is important!  Are there really two F2 reflecting layers, one around 300 km & one around 600 km?  Nope, there is only one at 300 km.  The other one is the same wave that reflected down from the F2, hitting the Earth’s surface, going back up & bouncing back down again.  This indicates a moderately good mirror.  We can count the number of multiple reflections & use that as a guide to how strong to expect the net signals to be.
So now I don’t need to depend on forecast models.  What I see on the Point Arguello ionosonde site is the real, actual foF2 at the time.

Now I am able to answer one of my original questions:  are the disparate predictions really that far off?  The quick answer is Yes.  Using data from mid-July 2015 to mid-September, I see that there is apparently quite a lot (one sigma on the difference O-C is 0.75 MHz during RN6/1 and 0.89 MHz during RN6/2) of variation in the ionosphere that is due to factors not accounted for by the model (such as maybe geomagnetic activity?).  These two diagrams show (in blue) the observed ionosonde foF2, as a function of what one of the models predicts and (in orange) the difference (observed minus calculated).

During this particular two-month period, there also seems to be a systematic difference.  During RN6/1 the real ionosphere better than the model by mean value of 0.39 MHz, or a median of 0.25 MHz; the difference between mean and median indicates there are more outliers on the high side than on the low side.  During RN6/2, the mean difference is even larger (mean 0.60 MHz & median 0.50 MHz).  The real ionosphere performed better than the model, especially during the second session (RN6/2).  I’ll be interested in watching it through the winter & spring, to see how this situation changes.

All this goes to show that there are some wonderful scientific data out there on the Information Superhighway to play with.  You don’t want to be shy about enjoying it!

The model that was used for this project is the “effective sunspot number” model used by DX Atlas software ( .  I picked it for the simple reason that it is the easiest to use & does not involve reading value off a tightly spaced chart.

In my next installment, I will write about how to read the ionograms, a process which I am just learning.  Just let me say, though, that (as one of the southern CA DRS stations), I am almost at the point of being able to look at the web & tell whether or not I’ll be able to get my NTSD MBO connection or not.  So, stay tuned (so to speak).

No comments:

Post a Comment