By shortwaves, we generally mean those wavelengths which occur with frequencies from about 3 to 30 mHz, or 100 to 10 meters. What makes these frequencies so special is the efficient use they make of the ionosphere as a reflective surface, which allows the waves to go beyond the horizon. Often the wave will bounce several times off the ionosphere and earth's surface before arriving halfway around the world. This bouncing causes phase distortion, one common source of selective fading on the shortwave bands.
As you may suspect, many doctoral theses have been written on the ionosphere and a thorough discussion is beyond my capability to write, and the reader's patience to read. Follows, then, a brief outline.
At the lower end of the shortwave spectrum, groundwave coverage is fairly good, on the order of a hundred miles with up to 1000 watts. As you go up in frequency, groundwave coverage decreases, so that at 27 mHz, where CB radio lives, it is only a few miles.
However, it is not groundwave that people use these frequencies for. Going back down to 3 mHz, we find that during the day the D-layer of the ionosphere absorbs most of the signal energy, so during daylight these frequencies have little to offer. At night, though, the D-layer disappears, so that these signals can reach the F-layer, and be propagated to distant points (this is also why AM signals only travel a few hundred miles during the day, but can travel thousands of miles at night).
As we approach 6 mHz, D-layer absorption tapers off to the point that signals bounce off the E-layer. This layer of the atmosphere, being relatively low, provides regional as opposed to world-wide coverage. You will find lots of hams here during the day, gossiping with their buddies up to five hundred miles away, but little international broadcasting. Again, these bands offer world-wide coverage at night.
As we approach 12 mHz, the E-layer is no longer an effective reflector of radio waves, and they begin to pass on through to the F-layer. This means that these frequencies are open during the day instead of at night, so as a general rule we use the following: above 12 mHz during the day and below 12 mHz during the night (actually, around local sunrise and sunset, anything can happen, and these are the best times to hear rare ones).
The F-layer is the workhorse of shortwave broadcasting. Because of its greater height, it allows the signals to be reflected farther. There are actually two distinct regions in the F-layer during the day; at night they converge.
Once we get to about 20 mHz, a new limitation comes into play. As we go higher in frequency, we become more and more dependent on high levels of solar activity to make the ionosphere reflective. Solar activity itself runs on an eleven year cycle. In 2001, we are just coming off the peak of the solar cycle, so these higher frequencies should still be very active for a few more years. Then the activity will lessen, dropping to a low point in about five or six years, only to start back up again as the cycle continues. Activity on these bands never actually stops, it just becomes more sporadic as solar activity lessens. Also, because of the increased need for solar activity to cause the F-layer to reflect radio waves at these frequencies, these bands generally offer the best propagation during the day. At night they often become inactive.
There is, of course, an upper frequency limit to ionospheric reflection. During the greatest peak in the solar cycle to occur since radio was developed, which peak occurred in the late 'fifties, users of frequencies above 50 mHz reported interference from the other side of the Atlantic. In the fall of 2001, I myself had the thrill of receiving an Atlanta VHF television station here in Central Texas, on a common pair of rabbit ears! Most years, though, the limit will be around 30 to 35 mHz.