By - Tom Williams, WA1MBA
The range of frequencies between about 2 and 10 GHz has a special characteristic - very low noise. Man made noise is pretty rare (except from other microwave uses), there is almost no static noise from spark sources, and galactic noise is fairly low. Above 3 GHz, the heat from the atmosphere and the earth begin to generate noise that can affect communications. All of this makes the microwave bands great for experimentation. Just as with any other electromagnetic wave (e.g. RF, light), a 10 GHz wave spreads out, interacts with different media - bouncing off things, and is collected at a receiving antenna. It is the "bouncing around" that intrigues most of the users of this band, and results in most of the available dx modes. Figure 1 depicts two of the common modes of microwave propagation - line of sight and tropospheric scattering.
The troposphere is the lower 50,000 or so feet of the atmosphere. Scattering is the process by which signals interact with the transmission medium in a way that causes the signals to go in directions different from the original direction. Usually, it is the interaction between the RF energy and particles that cause the scattering, but not always. The thing that causes all disruptions in the direction of a propagating RF field is a change in dielectric properties (resulting in a change in the propagation speed and, therefore, direction), often referred to by its effect, such as "change in dielectric constant" or "change in refractivity". It is this change that makes lenses work because this property is different between glass and air.
Scattering increases by a factor of the fourth power of the frequency, until the wavelength approaches the dimension of the particle. Once the particle is as big as 1/10th the wavelength this fourth power factor begins to reduce, and when the particle is the size of a wavelength and larger, the scattering efficiency remains constant (does not increase with frequency). If there is a set of particles (like rain) in the medium (air) the signals will end up scattered in many directions. One common misconception is that scattering is the same as reflection. Actually, the scattering objects re-radiate the energy that enters them. This helps explain why there is forward scattering, and why polarization is maintained in most scattering situations.
In the case of Tropo scattering, there are a number of familiar scatterers (particles). One is the Oxygen and Nitrogen molecules that constitute air. They can be thought of as scatterers in the vacuum that they occupy. As such, they are very small with respect to the wavelength of light, and because scattering increases with frequency, the sky is blue (blue being a higher frequency than red or yellow). The air molecules themselves are too small to be effective scatterers at any but visible wavelengths (representing frequencies of 500,000 GHz and above). Typical VHF and UHF scattering is caused (for the most part) by turbulence in the atmosphere. This turbulence is manifest in localized pressure differences with sizes that range from a few meters down to about 10 cm. Difference in pressure in a gas results in difference in refractivity, and so, turbulence causes scattering of VHF and UHF signals. Tropo scattering from turbulence does have an effect at 10 GHz but it is not much greater than on 1,296 MHz. There may be some contribution to microwave Tropo by airborne dust, but the author has not found a definitive reference on this topic.
We tend to use fairly narrow beam antennas on 10 GHz - often with beamwidths of five degrees or less. You can think of the transmitted signal like a light beam from a search light. These antennas tend to "illuminate" smaller volumes of the scattering medium than broader beamed VHF arrays. This results in interesting short term fading effects and significant variability in signal levels day to day when the primary propagation mode is Tropo scattering. It is not uncommon to get a 3 to 6 dB boost from a low flying large commercial aircraft that happens to pass through the illuminated volume. Technically, these signals are reflected, not scattered. The focused beam also allows one to efficiently illuminate a single thunderstorm a hundred miles away, making rain scatter dx a reality.
Water in the atmosphere clumps into a few ranges of particle sizes. The first is water vapor - which is individual molecules that have properties similar to other atmospheric gasses. Haze particles range from under .001 um (a thousandth of a micro-meter) to .01 mm and those in clouds range from .001 mm to .1 mm. At 10 GHz, it is not practical to get useful scattering from clouds and haze - the return power levels are far too low for communications. (Cloud radars are built, but they run frequencies between 35 and 95 GHz.) Rain droplets range in size from about 0.5 mm to 3 mm (drizzle to tropical downpour). Rain is in exactly the right range for scattering of 10 GHz signals. Drops are 1/60 to 1/10 of a wavelength - and indeed it is the heavier rain storms (with their larger rain drops) that give the best scattering returns. The author and others have used rain scatter for communications on 5 GHz, and have observed scattering on 3400 MHz, but with very low signal levels. Recall the fourth power law - at 5 GHz (about half the frequency) we would expect to receive 1/16 the signal (1/2 to the fourth power) - representing 12 dB less received signal level. At 3400 MHz (1/3 of the frequency) we would expect to receive 1/81 of the signal, or about 19 dB less signal than at 10 GHz.
Let's take a look at the pattern of energy emitted from the scattering volume. Directivity of the scattering from a population of spheres in a medium (rain in air) depends on polarization and the size of the spheres (relative to the wavelength). When the drops are on the order of 1/10 wavelength and smaller, the scattering plot is as shown in figure 2. In the case of rain and the 10 GHz band, these conditions are met quite well. View these plots just like an antenna plot as seen from above. The incoming signal (the "illumination") arrives from the bottom of the figure, and the scattering material is at the center. This plot (see figure 2.a) shows that for energy entering in a polarization perpendicular to the plane of incidence, the scattering is almost completely isotropic - the same in all directions. This would be the case for vertical polarization. In other words, imagine that the plot is for a vertical antenna, looking down on it. The energy arrives from the "south" side of the plot, and radiates from the scattering volume just like a vertical antenna would radiate the energy if it were located at the center of the scattering volume.
On narrow-band 10 GHz, however, we use horizontal polarization, which is parallel to the plane of incidence. Scattering for this condition follows the pattern shown in figure 2.b. This means that, in normal communications use of narrow-band 10 GHz during rain scattering conditions, the worst angle is when the rain event is at a point that causes the two stations to aim at 90 degrees relative to one another. At shallower and greater angles, the signals increase. If amateurs want to maximize the likelihood rain scattering, then either vertical polarization should be used or rain cells that do not cause this 90 degree relation should be sought - see figure 3.
On a number of occasions the author and other New England stations WB1FKF and W1RIL have measured rain scattered signals of significant amplitude and over surprising distances. In the case of signals between W1RIL and WB1FKF, where there is a fairly difficult obstruction (a nearby hill), rain scatter has offered over 10 dB and sometimes as much as 20 dB of signal boost. This can easily make the difference between not hearing the other station at all and being able to copy clearly.
One characteristic of rain scatter is Doppler shift. This phenomenon is manifest in a shift in frequency when the path between the transmitter and receiver is shortening or lengthening. Because the rain in a storm is being blown by winds, the scatterer is moving, the path length is changing, and the frequency is shifted. Usually, a CW note becomes quite "fuzzy" in its sound because the individual rain drops are moving at slightly different speeds and directions. Press here to listen to some rain scatter (200K sound file). The sound is very much like Aurora propagation on 2 meters - at times it can sound like strong noise. When looking for a rain scatter contact it is advisable to have the transmitting station send CW dashes, so that the receiving end can distinguish the noisy signal from the background noise.
Because the wind will have one predominate direction, stations will notice some Doppler shift, even though the signal sounds quite fuzzy. Sometimes the wind speed is high enough to move the signal out of the pass-band!
In May of 1995, W1RIL and WB1FKF pointed towards a developing thunderstorm with a round-trip distance of some 250 miles and noticed a Doppler shift of 4.1 KHz - which represents a wind/rain velocity of over 130 mph (see insert). As it turned out, this storm developed a devastating tornado. On yet another occasion, the author and these stations scattered their signals off a storm in New York with a round trip distance in excess of 350 miles, and with signals strong enough for a phone QSO. Other stations have reported similar extended dx from rain scattering.
Over the years, we have come up with various ways of trying out rain scatter. If you are lucky enough to have a sked soon after your local TV station weather report, just look at the radar map - it will be the most current and local of any sources. You are more likely to need an up-to-date map at other times of day, so if you have access to the Weather Channel you might get enough detail to determine the local storms and their paths. Another way of obtaining maps is to select weather map sources on the internet. These tend to update every few hours (the author has used http://www.intellicast.com/weather/usa/ and there are others available on his website, see below). You may find other useful URLs by searching for appropriate titles. Remember, when the two stations' paths to the storm are at 90 degrees, the signals will be weakest.
Snow scattering is very much like rain scatter, and at 10 GHz has approximately the same scattering effect for the same equivalent rainfall rate. However, most snowfall occurs at a slower equivalent rainfall rate than typical rain, so only the heavier snowstorms are likely to have significant snow scattering. Nonetheless, snow scattering has been observed by the author and other users of 10 GHz. At times is can be very effective at extending communications even though it is unlikely that hilltop expeditions would take advantage of this mode - its cold enough at some of these sites even in the Summer!
So, what does this information about propagation mean to a ham who wants to get into 10 GHz narrow-band? One thing is that a home station can be lots of fun, especially if there are other hams with home stations within 30 to 100 miles. It also means that portable operations can work in situations that would normally seem impossible - such as where there are mountains in the way. Rain may be a nuisance to operating, but it can be a welcome signal enhancer that lets you work around mountains and more than a hundred miles beyond normal range. The most important thing about 10 GHz narrow-band operation is to get on the air - join the fun on this fast growing part of the spectrum.
Any propagating wave will be received with a changed frequency if the path between the transmitter and receiver is changing length. This is obvious when listening to a train pass by, as it blows its horn. When it approaches, the pitch sounds higher than the pitch of the horn as perceived by the people on the train. When it goes by, its distance is not changing much and the listener hears it at the correct pitch. Then, after it passes, the pitch becomes lower than that heard by people on the train. Ask any amateur who has used the ham radio satellites. They will tell you that Doppler shift can be somewhat tricky to contend with, especially when having a three-way conversation, where different stations are experiencing different amounts of shift.
One way to understand this phenomenon is to simplify the situation by fixing the transmitter in one spot and moving the receiver (the reverse of the train example). The speed of the waves traveling through the medium is constant. Let's take as an example a transmitter operating at exactly 10 GHz (10,000,000,000 Hz). This emits waves with a wavelength of about 3 cm. If we move the receiver towards the transmitter at 300 cm per second (about 6 1/2 mph), then every second the receiver would receive 100 more waves than if it were not moving. One hundred waves in one second is 100 Hz. Thus, the receiver would report the received frequency as 10,000,000,100 Hz. As the closing speed increases, the frequency increases linearly. This is the same principal by which a police speed radar and a Doppler weather radar works. In the example mentioned in the text, two amateurs noticed a frequency shift of 4.1 KHz at 10,368.1 GHz. Where f is the transmitted frequency, Vr is the velocity (relative), c is the speed of light, the equation for the Doppler shifted frequency fd is:
The speed of light c is about 300,000,000 meters per second. At 10,368,100,000 Hz, with 4,100 Hz Doppler shift, round-trip speed of 118.6 m/s (or 265 mph). The actual wind speed in the storm would be half of this, or at least 133 mph (half because the path length is changing at twice the speed of the scatterer - the actual radar Doppler equation has a 2 in the denominator to account for this). I say "at least" because this speed represents the closing velocity relative to the two stations - because it is unlikely that the direction of the wind was exactly towards the stations, the actual wind speed was probably higher. During the ensuing tornado, the weather service reported a Doppler recording of over 250 mph.
http://www.wa1mba.org/ At this site you can find sounds of rain scattering contacts (even snow scatter), and a lot of links to other sites that have info about microwaves, and in particular, 10Ghz. From that site you can also find out about the North Texas Microwave Society, the San Bernadino Microwave Society, and the Mt. Airy VHF Club (called the "Pack Rats").
Tom Williams, WA1MBA, "Narrow-Band 10 GHz and Some Observations From New England", New England VHF Conference, August 1995, ARRL Publications.