Rain Scatter on Microwaves: Working Beyond the Horizon

For most radio amateurs rain is a nuisance: it damps the signal of a microwave link, it drips from the antenna feed and it brings thunder close to the mast. For a small community of microwave operators, however, a thunderstorm is an opportunity. The same raindrops that weaken a direct path will, in the right geometry, scatter a part of the signal in all directions, and two stations that cannot see each other can both point their dishes at a heavy shower and hear each other’s signal come back from it. This is rain scatter, a mode that has produced contacts of 100 to 400 kilometres on 10 GHz and neighbouring bands without any help from the ionosphere. This guide explains how it works, how to find the rain cell, what the signal sounds like and which equipment and modes make sense.

How Raindrops Scatter a Signal

When a radio wave meets a raindrop whose size is small compared with the wavelength, the drop acts as a tiny antenna. It absorbs a bit of the wave and re-radiates it in other directions. For small drops this scattering grows very quickly with frequency, in proportion to the fourth power of it in the simplest model known as Rayleigh scattering, so that the amount of scattered energy at 10 GHz is vastly higher than at 1 GHz. The drops of a heavy shower measure a few millimetres, comparable with the 3 centimetre wavelength of the 10 GHz band, so the simple model is only approximate, but the principle holds: the shorter the wavelength, the more efficient the scatter.

A single drop reflects almost nothing, but a rain cell contains an enormous number of them, spread over a volume of many cubic kilometres. The scattered signal from all the drops adds up, not in step with each other, but with random phases, to a signal that is weak and noise-like. If the two stations’ antennas intersect in the cell and the rain is heavy, the combined signal may be strong enough for a contact. This is the same idea as the radar echoes of precipitation, used by the weather services, and the analogy is useful for finding the cell. The weather radar picture, in effect, is the map of what your signal will hit.

Which Bands Work

Rain scatter has been used from 3 GHz up to 47 GHz, but the best compromise is found at 10 GHz and at 5.7 GHz. At the lower microwave bands the scattering is weaker, while at 24 GHz and above the rain also absorbs the signal severely on its way to and from the cell, so the shower that scatters the signal also hides it. This conflict, scatter increasing with frequency and attenuation increasing too, is the reason for the optimum. On the 10 GHz band, which has the benefit of small dishes and cheap equipment from satellite television and from the contest community, a path of 200 kilometres via a good rain cell is quite practical. The summary of microwave events in the article on the ARRL 10 GHz and up contest shows where operators go when the weather is right.

Finding the Rain Cell

The two stations do not point at each other. Each points at the same rain cell, which lies somewhere along or near the line between them and often not on the great circle path at all. Finding it requires a map of the precipitation, and the weather radar images that national weather services publish on the web are ideal: they show the intensity of the rain in colours, with a refresh every five minutes or so. On a radar picture you choose a cell that is both strong and in a place that both stations can aim at, with reasonable elevation angles. A cell closer to the middle of the path gives the shortest total distance and so the strongest signal.

The cell moves with the storm, usually at tens of kilometres an hour, and it also grows and decays. The dishes must follow it, which is done by hand with a compass and an inclinometer on a small dish, or with a motorised mount on a larger one. Both ends must also agree which cell they use, by chatting on a phone or on a lower band, and sometimes the operators swap to a new cell as the first one dies. The skill is to read the radar and anticipate the movement. The geometry of the paths and the horizon is related to the ideas in the article on the radio horizon, because a low elevation angle lets the beam reach the cell above the local obstacles.

Polarisation and the Signal

Raindrops are not spheres. Falling drops flatten into shapes with a horizontal long axis, and this affects the polarisation of the scattered signal. Many operators use the same polarisation at both ends, the one agreed locally, which for narrowband work on 10 GHz is usually horizontal in Europe, and they accept some loss. In forward scatter, the polarisation of the signal changes in ways that depend on the shape of the drops and the angle, so the agreement should be tested by trial on the day. The standard of the area decides the first choice, and an experiment settles the rest.

The signal itself is unlike any other. It is not a clean tone but a noisy, hissing signal, spread in frequency by the motion of the drops, which move up and down and sideways with the air. This Doppler spread is typically a few hundred hertz to a few kilohertz at 10 GHz, depending on the turbulence in the cell and the speed of the wind. A carrier that was transmitted as a single steady tone arrives as a band of noise, with a character like a rushing sound, and the strength varies from second to second. The Doppler physics behind this is described in the article on Doppler shift in amateur radio.

Modes That Work

Because the signal is spread in frequency, narrow modes lose much of their energy. Wideband modes that tolerate Doppler spread are therefore better than very slow CW. On 10 GHz, operators typically use SSB with a wide filter, FM with a deviation chosen to capture the spread, or CW at a modest speed that fits the spread. The weak-signal digital modes designed for scatter, in the WSJT family, are an option in some cases, as in the article on Q65 for EME and scatter, where wide variants of the mode cope with Doppler spread. The key is to choose a mode whose bandwidth is similar to or wider than the spread.

If the signal is too noisy to read as speech, a simple tone modulation, such as repeating a tone on and off, can show that the cell is working. Keep exchanges short: callsigns, a report and the locator, exactly as in other scatter modes. The signal level changes as the cell moves, and a contact can often be completed in the first minutes of a good cell, before it drifts or falls apart.

When and Where It Happens

Rain scatter is a seasonal mode. The cells that matter are convective: summer thunderstorms and heavy showers that build up on warm afternoons and evenings, with high rainfall rates over a compact area. Widespread steady rain, even when it is wet, contains smaller drops and a lower rate and gives much weaker scatter. Watching the forecast for convective activity, and the radar for the development of cells, tells you when to go out. Many contacts are made in a window of an hour or two before the storms reach the stations themselves, when the cell is still some distance away.

The effect is also helped by height. A station on a hill or a mountain has a clear horizon and can aim at a cell at a low elevation angle, which keeps the path short and avoids ground clutter. Microwave contest operators spend a lot of effort on portable locations for exactly this reason. For fixed home stations, a clear view toward the direction of the prevailing storm tracks matters more than the power of the transmitter.

Rain Scatter Compared With Other Mechanisms

Rain scatter is one of several ways to go beyond the horizon at microwave frequencies. Tropospheric ducting, described in the article on tropospheric ducting, gives much stronger signals over longer distances but depends on a stable atmosphere and happens rarely. Meteor scatter, covered in the guide to meteor scatter with MSK144, uses short bursts from ionised trails and works best on VHF. Rain scatter needs neither a special atmosphere nor a shower of meteors, only heavy rain at the right place, and so it is the practical choice for a summer evening when the thunderstorms arrive.

The mechanisms also differ in the signals they give. Ducting signals are steady and strong, meteor scatter is a series of short pings, and rain scatter is a continuous but noisy signal that can last as long as the cell does, often fifteen minutes to an hour. That duration is useful: a contact does not need to be rushed, and several stations in the area can work through the same cell.

A Typical Session

A typical session begins with a plan made the day before. Two operators agree on a time window, a frequency on the 10 GHz narrowband segment, a mode and a polarisation, and they exchange locators. On the evening they check the radar. When a promising cell appears, each operator calculates the bearing and elevation to a common point on the map and aims the dish. Then they call on a lower band to confirm that both are ready, one transmits a steady carrier and the other searches for it in a wide filter. The first sign is often an increase in noise at the right frequency, followed by recognisable tones. They exchange callsigns, reports and locators in short bursts, confirm by voice on the lower band and log the contact. If the signal fades, they look for a new cell and repeat.

Equipment and Practice

A workable 10 GHz rain scatter station has a transverter or a purpose-built transceiver, a power of 1 to 10 watts or more, a dish of 40 to 90 centimetres or a horn with moderate gain, and a stable frequency source. The path loss of several tens of decibels more than a direct path requires an efficient system, as discussed in the article on free-space path loss and link budgets. Keep the cables as short as possible, since losses at 10 GHz are high even in good cable, and use a feed at the dish.

Safety matters in a thunderstorm. The best cells are the ones with lightning, and an operator holding a dish on a hill in an electrical storm is in real danger. Set up a position that is safe, such as a car or a building with a dish on a tripod at a distance, and stop at the first sign of lightning nearby. The radiation exposure at the dish is small at low power, but obey the rules for power density at the front of the antenna. No contact is worth a lightning strike.

The Bottom Line

Rain scatter uses raindrops in a heavy shower to scatter microwave signals between two stations that both aim their antennas at the same rain cell. It works best on 10 GHz and 5.7 GHz, where scatter is strong and rain absorption is still tolerable, and delivers distances of 100 to 400 kilometres. Use weather radar images to choose a cell between the stations, follow it as it moves, expect a noisy signal spread by Doppler of a few hundred hertz to several kilohertz, and choose a wide mode. Keep exchanges short and keep away from lightning.

Frequently Asked Questions

What is rain scatter?

A propagation mode in which raindrops scatter a microwave signal, so that two stations beyond each other’s horizon can communicate by pointing their antennas at a common rain cell.

Which band is best for rain scatter?

The 10 GHz and 5.7 GHz bands are the best compromise, because scatter grows with frequency but so does absorption by the rain.

Where do I point the antenna?

At the rain cell, not at the other station. Use a weather radar image to choose a strong cell that both stations can aim at.

Why does the signal sound like noise?

Raindrops move in the air, so the reflected signal has a Doppler spread of hundreds of hertz to kilohertz. A steady carrier arrives as a band of hissing noise.

Which modes work best?

Wide-bandwidth modes such as SSB with a wide filter, FM or CW at a moderate speed. Narrow, very slow modes lose energy in the Doppler spread.

Is it safe to operate during a thunderstorm?

Not near the storm. Keep a safe distance, use a protected position and stop if lightning is close.