How to Diagnose Weather-Related Radio Dropouts

Back to Blog

First Checks: Confirm the Weather Correlation

Before touching hardware, prove the pattern is real, because a link can drop for reasons that only look weather-related. Pull the comms-failure or poll-timeout timestamps for the affected remote from the SCADA history and lay them over local weather records - rainfall, wind speed, and temperature - for the same period. A genuine correlation shows the outages clustering on rainy, windy, or hot-and-calm intervals rather than scattering randomly. If the outages do not actually line up with weather, you are chasing the wrong cause and should investigate power, interference, or the far end instead.

Identify which weather variable the failures track, because that single fact points at the likely physical cause. Drops that follow rainfall intensity suggest signal-level problems: either the link never had enough fade margin, or water is getting into the feedline. Drops that follow wind speed suggest a mechanical problem: an antenna mount that flexes or a connector that works loose under load. Drops that appear on hot, still afternoons and clear at night suggest an atmospheric effect. Sorting the outages by weather variable is the cheapest, fastest diagnostic you can run.

Where the radio logs received signal level over time, this is invaluable: overlay received level on the weather record too. A link with a slow received-level sag during rain, recovering afterward, is a fade-margin story; a link whose received level drops in sharp steps as the wind gusts is a mechanical story. This is the same correlate-symptom-to-cause discipline used in a cellular gateway dropout investigation, applied to a radio transport.

Rain-Correlated Drops: Fade Margin and Water Ingress

If outages track rainfall, the first suspect is insufficient fade margin. Rain adds attenuation on the path, and foliage in the Fresnel zone gets wetter and lossier when it rains, so a link that was only marginally above the receiver's sensitivity in clear weather dips below it in a downpour. Compare the clear-weather received level against the receiver sensitivity to see how much margin the link actually has; if it is thin, the fix is more margin, via higher-gain antennas, lower-loss coax, more height, or a repeater to shorten the hop. The concept and how much margin to carry is covered under fade margin.

The second rain suspect, and a very common one, is water in the coax or connectors. A pinhole in the jacket, an unsealed connector, or a missing drip loop lets rain wick into the feedline, and wet coax loses far more signal than dry, so received level collapses during and after rain and slowly recovers as it dries. Inspect every connector for corrosion or moisture, check the weatherproofing and the drip loop, and measure the feedline's return loss if you can; a feedline that reads fine when dry but degrades when wet is a leaking coax you should replace and reseal properly.

Distinguish the two by timing. A pure fade-margin problem tracks rain intensity closely and recovers as soon as the rain stops, because the path loss returns to normal immediately. A water-ingress problem lags: received level stays depressed after the rain ends and only recovers as the coax dries out over hours. That lag is the tell that sends you up the mast to the connectors rather than back to the link budget.

Wind-Correlated Drops: Mounts and Connectors

If outages track wind speed, the antenna or its feedline is almost certainly moving. A directional antenna that gusts even a couple of degrees off its peak loses signal, and a long-boom Yagi on an under-braced mount can swing enough in wind to drop below the fade margin at each gust. Inspect the mount for play, check that every clamp is torqued, and add bracing or guying if the mast visibly flexes. Re-peak the antenna afterward, following the antenna-aiming procedure, because a mount that has been working loose has probably drifted off its original aim.

Wind also works connectors loose over time, and a connector that is intermittently open under mechanical load produces exactly the sharp, gust-correlated received-level steps you may have seen in the trend overlay. Check that the coax is dressed and secured so wind cannot tug the connector, that there is a proper drip loop, and that connectors are tight and weatherproofed. A connector that reads fine at rest but crackles when you flex the cable is the mechanical equivalent of an intermittent wire and needs remaking.

Do not overlook the far end. Wind at the remote can flex the remote's antenna just as easily as wind at the master, and because the link is only as good as its worse direction, a remote whose antenna gusts off-peak drops the link even if the master end is rock-solid. If the master end checks out mechanically, inspect the remote's mount and feedline before concluding the wind correlation is unexplained.

Heat-and-Calm Drops: Atmospheric Ducting

If outages appear on hot, still afternoons and clear overnight, the cause is often atmospheric rather than anything on the tower. Temperature and humidity gradients in calm air bend radio waves, and under certain conditions the atmosphere sub-refracts, effectively raising the earth-bulge under your beam and stealing the Fresnel clearance you designed for. On a long or already-marginal path, an afternoon of adverse refraction can pull the link below the diffraction threshold even though nothing physical has changed. This is why path designs check a pessimistic effective-earth-radius factor, as discussed in verifying a radio path with Fresnel geometry.

Ducting can also cause the opposite problem: an inversion layer channels a distant signal into your receiver, raising the noise floor or letting a co-channel transmitter far away interfere for the duration of the inversion. If the received level looks fine during the outage but the link still fails, suspect interference rather than fading, and check whether the failures coincide with the calm, layered atmospheric conditions that support long-distance propagation. A spectrum check during an event, if you have the tools, will show a raised noise floor or an intruding carrier.

The durable fix for refraction-driven fades is margin and clearance, not a knob on the radio. If a path drops only in adverse atmospheric conditions, it was designed too close to the edge; adding height, gain, or a repeater to shorten the hop buys back the clearance and margin that the atmosphere occasionally borrows. For interference-driven ducting, a frequency change or a more directional antenna that rejects the off-path interferer is usually the answer.

When to Escalate

Escalate to a specialist path study when the correlation is clear but no single physical cause explains it, or when the fix requires changes beyond the site - a new tower, a licensed frequency change, or a repeater site acquisition. A marginal long path that fades under adverse refraction may genuinely need a re-engineered link rather than a field repair, and that is a design decision, not a maintenance one.

Escalate to spectrum coordination or the licensing authority if you suspect ducting-driven interference from another licensee, because resolving co-channel interference between operators is a regulatory matter, not something you can fix at your own antenna. Document the times, conditions, and any captured spectrum so the coordinator has evidence to work from.

For anything involving tower climbing, high-voltage-adjacent work, or lightning-protection systems, defer to qualified riggers and follow site safety procedures. Diagnosing the fault from the trend is desk work anyone can do; correcting a mount or feedline at height is not, and the diagnosis should hand a specific, well-evidenced task to the crew that goes up the tower.

Frequently Asked Questions