The configuration replays every convective season: at night, in cruise, a line of thunderstorms across the route, and on the screen two red echoes separated by a black corridor a few dozen nautical miles wide. The deviation of a few degrees seems obvious, it is requested, it is granted — and the aircraft passes through hail in the middle of what was, on the screen, only empty space. The corridor did not exist: it was the shadow of the first cell. Three physical mechanisms are enough to explain it, all known, and all forgotten at the precise moment the decision is made.
What the X-band measures, and what it ignores
The onboard radar does not see the cloud, turbulence, or updraft. It measures the energy returned by hydrometeors large enough to backscatter a centimeter wave, and nothing else. Everything else is deduction. The first consequence is arithmetic: in the domain where the equipment operates, the power returned by a drop varies as the sixth power of its diameter. Doubling the size of the drops multiplies the echo by sixty-four. A dense drizzle remains green when a downpour of large drops saturates the screen, even though both can produce the same water height on the ground.
The second consequence is less often drawn: the phase counts as much as the size. At equal diameter, a crystal or a dry hailstone returns about five times less than a water drop, because ice polarizes much less under the wave field. A hailstone coated with a film of water, on the other hand, behaves like a drop of its own size — hence the violent echoes of the wet hail cores, and the near-silence of the same dry hail a few thousand feet higher.
Yet, in cruise, it is exactly this icy summit that the beam crosses. The anvil, the column of crystals above the freezing level, the cirrus blown downwind: none of this paints. The reflective part of the thunderstorm lives below, where the water is liquid. A crew that leaves the tilt at zero at FL350 is not looking at a thunderstorm, it is looking at the little that the ice is willing to give back — and it concludes on the size of a cell based on its most discreet part.
The radar only sees large water drops, not ice. In cruise, it scans the icy top of storms, which reflects little, making the storm appear smaller and weaker than it really is.
An A320 at FL350 encounters a thunderstorm with a wet hail core at 15,000 feet. The radar at zero tilt shows a weak echo, misleading the crew to think the storm is less intense than it actually is, leading to a dangerous deviation.
Attenuation: the first cell eats the beam
The choice of a wavelength of a few centimeters is not arbitrary: it is the only one that allows, with an antenna fitting in a nose radome, a beam narrow enough to separate two cells at a useful distance. The price to pay is absorption. Liquid water absorbs and scatters more strongly as the frequency increases, and the wave pays twice: on the way to the target, on the way back to the antenna. Across a rainy core, the loss is counted in tens of decibels; what comes out the other side has little to do with what went in.
The screen does not display what it does not receive. It has no way of writing « I no longer measure anything » in place of the black, and the black of the shadow is the same black as that of the clear sky. Two practical effects follow, and the second is more insidious than the first.
The first is the shadow itself: a second cell aligned behind the first disappears, all the more completely the more intense the first is. In other words, the more dangerous the situation, the cleaner the corridor behind appears. The second effect distorts the reading of the cell itself: its near flank is measured at full energy, its far flank through the entire core. That flank always appears softer than it is. The eye, which naturally seeks the softest side to pass, is therefore systematically drawn to the least well-measured side.
The trap. The corridor does not close at the last moment: it was never open. And when the second cell finally appears — most often when the aircraft arrives abeam of the first and the beam no longer has to cross it — it is read as a rapid development and not as the end of a shadow. The wrong mental model then sets in at the worst moment: the crew believes they are witnessing an exploding convection, therefore unpredictable, when they are simply discovering what was masked from the beginning, and they start looking for a new escape route instead of bypassing an ensemble they already knew. What lies behind a red core is not free: it is unknown, and the unknown is treated as occupied.
The ground shadow test
There is a simple way to know if you are looking at a gap or a shadow, and it costs only a gesture: lowering the tilt to paint the ground echo. The terrain returns a strong and almost continuous signal; it constitutes a free reference source, distributed over the entire sector. A cell that attenuates the beam deprives of signal everything behind it, including the ground. It therefore carves out a black corner in the ground echo aligned with it, open towards the outside of the sector, and perfectly recognizable.
The reading then becomes binary. If the ground echo is present behind the gap, the beam reaches it: the gap is real, and the absence of precipitation echo there has meaning. If the ground echo fades behind the cell, the beam no longer reaches it: what is seen beyond is not information, it is a blind spot. No other manipulation gives this answer with as much clarity.
The limit of the method is geographical. Over the sea, the surface echo is weak at grazing angles, especially in calm seas, and the test loses most of its value. It is precisely there — oceanic, night, organized convection — that it would be most needed. On these sectors, other indicators must be relied upon: the persistence of a gap when the tilt is changed, its geometry relative to the cell axis, and the behavior of the corridor as it is approached. A real gap widens and becomes more precise; a shadow remains aligned with the cell that produces it and closes suddenly.
The right reflex. Before entering a gap, the tilt is lowered until the ground echo appears, and it is verified that it is present behind the gap, not just in front. A complete tilt sweep, from the lowest to the highest, takes about ten seconds and is done before the point of no return of the deviation, not during.
Tilt and gain: the two settings that create voids
The tilt is the setting that decides the slice of atmosphere observed, and its geometry is merciless: the beam opens by about a hundred feet per degree of opening and per mile of distance. A beam with an opening of about three degrees already covers several thousand feet at about fifty miles. Two opposite consequences. On the one hand, dilution: an intense but thin core occupies a fraction of the illuminated volume, and the average energy returned underestimates its violence. On the other hand, over-scanning: at zero tilt in cruise, the beam axis remains at the level of the aircraft, far above the liquid water, and the cell appears to be born as it is approached.
That is why a fixed tilt makes no sense. The tilt is used as a cutting plane that is moved: it is lowered to paint the ground to set the scale and flush out the shadows, it is raised to assess the height of the reflective summits, it is lowered again to find the core. The useful setting changes with the distance, with the flight level, and every time the longitudinal attitude is modified, even on a stabilized equipment.
The gain, on the other hand, is what gives meaning to the colors. In the calibrated position, each color threshold corresponds to a known reflectivity level; it is the only position where red means something comparable from one flight to another. Manually increasing the gain brings out weak returns, which helps to probe a suspicious shadow zone, but makes the scale misleading. Lowering it reveals the structure of the strongest core by erasing the rest. Both maneuvers are legitimate; what is not, is to forget them in position. A gain left low paints a calm sky above a line of showers, and nothing on the screen signals the anomaly.
What the automatic compensation recovers, and where it stops
Modern equipment applies path attenuation compensation: the receiver increases its amplification with distance, taking into account the reflectivity already crossed, to restore the true intensity to distant echoes. This works as long as there is something left to amplify. Beyond a certain cumulative attenuation, there is no signal left, only noise, and no processing creates information from nothing.
Good equipment indicates this. When the compensation reaches its limit, they mark the sector that has become unusable — band on the outer arc, particular color, hatching — and this mark must be read as an instrument, not as a decoration: it means « from here I am blind in this direction ». Many crews know it poorly, because it only appears in situations where there is something else to do than to question a symbol. Automatic volumetric scanning radars, which memorize several tilts and compose an image free of ground echo, eliminate the setting error, not the physics of absorption: the shadow remains, only the way to signal it changes.
Other sources that do not suffer from the same defect remain. Lightning detection detects the discharge, not the precipitation: it sees through the shadow and signals an active core where the radar no longer paints anything. Onboard relayed ground imagery works at much larger wavelengths, which rain attenuates much less; it most often sees the second cell, but with several minutes of delay and an age that is not uniform across the mosaic.
| Criterion | Airborne radar, centimetre band | Ground imagery relayed on board |
|---|---|---|
| What is measured | Backscatter from hydrometeors in a narrow beam, from the aircraft, in real time | Composite of several ground radars, seen from below and from multiple angles |
| Attenuation by rain | Strong, increasing with the intensity crossed; creates shadows | Much weaker at longer wavelengths; one cell rarely masks the next |
| Frozen top and anvil | Almost invisible: dry ice returns little | Also weakly reflective, but the mosaic reduces blind spots |
| Freshness | Real time | Several minutes, and a non-uniform age across the area |
| Useful resolution | Fine close to the aircraft, degraded at range by the beam spread | Uniform but gridded, without tactical detail |
| Defensible use | Tactical avoidance, heading decisions for the next few minutes | Strategy: choice of route, level, diversion |
Deciding: gradient, shape, margin, and side of bypass
Once the image is deemed usable, it is not the maximum color that dictates the decision, but the gradient. A transition from green to red over two or three miles signifies violent shear between the updraft and the precipitation column; the same red reached gradually over about fifteen miles describes a thick rainy area, unpleasant but not comparable. Shapes add to the gradient: scalloped edge, hook, finger protruding from the flank, notch under an echo overhang at altitude. All say the same thing — the cell is organized, it has a vertical structure, and it ejects.

- Gradient. A sharp change from green to red in a short distance indicates dangerous turbulence and hail.
- Shape. Hooks, fingers, or notches on a storm’s edge suggest it’s organized and ejecting hail.
- Margin. Always leave at least 20 nautical miles of space from the storm, more at higher altitudes.
- Side of bypass. Always pass upwind of the storm, never downwind, to avoid the anvil and hail.
The usual lateral margin is expressed in tens of nautical miles, and it must increase with the flight level, because the anvil and the projected hail extend far beyond the rainy core, which is painted. The bypass is done upwind of the cell, never downwind, where the anvil, the carried hail, and the clear-air turbulence are located. Passing above is the worst compromise: the top visible on the screen is not the top of the cloud but the top of the reflective part, the maneuvering margin in thrust is low at the ceiling, and the icy layer that seems harmless on the screen is not in reality.
| What the screen shows | Most likely mechanism | Conduite |
|---|---|---|
| Black corridor between two strong returns, with no ground return beyond | Attenuation shadow | Treat the whole as a single area and avoid it upwind |
| Strong return close by, nothing above, low tops | Rain core below the freezing level, frozen top not seen | Draw no conclusion about the real height of the tops from the display |
| Scalloped edge, hook or finger on one flank | Organised cell, probable hail ejection | Increase the margin, avoid the downwind side |
| Uniform pale green across the whole sector | Gain left in a low position, or tilt too high | Return to calibrated gain, scan in tilt before any decision |
| Ground return flooding the display ahead | Tilt too low for the range displayed | Raise the tilt, otherwise the useful layer stays masked |
What to open before the next flight
Three checks are better than a theoretical revision. The first concerns the installed equipment: does it signal an over-attenuated sector, and in what exact form? The manufacturer’s documentation states it, training often forgets it, and a symbol discovered at the moment it appears is useless. The second concerns the automatic mode: what it chooses as tilts, what it displays of the actually used tilt, and what it does with the ground echo it removes — because removing the ground echo from the image is also removing the indicator that allowed detecting the shadow.
The third concerns maintenance and everyone on board. An eroded radome, repaired out of specification, repainted with an unapproved product, or simply soaked with water attenuates the signal in all directions at once. The defect does not resemble a failure: it resembles good weather. No anomaly message accompanies it, and the only way to catch it is to compare what the screen paints with what is seen outside, or with what the weather briefing announced.
The gray area remains, and it must be named: no text says how to interpret a black corridor. Regulations require carrying a means of detection and avoiding dangerous phenomena, the operator’s documentation sets margins, but the reading of the echo is a matter of judgment and training. For those preparing for a type rating or line transition, this is a subject where the examiner expects a method stated aloud, not a distance memorized: saying what is being looked for, with which tilt, and what is concluded. The FL330 training courses and the experience shared on the forum actually revolve more around this method than the thresholds themselves.
This text describes mechanisms, it does not set any rules: the applicable specifications, the aircraft flight manual, and the operator’s approved documentation prevail over everything that precedes, even when they state the opposite.

Leave a Reply
You must be logged in to post a comment.