A radio horizon you can lift

Seven Thousand Towers

Everyone half-believes a forgotten phone in seat 23C could nudge a jet off course. The documented story is stranger: the FAA relaxed gate-to-gate use of airplane-mode devices in 2013, while the FCC cellular rule dates to 1991 and was concerned with the network below. Lift the phone and watch that ground-facing problem grow. Then reverse the arrow.

A transmitter does not gain power when it climbs. It gains horizon. At street height the curved Earth hides almost everything. At cruise, the same line of sight opens a disc hundreds of kilometres wide.

geometric horizon
369 km
radius of the visible disc
ground disc
427,000 km²
π times horizon radius squared
physical towers in disc
7,500
average-density estimate, not active connections
d = sqrt(2 × 6,371 km × 10.668 km)

The third number is a scale model, not a phone log. WIA counted 142,100 physical US cellular towers at the end of 2022. Spread across an assumed 8.08 million km² contiguous-US area, that is about 0.0176 tower per km². The bench multiplies that average by the geometric disc. Real routes cross cities, deserts, water and mountains; antenna direction, terrain, receiver sensitivity and network handoffs decide which sites a handset can actually reach.

That is why the careful claim is roughly 7,500 physical towers lie inside the ideal line-of-sight disc, not that a phone talks to all of them. The FCC's 1991 engineering concern was airborne handsets upsetting terrestrial frequency reuse by being visible far beyond one intended cell. Modern networks are not frozen in 1991: the FCC later considered relaxing the prohibition, and onboard picocells make cellular service possible without shouting at the ground.

The arrow reverses near the runway

The serious 2021 interference dispute was not cabin phone to cockpit. It was ground network to aircraft: new US C-band transmitters sat below the band used by radio altimeters, instruments that measure height above terrain and feed landing systems. RTCA's 2020 assessment found a risk of harmful interference; telecom interests disputed parts of the analysis. The FAA nevertheless imposed operational limits while equipment and deployments were assessed.

selected 5G edge
3.98 GHz
historical comparison
to altimeter band
220 MHz
4,200 MHz minus selected edge
altimeter band
4.2–4.4 GHz
200 MHz allocation width
4,200 MHz - 3,980 MHz = 220 MHz

Move the dial to the 2022 EU comparison: a 3.8 GHz cap left 400 MHz to the 4.2 GHz altimeter edge, and the European Commission reported no harmful-interference incident there at that time. Frequency separation is not the whole risk calculation. Transmitter power, unwanted emissions, antenna geometry and how well a particular altimeter rejects energy outside its band all matter. This control demonstrates only the sourced edge-to-edge subtraction.

The 2021 to 2022 US response had real consequences: FAA airworthiness directives restricted some radio-altimeter-dependent operations where C-band interference was identified, particularly low-visibility operations near airports. It does not follow that 5G nearly crashed an aircraft. The risk was managed, aircraft were cleared or upgraded, and the relevant exposure was low altitude near terrestrial transmitters, not cruise.

The check

35,000 ft → 10.668 km high → 368.7 km horizon → 427,100 km² → 7,512 towers

Every arrow above is recomputed in the browser from the selected altitude: feet to kilometres; d = sqrt(2Rh) with mean Earth radius 6,371 km; disc area πd²; then area times 142,100 / 8,080,000 towers per km². The offline verifier independently repeats the chain and also compares it with the exact spherical tangent formula.

The free choices are the flat-horizon approximation, mean Earth radius, a circular unobstructed horizon, physical-tower rather than site or sector count, contiguous-US area rounded to 8.08 million km², and uniform average density. Atmospheric refraction, terrain, water, tower clustering, antenna patterns, propagation loss, device power and network protocol are omitted. So the tower result is an order of magnitude, not precision.

What the record can and cannot say

Reports and bench tests have suspected portable-device effects on older avionics. That is not the same as a conclusively reproduced cause or an accident attribution. The FAA's 2013 review concluded that most commercial aircraft could tolerate PED interference and allowed operators to expand airplane-mode use. An exhaustive universal claim that a phone has never produced any interference would go beyond that evidence, so this page does not make it.

Could not verify from one authoritative, exhaustive database the absolute claim that no accident anywhere has ever been attributed to a passenger phone. The narrower finding supported here is that the cited FAA material did not establish ordinary PED use as an accident cause and judged most commercial aircraft PED-tolerant. No handset-to-tower contact count, universal transmit power, interference probability or cancellation total is claimed.