Why average noise hides the problem: counting loud planes instead

Official noise maps are drawn in averages: Lden in Europe, DNL in the US, LAeq,16h in the UK. Lden, DNL and Lnight explained sets out how each is worked out. They are useful for comparing airports and years, but they describe a level no one hears, and they can hide exactly what people complain about.

Why an average doesn't match what you hear

People hear planes one at a time. An average adds up the sound energy of all of them and spreads it over the day, and that has three side effects.

• The loudest flights dominate. The FAA's own example: one flight a day with a sound exposure level of 114.4 dB, or a hundred flights at 94.4 dB each, give the same 65 dB DNL. A place with one very loud plane and a place with a hundred can sit on the same contour line.
• More flights barely move it. Doubling the number of flights adds 3 dB to the average, about the smallest change most people notice in a sound, although there are now twice as many planes.
• Nobody hears the average. In the Australian government's assessment for Western Sydney's new airport, a sequence of flights whose loudest peaked at 75 dB averaged 55 dB. The peaks are what interrupt a conversation.

The wind adds a fourth. Planes land and take off into the wind, so a street under the approach can have planes all day for a week and then almost none. A yearly average blends the two into a middle that never happens.

Our sample reports show how wide the gap can be. At Kew Green, 10 km from Heathrow, there are 92 loud landings a day with westerly winds and 4 loud flights a day with easterly winds, 66 on an average day. At Winthrop, 0.8 miles from Boston Logan, there are 130 loud take-offs a day with northerly winds and 8 with southerly winds.

Number-above metrics: N65, N70 and N60

A "number above" count is what it sounds like: the number of flights whose peak level passes a threshold at a place in a given period. If 100 planes pass and 50 of them go above 70 dB, the N70 is 50. The CAA points out the advantage: it is a plain count, not a logarithmic decibel figure. Its limit is the other side of the same coin: a plane at 66 dB and one at 86 dB each count as one.

The thresholds are chosen for what the noise interrupts:

• N70, used by day in Australia. A plane at 70 dB outside makes about 60 dB inside a room with the windows open, enough to disturb a conversation.
• N65, the UK's daytime count. The CAA's 2014 Survey of Noise Attitudes preferred it to N70 because in many areas flights were already quieter than 70 dB on the ground, and fleets keep getting quieter.
• N60, used at night in both countries. A plane at 60 dB outside makes about 50 dB inside with the windows open, close to the level at which a sleeping person may wake.

What 65 dB sounds like puts these levels next to everyday sounds.

Where counts are used officially

• Australia. Accepted practice for assessing changes at airports is to look at the ANEF, the country's average-based measure, together with N70 over 24 hours and N60 at night and over 24 hours. The counts grew out of a federal discussion paper of 2000 on new ways to describe aircraft noise, written in response to the reliance on the ANEF when Sydney's third runway was assessed.
• United Kingdom. The 2017 air navigation guidance required new flight-path proposals to consider N65 by day and N60 at night, particularly for communities further from the airport, and to use them to tell communities what to expect. The Air Navigation Guidance 2026, in force from 1 October 2026, bases the assessment on LAeq averages and no longer names them. Heathrow still publishes N65, N70 and N60 maps every year.
• Germany. The night protection zone around an airport is set by an average and by a count: for new or expanded airports, it reaches out to the line where planes pass 53 dB(A) indoors six times a night, with 15 dB allowed for the walls and windows; at existing airports the level is 57 dB(A).
• The EU. The Environmental Noise Directive maps in averages, but lists peak levels for protecting sleep, and places with few noise events, among the cases where extra measures may be worth using.

What the counts show at Heathrow

Heathrow's 2025 noise report, prepared by the CAA, gives populations for both kinds of map:

The counts are for an average day (07:00 to 23:00) and night (23:00 to 07:00) across the year. They measure something different from the averages, so the rows are not a like-for-like comparison. But they show the scale: about 140,000 more people live where 50 or more planes a day pass 65 dB than inside the 54 dB contour, the level at which UK policy places significant annoyance. At night, four times as many people live where ten or more planes pass 60 dB as inside the 50 dB Lnight contour.

What the research says

Averages and counts each do a different job.

• Averages predict a community's annoyance well. In the CAA's 2014 Survey of Noise Attitudes, the summer-day LAeq,16h matched average annoyance best, and counts of planes above 65 dB and 70 dB less closely. The survey found no evidence that any other measure did better, but noted that residents struggle with time-averaged, logarithmic figures, and recommended counts as a supplement.
• Counts track annoyance too. In the CAA's 2023-24 survey, with nearly 30,000 responses around ten airports, the share of people highly annoyed rose from 12% where no planes passed 65 dB on an average summer day to 25% at 50 planes, 34% at 100, 39% at 200 and 43% at 300.
• Old averages understated the response. The FAA's national survey, published in 2021, found 60-71% of people highly annoyed at 65 dB DNL, where the older curve behind US policy predicted 12%.
• Numbers and breaks matter. The Australian assessment notes that it is now well accepted that the number of arrivals and departures and the periods of respite also shape how people react.

So counts do not replace averages for policy. What they add is a figure people can check against their own day: how many loud planes, and at what time.

The limits of counting

A count has blind spots of its own:

• The threshold is a cliff edge. A plane at 64.9 dB does not count towards N65; one at 65.5 dB counts the same as one at 85 dB.
• It does not show how loud. Two places can have the same N70 with planes at 70 to 75 dB at one and 80 to 85 dB at the other.
• Zero does not mean silence. Beyond a certain distance every plane is below 65 dB, but in the CAA's survey around one person in ten was still highly annoyed there.

The fullest picture combines the count with the typical peak level, the time of day and how the count changes with the wind.

How our reports count loud planes

A Flight Noise report covers one address, from every recorded landing and take-off of the last twelve months. It counts a plane as loud when it passes 65 dB outdoors, loud enough to interrupt a conversation, and gives the number of loud planes a day and at night, the runway split and the heights, with real days replayed plane by plane. The noise is calculated with the method airports use for their official maps (ECAC Doc 29), from ADS-B flight data via ADSB.lol. It is an independent estimate, not an official noise map.