Low-altitude aircraft will reshape noise exposure across global cities

1 Introduction

• Low-altitude aviation is developing as a globally competitive urban industry, with initiatives spanning passenger, cargo and public-service operations.Examples include eVTOL services, drone logistics, airspace trials and supporting infrastructure.Source: §1 · p. 1
• Urban noise is a health burden, while existing noise policy mainly addresses roads, railways and conventional airport aircraft.Low-altitude aircraft would add repeated aerial sound above streets, façades, courtyards and public spaces.Source: §1 · p. 1
• Drones and passenger eVTOLs differ from established transport sources in height, spectrum, directivity and timing, and may become regular sounds along scheduled routes.The study groups cargo delivery drones and passenger eVTOLs as low-altitude aircraft; multirotor drones commonly have prominent tones and high-frequency broadband components.Source: §1 · p. 1
• Urban form redistributes transport noise across streets, courtyards, façades and building heights, so low-altitude aircraft may expose spaces sheltered from ground traffic.The passage identifies street layout, building geometry and urban structures as factors shaping horizontal and vertical exposure.Source: §1 · pp. 1–2
• Because low-altitude aircraft exposure in dense cities has been less examined, the study asks how operations reshape horizontal and vertical exposure across ten city districts.It uses drone and passenger eVTOL configurations and assesses horizontal receiver planes and vertical façade grids.Source: §1 · p. 2

2.1 Spatial distribution of L-aircraft-induced noise increases

• About 65 dB(A) was reached directly beneath the route under high-flow passenger eVTOL operations, around 20 dB(A) above the low-flow delivery-drone scenario.The primary exposure footprint expanded with source strength and corridor flow.Source: §2.1 · p. 3, Figure 2 · p. 4
• Urban form shaped horizontal exposure: continuous blocks sheltered interiors, while openings and low-obstruction corridors provided clearer sound paths and larger local increases.Greater building coverage may reduce broad exposure while increasing contrasts between neighboring spaces.Source: §2.1 · pp. 4, 8
• Above 20 dB(A) increases occurred in some road-shielded London and New York courtyards, while already noisy corridors often changed less.The aircraft contribution can remain apparent in quiet spaces even where road noise dominates nearby areas.Source: §2.1 · pp. 4–5, Figure 3 · p. 5
• Noticeable increases faded within approximately 75 m in Tokyo but persisted 315–360 m in some cities, making a uniform route buffer uneven across cities.Open urban structures and quieter receptor areas allowed contributions to remain apparent farther from the corridor.Source: Table 1 · p. 7, §2.1 · p. 7

2.2 Vertical Façade Exposure

• Upper or intermediate floors near the 100 m route could be more exposed than lower and higher floors, with terrain bringing even low buildings closer to the corridor.The pattern became more pronounced as source level or corridor flow increased.Source: §2.2 · p. 8, Figure 5 · p. 9
• At comparable heights, façade orientation and nearby buildings created sharp contrasts, including a 76.6 versus 69.3 dB(A) maximum across adjacent Shenzhen façades.Narrow gaps and neighboring built masses also concentrated exposure over limited façade sections.Source: §2.2 · p. 8, Figure 5 · p. 9
• Noise intrusion occurred where blocks sheltered locations from road traffic but left them exposed to aircraft sound; Hong Kong façade ΔLden averaged 15.5 dB(A).The maximum Hong Kong façade ΔLden was 17.7 dB(A); ground-level increases exceeded 20 dB(A) in some locations.Source: §2.2 · pp. 8–9, Figure 5 · p. 9
• New York and London showed approximately 10–15 dB insertion loss over much of the spectrum, and P10 building height was the strongest correlate of attenuation.Morphology affected overall attenuation more strongly than spectral shape; these correlations were exploratory across ten districts.Source: §2.2 · p. 10, Figure 6 · p. 11

• Across ten districts, identical aircraft operations reshaped horizontal and vertical exposure differently through source strength, corridor flow, road noise and urban form.Source: §3 · p. 11
• More than 20 dB(A) increases occurred in some road-shielded courtyards, so route assessments should distinguish newly exposed spaces from already noisy areas.Source: §3 · p. 11
• Route assessments should combine horizontal maps, façade receivers and vertical sections with aircraft-only exposure, induced change, newly exposed area and spatial influence.Source: §3 · p. 12
• Routes can avoid courtyard openings and building gaps, while staggered openings, absorptive treatments, screens and roof geometry may reduce exposure inequalities.Source: §3 · p. 12
• Exposure varied non-monotonically with height, often peaking near flight altitude; route planning should consider shielding, façade orientation and terrain.Source: §3 · p. 12
• Spectral attenuation slopes were mostly below 1 dB per octave; averaging within frequency bands and ray tracing’s weaker low-frequency diffraction may contribute.Source: §3 · p. 12
• The simulations cover level flyovers at 100 m with harmonised source spectra and assumed flows, omitting flight phases, directivity, meteorology and local demand.Each city is represented by one 2 × 2 km district, and road inputs were standardised where detailed local data were unavailable.Source: §3 · p. 12

4.1 City Selection

• Ten central 2 × 2 km districts were selected from 49 aviation-relevant candidates after regional population/GDP screening, balancing morphology, geographic coverage and development stage.Each district was chosen where consistent building, road, terrain and traffic data were available.Source: §4.1 · p. 13
• The districts span four propagation-relevant urban forms, from discontinuous towers and continuous mid/high-rise blocks to fine-grained and open low-rise settings.These forms vary in building height, coverage and urban grain, which provide context for interpreting noise exposure patterns.Source: §4.1 · p. 14

4.2 L-aircraft

• Five aircraft spectra were converted to common one-third-octave bands; the DJI Matrice 600 and Joby eVTOL supplied the simulation inputs.Because measurement conditions differed across sources, spectral comparisons were limited to relative shape.Source: §4.2 · pp. 14–15

4.3 Traffic

• Road traffic was estimated from available traffic counts and AADT, grouped into three urban-road classes, with standardized road widths and approximate time-of-day profiles.Motorways and other grade-separated high-speed roads were excluded; where temporal distributions were unavailable, traffic was split 70% day, 20% evening and 10% night.Source: §4.3 · p. 16

4.4 Simulation Setting

• Three-dimensional simulations compared road-only, aircraft-only and combined noise scenarios using fixed geometry, 100 m cruise altitude, gridded maps and façade receivers.The study used a 55 dB(A) screening threshold and a 3 dB(A) change criterion; the 2 × 2 km domain was justified with a free-field propagation estimate.Source: §4.4 · pp. 16–17

• Across ten districts on all inhabited continents, identical low-altitude aircraft operations produced markedly different noise exposure across and within cities.The variation means low-altitude-aircraft-only maps or single-height assessments cannot capture the resulting exposure patterns.Source: §5 · p. 17
• Some relatively quiet semi-enclosed spaces experienced aircraft-induced increases above 20 dB(A), despite modest district-wide changes.The horizontal inequalities reflected existing road-noise conditions and urban morphology, which also strongly influenced overall attenuation.Source: §5 · p. 17
• Exposure varied with height and acoustic visibility to the flight route, creating inequalities between storeys, façades and buildings at comparable heights.Route assessment should distinguish newly exposed from already exposed areas and combine horizontal maps, façade receivers and vertical sections.Source: §5 · p. 17

• Custom Python scripts for receiver matching, exposure metrics, morphology analysis and figure generation are available from the corresponding author upon reasonable request.Source: Code availability · p. 17
• CadnaA is proprietary software and is not redistributed.Source: Code availability · p. 18

• Tianjing Feng and Jian Kang are listed at the Institute for Environmental Design and Engineering, The Bartlett, University College London.Source: Authors and Affiliations · p. 18
• Feng conceptualised, implemented, analysed and drafted the study; Feng and Kang interpreted and revised it, while Kang supervised and administered the project.Source: Authors and Affiliations · p. 18
• The operational scenarios are Q25 (25 L-aircraft h⁻¹ daytime, 13 evening) and Q150 (150 L-aircraft h⁻¹ daytime, 75 evening), with no night-time operations in either scenario.Source: Authors and Affiliations · p. 18