Using experience sampling method to study railway noise annoyance: a review

Marjolaine Roux, C. Philipps-Bertin, Olivier Chiello, Pierre Aumond, Arnaud Can, Anis Kacem, Guillaume Lemaître2024 · DOI

TL;DR

Current regulatory noise metrics fail to fully explain why high-speed trains (HST) are more annoying to residents than conventional rail. This paper reviews traditional study methods and proposes the Experience Sampling Method (ESM)—a real-time, "in-the-moment" diary approach—as a vital tool to bridge the gap between controlled laboratory precision and real-world ecological validity.

What Problem Does This Paper Address?

As high-speed rail networks expand, public acceptance is declining due to noise annoyance. Interestingly, high-speed trains (>250 km/h) are often perceived as more annoying than conventional trains, even when their average energy levels ($L_{DEN}$) are identical.

The authors argue that current indicators are too blunt; they average sound pressure over hours, missing the "fine-grained" acoustic signatures that drive human irritation, such as:

  • Suddenness: How quickly the noise ramps up.
  • Spectral Content: The specific low-frequency "rumble" of aero-acoustic sources.
  • Temporal Fluctuations: The rhythm and duration of a pass-by.
  • Event Density: How often the trains pass.

Existing research is split between social surveys (which are realistic but rely on faulty long-term memory) and lab experiments (which are precise but fail to capture the context of a person’s home life and activities).

Methods

The paper provides a comparative review of established methodologies before outlining a framework for a new ESM-based approach:

  1. In-situ Social Surveys: Retrospective questionnaires asking residents about their annoyance over the last 12 months.
  2. Laboratory Experiments:
    • Psychoacoustic approach: Participants focus intently on sounds and rate them.
  3. Cognitive/Distracted approach: Participants perform a task (like reading) while noise is played to measure "disturbance."
  4. Experience Sampling Method (ESM): The authors propose adapting this "diary" method for railway noise. In ESM, participants use a mobile device to report their annoyance in situ at various times while outdoor recording equipment captures the exact noise exposure for that specific moment.

The authors specifically detail three ESM protocols:

  • Interval-contingent: Reports at fixed times (e.g., every 2 hours).
  • Signal-contingent: Reports prompted by random notifications.
  • Event-contingent: Reports triggered immediately after a specific noise event (the preferred method for capturing the impact of individual train pass-bys).

Datasets

As this is a review and methodology proposal, the paper does not introduce a new primary dataset. However, it synthesizes findings from:

  • French Regulatory Thresholds: Comparing limits for "HT" (High-Speed) lines vs. conventional lines (see Table 1 in the paper).
  • The Miedema and Oudhsoorn Dose-Response Relationships: A foundational meta-analysis of transportation noise.
  • Aircraft Noise ESM Studies: Referencing the COSMA and NORAH studies as a blueprint for how ESM can be applied to transportation.

Key Insights and Ideas

  • The "One-Third" Rule: Social surveys suggest that acoustic factors only account for about one-third of the variance in annoyance. Another third is attributed to non-acoustic factors (social, individual, situational), and the final third remains unexplained.
  • Aero-acoustic Dominance: At speeds over 250 km/h, noise is dominated by air turbulence rather than wheel-rail contact, leading to more intense low-frequency content that current dBA-weighted metrics might undervalue.
  • The Memory Bias: Traditional surveys suffer from "retrospective bias." ESM eliminates this by capturing the "acute" annoyance before the participant has time to cognitively or emotionally regulate their response.
  • Ecological Validity: Unlike a lab, ESM captures annoyance while a person is actually living their life—cooking, watching TV, or trying to sleep—which significantly changes how noise is perceived.

Key Results

While the paper is a review, it highlights critical quantitative benchmarks in the field:

  • Threshold Disparity: In France, the $L_{DEN}$ threshold for high-speed lines is 68 dBA, compared to 73 dBA for conventional lines. This 5 dB difference is a regulatory acknowledgement that HST noise is inherently more disruptive.
  • Event Importance: Research by Lambert et al. showed that the number of train passages is a significant contributor to annoyance, particularly in the morning and evening, a factor often smoothed over by 24-hour averages.

Limitations and Future Work

The authors acknowledge that ESM is a "demanding exercise." It is resource-intensive for researchers and carries a high "participant burden," which can lead to drop-outs or careless responding.

Future Work: The authors are currently designing a pilot study to implement the ESM framework for HST lines. Key challenges they aim to solve include:

  • Determining the optimal number of daily notifications to maintain participant motivation.
  • Developing micro-incentive strategies to ensure truthful, real-time reporting.
  • Ensuring data privacy compliance (GDPR) while syncing outdoor noise recordings with indoor subjective reports.

Summary updated

Abstract

The goal of this article is to report on the development of a study of the acoustic factors driving the short-term annoyance experienced by neighbors of high-speed train lines.In fact, annoyance caused by high-speed trains (> 250 km/h) is not completely modeled by current indicators (e.g., LDEN).For example, the suddenness, spectral content, temporal fluctuations of pass-by noises, and the density of train passages seem to significantly impact annoyance in a way that is not fully understood [1].To investigate these aspects, we first review different approaches used to study annoyance caused by transportation noise.For railway noise, two main approaches are reported in the literature: insitu social surveys exploring the contributions of acoustic and non-acoustic factors to long-term annoyance; laboratory experiments, in which controlled stimuli (individual pass-by noises) are played back to participants who rate their annoyance.Another approach is also used in transportation noise studies, but not yet for railway noise: diary, or experience sampling methods, whereby neighbors report their annoyance, at home, at different times over a longer time span (usually several weeks) while noise exposure is simultaneously recorded [2].Such in situ approach allows experimenters to focus on the precise characteristics of each pass-by noise and consider annoyance in the context of participants' real environment
BibTeX
@misc{roux2024using,
  author = {Marjolaine Roux and C. Philipps-Bertin and Olivier Chiello and Pierre Aumond and Arnaud Can and Anis Kacem and Guillaume Lemaître},
  title = {{Using experience sampling method to study railway noise annoyance: a review}},
  year = {2024},
  doi = {10.61782/fa.2023.0808},
}
Figure