Gladiatorial games, part 11

(… continued from Gladiatorial games, part 10, posted on 21 Jan 2016).

Current HS2 noise policy – as set out in Information Paper E20 – is that noise mitigation for any location is only considered if the noise from HS2 is predicted to exceed the Lowest Observed Adverse Effect Level (LOAEL). The choice of an appropriate value for LOAEL is, accordingly, critical. The Promoter has prescribed level values for LOAEL – one for daytime and two for night are specified in Table 1 in Appendix B of E20 – but it appears that there is no consensus amongst acousticians as to what constitutes an appropriate value for LOAEL. This was made all too apparent by the fundamentally different interpretations of LOAEL and its application that were presented to the HS2 Select Committee by the Promoter’s acoustics expert, Rupert Thornely-Taylor, and, Rick Methold, the expert giving evidence on behalf of the Chiltern District Council (CDC) last November (see footnote 1).

You have to have some sympathy with the acousticians – this is difficult territory. There is no such thing as a “standard” human response to noise pollution. When subjected to an identical noise environment, people’s reactions may be poles apart; what some may find intolerable, others may barely notice, and reactions may change over time as people get accustomed to the noise – an effect known as habituation. In determining the effect that a source of noise pollution will have, predicting the sound level at a given location, referred to as the “impact”, is reliant upon a good understanding of the physics of sound propagation, whereas appreciating the “effect” that the resulting noise will have on the listener requires psychology to be applied.

As a preliminary to reporting on discussions on the meaning of LOAEL that have taken place before the HS2 Select Committee, I want to prepare the ground by inviting you to consider an imaginary experiment. Let’s assume that we have a recording of a source of noise that we can play back through speakers, with the ability to change the volume of playback in small steps – let’s say by 1dB at a time. Imagine that we have assembled a crowd of people, perhaps a hundred, or so, of them, in a large room front of those speakers, and we increase the noise level, step by step. The instruction that each member of the subject group has been given is to leave the room when they first find the noise level “annoying”. As the experiment progresses, we record the number of subjects leaving the room at each level step in noise, until the room is empty.

Now I’m not suggesting for one minute that any researcher contemplating investigating the human annoyance response to noise would design any experiment as crude as this. A real-world experiment would have to be far more sophisticated in order to avoid, amongst other undesirable distortions, the results being skewed by psychological influences such as habituation and herd instinct. This is why I have termed it “imaginary”, but I hope that, crude as it is, it will serve to illustrate what follows.

In order to analyse the outcome of this imaginary experiment, it would be informative to construct a bar chart, with the bars spread along the horizontal axis in single dB intervals and the vertical axis representing the number of people who left the room at each step, expressed as a percentage of the sample. If the height of the bars are then replaced by a best-fit curve, we will have something that looks like the image below.

Normal_distributionThis is what the Promoter’s expert witness on acoustics, Rupert Thornely-Taylor, described to the Select Committee as “a bell curve” (see footnote 2); mathematicians refer to it as a “normal distribution curve”. Its shape is defined by two parameters: the mean (μ), which sets the position along the horizontal axis, and the standard deviation (σ), or more correctly the variance which is the square of the standard deviation, which determines the spread, or “fatness” of the curve.

Since this was the result of an imaginary experiment, we can assume that there are no sampling or other experimental errors and the curve is an accurate representation of the characteristics of the population.

A very useful feature of the curve is that the total area under it to the left of the point on the horizontal axis that corresponds to a particular noise level represents the percentage of the population that is annoyed by noise at, or below, that level. If we consider, for example, that the noise is at the level that is two standard deviations below the mean (μ – 2σ), then only 2.5 per cent of the population will have been annoyed by that noise. At two standard deviations above the mean (μ + 2σ) the percentage annoyed will be 97.5 per cent, and at the mean level (μ) 50 per cent will be annoyed, so the mean is also the median.

The mean and the standard deviation of the curve will vary with the circumstances of the experiment, for example:

  • The nature of the noise. The susceptibility of people to noise measured at any one particular level may be influenced considerably by certain characteristics of the noise, such as whether it is continuous or intermittent, the frequency distribution of the noise energy and whether it is perceived as harmonious or jarring. So different transportation noise sources, such as road traffic, aircraft, conventional trains and high speed trains will require to be represented by different bell curves.
  • The description of the effect. So, for example, we would have seen a different curve if we had asked our experimental sample to indicate only when they were “very annoyed” by the noise.
  • The degree of habituation of the subjects to noise. People who are used to town living, for example, are likely to be less sensitive to noise than country folk.

In the real world, sensitivity to noise of a subject will also depend upon the circumstances that s/he finds themselves in, such as whether s/he is working, at leisure, outdoors/indoors, or trying to sleep.

With all this potential for variation, it is hardly surprising that deciding on an appropriate level for the LOAEL threshold is far from obvious, and was the subject of considerable difference between the two experts.

In the next part I will examine what help national noise policy is able to offer in this respect.

(To be continued …)


  1. Mr Methold appeared as an expert witness for Chiltern District Council (CDC): the evidence presented on behalf of CDC occupied the whole of the morning session held on Wednesday 4thNovember 2015 (video) and the Promoter’s response, including Mr Thornely-Taylor’s evidence was given in a shorter than usual afternoon session (video).
  2. See paragraph 105 of the transcript of the afternoon session of the HS2 Select Committee that was held on Wednesday 4thNovember 2015.


The illustration of a normal distribution curve has been extracted, with thanks, from the Wikipedia article Normal distribution.

I wish to thank Michael Woodhouse for his suggestions and comments, which I have found invaluable in preparing this series of blogs.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: