| [Download PDF]
|Year : 1999 | Volume
| Issue : 3 | Page : 11--16
Non-acoustical factors in environmental noise
Ian H Flindell1, Pieter Jan M Stallen2,
1 Institute of Sound and Vibration Research, University of Southampton, UK
2 Stallen & Smit, Gentiaanstraat 13, 6813 ES ARNHEM, The Netherlands
Ian H Flindell
Institute of Sound and Vibration Research, University of Southampton, UK
|How to cite this article:|
Flindell IH, Stallen PM. Non-acoustical factors in environmental noise.Noise Health 1999;1:11-16
|How to cite this URL:|
Flindell IH, Stallen PM. Non-acoustical factors in environmental noise. Noise Health [serial online] 1999 [cited 2020 Aug 8 ];1:11-16
Available from: http://www.noiseandhealth.org/text.asp?1999/1/3/11/31717
The Netherlands Directorate-General of Civil Aviation (DGCA) organised a symposium entitled 'Significance of non-acoustical factors' in Den Haag on September 30th and October 1st 1998. The symposium was held mainly because, as Mr Jacques Remmen (Deputy Director of the DGCA and the first speaker) explained, noise issues were becoming a matter of increasing importance for aviation in the Netherlands and that the situation around Schiphol (Amsterdam) Airport had become critical because of it. The DGCA had increasingly perceived a clear gap between established theories of noise annoyance and the actual effects as demonstrated by the numbers of complaints received. They were beginning to feel that effective noise management was not just about noise levels and noise limits expressed in simple decibel terms, but that there were many other factors involved which they also wanted to understand. The purpose of the symposium was to assess the extent of scientific agreement on the significance of non-acoustic factors; to encourage debate on some of the wider issues; and if possible, to help devise better solutions than have existed in the past.
The number of delegates invited to attend the symposium was purposely kept small so as to facilitate a free-ranging debate. Invited presentations were made by Dr Pieter Jan Stallen, Peter Smit and Rogier Noyon, The Netherlands; Prof. Soames Job, Australia; Dr Sandford Fidell, USA; and Dr Ian Flindell, UK. Other participants included Prof. Rainer Guski, Germany; Prof. Jim Kahan, Fokke van der Ploeg, Dr Eric Lebret, Dr Henk Meidema, Ronald de Jong, The Netherlands. The symposium was also attended by a number of civil servants from the relevant Ministries. The discussions were wide-ranging and identified a great number of relevant non-acoustical factors. There was general agreement that many of these factors can be of considerable importance in understanding reported annoyance responses.
The purpose of this paper is to introduce the subject of non-acoustical factors and to touch on some of the wider issues debated at the symposium but not necessarily represented by the papers reproduced here.
Non-acoustical factors: an inventory
The main focus of the symposium was noise annoyance effects attributable to environmental or community noise. There are other outcome variables of interest, chiefly sleep disturbance and the possibility of non-auditory health effects, but even these might be expected to be influenced by a respondent's state of mind, which in turn might well be influenced by nonacoustic as well as acoustic input variables.
The starting point for the symposium was a review of the many factors found or not found to influence reported annoyance across the 136 surveys critically examined by Fields (1993). Following on from the invited presentations and the subsequent discussions a list of relevant factors was drawn up (in no particular order) as follows:
Perceived predictability; this refers to the perceived likelihood of future noise level increases (or decreases)
Perceived control; either by the individual or by others, e.g. by airport management (see also Stallen, this issue);
Trust and recognition; the latter refers to perceptions that impacts are recognised by authorities.
Voice; this refers to the extent to which people are able to speak to and be listened to by authorities;
General attitudes; such as awareness of economic and social benefits of the undertaking generating the noise such as an airport, awareness of noise control action costs, fear of crashes, etc.;
Personal benefits; employment at the airport or industry concerned could be one example;
Compensation; noise insulation or house purchase schemes could be considered as typical examples;
Sensitivity to noise; it is widely recognised that there are individual differences in sensitivity to noise (see also Job, this issue);
Home ownership; home owners might be concerned about effects on the value of their property;
Accessibility to information; this could be seen as a means of influencing general attitudes, but it can also affect the extent to which authorities are perceived to be taking an interest in the noise exposed community;
Understanding; this refers to all relevant issues, it is quite possible for general attitudes to be based on irrelevant or incorrect information;
Personal benefits, compensation and home ownership can be determined objectively, but these three factors can still have a psychological effect which can be extremely dependant on many of the other factors listed above. All the other non-acoustical factors are inherently subjective and this has created difficulties in the past for any regulatory authority wishing to take them into account. The problem lies in the extent to which practical policies affecting one or more of these non-acoustical factors might be seen to lead to predictable outcomes in the same sense that the effects of engineering noise control can usually be measured in terms of some more or less easily quantified reduction in noise level.
There was a considerable amount of discussion at the symposium on the extent to which these and other factors (the list was not intended to be exhaustive) might be influenced by alternative noise management policies. Many regulatory authorities take a number of non-acoustical factors into account already (see Flindell and Witter, this issue). However, there are a number of non-acoustical factors which regulatory authorities might choose to leave well alone. Individual differences in noise sensitivity, for example, can explain a large proportion of total variance in annoyance responses, but the extent to which regulatory authorities can or should take these differences into account is debatable. Those components of individual differences in noise sensitivity which are purely idiosyncratic might justifiably be treated differently from those components which research might show can be associated with demographic or health status variables. If there is a choice, and in the case of transport noise management this is sometimes a possibility, should noise exposure be concentrated on one particular section of the community or another, perhaps depending on the relative proportions of residents expressing different degrees of noise sensitivity in different areas, or should everyone be entitled to expect the same treatment, regardless of differences in individual sensitivity? It is a matter for scientific research to determine the 'facts' of any such differences, but it then becomes an entirely political issue when the interests of one section of society must be balanced against another.
What does 'reported annoyance' mean?
The term 'reported annoyance' is used here to describe the type of general adverse response to community noise as measured by an appropriate questionnaire scale or similar instrument. 'Reported annoyance' need not necessarily be associated with actual or acute annoyance whereby the respondent might be motivated to take some kind of action or is otherwise in a highly emotional state at that time, but instead, is normally used as a catch-all for general feelings of dissatisfaction or 'bother'.
Of course, there is an obvious difficulty here, in that the standard questionnaire items used in noise research could potentially be understood by different people to mean subtly different things in different situations. A considerable amount of effort has been devoted over many years to the development of dose-effect relationships relating input (noise exposure) variables to outcome (annoyance) variables by some appropriate empirical model (For the most recent example of this type of research, see Miedema and Vos, 1998). All such dose-effect relationships are intended for the purpose of assessing relative noise impact when the degree of physical noise exposure is known. It is well known that there is often considerable variation between the dose-effect curves found for different studies and for different types of noise source, but the interpretation of these differences is not just a matter of statistical analysis. If the outcome variables used cannot be precisely interpreted, then this must be an important issue whenever decisions are made to focus noise management on engineering noise control or on the reduction of noise effects or on some combination of the two.
In social surveys, reported annoyance can be clearly defined in terms of the actual questionnaire items used, and this underlies current international efforts to promote standardisation in this area (see Fields et al, 1997, and current efforts by Team 6 of the International Commission on the Biological Effects of Noise (Hatfield, 1999)). If the same questionnaire wording is used in all studies, then this clearly removes an unnecessary source of additional variability. On the other hand, the standardisation of the questionnaire items used does not overcome the possibility of alternative interpretations referred to above. For example, selecting a particular point on any noise annoyance scale could be related to an unknown extent to that individual's overall perception of their general quality of life or to their overall perception of the general environment in which they live. It is well-known that respondents will often describe the area in which they live as being reasonably or very 'quiet' while at the same time reporting on a subsequent questionnaire item that aircraft noise is very troublesome for them. An over-emphasis on the results obtained using either the first or the second style of question could be misleading in policy terms. From a psychological point of view it is perfectly plausible that an individual could perceive their overall environment as being entirely satisfactory whilst still resenting any particular intrusions into it.
Intermittent environmental noise events could easily be perceived as unwanted intrusions which do not necessarily cause any permanent change in the environment into which the intrusion occurs, and indeed, as soon as the event has passed it leaves no permanent trace (in this respect noise is unlike any other pollutant).
Moreover, if noise events are perceived as taking place externally to the local environment, this also raises the question as to whether respondents are really judging the effect on them as individuals (proximal judgement) or the noise magnitude of the event at source (distal judgement). For example, an aircraft event reported as 'noisy' or 'annoying' could either be having a significant effect on the state of mind of the respondent or it might be perceived as a distant object not particularly contributing to an altered state of mind but instead as an object which must be generating considerable sound power at source.
What does 'exposure' or noise 'dose' mean?
A similar issue is that of defining the input variables for use in any dose-effect relationship. Regulatory noise policy is almost invariably focused on physical noise levels whereas effective noise management usually involves non-acoustic factors as well. The extent to which non-acoustic factors are taken into account generally relies either on the degree of enlightenment adopted by the regulatory authority or on the degree of flexibility for adaptation to different circumstances permitted by any written regulations in force at the time. The main problem here, and it is almost universally recognised, is that noise exposure level never accounts for more than a small proportion of the total variance of any outcome variable considered (see Guski, this issue). Different authors disagree as to the precise amount of variance accounted for, and partly this is because the relevant variables are often defined in different ways, but it is clear that there are many other contributory factors than noise level alone.
There are really only two conventional approaches to dealing with this problem. First, the physical noise exposure indicator is carefully selected so that it most closely represents those aspects of the acoustic environment which are considered to be the most relevant to the outcome variables considered. Examples of this first approach are the use of the A-frequency weighting to approximate to the assumed frequency response of the ear, and the so-called equal-energy trade-off implied by the use of LAeq. Both the A-weighting and the equalenergy trade-off are constantly criticised in the literature for not quite doing the job they were designed to do, but there really is nothing better available for general use.
However, because noise exposure measured in LAeq still explains only a small proportion of the total variance in the outcome variables (however defined) a second approach is required. This involves investigating and then applying corrections for all other input variables found to be relevant in terms of decibel equivalent values. The amount by which the level of reported annoyance changes because some additional factor is present can be represented by the amount by which the physical noise level would have to change to give the same changed reported annoyance. This difference in physical noise level is known as the decibel equivalent of the effect of the factor under consideration. There are problems with this approach. First, it is unusual for the effect of change in either physical noise level or in the presence of other (non-acoustic) factors to be included in any research studies. Typically, changes in other variables are related in terms of equivalent changes in physical noise level by comparing different levels of reported annoyance for static or steady-state conditions using cross-sectional research designs. A more complete understanding of how people respond to change requires longitudinal research addressed to dynamic situations. Secondly, the list of other input variables is very long and probably, different variables are more or less important in different situations depending on local circumstances. Thirdly, there has been insufficient debate as to precisely which 'other input variables' should be taken into account in noise regulations and assessment methods and which should be left alone as being matters of individual preference or as matters more appropriately dealt with as part of a political process. Although we cannot offer any definite solutions to these particular problems at this time, it is becoming increasingly clear that the conventional decibel equivalent approach to dealing with non-acoustic factors may not be optimum in the longer term.
How much noise is too much noise?
This question is fundamental to all noise policy decisions, yet in the field of environmental and community noise, most researchers avoid it if they can. Environmental noise targets or limits could be set in terms of physical noise levels, or they could be set in terms of outcome variables, such as the percentage of respondents expected to be 'highly annoyed' by the noise. Estimating the cost of noise control down to some defined level of physical noise exposure can (in theory at least) be achieved through what are largely engineering methods, whereas predicting what this would mean in terms of the real impacts on people can be far more difficult. Unfortunately, while researchers usually appreciate the difference between these two concepts, it is not always obvious to regulatory authorities who can sometimes be forced to make important decisions on the basis of what might be completely inadequate information.
There have been lots of ideas but no consensus as to the best way to deal with the question of 'how much is too much noise?'. As soon as we begin to consider the role and importance of nonacoustic factors, the question becomes even more difficult to answer, and indeed there probably is no simple answer anyway. On the other hand, few people would argue with the proposition that there is too much environmental noise and having a bit less of it would be a good thing, providing that the economic and social costs of noise control are affordable. Should we wait until we have a precise predictive model of the causes of reported noise annoyance or should we take action now?
Perhaps 'how much is too much noise?' is the wrong question. What about; 'which of these two options is best?'. As in any other field of business, effective noise management is a matter of informed choice. In any static situation, if there is no possibility of any change, then noise assessment may be interesting but it is completely irrelevant. If there is some possibility of change then the regulatory authority or the decision maker needs the right information on the likely costs and benefits of that change in order to make the right decision. Effective noise management is always about choices. 'Should we introduce porous asphalt on this road? - what are the costs and what are the benefits? Should we install a silencer on this industrial plant? - how much will it cost and how many people will benefit from less noise? If the only thing that will change is the physical noise level, then this alone can be a good indicator of how reported annoyance might change, either up or down, but if other factors change as well, then these also may need to be taken into account. In almost all situations like these, the way that the problem is handled and the way that the decision is taken could be as important or more important than the actual noise level reduction achieved.
There are no simple conclusions once we seriously begin to study the role and importance of non-acoustic factors in environmental noise, other than to note that human response depends as much on the listener's state of mind as it depends on physical noise levels alone. How can we manage the listener's state of mind? Should we try? In the past, the importance of this factor has often first been recognised and then put back on the shelf as being too difficult to deal with. This approach has left a legacy of a complex system of physical noise indicators, many of which are perfectly adequate descriptors of physical noise levels, but no matter how well they are 'adjusted' or 'corrected' (Gezondheidsraad, 1997) they can never fully explain observed variance in human response. Job (Job, 1999) mentions the 'pervasive engineering focus on silence' as the most dominant aspect of current regulatory approaches to noise control. Of course, the main focus of engineering noise control is physical noise level reduction where this is both beneficial and cost-effective, and physical noise indicators are clearly the best way of measuring progress (or regression) in this field. However, we cannot leave the decisions on noise control policy just to the engineers, and effective noise management will almost invariably involve more than just engineering noise control.
The purpose of the Den Haag symposium was to encourage debate. The main purpose of publishing this special issue of the journal is to encourage still wider debate. We do not expect everyone to agree with the points made in our brief overview, and indeed, we may be open to persuasion on some of the points ourselves. All we can really ask is for people to think a bit more deeply about some of the problems we have attempted to highlight. Effective noise management is more often going to be about choosing between alternatives than about counting the percentage 'highly annoyed' and this is where future research should be addressed. We think this is important because noise control can be difficult and it can be expensive and it is therefore essential that it should be focused precisely where it will do the most good. Just considering dose-effect curves based on physical noise levels alone will not always achieve this.
|1||Fields, J. M. (1993) Effect of personal and situational variables on noise annoyance in residential areas, Journal of the Acoustical Society of America, 93, 2753-2763.|
|2||Fields, J. M., et al. (1997) Guidelines for reporting core information from community noise reaction surveys, Journal of Sound and Vibration, 206(5), 685-695.|
|3||Gezondheidsraad (1997) Assessing noise exposure for public health purposes, Report of a Committee of the Health Council of the Netherlands, Report No. 1997/23E, Gezondheidsraad, Den Haag.|
|4||Hatfield, J. (1999) Review of Noise Effects '98 (The 7th ICBEN conference), Noise and Health, 2, 80-82.|
|5||Job, R. F. S. (1999) Review of Internoise '98 conference, Noise and Health, 2, 78-79|
|6||Miedema, H. M. E. and Vos, H. (1998) Exposure-response relationships for transportation noise, Journal of the Acoustical Society of America, 104(6), 3432-3445.|