Oslo traffic study – part 1: an integrated approach to assess the combined effects of noise and air pollution on annoyance

https://doi.org/10.1016/S1352-2310(00)00304-6Get rights and content

Abstract

Vehicular traffic is a common important source of air pollution, traffic accidents, road traffic noise as well as other environmental exposures. The relationship between each of these exposures and their respective impacts are nevertheless most often studied separately. An integrated alternative approach was adopted in the Oslo traffic study to allow people's environmental annoyances to be studied relative to the indicators of air pollution, road traffic noise and residential traffic. These annoyances include annoyance with the smell of exhaust, with dust and grime, feeling insecure in traffic and being annoyed with road traffic noise. A hypothesis was that multiple exposures typical in city areas have combined impacts – that people exposed to both air pollution and road traffic noise will be more annoyed than in the respective single-exposure situations. Three environmental studies in 1987, 1994 and 1996 each year comprising about 1000 respondents after a response rate of 50%, serve as before–after studies of two tunnel projects. Personal interviews were utilised in the before study in 1987 and telephone interviews in 1994 and 1996. Exposure indicators for air pollution as well as road traffic noise and residential traffic levels were produced for each respondent by comprehensive environmental modelling. Exposure–effect logistic regression models for the probability of people being highly annoyed by the smell of exhaust and by road traffic noise, respectively, were estimated. The results indicate that the higher the road traffic noise levels people are exposed to, the more likely they are to be highly annoyed by exhaust smell at a specified air pollution level. The higher air pollution levels people are exposed to the more likely they are to be annoyed by road traffic noise at a specified noise level. Modifying factors were controlled for.

Introduction

In addition to road traffic noise, insecurity and severance, an increasing number of studies are inquiring into the perception, annoyance, coping behaviour and subjective health effects of air pollution (Evans et al., 1988; Forsberg et al., 1993, Forsberg et al., 1997a; Kolbenstvedt et al., 2000; Lercher and Kofler, 1996; Lercher, 1996; Williams et al., 1992; Winneke et al., 1996). A common aspect of these studies is that dust/grime and the smell of exhaust can be perceived (Forsberg et al., 1997b), to affect behaviour, and cause problems at levels well below current air pollution limits and at levels often associated with urban traffic.

Most often annoyance reactions are studied with respect to each environmental exposure separately without taking other environmental exposures into account. However, air pollution, road traffic noise and insecurity might all be regarded as ambient stressors (Campbell, 1983) that can have combined impacts. This paper investigates whether there are combined effects of air pollution and road traffic noise on annoyance with the smell of exhaust and with road traffic noise, respectively. Are people exposed to higher levels of road traffic noise more inclined to be annoyed with air pollution than those exposed to lower levels Vice versa, are people exposed to higher levels of air pollution more inclined to be annoyed with road traffic noise than those exposed to lower levels? We follow a three-step approach:

  • 1.

    The correlations between different types of exposure indicators are calculated as a simple means of describing the geographical correspondence between different types of exposure. We also describe the accumulation of different types of annoyances the respondents report as a function of increasing levels of exposure to air pollution and road traffic noise, respectively.

  • 2.

    In step 2 of the investigation, logistic regression models are estimated in order to determine whether people being annoyed by one type of environmental exposure have a higher probability of being highly annoyed by another type of exposure (while controlling the exposure level of the latter type of exposure). These models control for demographic factors and other possible confounding factors such as some people being inclined to complain irrespective of exposure, and for people being generally sensitive to environmental exposures. Logistic regression models are described briefly in part 2 (Clench-Aas et al., 2000).

  • 3.

    In step 3 of the investigation, an exposure–effect model for people being highly annoyed by the smell of exhaust is tested for the inclusion of the air pollution indicator and thereafter an indicator of road traffic noise. An exposure–effect model for people being highly annoyed by road traffic noise is likewise tested first for the inclusion of the road traffic noise indicator, and secondly for the additional inclusion of the air pollution indicator as an independent variable.

In the second part of the Oslo traffic study, Clench-Aas et al. (2000) describe the air pollution modelling in more detail, describe briefly the use of logistic regression models and report on the health effects.

Section snippets

Subjects and dependent measures

Three environmental studies were undertaken in the autumns of 1987, 1994 and 1996. The surveys functioned as before and after studies of two separate tunnel projects alleviating a centrally located urban area in Oslo of through-traffic. In 1987, personal interviewing took place in 8 sub-areas. In 1994 and 1996 telephone interviews were undertaken in 14 areas, including the original 8. The response rate was approximately 50% in three surveys (resulting n=1028, 1140, 1097). The sub-areas were

Step 1: Correlations exposure indicators and the accumulation of annoyances

Due to changes in traffic, different mean temperatures during the 3-month period between different survey years and the two environmental tunnels, correlations between road traffic noise and air pollution indicators were moderate. The 24 h equivalent sound pressure levels correlated 0.46 with the 3-month mean of NO2 and 0.48 with the logarithm of these values. Correlations between 24 h equivalent sound pressure levels and 3-month average levels of PM10 and PM2,5 were 0.34 and 0.39, respectively.

Discussion

People liable to experience negative emotionality – negative affectivity – are known to complain about problems not reflected in objective health status or in exposure (Watson, 1984; Watson and Pennebaker, 1989). For research using people's self-reported health or degree of annoyance, this may confound results. If a number of people belonging to this group are reporting that they are both highly annoyed by road traffic noise and highly annoyed by air pollution, the results in Fig. 3 may be due

Assessment of traffic measures and other changes

Between 1987 and 1996 road traffic volumes, road traffic noise levels, and air pollution exposure were all greatly reduced. The combined effects of these reductions has previously been hypothesised to be the cause of larger noise annoyance reductions than that which follows from exposure–effect relationships only considering noise exposure (Klæboe et al., 1998). The exposure-effect relationship estimated for road traffic noise annoyance reported in Table 2 shows such an additive effect of

Conclusion

With data well-suited for the purpose of separating the effects of air pollution and road traffic noise, these environmental effects have been shown to have substantial combined impacts. Exposure-effect models for noise and air pollution annoyance only including noise and air pollution indicators, respectively, will give misleading results. We have not provided a comprehensive alternative model for environmental exposures having combined effects.

In order to establish more robust exposure-effect

Acknowledgements

We thank The Research Council of Norway and the Public Roads Administration for their support of this research.

References (16)

There are more references available in the full text version of this article.

Cited by (0)

View full text