Effects of Aircraft Noise: Research Update on Select Topics (2008)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Annoyance remains the single most significant effect associ- ated with aviation noise. Community annoyance is the aggregate community response to long-term, steady-state exposure conditions. However, to adequately support government noise policy-making efforts, it is necessary to synthesize the large amount of data contained in journal articles and technical reports to develop a useful exposure- response relationship. In his seminal journal article, Schultz (1978) reviewed data from social surveys concerning the noise of aircraft, street and expressway traffic, and railroads. Going back to the original published data, the various survey noise ratings were trans- lated to Day-Night Average Noise Level (DNL) and, where a choice was needed, an independent judgment was made as to which respondents should be counted as “highly annoyed.” According to Schultz “. . . the basic rule adopted was to count as ‘highly annoyed’ the people who responded on the upper 27% to 29% of the annoyance scale . . .” (Schultz 1978). For decades, environmental planners have relied heavily on the Schultz Curve for predicting the community annoyance produced by noise from transportation noise sources. Notwith- standing the methodological questions, errors in measurement of both noise exposure and reported annoyance, data interpre- tation differences, and the problem of community response bias, Schultz’s recommended relationship has historically been the most widely accepted interpretation of the social survey literature on transportation noise-induced annoyance. Beginning with the publication of this original exposure- response curve, work has continued in many countries to conduct new field studies, develop databases with the results of dozens of new social surveys, and explore whether sepa- rate curves are needed to describe community responses to aircraft, street traffic, and railway noise. Based on an updat- ing of the Schultz curve by Fidell et al. (1991), the Technical Section of the Federal Interagency Committee on Noise (FICON), stated in 1992 that there were no new descriptors or metrics of sufficient scientific standing to substitute for the present DNL cumulative noise exposure metric (Federal Interagency Committee on Noise 1992). The dose-response relationship, as represented by DNL, and the percentage of persons “highly annoyed” remains the best available approach for analyzing overall health and welfare impacts for the vast majority of transportation noise analyses. In later years, Fidell goes on to criticize the use of this type of simplistic 10 curve such as the one in FICON, in light of the high data vari- ability, the effect of low- and high-noise exposure levels on the curve fit, and the lack of consideration of other variables in community response to noise. A comprehensive review and critique of the Fidell et al. up- date was later published by Fields (1994) that raises questions about the use of the synthesis data to develop the commonly used annoyance/DNL dose-response relationship. The report arrives at several conclusions, including “the curve is NOT a measurement of the relationship between DNL and the percentage of the population that would describe themselves as ‘highly annoyed’” and “if it is necessary to estimate the dose/response relationship . . . a single constituent survey pro- vides a better estimate” (Fields 1994). Fidell et al.’s expansion of the existing community annoyance research database and their revised prediction curve provided a considerable extension of the original Schultz meta-analysis (Fidell et al. 1991). However, because there were several debatable methodological issues involved in this update, Finegold et al. (1994) reanalyzed the Fidell et al. data focusing primarily on the choice of screening criteria for selecting which studies to include in the final database and the choice of a data fitting algorithm. Using the new data set, a new logistic fit curve as the prediction curve of choice was developed and adopted by FICON in 1992 for use by federal agencies in aircraft noise- related environmental impact analyses (Federal Interagency Committee on Noise 1992). It was also adopted as part of the American National Standards Institute (ANSI) Standard on community responses to environmental noises (Acoustical Society of American 2006). Finegold et al. (1994) showed that if the data are broken down into separate curves for various types of transportation noises (aircraft, roadway, and rail noise) aircraft noise appears to be more annoying at the same DNL than road or rail noise. Over the past decade, Miedema and Vos (1998) have compiled the most comprehensive database of community annoyance data yet available, and several studies have been published on the results of their analyses. It is a comprehen- sive review of an issue—separate, non-identical curves for aircraft, road traffic, and railway noise—that has been the subject of much debate since Shultz published his data in 1978. Caution should be exercised, however, when drawing CHAPTER THREE ANNOYANCE AND AVIATION NOISE

11 conclusions about the state of knowledge regarding the rela- tionship between various transportation noise sources and community annoyance. The European Commission position on annoyance is based on a report recommending the percentage of persons “highly annoyed” be used as the descriptor for noise annoyance. Similar to Miedema and Vos (1999) the report distinguishes between aircraft, road, and rail traffic noise; recommends use of a separate pair of curves (“annoyed” and “highly annoyed”) for each; and clearly shows a tendency to treat aircraft, road, and rail noise as unique when estimating population that will be “annoyed” or “highly annoyed” by noise (Miedema and Vos 1999). In their 1999 paper, Miedema and Vos further studied the effects of demographic variables (sex, age, education level, occupational status, size of household, dependency on the noise source, use of the noise source, etc.) and two attitudi- nal variables (noise sensitivity and fear of the noise source) on annoyance. The results are very interesting and suggest that fear and noise sensitivity has a large impact on annoy- ance. Additionally, in a 2002 report by Fidell et al., it is sug- gested that a good part of the excess annoyance is attributable to the net influence of non-acoustic factors. Some of the most interesting research comes from Fidell’s “The Schultz Curve 25 Years Later: A Research Perspective” (2003). It presents the argument that although federal adoption of an annoyance-based rationale for regulatory policy has made this approach a familiar one, it is only one of several historical perspectives, and not necessarily the most useful for all purposes. This tutorial article traces the development of the dosage-effect relationship on which FICON currently relies and identifies areas in which advances in genuine understanding might lead to improved means for predicting community response to transportation noise. It provides an important sum- mary of how the annoyance synthesis was developed, and the inherent weakness of the DNL/dose-response relationship that was developed. Fidell is highly critical of U.S. policy that relies solely on the synthesized dose-response relationship. Fidell and Silvati (2004) identified shortcomings of a fitting function endorsed by FICON for predicting annoyance in pop- ulations exposed to aircraft noise that are well-understood and well-documented. The authors argue that the U.S. National Environmental Policy Act (NEPA) (1969) requires the use of the best-available technology for disclosure of noise impacts of major federal actions, even though reliance on the FICON curve for meeting NEPA requirements does not use the best available technology. To summarize, significant research has occurred since the 1985 aviation effects report was published. Although no current research suggests there is a better metric than DNL to relate to annoyance, there still remains significant contro- versy over the use of the dose-response annoyance curve first developed by Schultz and then updated by others. Further, investigations that report a distinct percentage of the popula- tions that will be highly annoyed at a given DNL may be incorrectly interpreted as to having a more precise meaning than should be taken from the data. Lastly, a relatively new concept is that more research tends to support the idea that dose-response curves are different for aircraft, road, and rail noise sources. Areas of research that remain to be investi- gated include the relationship between single-event noise levels and annoyance. The expanding use of airport noise monitoring systems, flight-tracking systems, and geographic information systems may make the evaluation of annoyance and single-event noise rich for examination.