2

Recent Findings

The data from which Schultz (1978) and others have inferred dosage-response relationships between nonimpulsive sound exposure and the prevalence of annoyance were produced from social surveys. The small set of recent findings (since 1981) on the annoyance of impulsive noise that are similar in nature to those analyzed by Schultz is reviewed below. Discussion of the earlier studies may be found in CHABA (1981).

Since quantitative information about community response to high-energy impulsive noise is in short supply, it is reasonable to consider supplementary sources to aid development of a dosage-response relationship. Such information is also reviewed below.

SOCIAL SURVEY FINDINGS

Figure 2 summarizes the information presently available about the relationship between high-energy impulsive noise exposure and the prevalence of annoyance. The 52 data points displayed in the figure are derived from:

  • Borsky's (1965) Oklahoma City study (9 data points);

  • Two surveys of residential reactions to artillery noise (Schomer, 1985) (7 data points from Ft. Lewis and 6 data points from Ft. Bragg);

  • A study by the National Aeronautic and Space Administration (NASA) of reactions to sonic booms (Fields et al., 1994) (10 data points); and

  • A Swedish study assessing the effects of heavy weapons firing at shooting ranges (Rylander and Lundquist, 1996) (20 data points).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 6
Community Response to High-Energy Impulsive Sounds: An Assessment of the Field Since 1981 2 Recent Findings The data from which Schultz (1978) and others have inferred dosage-response relationships between nonimpulsive sound exposure and the prevalence of annoyance were produced from social surveys. The small set of recent findings (since 1981) on the annoyance of impulsive noise that are similar in nature to those analyzed by Schultz is reviewed below. Discussion of the earlier studies may be found in CHABA (1981). Since quantitative information about community response to high-energy impulsive noise is in short supply, it is reasonable to consider supplementary sources to aid development of a dosage-response relationship. Such information is also reviewed below. SOCIAL SURVEY FINDINGS Figure 2 summarizes the information presently available about the relationship between high-energy impulsive noise exposure and the prevalence of annoyance. The 52 data points displayed in the figure are derived from: Borsky's (1965) Oklahoma City study (9 data points); Two surveys of residential reactions to artillery noise (Schomer, 1985) (7 data points from Ft. Lewis and 6 data points from Ft. Bragg); A study by the National Aeronautic and Space Administration (NASA) of reactions to sonic booms (Fields et al., 1994) (10 data points); and A Swedish study assessing the effects of heavy weapons firing at shooting ranges (Rylander and Lundquist, 1996) (20 data points).

OCR for page 6
Community Response to High-Energy Impulsive Sounds: An Assessment of the Field Since 1981 FIGURE 2 Social survey information available to date about the prevalence of annoyance associated with community exposure to high-energy impulsive sounds. The data displayed in Figure 2 are tabulated in Appendix B. CDNL values in these tabulations are estimates of outdoor noise levels, rather than the at-ear values generally reported for laboratory measurements. The positions along the abscissa of Figure 2 of the Oklahoma City data and the Ft. Lewis and Ft. Bragg data have been adjusted from their previously reported values (CHABA, 1981; Schomer, 19851). The magnitude of these adjustments is modest for the Oklahoma City sonic boom data, but somewhat larger for the Ft. Lewis and Ft. Bragg data. The studies producing the new field data are briefly described below. A NASA-supported social survey of the annoyance of sonic booms (Fields et al., 1994) was recently conducted in five small communities near the U.S. Air Force's Nellis Range in eastern Nevada. Personal interviews were obtained from 1   All other things being equal, microphone placement can affect measurements of sound exposure created by outdoor sources by as much as 6 dB. A microphone mounted near a reflecting surface measures both the incident and the reflected energy, producing a so-called pressure doubling. Free field sound pressures (those measured at some distance from any reflecting surfaces) are lower numerically than measurements made at a building facade or with a ground plane microphone. Sonic booms and blast noises are generally measured with surface-mounted microphones. For the sake of consistency, other measurements analyzed in this report have been converted to equivalent surface measurements (pressure doubled). As described in Appendix B, this was accomplished by adding 4.5 dB to known free field measurements to approximate the pressure-doubled values.

OCR for page 6
Community Response to High-Energy Impulsive Sounds: An Assessment of the Field Since 1981 1,042 respondents exposed to sonic booms from Air Force training flights originating from an air base 50 to 150 miles from the communities. Noise monitors capable of storing sonic boom waveforms were used to measure daily sonic boom sound levels in terms of DNL and CDNL over 6-month periods prior to each of two sets of interviews. Sonic booms occurred at average rates ranging from 1 boom per day to 1 boom per 10 days. DNL values ranged from 20 to 41 dB, whereas CDNL values ranged from 36 to 58 dB. Although the accuracy of the acoustical measurements reported by Fields et al. of the levels of lower-level impulses may be in some doubt, the measurements of higher-level impulses are not in doubt. Of all the social survey findings shown in Figure 2, the findings of Fields et al. suggest the greatest sensitivity of communities to high-energy impulsive sound exposure. Rylander and Lundquist (1996) conducted a mail survey in 20 areas exposed to heavy weapons firing noise from eight shooting ranges in Sweden. Impulsive sound levels in these areas were estimated from knowledge of the types of weapons fired and assumed acoustic propagation over different types of terrain. Rylander and Lundquist analyzed returned questionnaires from a total of 1,483 respondents (from 22 to 143 respondents in the various areas). Total firings ranged from 950 to 16,100 shots per year at the various sites. No direct measurements of impulsive sound levels were made, but predicted CDNL values ranged from 42 to 67 dB. The findings reported by Rylander and Lundquist contain the greatest within-study variability of all of the social survey data sets. While some of the studies exhibit fairly good correlation between annoyance and sound level, taken as a whole, the correlation for the entire set of data using CDNL as the noise descriptor is quite low. LABORATORY AND CONTROLLED-EXPOSURE FIELD STUDIES Several studies in which individual impulses were presented under controlled conditions for immediate judgment of their short-term annoyance comprise a body of “new” (since 1981) nonsocial survey information about the annoyance of high-energy impulsive sounds. For reasons noted in Appendix A, annoyance judgments solicited in such studies cannot be directly compared with information about the prevalence in communities of annoyance with long-term impulsive noise exposure. Findings of laboratory and controlled-exposure field studies may be helpful, however, in reaching decisions about the form and slope of fitting functions for the limited amount of social survey information at hand. Some of these laboratory and controlled-exposure field studies suggest that the rate of growth of annoyance with the level of impulsive sounds is greater than that for nonimpulsive sounds, whereas others do not. Comparisons of the findings of these studies are complicated by differences among studies in acoustical measurement procedures, experimental methods, low-frequency content of test signals, listening environments, and other factors (see Appendix A).

OCR for page 6
Community Response to High-Energy Impulsive Sounds: An Assessment of the Field Since 1981 FIGURE 3 Relationship between equally annoying impulsive and nonimpulsive noises. Source: Schomer, 1994. Studies Finding a Greater Rate of Increase in Annoyance for Impulsive Sounds The U.S. Army Construction Engineering Research Laboratory has sponsored several paired-comparison experiments2 to empirically compare the annoyance of individual blast sounds and nonimpulsive (motor vehicle or artificially produced) sounds. These controlled-exposure experiments included some in which annoyance judgments were solicited from test subjects in actual houses, in response to actual blast sounds, from explosive sources at realistic distances from the test houses. Figure 3 summarizes equal annoyance judgments, using symbols and a solid 2   Test subjects in a paired-comparison study indicate on a trial-wise basis which of two signals, an invariant “standard” sound and a “comparison” sound that varies in some fashion (often in level) from trial to trial, is the more annoying. This method yields data with considerable face validity, since it neither requires test subjects to describe their sensations in arbitrary units or on unfamiliar scales, nor forces them to classify their sensations into absolute categories of uncertain meaning.

OCR for page 6
Community Response to High-Energy Impulsive Sounds: An Assessment of the Field Since 1981 FIGURE 4 Relationship between annoyance of blast sounds and motor vehicle passby. Source: Schomer, 1994. Note: Blast noise CSEL values in this figure are “pressure-doubled” measurements, whereas vehicle noise ASEL values are “free-field” measurements. line,3 made of blast sounds by about 550 test subjects in experiments conducted at Grafenwöhr (Germany), Munster (Germany), and Aberdeen Proving Ground (Maryland) (Schomer, 1994). The comparison signal in all cases in Figure 3 was a 450-ms long, band-limited (200-1,500 Hz) white noise. Acoustical measurements were made indoors near the subjects' ears. Two aspects of Figure 3 are noteworthy: the good agreement among studies and the slope of the relationship between judged annoyance and CSEL. For equal annoyance, a 1 decibel increase in CSEL of blast sounds in this data set corresponded to about a 2 decibel increase in A-weighted sound exposure level (abbreviated as ASEL, and symbolized LAE) of the equally annoying standard signal. At the point of subjective equality of annoyance with a white noise comparison sound, the CSEL of the blast sounds studied was equal to the ASEL value at about 104 dB. Figure 4 summarizes judgments made by 500 test subjects who compared the annoyance of blast sounds with those of motor vehicle passbys. The data plotted in Figure 4 indicate that a 1 decibel increase in CSEL of a blast sound (as mea- 3   The meaning of the dashed line is discussed below in connection with Eq. 3.

OCR for page 6
Community Response to High-Energy Impulsive Sounds: An Assessment of the Field Since 1981 sured outdoors) corresponds to about a 2 decibel increase in the ASEL of an equivalently annoying nonimpulsive sound. The CSEL of a blast sound was judged equally annoying to the ASEL of a vehicle passby at about 103 dB. Studies Finding Little or No Difference in Rate of Increase in Annoyance Between Impulsive and Nonimpulsive Sounds Several technical reports and memoranda by Leatherwood, McCurdy, Shepherd, Sullivan and their associates at NASA Langley Research Center (Leatherwood et al., 1991; Leatherwood and Sullivan, 1992a,b,c,d, 1993a,b, 1994a,b) describe a set of subjective judgment studies conducted in a small sonic boom simulator and in a larger listening room. Annoyance judgments were solicited by magnitude estimation methods, in which test subjects were asked to assign a numeric value to the judged loudness and/or annoyance of signal presentations. NASA's studies focused on evaluating differences in the ability of various frequency weighting networks to predict the annoyance of impulsive sounds, and on the effects of rise times and phase differences within a restricted ensemble of sonic boom-like signals. Leatherwood and Sullivan (1994a,b), however, developed a relationship between the CSEL of sonic booms and the ASEL of equally annoying aircraft flyovers, as heard both in the sonic boom simulator and in the larger listening room. As is evident from Figure 5, the rates of growth of annoyance with signal FIGURE 5 Comparison of rates of growth of annoyance of aircraft flyovers and simulated sonic booms. Source: Leatherwood and Sullivan, 1994a.

OCR for page 6
Community Response to High-Energy Impulsive Sounds: An Assessment of the Field Since 1981 presentation level are very similar for the sonic booms and the aircraft flyovers. Inferential tests performed on the magnitude estimation data showed no reliable differences in the growth of loudness of sonic booms and flyovers of much longer duration. The information summarized in Figure 5 is similar to that shown in CHABA (1981: Figure 3). CSEL values of actual sonic booms were about 5 dB lower numerically than ASEL values of subsonic flyovers when judged equally annoying in a set of direct comparisons (Kryter et al., 1968). In other tests (McCurdy, 1994) involving judgments of sonic booms that were conducted in a small sonic boom simulator, a slight but reliable improvement was found in the predictability of loudness and annoyance judgments in A-weighted level, loudness level, and perceived noise level units with respect to prediction in peak overpressure or C-weighted units. Table 1, containing information abstracted from a number of NASA reports, compares the rates of growth of magnitude estimates of loudness and annoyance for different classes of test signals, as measured in terms of C-weighted and loudness level units. These differences in slopes are generally minor, centering on about 0.4 for impulsive signals of lesser low-frequency content and about 0.5 for impulsive sounds of greater low-frequency content. Both of these values are notably greater than the often cited value of 0.3 for the rate of growth of annoyance with level for nonimpulsive sounds (Stevens, 1972). TABLE 1 Rates of Growth (Exponents of Magnitude Estimates) of Annoyance and Loudness Judgments for Various Signals with Presentation Level Judged Quantity Exponential Growth Rate of Judged Quantity with CSEL Exponential Growth Rate of Judged Quantity with Stevens MK VII Signals of lesser low-frequency content:     Loudness (outdoor N-waves) 0.42 0.42 Loudness (outdoor shaped simulated sonic booms) 0.40 0.39 Annoyance (outdoor N-waves) 0.39 0.40 Annoyance (outdoor shaped simulated sonic booms) 0.37 0.33 Signals of greater low-frequency content:     Loudness (indoor N-waves) 0.53 0.47 Loudness (indoor shaped simulated sonic booms) 0.59 0.49 Annoyance (indoor N-waves) 0.47 0.47 Annoyance (indoor shaped simulated sonic booms) 0.55 0.46

OCR for page 6
Community Response to High-Energy Impulsive Sounds: An Assessment of the Field Since 1981 EVALUATION OF AVAILABLE INFORMATION Variability in Social Survey Findings Variability is perhaps the most distinctive feature of the field data (Figure 2) on the prevalence of annoyance associated with high-energy impulsive sound exposure. Similar proportions of different communities describe themselves as annoyed by impulsive noise that varies in level by as much as 30 dB. This variability may reflect error in noise exposure estimates (whether based on predictions or actual measurements); poor association between annoyance and the noise metric of choice; the influences of nonacoustic variables, such as familiarity of the subjects with the high-energy sound exposure; or some combination of these and other effects. Regardless of the origin of this variability, its net effect is to diminish the persuasiveness of any single approach to deriving a dosage-response relationship based on curve-fitting considerations alone. Findings of Controlled-Exposure Studies The findings of controlled-exposure studies are not fully consistent with one another. For example, some laboratory studies of the annoyance of small arms fire suggest that impulsive noise is more annoying than nonimpulsive noise, but that the disparity in annoyance decreases with level. Other data, including results of controlled-exposure studies conducted in field settings that produced realistic levels of low-frequency energy characteristic of longer-duration impulses, indicate that the annoyance of impulsive noise grows at a greater rate than that of nonimpulsive noise. Schomer (1994), for example, notes that the results of a laboratory study conducted by Young (1976) resemble those noted in Figure 3 and Figure 4 with regard to the slope and the level at which the ASEL and CSEL of nonimpulsive and impulsive sounds are judged equally annoying. 4 Young (1976) studied the annoyance of simulated artillery firing sequences by instructing test subjects to estimate the magnitude of their annoyance with various artillery-like sounds, recorded aircraft flyover sounds, and other sounds. Schomer (1977, 1978a,b) has further analyzed and reported on these data. According to Schomer (1994), the linear regression of CSEL on the logarithm of Young's response data for simulated blast sounds is given by: log (magnitude of annoyance) = 0.0576 × (LCE − 70.8). (1) Schomer reports the following relationship for the comparable regression of ASEL on Young's response data for aircraft sounds: 4   A subset of this information is also discussed in CHABA (1981).

OCR for page 6
Community Response to High-Energy Impulsive Sounds: An Assessment of the Field Since 1981 log (magnitude of annoyance) = 0.0331 × (LAE − 43.8). (2) These relationships may be manipulated algebraically to yield the following relationship between the annoyance of impulsive and nonimpulsive sounds as measured in units of CSEL and ASEL, respectively: LCE = 0.575 × LAE + 45.6. (3) This relationship between noise levels (measured indoors) is shown as a dashed line in Figure 3. Schomer also notes that Kryter et al.'s (1968) observations about the annoyance of controlled exposure to actual sonic booms yield a relationship similar to that observed by Young. According to Kryter, “the unacceptability of sonic booms, as a function of intensity, increases at about twice again as fast a rate as does the unacceptability of the noise from subsonic aircraft. ” Judgments of the relative unacceptability of subsonic aircraft flyovers and sonic booms in Kryter et al.'s (1968) Edwards Air Force Base study showed that an increase of 12 PNdB in the intensity of sound produced by a subsonic aircraft was equivalent to about a 6 dB increase (from 1 to 2 pounds per square foot) in the intensity of a sonic boom. FIGURE 6 Levels of sonic booms and aircraft flyovers judged equally annoying. Source:   Kryter et al., 1968.

OCR for page 6
Community Response to High-Energy Impulsive Sounds: An Assessment of the Field Since 1981 This interpretation of Kryter's observation is potentially misleading, however, since the comparison is between a duration-corrected noise metric (CSEL) for sonic booms and a maximum level measure (PNL) for the equally annoying aircraft sounds. Figure 6 shows the comparisons made between the CSEL of the sonic boom and ASEL for the equally annoying subsonic aircraft flyover. Both of the latter metrics are duration corrected and have unity slope. Summary of the Findings of Controlled-Exposure Studies Figure 7 compares the results of regression analyses linking annoyance judgments with signal presentation levels in the most relevant controlled-exposure studies. The plotted range of each regression line is limited to the range of signal FIGURE 7 Comparison of rates of growth of annoyance with at-ear level for impulsive and nonimpulsive sounds reported in controlled-exposure studies.

OCR for page 6
Community Response to High-Energy Impulsive Sounds: An Assessment of the Field Since 1981 presentation levels in the study from which it was derived. The noise levels plotted have all been adjusted for indoor (at-ear) listening conditions by assuming an average noise reduction for low-frequency noise of 5 dB (difference in CSEL outside and inside a structure) and 15 dB for the higher-frequency noise associated with aircraft flyovers (difference in ASEL outside and inside a structure). The figure shows that only at lower noise levels (such as those studied by Schomer) do A-weighted sound levels of nonimpulsive sounds increase at a rate of 2 dB for every 1 dB of the impulsive noise of the same judged loudness or annoyance. At levels greater than 85 dB ASEL for nonimpulsive sounds and 95 dB CSEL for impulsive sounds, the slope of the relationship steepens to approximate an equal increase in both impulsive levels (measured in CSEL) and non-impulsive levels (measured in ASEL) to maintain equal judged annoyance and loudness for the two types of sounds. It is unclear from the limited data available whether this dependence on level of the relative slopes of increases in annoyance for impulsive and non-impulsive sounds is substantive, whether it is an artifact of the absence of rattle in test signals presented for judgment at higher levels, or whether it is due to other differences in study design.