Assessing Community Annoyance of Helicopter Noise (2017)

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.

23 3.1 Introduction This chapter describes site selection and measurement methods. Section 3.2 discusses the survey site selection process. Criteria used to assess the suitability of survey sites are presented along with the sites considered and a recommendation for survey sites for the study. Section 3.3 describes the questionnaire along with discussions of its form and organization as well as of inter- viewing methods. Section 3.4 is a general discussion of the role of sample size in social survey design. Noise measurement methods are described in Section 3.7, along with specific discussion of sample size concerns. 3.2 Survey Site Evaluation 3.2.1 Overview of Survey Site Selection Process Survey site selection is complicated by the fact that there is no such thing as generic “heli- copter noise.” Acoustic emissions of helicopters vary much more with flight regime than do those of fixed-wing aircraft. Sites exposed to sideline noise from straight and level flight have considerably different acoustical experiences than those near landing pads that can experience prominent blade slap from steeply descending helicopters. Sites on either side of the flight path can experience different acoustical exposures due to the directionality of BVI impulsive noise and tail rotor noise. Some sites may be exposed to relatively short overflights, while others may experience prolonged exposures from hovering, orbiting, or otherwise maneuvering helicop- ters. The selected sites should provide as wide a range of aircraft noise exposures as possible. The primary consideration for survey sites is that the residents must be exposed to appre- ciable amounts of civil helicopter noise and, where possible, fixed-wing aircraft noise. If only a small portion of an exposed population is annoyed by aircraft noise, or is only slightly annoyed by it, then unreasonably large numbers of interviews may be necessary to demonstrate that population proportions of annoyance differ significantly from zero. Further, it may not be possible to perform a credible dosage-response analysis if annoyance prevalence rates are low. As a generality, a large number of survey responses over as wide a range of helicopter flight regimes and nonmilitary noise levels is preferred. To maximize the potential for responses, thou- sands of households should be eligible for interviews at a site. Further, individual sites should be exposed to as great a variety of aircraft types as possible. If a site is overflown only by a small number of aircraft types, such as a small tour helicopter or a large military rotorcraft, it may be difficult to generalize any findings beyond those aircraft types. C H A P T E R 3 Site Selection and Opinion Survey Methods

24 Assessing Community Annoyance of Helicopter Noise One of the primary goals of the project is to determine the relative annoyance of exposure to rotary- and fixed-wing aircraft. Residents eligible for interview would ideally be exposed to noise from both forms of aircraft, if possible. Further, the magnitude of residential noise exposure levels of the two forms of aircraft noise should be roughly comparable to support straightforward analyses and inferences. In addition to the characteristics described above, the survey sites should preferably lack any features that preclude or complicate collection and processing of interview and acoustic informa- tion. For example, unambiguous aircraft noise exposure measurements require that non-aircraft noise levels at measurement sites not approach or exceed aircraft noise levels. To facilitate valid measurement of cumulative (average annual day) exposure metrics, aircraft operations should have little seasonal variability. Neighborhoods with large proportions of non-English speaking households can increase the cost and complexity of administering questionnaires. Detailed radar data and helicopter performance state data will be needed to provide an accurate basis for noise modeling. The following sections describe the site selection process. The primary, secondary, and survey optimization criteria used to select sites are discussed in Section 3.2.2. Section 3.2.3 presents the locations that were considered and discusses sites that satisfy the primary survey site criterion. A comparison of the potential survey sites relative to the selection criteria is presented in Sec- tion 3.2.4. Finally, Section 3.2.5 presents the recommended sites along with a discussion of the rationale for selecting them. 3.2.2 Survey Site Selection Criteria Selection of survey sites was accomplished in several steps. The primary criterion—sufficient civil helicopter overflights of residential neighborhoods—was used to develop an initial list of potentially acceptable sites. Secondary criteria were used to evaluate the acceptability of these potential sites to provide high quality data required for the analysis. Sites that were clearly unable to meet the secondary criteria were not considered further. The sites that were at least minimally acceptable were then compared and summarized in Table 3-1. The primary criterion for selection of survey sites was sufficient rotary-wing aircraft overflight of residential land. Four general types of areas were believed likely to satisfy the primary selec- tion criterion: those near commercial airports, neighborhoods near military airfields that are also exposed to noise from civil aircraft operations, neighborhoods near hospitals, and areas near civil heliports. 3.2.2.1 Secondary Criteria for Selecting Interviewing Sites Secondary criteria were used to further appraise the sites satisfying the primary criterion. Table 3-1 contains a list of the secondary criteria along with their relative importance and a summary discussion of each. The following paragraphs discuss secondary selection criteria in greater detail. The first of the secondary site selection criteria is the absence of any conditions that would unnecessarily increase the cost or complexity of data collection. Increased sampling, interview- ing, and acoustic measurement costs required for sites outside the contiguous 48 states were considered unjustifiable. While noise measurements were made concurrently with interviewing, noise modeling was required to quantify noise exposures at each site. The noise measurements were used to validate and improve the accuracy of modeled noise levels. Reliable radar data for aircraft operations in the week before and during interviewing was also needed and acquired.

Site Selection and Opinion Survey Methods 25 The noise exposure levels from aircraft overflights must engender measureable annoyance prevalence rates. Both the absolute level of the exposure from single overflights and numbers of overflights are important. In addition, each site must have sufficient aircraft noise expo- sure to result in an annoyance prevalence rate that can be detected by a reasonable number of interviews. Similarly, background noise levels (those due to non-aircraft noise sources) must not be so great that they mask single-event aircraft noise levels. Readily generalizable findings of the social survey require exposure to a variety of aircraft types and flight regimes. Sites with little variability in types of aircraft overflights were thus undesirable. Sites with high seasonal variability in aircraft operations and noise exposures were also undesirable. Such sites would result in misleading cumulative (annual average day) noise exposure metrics. Further, high seasonal variability could unreasonably constrain inter- viewing schedules. Likewise, special events such as parades and large sporting events with extensive helicopter activity provide only short exposures and are not the focus of this study. 3.2.2.2 Optimizing Social Survey Design Potential survey sites that satisfied both the primary and secondary selection criteria were then compared with respect to criteria for optimizing the design of the social survey. These criteria are listed in Table 3-2 along with their relative importance and a brief summary. 3.2.3 Sites Considered The primary criterion for selection of interviewing sites was sufficient rotary-wing aircraft overflights of populated areas. Areas that satisfy this basic requirement are typically found around civil airports, military airfields, heliports, and hospitals. Table 3-3 lists facilities that satisfy the primary criterion. CRITERION IMPORTANCE DISCUSSION Survey Feasibility/Cost Very High Survey sites must be suitable for both noise measurement and interviewing. Higher costs for sites outside the continental United States are not justifiable. Availability of Radar Data and Performance State Data Very High Radar data is essential for accurate and meaningful noise modeling. Performance state will be based on noise model profiles. Aircraft Noise Exposure Levels High Low noise exposures are likely to produce small annoyance prevalence rates and require larger sample sizes. Background Noise Levels High Aircraft noise should not be masked by other community sources. Fleet Mix Moderate Small variability in the fleet of aircraft limits the generalizability of the findings. Seasonality Moderate Highly seasonal operations may result in misleading cumulative (average annual day) exposure metrics and constrain schedules. Availability of Complaint Records Moderate Complaint information can be helpful for analytic purposes. (A recent D.C. Court of Appeals ruling on regulation of helicopter noise was largely based on complaints.) Predominant Language Moderate Neighborhoods with predominantly non-English speaking households increase complexity and cost of social surveys. Table 3-1. Secondary criteria for site selection, ranked by importance.

26 Assessing Community Annoyance of Helicopter Noise 3.2.3.1 Civil Airports Figures 3-1 through Figure 3-4 show published helicopter routes for Van Nuys Airport, Tor- rance Airport, Long Beach Airport, and Las Vegas Airport. Actual helicopter routes for Las Vegas, derived from radar tracking, are shown in Figure 3-5. The Long Beach and Las Vegas figures also show residential land uses (red-shaded) in the areas around the airport. Published helicopter routes in the region around Reagan National Airport are shown in Figure 3-6. Heliports reviewed include the numerous heliports in the Washington, D.C., area, Manhattan, New York, and Paulus Hook, New Jersey. Figure 3-6 and Figure 3-7 show heliports and pub- lished helicopter routes in the Washington, D.C., area. Figure 3-7 shows the Georgetown/Northern Arlington area in detail. This area of D.C. is exposed to helicopter operations over the river as well as fixed-wing aircraft from DCA that also fly over the river, albeit at higher altitudes. Residential land uses are shaded in red. Mixed-use land uses that include residential uses are shaded in orange. Figure 3-8 shows radar tracks for aircraft operations in this area. The aircraft altitudes are shown to distinguish helicopter operations from fixed-wing aircraft approaching and departing from Reagan National Airport. Aircraft at altitudes below 600 feet in Figure 3-8 are helicopters, while those above 600 feet are fixed-wing aircraft. CIVIL AIRPORTS MILITARY FIELDS HOSPITALS* HELIPORTS Van Nuys, CA (VNY) Long Beach, CA (LGB) Torrance, CA (TOR) Las Vegas, NV (LAS) Reagan National, D.C. (DCA) Anchorage, AK (ANC) Kahului, Maui, HI (OGG) Hilo, Hawaii, HI (ITO) Lihue, Kauai, HI (LIH) Camp Pendleton MCAB, CA Miramar MCAS, CA Ft. Rucker, AL Ft. Eustis, VA Edgewood Arsenal, MD 29 Palms MCB (Joshua Tree), CA San Francisco General, CA UCLA Medical Center, CA Massachusetts General, MA Manhattan, NY East 34th Street, NY MetLife Building, NY West 30th St., NY Paulus Hook (Jersey City), NJ Hamptons, NY Boston Harbor, MA Washington, D.C., heliports *Additional hospitals with helicopter noise issues were reviewed for consideration but excluded because they had less than one flight per day on average. The three hospitals noted have near-daily operations Table 3-3. Initial list of sites considered. CRITERION IMPORTANCE RATIONALE FOR CRITERION Mix of Exposure Levels Very High Wider range of noise exposures provides more defensible, credible, and generalizable dose-effect relationships. Mix of Helicopter Type and Operational Regimes High Helicopter noise is highly variable in character and dependent on both helicopter type and flight regime. The greater the range in these factors, the more generalizable the results. Mix of rotary-wing and fixed-wing aircraft High Sites exposed to both fixed-wing and helicopter overflights will allow for a direct comparison of annoyance rates. Relative rotary-wing and fixed-wing exposure levels Moderate/High Smaller disparities between rotary- and fixed-wing aircraft noise exposures simplify study design and reduce the need for statistical measures to compensate for large disparities. Use of unique transponder (XPNDR) Codes Moderate The use of unique XPNDR codes facilitates identification of aircraft type. Table 3-2. Survey optimization criteria by importance.

Site Selection and Opinion Survey Methods 27 Figure 3-1. Van Nuys Airport. Figure 3-2. Torrance Airport.

28 Assessing Community Annoyance of Helicopter Noise Figure 3-3. Long Beach Airport.

Site Selection and Opinion Survey Methods 29 Figure 3-4. Las Vegas International Airport. Figure 3-5. Las Vegas helicopter radar tracks.

30 Assessing Community Annoyance of Helicopter Noise Figure 3-6. Greater Washington, D.C., helicopter routes. Figure 3-7. Georgetown, Washington D.C.

Site Selection and Opinion Survey Methods 31 Figure 3-9 and Figure 3-10 show the published helicopter routes and nearby land uses in the immediate vicinity of the Manhattan and Paulus Hook helistops. Residential land uses are shaded in red. Mixed-use land uses that include residential uses are shaded in orange. 3.2.4 Site Evaluation Table 3-4 summarizes the considered sites’ characteristics relative to the selection criteria. The type of facility is presented along with information relevant to the primary, secondary, and Figure 3-8. Georgetown, Washington, D.C., radar flight tracks.

32 Assessing Community Annoyance of Helicopter Noise Figure 3-9. Manhattan heliport. Figure 3-10. Paulus Hook heliport.

Site Selection and Opinion Survey Methods 33 optimization criteria. The table shows approximate number of daily helicopter operations, along with information about the presence or absence of overflights of residential neighborhoods. The availability of radar tracks and a characterization of the background noise levels in the vicinity of the site are also shown. A characterization of the mix of aircraft types at each location, and the use of unique XPNDR codes, is used to evaluate the optimization criteria. The final column of Table 3-4 indicates whether further consideration was warranted for each of the sites considered. 3.2.5 Site Recommendations Site visits were conducted at Long Beach, Las Vegas, Washington, D.C., Van Nuys, and Torrance. Of these, Long Beach, Las Vegas, and Washington, D.C., were selected for the social surveys. 3.3 Questionnaire The social survey was intended to test as many of the hypotheses as feasible, as described in Chapter 2 about the annoyance of helicopter noise at three interviewing sites. The hypotheses con- cern community reactions to various aspects of helicopter noise exposure and required detailed acoustic and aircraft position (“radar”) information for testing. Some hypotheses required analy- ses of explicit questions about the nature of annoyance with helicopter noise. Other hypotheses SITE PRIMARY CRITERIA SECONDARY CRITERIA OPTIMIZATION CRITERIA F A C IL IT Y T Y P E N U M B E R O F D A IL Y O P E R A T IO N S R E S ID E N T IA L O V E R F L IG H T S R A D A R T R A C K A V A IL A B IL IT Y B A C K G R O U N D N O IS E M IX O F A IR C R A F T T Y P E U N IQ U E X P N D R C O D E R E C O M M E N D E D F O R F U R T H E R C O N S ID E R A T IO N Van Nuys, CA (VNY) Airport Unknown Yes Yes Acceptable Good Yes Yes Long Beach, CA (LGB) Airport 34 Yes Yes Acceptable Good Yes Yes Reagan National, D.C. (DCA) Airport ~35 Yes Yes Acceptable Excellent Yes Yes Las Vegas, NV (LAS) Airport 237 Yes Yes Acceptable Very Good Yes Yes Kahului, Maui, HI (OGG)1 Airport Unknown Yes Yes Acceptable Poor No No Hilo, Hawaii, HI (ITO) 1 Airport Unknown Yes Yes Acceptable Poor No No Lihue, Kauai, HI (LIH) 1 Airport Unknown Yes Yes Acceptable Poor No No Anchorage, AK (ANC) 1 Airport Unknown Yes Yes Acceptable Poor No No Torrance, CA (TOR) Airport Unknown Yes Yes Acceptable Poor Yes Yes Camp Pendleton MCB, CA Military Unknown No Unknown Acceptable Poor No No Miramar MCAS, CA Military Unknown Yes Yes Acceptable Poor No No Ft. Rucker, AL Military Unknown Yes Unknown Acceptable Poor No No Ft. Eustis, VA Military Unknown Yes Unknown Acceptable Poor No No Edgewood Arsenal, MD Military Unknown Yes Unknown Acceptable Poor No No 29 Palms MCB (Joshua Tree), CA Military Unknown No Unknown Acceptable Poor No No San Francisco General2 Heliport None No Yes Excessive Good No No UCLA Medical Center Heliport Low Yes Yes Excessive Good Yes No Massachusetts General Heliport Low Yes Yes Excessive Good No No Manhattan Heliport Unknown No Excessive No East 34th Street Heliport Unknown Yes Excessive No MetLife Building Heliport Unknown Yes Excessive No Hamptons Heliport Unknown Yes No Acceptable Good No No Paulus Hook (Jersey City) Heliport 03 No n/a Excessive n/a n/a No Boston Harbor Heliport Unknown Yes High No 1Eliminated from consideration due to travel costs. 2Conditional Use Permit for heliport not approved. 3The helipad owner has recently ceased all operations at this facility. It is not known if, or when, they will resume. Table 3-4. Survey site evaluation summary matrix.

34 Assessing Community Annoyance of Helicopter Noise could be evaluated simply by comparing dosage-effect relationships constructed with different noise metrics, or other variables, as independent (predictor) variables. 3.3.1 Form and Organization of Questionnaire An ISO Technical Specification (15666:2003 “Acoustics—Assessment of noise annoyance by means of social and socio-acoustic surveys”) offers general recommendations for the order and wording of transportation noise annoyance questionnaire items. The recommendations are intended to facilitate meta-analysis and interpretation of survey findings, not to further specific research goals. All of the Technical Specification’s recommendations are merely informative, and are quali- fied by provisions that they not conflict with survey goals. The ISO specification explicitly states, “specific requirements and protocols of some social and socio-acoustic studies may not permit the use of some or all of the present specifications. This Technical Specification in no way lessens the merit, value or validity of such research studies.” The suggested organization of the present questionnaire follows that of many prior studies of the prevalence of annoyance with aircraft noise exposure in airport neighborhoods. 3.3.2 Questions for All Interviewing Sites Table 3-5 shows the complete questionnaire. Instructions to interviewers that are not posed to respondents are shown in italic blue or red: questions posed to respondents are in black. The interview was introduced as a study of neighborhood living conditions, not as one of the annoy- ance of exposure to helicopter noise. This approach reduces the likelihood that respondents will either grant or refuse an interview, or bias their responses to questionnaire items, based on foreknowledge of the purpose of the study. Item 1 was intended to confirm eligibility for interview. Respondents who did not confirm residence at the household street address (e.g., guests, relatives, household employees, etc.) were not eligible for interview, but were asked whether and when an adult resident would be available for interview. The response coding provides information for a test of a potential relationship between duration of residence and degree of annoyance with aircraft noise—an indirect measure of adaptation. Items 2 and 3 were included for the sake of consistency with the introduction of the study as one of neighborhood living conditions. They also provided an opportunity, prior to any men- tion of noise-related concerns, for spontaneous mention of helicopter noise as the least-favored aspect of neighborhood living. Items 4 and 4A introduced respondents to the closed category absolute judgment scale used in all subsequent items for expressing degrees of annoyance with noise exposure. Item 5 was the first explicit mention of noise as a neighborhood living condition of interest. Items 6 and 6A sought information about the frequency of notice of helicopter noise in the week preceding interview. Items 7 and 7A inquired about the degree of annoyance of helicopter noise. Several variant sets of questionnaire items could follow Item 7A, depending on the suitability of noise exposure and other site-specific circumstances. These included: • Variant 1: Assessment of relative annoyance of exposure to fixed- and rotary-wing aircraft noise, intended for administration at sites exposed to both types of flight operations. • Variant 2: Assessment of relative contributions of different aspects of helicopter noise for sites exposed to BVI (“blade slap”), thickness, blade-wake interaction, and ducted fan tail rotor noise, intended for administration at sites exposed to noise of diverse helicopter operations.

Site Selection and Opinion Survey Methods 35 Item 1 How long have you lived at (street address)? Response/Coding Categories: don’t live at this address (0, ask to speak with resident, schedule a callback, or terminate interview), less than 1 year (1), at least 1 year but less than 2 years (2), 2 to 5 years (3), 5 to 10 years (4), more than 10 years (5), don’t know (6), refused (7) Item 2 What do you like best about living conditions in your neighborhood? Record verbatim response (coding per optional post hoc content analysis) Item 3 What do you like least about living conditions in your neighborhood? Record verbatim response, code as “aircraft noise-related” (1) or “non-aircraft noise-related” (2) Item 4 Would you say that your neighborhood is quiet or noisy? Response/Coding Categories: quiet (0), quiet except for aircraft (of any kind) (1), noisy (2), don’t know (5), refused (6), skipped (7) If respondent answers “noisy,” ask Item 4A; if any other response to Item 4, ask Item 5 next Item 4A Would you say that your neighborhood is slightly, moderately, very , or extremely noisy? Response/Coding Categories: slightly (1), moderately (2), very (3), extremely (4), don’t know (5), refused (6) , skipped (7) Item 5 While you’re at home, are you bothered or annoyed by street traffic noise in your neighborhood? Response/Coding Categories: no (0), yes (1), don’t know (5), refused (6) If respondent answers yes to Item 5, ask Item 5A; if any other response to Item 5, ask Item 6 next Item 5A Would you say that you are slightly, moderately, very, or extremely annoyed by street traffic noise in your neighborhood? Response/Coding Categories: slightly (1), moderately (2), very (3), extremely (4), don’t know (5), refused (6), skipped (7) Item 6 While you were at home last week, did you notice noise made by helicopters? Response/Coding Categories: no (0), yes (1), don’t know (5), refused (6) If respondent answers yes to Item 6, ask Item 6A; if any other response to Item 6, ask Item 7 next Item 6A About how often did you notice noise made by helicopters while you were at home last week? Would you say you noticed noise made by helicopters less than once a day, about once a day, a few times a day, or at least several times an hour while you were at home last week? Response/Coding Categories: less than once a day (1), a few times a day (2), several times or more per hour (3), don’t know (5), refused (6), skipped (7) Item 7 While you were at home last week, did noise made by helicopters bother or annoy you? Response/Coding Categories: no (0), yes (1), don’t know (5), refused (6) If respondent answers yes to Item 7, ask Item 7A; if any other response to Item 7, ask Item 8 next Item 7A Would you say that you were slightly, moderately, very, or extremely annoyed by noise made by helicopters while you were at home last week? Response/Coding Categories: slightly (1), moderately (2), very (3), extremely (4), don’t know (5), refused (6), skipped (7) Item 8 While you were at home last week, did you notice noise made by aircraft other than helicopters? Response/Coding Categories: no (0), yes (1), don’t know (5), refused (6) If respondent answers to Item 8, ask Item 8A; if any other response to Item 8, ask Item 9 next Item 8A About how often did you notice noise made by aircraft other than helicopters while you were at home last week? Would you say you noticed noise made by aircraft other than helicopters less than once a day, about once a day, a few times a day, or at least several times an hour? Response/Coding Categories: less than once a day (0), once a day (1), a few times a day (2), several times an hour or more (3), don’t know (5), refused (6), skipped (7) (continued on next page) Table 3-5. List of questionnaire items.

36 Assessing Community Annoyance of Helicopter Noise If respondent answers yes to Item 9, ask Item 9A; if any other response to Item 9, ask Item 10 next Item 9A Would you say you were slightly, moderately, very, or extremely annoyed by noise made by aircraft other than helicopters while you were at home last week? Response/Coding Categories: slightly (1), moderately (2), very (3), extremely (4), don’t know (5), refused (6), skipped (7) Item 10: While you were at home last week, did you notice repeated pounding or slapping noises made by helicopters? Response/Coding Categories: no (0), not home last week (1), yes (2), don’t know (5), refused (6) If respondent answers yes to Item 10, ask Item 10A; if any other response to Item 10, ask Item 11 next Item 10A Would you say that you were slightly, moderately, very, or extremely annoyed by thumping or slapping noises made by helicopters while you were at home last week? Response/Coding Categories: slightly (1), moderately (2), very (3), extremely (4), don’t know (5), refused (6), skipped (7) Item 11 While you were at home last week, did you notice buzzing noises made by helicopters? Response/Coding Categories: no (0), not home last week (1), yes (2), don’t know (5), refused (6) don’t know, refused If respondent answers yes to Item 11, ask Item 11A; if any other response to Item 11, ask Item 12 next Item 11A Would you say you were not at all, slightly, moderately, very, or extremely annoyed by buzzing noises made by helicopters while you were at home last week? Response/Coding Categories: slightly (1), moderately (2), very (3), extremely (4), don’t know (5), refused (6), skipped (7) Item 12: While you were at home last week, did you notice whining or tonal noises made by helicopters? Response/Coding Categories: no (0), not home last week (1), yes (2), don’t know (5), refused (6), skipped (7) If respondent answers yes to Item 12, ask Item 12A; if any other response to Item 12, ask Item 13 next Item 12A Would you say you were not at all, slightly, moderately, very, or extremely annoyed by whining or tonal noises made by helicopters while you were at home last week? Response/Coding Categories: slightly (1), moderately (2), very (3), extremely (4), don’t know (5), refused (6), skipped (7) Item 13 Did helicopters make vibrations or rattling noises in your home last week? Response/Coding Categories: no (0), not home last week (1), yes (2), don’t know (5), refused (6), skipped (7) If yes to Item 13, ask Item 13A; if any other response to Item 13, ask Item 14 next Item 13A Would you say that you are slightly, moderately, very, or extremely annoyed by vibrations or rattling noises in your home that are made by helicopters? Response/Coding Categories: slightly (1), moderately (2), very (3), extremely (4), don’t know (5), refused (6), skipped (7) Item 14 About how often do you notice vibrations or rattling noises in your home that are made by helicopters? Do you notice vibrations or rattling noises about once a week, once a day, or several times a day? Response/Coding Categories: once a week or less (0), once a day (1), several times a day (2), don’t know (5), refused (6), skipped (7) Item 15 Has any member of your household ever called or written to the airport to complain about noise made by helicopters? Response/Coding Categories: no (0), yes (1), don’t know (5), refused (6), skipped (7) If yes to Item 15, ask Item 15A; if any other response to Item 15, terminate interview Item 15A About how many times has a member of your household complained about helicopter noise in the last year? Has someone in your household complained just once, a few times, or many times over the last year? Response/Coding Categories: once (1), a few times (2), many times (3), don’t know (5), refused (6), skipped (7) Item 9 While you were at home last week, did noise made by aircraft other than helicopters bother or annoy you? Response/Coding Categories: no (0), not home last week (1), yes (2), don’t know (5), refused (6) Table 3-5. (Continued).

Site Selection and Opinion Survey Methods 37 • Variant 3: Assessment of annoyance due to secondary emissions (vibration and rattle) excited by BVI noise. • Variant 4: Assessment of predictability of annoyance from complaint information, particu- larly for sites with reliable complaint databases. 3.4 Description of Questions 3.4.1 Questions for Direct Comparison of Relative Annoyance of Exposure to Fixed- and Rotary-Wing Noise At sites for which it was possible to directly compare the relative annoyance of exposure to fixed- and rotary-wing aircraft, Items 8 and 9 follow the initial several items. Items 8 and 8A sought respondents’ opinions about the frequency of notice of exposure to noise of fixed-wing aircraft operations. The term “aircraft other than helicopters” was preferred because it would be easier for some respondents to understand than “fixed-wing” aircraft. The wording and coding of these items parallel those of Items 6 and 6A. Likewise, Items 9 and 9A parallel Items 7 and 7A. The similarity of wording and coding of these items were intended to support comparisons of the frequency of notice and degree of annoyance of fixed- and rotary-wing aircraft noise. 3.4.2 Questions for Assessing Relative Annoyance of Exposure to Various Forms of Helicopter Noise Items 10 through 12 were posed to respondents at sites exposed to noise from helicopter operations that generate more than one form of noise, and/or to operations of a mixed fleet of helicopters that includes some equipped with shrouded rotors (Fenestron) and some with open counter-torque rotors. 3.4.3 Questions for Assessing Annoyance of Helicopter-Induced Rattle and Vibration Items 13 and 14 were posed to respondents at sites exposed to blade slap noise. 3.4.4 Questions for Assessing Relationship Between Helicopter Noise Complaints and Annoyance Items 15 and 15A were intended to reveal potential relationships between helicopter annoy- ance prevalence and complaint rates, as well as potential relationships between helicopter com- plaint rates and noise exposure levels. 3.4.5 Target Population and Preparation of Sampling Frames The survey was intended to provide unbiased information about the relative annoyance of exposure to nonmilitary, fixed- and rotary-wing aircraft noise in adult residential populations. In practice, the population of interest is confined to geographic areas within relatively short ranges of aircraft flight routes and civil helipads. Opinions of the general population exposed only to occasional overflights and/or to low levels of fixed- and rotary-wing aircraft noise were of second- ary interest. By definition, an unbiased sample of any target population requires that each member of the target population have an equal opportunity of contributing opinions to the survey. This means, among other things, that respondents cannot self-select for participation in the survey. It also means that inexpensive methods for compiling a sampling frame (an exhaustive and current

38 Assessing Community Annoyance of Helicopter Noise enumeration of every person eligible for interview) are inappropriate for present purposes. These include constructing sampling frames from citywide voter registration, countywide tax assessor information, and other wide-area public records, not to mention random digit dialing of all numbers within a telephone exchange. Reverse telephone directories were common sources of sampling frames in the era when land- line telephone subscription was effectively universal. In recent years, rates of unlisted telephone numbers have become so high, and cell phone-only telephone subscription so widespread, that it has become difficult to rely on public information for such purposes.12 The Telephone Consumer Protection Act of 1991, as amended, further complicates and increases the cost of telephone- based interviewing. 3.5 Potential Interviewing Methods Three common methods of conducting interviews about opinions and reactions to aircraft noise exposure are by telephone, mail, and in person (face to face).13 As summarized in Table 3-6, each method is characterized by unique sets of advantages and disadvantages. These must be balanced against study goals. The questionnaire was administered by telephone to a sample of landline and cell phone subscriber households located within areas defined by the vertices of Feature PERSONAL (FACE TO FACE) POSTAL TELEPHONE Interview Completion Rate High Low Historically high; recently low Relative Cost of Data Collection High Low to moderate, depending on follow-up methods for nonresponse Intermediate (depends on sample incidence rate and numbers of callbacks) Duration of Data Collection Moderate (at least several days, dependent on field logistics) Long (weeks), vulnerable to shifts in opinions due to external events (e.g., aircraft crashes, current events) Short (several hours per day over the course of 3 or 4 days, depending on callback scheduling) Efficiency of Data Collection (Cost per Interview, Including Data Entry) Greatest in high population density settings Independent of population density Independent of population density Common Limitations High training costs, limited field supervision, costly to administer over wide areas No knowledge of respondent identity; loss of control over order of questioning; biased toward more literate respondents Questionable representation of younger, single, lower socioeconomic and less educated respondents, possible ethnic and racial biases Most Appropriate for… Administration of lengthy interviews to relatively small numbers of respondents in small, densely populated geographic areas about complex or sensitive matters Settings in which duration, temporal specificity, confirmation of the identity of respondents, and supervision of the interviewing process is not critical Representing residential response to noise exposure in large populations in short, well-defined time periods, with tight control over data collection Difficulty of Constructing a Sampling Frame Low (for example, field workers may be instructed to flip a coin or solicit interviews at every nth door or floor in an apartment building, or at every nth street address in an area of dense single-family detached dwellings) Moderate (currency of sampling frame is difficult to maintain in high- turnover rental areas) Moderate (workarounds required for high rates of unlisted telephone numbers and for cell phone-only users) Interview Quality Control Low (little effective real-time field supervision; slow tracking of response rates and callback success; difficulty in managing release of sub- samples and scheduling additional interviews) None; lengthy delays in administration and tracking of survey progress High (real-time supervision of interviewing possible; immediate tracking of sample incidence and refusal rates and scheduling of callbacks; possibility of conversion of refusals) Knowledge of Respondent Identity High None Intermediate Control Over Order of Complete None Complete Questioning Table 3-6. Comparison of relative advantages and disadvantages of alternate interviewing methods.

Site Selection and Opinion Survey Methods 39 polygons enclosing geographic areas with reasonably homogeneous aircraft noise exposure. The selection of telephone interviewing was based on the following factors: 1. The costs of making field measurements for prolonged periods to correspond with the period of questionnaire items (“While you’ve been at home during the past week. . . .”); 2. The need to control the order of presentation of questionnaire items; 3. The lack of necessity for lengthy and/or sensitive personal information; and 4. Overall data collection costs, except possibly at some (urban, high-density residential) sites, at which in-person (face-to-face) interviewing might be cost-effective. 3.6 General Discussion of Sample Size Constraints This section presents background information about the role of sample size in social survey design. A more specific discussion of sample sizes required to test the hypotheses of current interest is included in the mock data analysis section. The size of the population exposed to rotary-wing aircraft noise is a basic issue affecting study design and site selection. Larger sample sizes reduce the uncertainty of estimates of annoyance prevalence of rates for a given cumulative sound level exposure. They also reduce uncertainty about equivalent shifts, in decibels, of the dose-response curve that reflect nonacoustic influ- ences on annoyance prevalence rates. Smaller uncertainties, in turn, permit more reliable esti- mates of smaller differences in community tolerance to a noise source. Smaller sample sizes have the opposite effect. A basic decision must be made before final site selection regarding the minimal magnitude of effect of current interest, since it may not be realistic to seek evidence of small differences in annoyance rates at some sites. As a general- ity, surprisingly few (50–100) interviews may suffice to detect large differences between the annoyance of exposure to fixed- and rotary-wing aircraft, while surprisingly many (several hun- dred, if not more) interviews may be needed to detect small differences. In practice, the number of respondents and the size of expected differences in annoyance prevalence rates are the major factors affecting site selection. Annoyance prevalence rates may be expected to change by about 1% (near asymptotes of dosage-response relationships) to 3% (in the linear portion of dosage-response relationships) per decibel of noise exposure. If differences in annoyance prevalence rates between interviewing sites with exposures differing by only 3 to 5 dB must be detected, then 95% confidence intervals of about 2% to 3% are required. About 200 to 300 completed interviews are usually sufficient to achieve such confidence inter- vals. Roughly estimated, about half of the households in a sampling frame are likely to have unlisted telephone numbers, or cannot be reached with reasonable numbers of callbacks. Another half of the eligible respondents with listed telephone numbers may refuse to grant interviews. Working backward from confidence intervals of the desired widths, several thousand households must be eligible for interview by address-based landline telephone at a given interviewing site. Residential neighborhoods with uniform low-density housing (e.g., single-family detached dwellings on large lots) may therefore not be optimal as interviewing sites. Levels of exposure to helicopter noise may vary considerably across such sites, unless they extend for distances as great as miles parallel to well-defined helicopter flight corridors. 3.6.1 Size of Expected Differences in Annoyance Prevalence Rates due to Rotary- and Fixed-Wing Aircraft Noise Figure 3-11 shows a set of dosage-response relationships between cumulative noise expo- sure levels and percentages of respondents describing themselves as highly annoyed by aircraft

40 Assessing Community Annoyance of Helicopter Noise noise exposure. These curves are derived from the assumption that annoyance is most effectively predicted from the “effective” (duration-adjusted) loudness of noise exposure, as described by Fidell et al. (2011) and Schomer et al. (2012). The separations between dosage-response curves reflect varying degrees of community tolerance for noise exposure. For example, at a noise expo- sure level of Ldn = 65 dB in a community 6 dB less tolerant of helicopter than fixed-wing aircraft noise, an additional 15% of the population may be highly annoyed by helicopter noise than by fixed-wing aircraft noise. The curve reflecting the grand mean of annoyance judgments made by 75,000 social survey respondents at about 540 interviewing sites is the one in the middle (shown with filled red circle plotting symbols). The other curves are for communities that are either more or less tolerant than average of aircraft noise exposure. If helicopter noise is truly more annoying than fixed- wing aircraft noise on a decibel-for-decibel basis, then the annoyance of helicopter noise should be displaced from the mean curve shown in Figure 3-11. The amount of displacement from the mean curve is a decibel-denominated measure of the size of the effect of differential tolerance for the noise of fixed- and rotary-wing aircraft noise. 3.6.2 General Examples of Sample Size Requirements Figure 3-12 illustrates the effects of sample size (number of completed interviews) on the pre- cision of estimation of the prevalence of high annoyance. Precision of measurement of a bino- mial proportion, such as the proportion of a population highly annoyed by rotary-wing aircraft noise, is expressed in Figure 3-12 in terms of the widths of confidence intervals constructed around observed proportions. For moderate or greater sample sizes, the upper bound of the 95% confi- dence interval is the observed proportion plus 1.96(pq/n)1/2, while the lower bound of the 95% confidence interval is the observed proportion minus 1.96(pq/n)1/2, where p is the percent highly annoyed, q is the percent not highly annoyed, and n is the number of completed interviews.14 Figure 3-12 shows that over the range of annoyance prevalence rates of present interest, confi- dence intervals for estimates of proportions of respondents highly annoyed are smaller than 1%, Figure 3-11. Hypothetical differences in annoyance prevalence rates in communities with greater or lesser degrees of tolerance for noise exposure due to fixed- and rotary-wing aircraft.

Site Selection and Opinion Survey Methods 41 for samples of n = 200 and greater. The figure also shows that a point of diminishing returns in reduction in confidence interval width is reached at a sample size of about 200. Since the preci- sion of measurement is proportional to the square root of the sample size, further doublings of sample sizes yield only a factor of the square root of 2 (~1.4) improvements. In other words, impractically larger sample sizes would be required to reduce the widths of confidence intervals by useful amounts. It is therefore apparent that interviewing sub-sites would preferably be able to yield at least 200 completed interviews. Since not every household at a potential interviewing site can be contacted, nor is necessarily willing to grant an interview, a useful interviewing site must contain at least several multiples of 200 households. If the sample incidence rate is as great as 50% (that is, if half of the sampling frame can be reached and is willing to grant an inter- view), then the minimum number of households at a site should be 400. If the sample inci- dence rate is lower, the minimum number of households at a site must be correspondingly greater. 3.7 Noise Measurement Methods The social survey was accompanied by field noise measurements and INM estimates of heli- copter noise levels. Noise measurements and recording of aircraft flight tracks started 1 week prior to the first date of interviewing and continued for the remainder of interviewing. The duration of interviewing was expected to be 3 to 4 days, but was extended in some cases to permit additional callbacks to yield adequate numbers of completed interviews. Figure 3-12. Widths of 95% confidence intervals around a range of binomial proportions, as a function of sample size.

42 Assessing Community Annoyance of Helicopter Noise The basic noise measurement instrumentation was the Larson Davis 824 precision Class 1 sound level meter. [Class 1 refers to the International Electrotechnical Commission’s (2005) highest specification for precision sound level meters 2005]. Broadband audio recordings were made with a Zoom H2 digital recorder connected to the Direct Output of the L-D 824. The audio recorders use SD memory cards to store the audio signal in a standard audio WAV file format. The broadband audio files stored 24-bit samples at a rate of 44.1 kHz. The goal of the field measurements was to continuously document simultaneous measure- ment of sound pressure levels in A- and C-weighted decibel units, along with one-third octave band sound pressure levels, and broadband audio for the duration of the measurements. The broadband audio recordings allowed for manual identification of noise sources and also pre- served the noise environment near respondents’ homes for further analysis. Table 3-7 identifies the noise metrics recorded during the measurement survey. The acoustic measurements for Long Beach and Las Vegas were made simultaneously at four monitoring sites spaced throughout the survey area. The measurement sites were selected to col- lect data as nearly directly beneath the flight tracks and to the sideline of the corridors. Field measurements of actual noise exposure were calibrated and supplemented INM-based estimates of aircraft noise exposure. The noise measurement data were used to calibrate INM predictions so that exposure predictions could be generated for each household that completed an interview. This was done by using INM to create a grid of points or INM “location points” for each noise metric of interest. The field measurements were used to create a decibel dif- ferential between predicted and measured values at the four measurement points and at INM grid or location points. This grid was used to estimate noise exposures at the homes of the social survey respondents. The longitude and latitudes of respondents’ homes were coded in the sampling frame. FREQUENCY WEIGHTING TIME AVERAGING SLOW FAST IMPULSE Leq A X X X X C X X X X 1/3 Octave (12.5 to 20 kHz) X Audio (24 bit, 44.1 kHz) Not Applicable Table 3-7. Noise metrics simultaneously measured.