Evaluating the Use of Spatially Precise Diurnal Population Data in Aviation Noise Studies (2020)

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.

34 5 Creating a Spatiotemporal Aviation Noise Study with Examples 5.1 Introduction This section presents practices and guidelines we developed for airport professionals, policymakers, and aviation noise consultants regarding the use of spatiotemporal population data. The first step is to ask how spatiotemporal population data is likely to help understand and communicate the noise impact situation around the airport in question compared to traditional methods. Spatiotemporal population data can reveal areas of potential noise impact and sources of noise complaints if any of the following situations apply: • There are operations close to areas with high concentrations of daytime population. Urban downtowns tend to see large increases in the daytime population, while population tends to be more spread out across a metro area during the nighttime hours. • There are notable differences in the type of operations or routes aircraft tend to fly between the daytime and the nighttime. Examining day and night noise impacts separately may reveal information about noise impacts that are obscured when using 24-hour averaged noise. The value of spatiotemporal noise analysis may apply even when the differences in noise impacts do not meet regulatory standards for significant noise. Indeed, a large majority of noise complaints come from locations outside the 65 dB DNL (ACRP 2009, Synthesis 16). The procedure for setting up a spatiotemporal noise study includes the following five steps. Detailed guidance on each step is given below: 1. Choose the population data 2. Choose the noise metric 3. Prepare the flight data 4. Model the noise 5. Analyze the noise We end this section with practical guidance to be consistent with existing noise analysis practices including NEPA studies, Part 150 studies, and other research studies. 5.2 Choose the Population Data The various population data sets we examined included explicit daytime and nighttime population (LandScan USA), or daytime only (ESRI daytime). Both sets aimed to capture the total population in an area. LODES captured the number of workers’ movements between residential and workplace locations. The U.S. Census measured residential population, which generally corresponds to nighttime operations. The data set to use must be best applicable to the research question at hand. In the examples here, we applied the LandScan USA data for day and night since it had the highest resolution available and was the only data set explicitly separated into day and

35 night. The ESRI daytime population set can be used to study daytime noise exposure only. Census data, with its implicit nighttime bias since it is based on residential population, may be used as a proxy for nighttime population. 5.3 Choose the Noise Metric Each operation has a corresponding SEL. These are summed to calculate total noise energy. The SEL metric describes the total noise energy from all events if compressed into one second. Various time-averaged metrics then describe the equivalent sound level averaged over longer periods of time. The most common noise metric, the DNL, represents a day’s worth of operations averaged over an entire day. For example, there are 86,400 seconds in a day, so a total SEL level of 115.0 dB would translate to 65.6 dB DNL. The difference between the two, 49.4 dB, is simply 10*log(1/86,400). 5.3.1 Time-Averaged Metrics for Different Lengths of Time Day Equivalent Sound Level (LAeqD) and Night Equivalent Sound Level (LAeqN) metrics average noise over the 15 daytime hours and 9 nighttime hours, respectively. Both metrics are available by default in AEDT. Users can also define their own custom metrics for the desired period of time. The user must simply ensure the time-averaging constant matches the same time period covered by the set of flight operations to be evaluated (see Table 1). Table 1: Key Exposure-Based Metrics (Source: AEDT Technical Manual) 5.3.2 Operations Weighting for Noise-Sensitive Periods Common time-averaged metrics provide extra weightings to certain operations during noise- sensitive periods. The DNL metric counts each nighttime operation (from 10:00 pm through 6:59 am) as equivalent to 10 daytime operations of the same type. If all flight operations in a given flight set were conducted at night, they would have DNL contours 10 dB higher than the same flight operations in the daytime. That is because to the time-averaging adjustment from SEL to DNL would be 10*log(10/86,400) for nighttime as opposed to 10*log(1/86,400) for daytime. CNEL has the same extra weighting for nighttime operations as DNL, but also weights evening operations (7:00 pm through 9:59 pm) as equivalent to 3 daytime operations. CNEL is used instead

36 of DNL in California. By definition, LAeqD and LAeqN do not include any extra weightings for noise-sensitive periods. The user, therefore, must be mindful when interpreting the time-averaged metrics. As noted under the limitations of the current tools and practices in section 4.3, further research may inform the appropriate weighting to use for time periods designated as noise- sensitive, and a fully-developed activity-based model may identify multiple noise-sensitive times or places. Adding 10 dB to the LAeqN results is a recommended conservative starting point for assessing the impact of nighttime operations in isolation, but may over-estimate the impact compared to DNL, which dilutes 9 hours of nighttime operations over a full 24-hours. Any desired noise weightings can be applied in the default noise metrics shown in Table 1, creating a custom metric, or by modifying the flight data. Section 5.4 discusses how to change the operations weighting in the flight data. If weightings are applied in the flight data, the user should not use a metric that includes weightings to avoid double-counting. Examples of metrics without weightings include LAeqD, LAeqN, and LAEQ. One situation where a user may want to modify the flight data is when computing LAeqD for comparison to CNEL by adding a 3x weighting to evening operations. Supplemental noise metrics are available to describe the noise situation at a particular location in more detail. LMAX provides a measure of the maximum sound level experienced over a particular period of time. Number-of-events Above a Specified Level (NA) provides a count of the number of operations above a particular noise level, and Time Above Specified Level (TA) can describe the amount of time spent above a particular noise level. 5.4 Prepare the Flight Data The next step is to prepare the AEDT study and separate flight operations according to the time periods to be analyzed. In these examples we separated flights according to daytime (7:00 am – 9:59 pm) and nighttime (10:00 pm – 6:59 am) operations groups in AEDT, using the same definitions of day and night periods familiar to noise practitioners while using the DNL metric. We then combined the operation groups into three annualizations for day, night, and combined day and night (24-hour) operations. One way to account for noise weightings for noise-sensitive periods is to modify the operations count in the underlying flight data. For example, after separating daytime and nighttime operations, extra weightings can be applied to evening operations to simulate the extra weightings used in the CNEL metric. A processing script can be used to modify the operations count for evening events in the input ASIF (AEDT Standard Input File) data file to be loaded into AEDT. In Figure 10, the operations count is multiplied by three for an arrival operation in the 8 pm hour based on the <onTime> field. When LAeqD noise is computed for this operation, it will be given three times the usual weighting to account for extra sensitivity to evening operations.

37 Figure 10: Modified Operations Count in ASIF File 5.5 Model the Noise There are two main methods for using the population data: 1) compute noise directly on the population data set by importing into AEDT as a point receptor, or 2) compute noise on a regularly spaced grid and analyze the population impact using GIS software. We experimented with both approaches and determined that computing noise directly on the population points, while offering the most accuracy for noise levels at the affected points, can greatly increase AEDT run times, especially if the number of points is in the tens of thousands. Run times can be reduced by using multiple machines; the examples in the research ran AEDT on one machine. A recommended first step is to model noise onto a coarse regularly spaced grid covering a wide area around the airport, such as 0.5 nautical mile (nmi) spacing, to identify the general areas affected by noise. Once satisfied that the flight data is processing correctly, create a denser grid covering the areas affected by noise above the desired significance threshold, for example, 40 dB. A grid spacing of 0.2 or 0.1 nmi is usually sufficient. To minimize the number of receptor points and achieve the best AEDT run times, it may be worthwhile to use GIS software to create a contour of the 40-dB noise area on the coarse grid and clip the dense grid to the contour. The clipped grid can be reimported into AEDT as a point receptor set. For maximum accuracy of the population count for each noise interval, the user may wish to produce a point receptor set based on the population data. Clip the population set to the desired contour level in GIS software and filter out any zero population and “no data” points in the

38 population set. Then transform the population set into an ASIF file. An example is shown in Figure 11. The file must include the set name, latitude, longitude, population count, and a unique ID for the blockID field. The resulting XML file can be imported into the AEDT study as a “partial study” and will be recognized as a point receptor set. AEDT run times will increase with the number of receptor points in the study. Figure 11: Sample AEDT Point Receptor File 5.6 Analyze the Noise AEDT includes capabilities to produce noise impact tables and maps. But GIS software offers more flexibility to explore and visualize the data. GIS software, such as QGIS or ESRI’s ArcGIS can be used to draw contours (from grid receptor sets) and overlay the population sets and compute the noise exposure level in each contour band. GIS software also allows for the creation of custom maps to communicate the impacts from noise and population movements. Two demonstration analyses are presented below. The first is a sample Part 150 noise study for the Las Vegas area and the second is a NextGen research study of the San Francisco area from a JPDO/IPSA NextGen portfolio assessment. 5.6.1 Las Vegas Area Optimization Environmental Assessment (EA) Example Study The Las Vegas Area Optimization EA was conducted over a period of three years and concluded with the final EA that was released in the later part of 2012. Key characteristics of this study include: • The original study included McCarran International (LAS), Henderson (HND), and North Las Vegas Airport (VGT). Only operations from LAS were used for this example. • A 2012 baseline scenario and an RNAV-RNP 2017 optimized scenario. These are referred to as the “baseline” and “alternative” scenarios in this example.

39 • AAD operations representing 570 arrivals and 570 departures. • For LAS, we modeled 1,840 arrival backbones and 917 departure backbones. To simplify the data for the AEDT runs, we reduced the number of unique flight tracks and the number of unique flight operations. Flight operations were then scaled up to simulate the original level of traffic. The sample study therefore does not show actual noise conditions around the airport. 5.6.1.1 Split Operations and Population into Day and Night Looking first at the population data, the highest concentrations of daytime population can be seen in the urban downtown. Figure 12 shows the daytime population in the Las Vegas area in the LandScan USA data set. In this case, the highest number of people with potential exposure to noise would be northwest of the largest airport, LAS. Given the very high concentration of people in these areas in the daytime, the total number of people exposed to daytime noise will be highly dependent on the operations along tracks closest to them. For example, operations departing from runways 1L or 1R and turning west potentially have the largest impact on these areas when the baseline and alternative scenarios are compared in a moment. Figure 12: Las Vegas LandScan USA Daytime Population

40 Figure 13 shows the LandScan USA population data for the same region at night. The population is notably more spread out across the metro area and the number of people in the downtown areas northwest of the airport have dispersed. The highest remaining concentrations are further north of LAS. Figure 13: Las Vegas LandScan USA Nighttime Population Looking at the flight tracks over the area, Figure 14 shows the baseline and alternative cases with daytime operations only. In this example, the alternative case generally shows less dispersion in the tracks. Tracks are from LAS operations only. Figure 14: Las Vegas (LAS) Baseline Daytime Tracks: Baseline (left); Alternative (right)

41 Figure 15 shows noise impact for daytime operations only (7:00 am through 9:59 pm). The LAeqD metric averages noise over 15 daytime hours. Figure 15: Las Vegas (LAS) Day Contours: Baseline (left); Alternative (right) In Figure 16 the 55-60 dB LAeqD contours are compared for the baseline (“No Action,” in red) and the alternative (“Optimized,” in blue). When combining daytime population and the daytime baseline and alternative contours, it can be seen that the change in configuration helps avoid the highest concentration of daytime population just northwest of the airport, as seen previously in Figure 12. Indeed, the population inside the 55-60 dB contours is reduced by nearly 83,000 people, as will be shown in the population discussion later in Table 2. This is one possible way spatiotemporal data can help in noise abatement planning. Moving tracks may not always be possible given other operations constraints but can be one factor to consider. Figure 16: Baseline (No Action, Red) and Alternative (Optimized, Blue) Daytime 55-60 dB LAeqD Contour and High Population Areas

42 Figure 17 shows the baseline and alternative tracks for nighttime operations only. Since the nighttime population is less concentrated than in the daytime and is more spread out across the city, the change in configuration is less dramatic. The population change shown in Table 2 therefore shows a smaller difference between the baseline and the alternative. Figure 17: Las Vegas (LAS) Baseline Nighttime Tracks: Baseline (left); Alternative (right) Figure 18 shows nighttime noise impact only (10:00 pm through 6:59 am). The LAeqN metric averages noise over 9 nighttime hours. By definition, LAeqN contains no additional penalty to account for the sensitivity of nighttime operations. Therefore, these contours were expanded to show very low levels of noise, down to 40 dB. A conservative approach for interpreting the severity of LAeqN noise impact would be to consider noise levels up to 10 dB lower than usual. For example, when concerned about noise levels in the 55-60 dB range in DNL, look at the 45-50 dB exposure in LAeqN. Figure 18: Las Vegas (LAS) Night Contours: Baseline (left); Alternative (right)

43 Table 2 compares day versus night operations on the separate daytime and nighttime operations. The LAeqN metric, by definition, does not contain any added penalty to account for the sensitivity of nighttime operations. Examining the noise bands 10 dB below the daytime noise level is a conservative way of accounting for this sensitivity. Table 2: LandScan USA Day vs. Night Population LandScan USA (2018) Day LandScan USA (2018) Night LAeqD Baseline Alternative Change LAeqN Baseline Alternative Change <40 780,730 802,237 21,507 <40 1,150,684 1,340,977 190,293 40-45 243,885 329,720 85,835 40-45 394,298 256,163 -138,135 45-50 261,976 307,657 45,681 45-50 193,945 186,844 -7,101 50-55 352,328 324,091 -28,237 50-55 94,083 50,002 -44,081 55-60 201,981 118,987 -82,994 55-60 9,722 9,028 -694 60-65 87,847 50,917 -36,930 60-65 537 255 -282 65-70 15,196 10,283 -4,913 65-70 0 0 0 70-75 676 727 51 70-75 0 0 0 >75 0 0 0 >75 0 0 0 1,944,619 1,944,619 1,843,269 1,843,269 5.6.1.2 Full-Day Operations and Synthetic 24-Hour Population The contours below compare the noise exposure based on the traditional 24-hour averaged noise using DNL. The 10-dB nighttime penalty is implicit in the DNL metric. These contours were expanded to show very low levels of noise, down to 40 dB DNL. This is to facilitate comparisons with the day versus night contours, which contain no such noise penalties. Figure 19 shows DNL impact for the baseline and alternative examples. Figure 19: Las Vegas (LAS) 24-Hour Contours: Baseline (left); Alternative (right) compare population exposed to noise by contour band. Table 3 compares 24-hour population exposed to DNL noise using a weighted 24-hour average of the LandScan USA data compared to the traditional Census based data. Since Census data represents residential population, it tends to have an inherent nighttime bias. The 24-hour averaged population allows for a direct comparison of noise impact using the familiar DNL metric while still accounting for some day-to-night movement of population. The DNL metric includes 15 hours of daytime noise and 9 hours of

44 nighttime noise. A synthetic 24-hour average can be computed using the formula in Equation 1 below, where D is the daytime population and N is the nighttime population for each point in the population data set. Synthetic 24-hour population = (15*D + 9*N) / 24 Equation 1: Synthetic 24-Hour Averaged Population The 24-hour weighted data (Table 3) does indicate a higher number of people exposed to noise above 55 dB, about 160,000 extra people in the baseline case, and 80,000 extra in the alternative case compared to Census data, and more than 18,000 extra people (baseline) and about 9,500 extra people (alternative) in areas at or above 65 dB. This particular example therefore shows that applying spatiotemporal data that accounts for the diurnal movement of people reveals a higher number of people exposed to noise. Using spatiotemporal data in this case reveals the larger number of people exposed to noise in the urban downtown area which has high concentrations of daytime population. Airports close to downtown areas may see a similar pattern. Each airport’s noise situation is unique. Applying spatiotemporal data at another airport may show different results. As noted in the limitations of the approach, the presence of people in a particular daytime band does not describe their level of noise sensitivity in those locations. An activity-based model, which includes information about both location and noise sensitivity at a given time, could help interpret the significance of these results. Land use data, such as parcel data showing building type, could be one such source of this information. It could potentially also identify building types that may benefit from remediation efforts, such as soundproofing materials. Table 3: Weighted 24-Hour Population vs. Census LandScan USA (2018) Weighted 24-Hour Census (2018 estimated) DNL Baseline Alternative Change DNL Baseline Alternative Change <40 586,073 774,568 188,495 <40 714,661 914,566 199,905 40-45 345,831 277,503 -68,328 40-45 298,557 293,543 -5,014 45-50 278,957 304,739 25,782 45-50 344,398 293,231 -51,167 50-55 335,207 322,859 -12,348 50-55 326,887 237,296 -89,591 55-60 245,350 158,716 -86,634 55-60 144,991 103,466 -41,525 60-65 90,312 55,299 -35,013 60-65 49,664 40,028 -9,636 65-70 21,929 11,986 -9,943 65-70 6,404 3,431 -2,973 70-75 2,948 943 -2,005 70-75 8 7 -1 >75 6 0 -6 >75 0 0 0 1,906,613 1,906,613 0 1,885,570 1,885,568 5.6.2 San Francisco Research Study, JPDO/IPSA NextGen Portfolio Assessment Between 2009 and 2014, the JPDO conducted NAS-wide studies for the IPSA to model the benefits of NextGen capabilities such as new technology or high-density operations. This demonstration uses the study data for the San Francisco area as baseline conditions and contains operations from San Francisco International (SFO) and Metropolitan Oakland International (OAK) airports. We simplified the study data to facilitate faster run times in AEDT running on a single machine. As with the Las Vegas study, this was done by selecting a representative aircraft to fly each of the flight tracks and scaling up the result to equal the number of original flight operations in the study.

45 The flight tracks represent the paths taken by actual traffic; however, the noise levels shown do not represent actual noise conditions at the airport. The San Francisco study consists of about 250 arrival backbone tracks and 470 arrival Average Annual Day (AAD) aircraft operations, 120 departure backbone tracks, and about 500 departure AAD aircraft operations. The San Francisco example, focused on SFO, is intended to show flight tracks, noise contours, and day versus night population all together. It also compares results from the different population data sources evaluated in this research. Figure 20 shows noise contours for daytime operations only. Noise is shown in modified LAeqD: the noise values shown include all operations from 7:00 am through 9:59 pm with evening operations from 7:00 pm through 9:59 pm weighted as three operations. This was done by altering the operations count in the study’s input ASIF file, as described in section 5.3.2. Figure 20: SFO Modified LAeqD with Evening Noise 3x Weighting, LandScan USA Day Population

46 Figure 21 shows noise contours for nighttime operations only. Noise is shown in LAeqN, which is the 9-hour weighted noise for operations from 10:00 pm through 6:59 am. By definition, the LAeqN metric does not include a 10-dB addition for night operations used in computing the DNL metric. As in the previous airport example, a conservative way to interpret noise severity is to consider noise bands 10 dB less than usual. Figure 21: SFO LAeqN Noise, LandScan USA Night Population

47 The LODES workplace population set does not include all persons but is a measure of people at a workplace by Census block location. Like the LandScan USA data set, LODES shows that there tend to be a few concentrated areas of high population density around workplaces. LODES does not specify daytime or nighttime movements, but the workplace population generally corresponds to daytime location. Figure 22 shows LODES workplace population and LAeqD noise from daytime operations. Figure 22: LODES Working Population and Modified LAeqD Noise

48 Similar to the LandScan USA nighttime population set, the LODES residential population shows fewer concentrations of very high population density, and the population is more spread out across the metro area. Figure 23 shows LODES residential population and LAeqN noise from nighttime operations. Figure 23: LODES Residential Population and LAeqN Noise

49 The ESRI Daytime population set shows another estimate of the areas with greatest population densities during the daytime. Figure 24 shows ESRI Daytime population and modified LAeqD noise from daytime operations. Figure 24: ESRI Daytime Population and Modified LAeqD Noise Of the data sets that are intended to show total population, the LandScan USA, ESRI daytime, and U.S. Census values are generally comparable over the study area. The differences between population sets for individual noise contours may vary. There is insufficient information about the data sets to explain differences in population values over very small areas. The LODES data set, which only counts working population, is noticeably smaller than the general population sets (Table 4 and Table 5).

50 Table 4: Sample San Francisco Day Populations Modified LAeqD LandScan USA LODES Workers ESRI Daytime <40 44,598 20,911 43,580 40-45 60,889 29,438 86,644 45-50 116,887 48,556 127,362 50-55 106,980 61,054 88,471 55-60 83,047 53,957 83,119 60-65 28,196 10,673 13,603 65-70 3,272 2,355 846 70-75 976 4,857 0 >75 4 0 0 444,849 231,801 443,625 Table 5: Sample San Francisco Night Populations LAeqN LandScan USA LODES Residential Census 2018 est <40 112,473 57,016 40,567 40-45 113,337 51,376 77,946 45-50 90,874 46,235 134,203 50-55 80,674 43,253 110,519 55-60 44,495 20,431 80,508 60-65 6,869 3,579 12,181 65-70 453 297 1,082 70-75 0 10 0 >75 0 0- 0 449,175 222,197 457,006 5.7 Existing Policy and Practices Consistency In this step, we have checked for consistency between existing regulations, policies, guidelines, and practices, and the use of these new potential analytic procedures. We have also identified which regulations, policies, guidelines, and practices would be affected when using diurnal data. Ultimately, any practices and guidelines developed will have to conform to existing regulations, policies, guidance, and practices. We began this process by grouping existing regulations, policies, guidelines, and practices into three categories: 1) current policies, 2) additional guidance and practices documents, and 3) related DOT agencies’ policies and guidance. • Current policies (laws, regulations, and orders) and guidance (desk guides) that govern environmental review of actions related to aviation and potentially affecting surrounding communities. o NEPA (42 United States Code [U.S.C.] §§ 4321-4335) o Council on Environmental Quality (CEQ), Title 40, Code of Federal Regulations (CFR), Parts 1500-1508, Regulations for Implementing the Procedural Provisions of the National Environmental Policy Act (CEQ Regulations) o DOT Order 5610.1C, Procedures for Considering Environmental Impacts

51 o FAA Order 1050.1, Environmental Impacts: Policies and Procedures with its Desk Reference o FAA Order 5050.4, NEPA Implementing Instructions for Airport Actions with its Desk Reference o Title 14 CFR, Part 150, Airport Noise Compatibility Planning with FAA Advisory Circular 150/5020-1, Noise Control and Compatibility Planning for Airports o FAA Order JO 7400.2, Procedures for Handling Airspace Matters, Chapter 32, Environmental Matters • Additional guidance and practices documents that assist in the aviation-related environmental review process. o International Civil Aviation Organization (ICAO)/Committee on Aviation Environmental Protection (CAEP) o Federal Interagency Committee on Noise (FICON), Federal Agency Review of Selected Airport Noise Issues o Civil Air Navigation Services Organization (CANSO) o Aviation Sustainability Center (ASCENT), also known as the FAA Center of Excellence (COE) for Alternative Jet Fuels and Environment, and the Partnership for AiR Transportation Noise and Emissions Reduction (PARTNER) COE o FAA’s Community Involvement Manual o Similar Past or Ongoing ACRP Projects o FAA’s Air Traffic Community Involvement Plan • Related DOT agencies and their environmental policies and guidance. o Federal Highway Administration (FHWA), https://www.fhwa.dot.gov/environment/noise/ o Federal Railway Administration (FRA), Procedures for Considering Environmental Impacts, 64 FR 28545, 14(n)(3) Noise and vibration. The relationships among the various policies and practices are shown in Figure 25. Additionally, Table 6 through Table 8 in Appendix A include the specific references examined for each.

52 Figure 25: Relationship of Policies and Practices In examining the various policies and practices documents, we specifically were looking for references and applications related to the use of population centroid data. As stated previously, the applicability of potentially employing diurnal population data was found to be potentially beneficial in NEPA-type studies, Part 150 studies, and for research studies. Of those, NEPA and Part 150 studies are directly related to existing polices and guidance we examined. For both, DNL is the mandated metric. They both also use some form of population data in the analysis, in the determination of potential impacts, and in consideration during the decision-making process. Next, we considered how the use of diurnal population data could affect the policy and guidance. It is feasible that diurnal population data could be used as a supplement to, or instead of, static population data. For instance, instead of reporting just the number of people located in their National Environmental Policy Act (NEPA) (42 United States Code [U.S.C.] §§ 4321-4335) Council on Environmental Quality (CEQ), Title 40, Code of Federal Regulations (CFR), Parts 1500-1508 Department of Transportation (DOT) (Order 5610.1) Federal Highway Adm. (FHWA) (https://www.fhwa.dot.gov/environment/noise/) Federal Aviation Adm. (FAA) (Order 1050.1) Office of Airports (Order 5050.4) Air Traffic Organization (Order 7400.2, Chap. 32) International Civil Aviation Org., Committee on Aviation Environmental Protection (ICAO/CAEP) Federal Interagency Committee on Noise (FICON) Federal Agency Review of Selected Airport Noise Issues Civil Air Navigation Services Organization (CANSO) ASCENT/PARTNER Centers of Excellence (COEs) Title 14 CFR part 150, Airport Noise Compatibility Planning Airport Cooperative Research Program (ACRP) Federal Railroad Adm. (FRA) (Procedures for Considering Environmental Impacts, 64 FR 28545, Section 14(n)(3)) Air Traffic Community Involvement Plan FAA Community Involvement Manual

53 residences (basically in the evening), the number of people affected at various locations during different times of the day could be reported. This would give decisionmakers the ability to determine where those same people might be throughout the day. Aviation noise practitioners cover a fairly wide range of noise impact assessments. Within the range of use cases that utilize population centroid data, we find it useful to group them under four broad categories: 1) studies supporting NEPA, 2) noise compatibility planning, 3) other analyses/activities, and 4) research efforts. In this subsection, we discuss the applicability of potentially employing diurnal population within each of these four broad categories. 5.7.1 NEPA Studies Guidance NEPA decisions are generally made based on the determination of exceeding the noise significance threshold at population centroids and an increase in the number people impacted by the increase in noise. This policy is established in FAA Order 1050.1F, Appendix B, Section B-1.4, Environmental Consequences. As stated above, if diurnal population data is determined to be feasible for NEPA-type studies, it could be used as supplemental data to the standard/mandated noise analysis. Supplemental data and analyses are beneficial to inform the federal decisionmaker and to assist in the public’s understanding of the potential noise impact. This supplemental analysis option is in line with the following standard policy and guidance: • FAA Order 1050.1F, Desk Reference, Section 11.4, Supplemental Noise Analysis • FAA Order 5050.4B, Desk Reference, Chapter 17, Noise • FICON, Federal Agency Review of Selected Airport Noise Issues, Sections 3.2 and 3.3 Additionally, if the use of diurnal population data is determined feasible for use instead of static population data, this is a more challenging use. If used instead of static data, this would have implications for the determination of “significant noise impact.” This would require a change to the policies established in FAA Order 1050.1F (see Order 1050.1F, Exhibit 4-1, Noise and Noise- Compatible Land Use) and implemented in Orders 5050.4 and 7200.2, Chapter 32. These changes would require not just research, but validation, potential public comment and input, and DOT and CEQ concurrence and approval. Of the listed impact categories, Noise and Noise-Compatible Land Use are those that use population centroid data and could potentially benefit from inclusion of diurnal population data. Order 1050.1F, Appendix B, Section B-1.3, Affected Environment, describes related requirements to provide appropriate context in the use of this data: • DNL contours or noise grid points showing existing aircraft noise levels. Noise exposure contours must include DNL 65, 70, and 75 dB levels (additional contours may be provided on a case-by-case basis). Noise grids are sized to cover the study area for noise analysis. Multiple grids may be created, but at least one grid consists of population centroids from the U.S. Census blocks. • The number of residences or people residing within each noise contour where aircraft noise exposure is at or above DNL 65 dB; or for a larger scale air traffic airspace and procedure action, the population within areas exposed at or above DNL 65 dB, at or

54 above DNL 60 but less than DNL 65 dB, and at or above DNL 45 dB but less than DNL 60 dB. Additionally, the analysis of potential EJ and School Learning impacts may also benefit from the use of diurnal population data. 5.7.2 NEPA Significance Thresholds The FAA uses thresholds that serve as specific indicators of significant impact for some environmental impact categories. FAA-proposed actions that would result in impacts at or above these thresholds require the preparation of an EIS unless impacts can be reduced below threshold levels. As noted above, the NEPA process requires the assessment of noise impacts at population centroids and, hence, they are used in noise studies supporting EAs and EISs. The FAA has determined that the significant noise threshold has been exceeded when a federal action would increase noise by DNL 1.5 dB or more for a noise-sensitive area (identified by those population centroids) that is exposed to noise at or above the DNL 65 dB noise exposure level, or that will be exposed at or above the DNL 65 dB level due to a DNL 1.5 dB or greater increase, when compared to the no-action alternative for the same timeframe (see Order 1050.1F, Exhibit 4-1, Significance Determination for FAA Actions). Typical NEPA analyses are conducted for regional airspace traffic flow changes, airport changes such as a new runway, or new flight procedure implementation. Typical noise analyses compare the changes in people impacted by significant noise between a no-action scenario and one or more alternative scenarios. Noise is modeled at census population centroids with total number of people impacted by the significant noise reported, as shown in Figure 26.

55 Figure 26: Representative Noise Impact Graph Depicting Significant Changes For these studies, DNL is the mandated metric, although supporting or supplemental noise analyses may be conducted using additional metrics. Additionally, if appropriate guidelines and technical procedures are developed, it is feasible that diurnal population data could be used instead of, or as a supplement to, static population data. Instead of reporting just the number of people located in their residences (basically in the evening), this would give decisionmakers the ability to determine where those same people may be throughout the day. If used instead of static data, this would have implications for the determination of “significant noise impact.” However, if used as supplemental data, it could assist the public in understanding the potential impact of these federal actions. 5.7.3 Land Use Compatibility Planning (“Part 150” Studies) The basic purpose of 14 CFR Part 150 is to promote a planning process through which an airport operator can examine and analyze the noise impact created by the operation of their airport, as well as the costs and benefits associated with various alternative noise reduction techniques. In addition, the responsible impacted land use control jurisdictions can examine existing and forecast areas of noncomputability and consider actions to reduce noncompatible uses. While Part 150 studies do not mandate the use of population centroids, aviation noise practitioners have often included population centroids in order to provide estimates of the number of people residing within the required 65 dB, 70 dB, and 75 dB DNL contours (see 14 CFR Part 150, Part B, Section A150.101(a)). This is done to support the development of a NEM for current and proposed conditions and to identify adverse impacts early in the process.

56 As with NEPA studies, this process may be able to utilize diurnal population data as a supplement or instead of static population data to better understand the impacts. However, if it is determined to be feasible for use instead of static population data, as with NEPA, policy and regulatory decisions will need to be undertaken which would also require not just research, but validation, public comment and input, and DOT and CEQ concurrence and approval. 5.7.4 Other Analyses/Studies As stated previously, the most promising uses of diurnal population data are potentially in the field of aviation noise research. Similar to noise modeling for NEPA and Part 150 processes, typical research efforts examine the noise impact differences between two or more scenarios. As these are research efforts, there is greater leeway to explore the use of diurnal population data instead of static population data and present the findings in new and innovative ways. It is these types of research studies that can provide the data necessary to justify the use of diurnal population data in NEPA and Part 150 studies. Just as the research that has previously been conducted by FICON, ACRP, and the COEs has been used to make adjustments and modifications to noise analysis procedures and practices, research in this area can become the foundation for revised policies and practices in the future.