Improving AEDT Noise Modeling of Mixed Ground Surfaces (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.

5-1 CHAPTER 5. VALIDATION DATA A listing of data that can be used to validate the methods above and study the sensitivity of their application is presented in this chapter. The methods to estimate the ground effect that come from the sensitivity studies described in Chapter 6 will be validated with this data. A second use of this data will provide information for future studies. 5.1. Topography and National Land Cover Database The National Land Cover Database (NLCD) with classification of land cover in the continental United States is available at www.mrlc.gov. The classification was performed with satellite imagery with a 30 m resolution. The land cover classifications are listed in Table 5. The website allows the downloading of this land cover data and allows for downloading National Elevation Data (NED) for the same requested area free of charge. The BASEOPS program which is part of the NOISEMAP suit of programs can be used to format these data into the elevation and impedance file formats used by AAM and NMAP. These file formats were also suggested as part of the weather and terrain study by Plotkin et al. (2013). The flow resistivity values that BASEOPS associates with the different land cover categories are estimates based upon a compilation of measurements of similar land cover types (Page, et al., 2008). No direct measurements or validation of the associated flow resistivity values as a function of the NLCD ground cover has been performed. The flow resistivity values in Table 5 are estimates based on the ground type similarity to other ground types whose flow resistivity has been measured. Because the ground handling algorithms in NMAP are concerned with whether the surface is hard (greater than 1000 kPa s/m2) or soft (less than or equal to 1000 kPa s/m2) the accuracy of these classifications has not been scrutinized. BASEOPS allows the user to reclassify the flow resistivity for each land cover type. A method in keeping with the original intent of Rasmussen’s binary treatment of the surface as hard or soft is to identify all water and wetlands as 106 kPa s/m2 and everything else as 225 kPa s/m2. A study of ground cover around all airports in the continental US was made to determine which had the largest ratio of water to land within a 2 nmi radius of the airports’ reference points. For these airports, the areas of each of the classifications of land cover were calculated within a 2, 7, and 14 nmi radius centered on the airports’ reference points. Based on this information, together with the availability of a local noise monitoring system, and including suggestions from the ACRP panel and A-21 PWT members, ground cover data was obtained from DCA, BOS, PHL, SAN, LHR, SFO, OAK, and PDX. A tool has been developed that reads in elevation and impedance data files processed from the NLCD. The processing from the NLCD was performed using BASEOPS which associated the flow resistivity values with the land cover classifications. As an example, the impedance data available around the Port of Portland International Airport (PDX) is shown in Figure 27. The graph at the bottom of the figure shows the elevation profile to either side of a point on a track (marked by a large pink dot in the figure) extending from runway 28R. Like the area in the top of the figure, the flow resistivity values are colored from blue to green representing the maximum flow resistivity (water at 100000 kPa s/m2) to the minimum (forest at 50 kPa s/m2), respectively. This tool will be used to identify the profile between aircraft and receivers for the data sets described in the following section on airport data and can be expanded to calculate the Fresnel zone average area impedance as discussed above. One possible use of this data is the method of classification for hard and soft surfaces in NMAP discussed above as well as isolating land and water in the definition of the flow resistivity. It is expected that the procedure for reducing an area in the NLCD will be included in an AEDT manual update if the file types used by AAM and NMAP are allowed to be used; otherwise, the format of the files is not proprietary and is well described in the NOISEMAP documentation. The Swedish National Space Board have sponsored work

5-2 (Sohlman et al.,2004) on a similar effort to use satellite imagery to classify ground cover and identify flow resistivity. Their aim was to classify data according to the surface types available in Nord2000. TABLE 5 National Land Cover Database Ground Cover Classifications and Associated BASEOPS Flow Resistivity Estimates NLCD2011 Land Cover Classifications Estimated Flow Resistivity (kPa s/m2) 11. Water 100000 12. Perennial Ice Snow 20000 21. Developed, Open Space 225 22. Developed, Low Intensity 10500 23. Developed, Medium Intensity 19500 24. Developed, High Intensity 25500 31. Bare Rock/Sand/Clay 3000 32. Unconsolidated Shore (such as silt) 650 41. Deciduous Forest 50 42. Evergreen Forest 50 43. Mixed Forest 50 51. Dwarf Scrub (Alaska Only) 200 52. Shrub/Scrub 50 71. Grasslands/ Herbaceous 225 72. Sedge/Herbaceous (Alaska Only) 225 73. Lichens (Alaska Only) 225 74. Moss (Alaska Only) 225 81. Pasture/ Hay 225 82. Cultivated Crops 200 90. Woody Wetlands 100000 91. Palustrine Forested Wetland 90000 92. Palustrine Scrub/Shrub Wetland 90000 93. Estuarine Forested Wetland 90000 94. Estuarine Scrub/Shrub Wetland 90000 95. Emergent Herbaceous Wetlands 90000 96. Palustrine Emergent Wetland (Persistent) 90000 97. Estuarine Emergent Wetland 90000 98. Palustrine Aquatic Bed 90000 99. Estuarine Aquatic Bed 90000

F5.2. Airp A that come for this p aircraft o operation with the IGURE 27. ort Data irport noise from the sen roject from s peration, the . The noise d sound exposu Flow resisti data obtained sitivity studi uch systems tracking da ata required re level. M vity and elev from airpor es of the diff or specialize ta for the ai are the maxi ore robust d ation contou t noise moni erent method d measurem rcraft, and t mum A-weig ata sets wou rs around P tors will be u s reviewed. ents are the he meteorolo hted sound l ld include t DX - profile sed to valid At a minimum noise measu gical data a evels for the he one-third along pink ate the algor , the data n red from a k t the time o operation tog octave band 5-3 line. ithms eeded nown f the ether time

5-4 history at the monitor during the aircraft overflight. The tracking data has to have at least the position of the aircraft relative to the monitor at the closest point of approach. Higher fidelity tracking data would have the position and speed of the aircraft as a function of time for the entire operation. Meteorological data should have at least the pressure, temperature, relative humidity, and vector wind speed for the hour of the operation. Exceptional meteorological data sets would have a measure of the atmospheric profile (i.e. these variables as a function of altitude). The type of aircraft, operation (departure or arrival), and runway used are also needed. Noise monitoring and flight track data was obtained from the following airports:  Washington National Airport (DCA) (2002-2003, Arlington, USA)  Boston Logan International Airport (BOS) (1999, Boston, USA)  Portland International Airport (PDX) (2015- 2016, Portland, USA)  Philadelphia International Airport (PHL) ( 2007, Philadelphia, USA)  Gatwick International Airport (LGW) (1996-1999, London, UK)  London Heathrow Airport (LHR) (2015, London, UK)  Oakland International Airport (OAK) (2016, Oakland, USA)  San Francisco International Airport (SFO) (2016, San Francisco, USA) 5.2.1. Washington National Airport (DCA) There are two sets of measurements at DCA that can be used as part of the validation in this project. The first was conducted in conjunction with a Part 150 study (RICONDO, 2004). Monitors were deployed around the airport as depicted in Figure 27. Aircraft flying over the ILS monitor in the figure could be used to isolate the effect of water and mixed-impedance surfaces on aircraft noise propagation by comparing the noise levels at the WGC and A-IT monitors. The noise monitors were deployed from October 17 to October 31, 2002. During that time 349 aircraft operations that utilized the main runway (01-19) were selected for study. The noise monitors were set to gather event data that included the A-weighted maximum level (Lmax), sound exposure level (SEL), and duration of the events. Radar tracks were collected during the study. Full position data as a function of time is included along with the aircraft type, origin/destination, and runway utilization.

O by Downi 2002 and handling monitors closest at 28. The N O locations. airport co well as we the days f campaign ther data inv ng, et al. (20 Spring 2003 ground impe each were de the embankm monitors w ne-third octa Only two of llected spectr ather data fr or the events s is presented FIGUR olving aircra 04). The pro . The objec dance. A refe ployed in a l ent. The mo ere only depl ve band time the three mo al data. Trac om a station are displaye in Appendix E 28. Nois ft operations ject studied tive was to d rence monit ine normal to nitor location oyed for the F history data nitors deploy king data wi set up at the m d in Table 6. E. e monitors d at DCA that aircraft prop etermine the or was deplo the river wi s RF, S1, 2, all 2002 mea was collecte ed in each o th aircraft po onitor array The format eployed arou can be used agation over accuracy o yed on the a th the furthes and 3, and N surements an d on both si f the arrays o sition as a fu s. The numb of the data f nd DCA. for this proje water for the f NOISEMA irfield, and t t 100 ft from 1, 2, and 3 ar d not utilize des of the ri n the side of nction of tim er of events iles from the ct is from a Navy durin P’s algorithm wo arrays of the shore an e shown in F d in the analy ver at the mo the river fro e was collec in the analys two measure 5-5 study g Fall s for three d the igure sis. nitor m the ted as is and ment

5-6 TABLE 6 Measurement Dates and Daily Event Count Date Events Fa ll 10/09/2002 14 10/22/2002 46 Total 67 Sp ri ng 5/20/2003 52 6/9/2003 14 6/10/2003 8 6/11/2003 64 6/12/2003 65 6/16/2003 82 Total 282 Total 349 5.2.2. Boston Logan International Airport (BOS) A noise measurement study was conducted at Logan International Airport in 1999 in Boston, Massachusetts to examine the applicability of mathematical models of lateral sound attenuation available at that time (Senzig, et al. 2000). The algorithm in question was SAE-AIR-1751 (SAE, 1981), and the analysis associated with this study lead to the development of SAE-AIR-5662 (SAE, 2006). One of the objectives of the Logan Study was to assess the accuracy of the new ground effects regressions developed for future inclusion in the FAA’s INM. Since Logan is a coastal airport, the study concentrated on the assessment of the ground effects regressions developed for propagation over an acoustically hard surface. Simultaneous acoustic measurements were conducted at three locations along the Winthrop, Massachusetts shoreline. As Figure 28 shows, these locations consisted of a centerline reference site located directly underneath the nominal flight track for departures and approaches on Runway 9/27 (Point 1), a second site located about 915 m (3,000 ft) to the north of the nominal track (Point 2), and a third site located on Snake Island in Boston Harbor (Point 3). Two microphones were deployed at each location. The first microphone was installed at a height of 1.5 m (5 ft), while the second microphone was installed directly above the first microphone at a height of 4.5 m (15 ft). The microphone assemblies at sites 2 and 3 were directly adjacent to the shoreline. Thus, both sites had the same effective ground impedance. The measurements took place over seven days in the summer of 1999. A total of 339 departure events were measured (see Figure 29), and 237 of those events were for six primary aircraft, which were the focus of the original analysis. This data set was reduced down to 45 events after applying meteorological and climb gradient criteria, and five different data points associated with different elevation angles were measured for each event (see Figure 30). The data in these figures were corrected for spherical spreading, atmospheric absorption, and ground effect. The data are referenced to the under- track microphone (1 in the figure). The dependent axis is labeled ‘residual’ in the figures because it is believed to be synonymous with the engine installation effect, since neglecting meteorological effects, all other physical effects in the measured data have been accounted for in the correction process. The measurement data consisted of acoustical data, video data, and meteorological data. For each measurement event the acoustic data comprise timestamped one-third octave band spectral time histories

of the sou position; pressure, further pr water sur presented nd pressure and the mete wind speed a ocessed and face: Logan here. The pr level; the vi orological da nd direction. analyzed in “ airport study ocessed data FIGURE deo data com ta comprise t These data Lateral atten ” (Fleming, were used in 29. Noise m prise video imestamped are presented uation of air 2002), only t the developm onitors 1, 2 recorded in temperature, in Appendix craft sound l he raw data ent of SAE- , and 3 deplo the field to relative hum B. Althoug evels over an for the 45 s AIR-5662 (S yed at BOS. determine ai idity, atmosp h, these data acoustically elected even AE, 2006). 5-7 rcraft heric were hard ts are

FIG FIGURE 30 URE 31. A . Acoustic d coustic data ata from 33 from the fin 9 departure al 45 depart events at BO ure events a S. t BOS. 5-8

5.2.3. Po T (ANOMS www.bksv aspx. Th weather d T monitors, monitorin the same history da included i format ou noted in A 5.2.4. Ph A locations As part of length of There we measurem the West. rtland Inter he airport ha ). Addit .com/Produc e system co ata with aircr he data from and weather g stations 10 format noted ta from those n both data s tlined in App ppendix E. iladelphia s part of a shown in Fig the measure the measurem re over 4,0 ent period. Using the da national Air s installed B ional infor ts/Environm mbines data aft tracking a PDX includ . The spec 1, 103, 104, 1 in Appendix monitors we ets. The tim endix E (DC FIGURE 32 Internation Part 150 upd ure 33. Data ments aircraf ents. The n 00 matched Monitor P4 w ta from mon port (PDX) rüel & Kjæ mation as entManagem from 10 no nd flight plan es tracking i ific data pro 07, and 109 D (LHR, 2 re collected e history dat A). The airc . Noise Mo al Airport ate at PHL was acquired t tracking da oise monitor aircraft ope as located a itor P2 as a r r’s Airport N to the entSolutions/ ise monitori data. nformation, vided has co shown in Fig 015). One-t from 8 to 15 a from the n raft tracking nitors Numb (PHL) (Wyle, 2009 at the sites ta was acquir s were set to rations with cross the rive eference, the oise Monito ANOMS AirportEnvir ng terminals Lmax, SEL, a rrelated trac ure 32. The hird octave b July 2016. W oise monitor data are redu ered 101-11 ), noise mo shown Novem ed together w record one-s noise even r and South lateral attenu ring and Ma capability onmentMana (NMTs) co nd duration o k and noise ANOMS que and sound p eather data s are reduced ced to the RA 0 Around PD nitoring was ber 10 to N ith hourly w econd, A-we ts at monito of the flight ation may be nagement Sy is available gement/ANO llecting nois f the noise data at the ry returned d ressure level from the airp to the zwee T file forma X conducted a ovember 17, eather data f ighted levels r P4 durin tracks depart defined usin 5-9 stem at MS. e and at the noise ata in time ort is k file t also t the 2007. or the (Leq). g the ing to g the

measurem E. The n weather d T humidity, the docum ents at P4. T oise data hav ata are contai he meteorolo wind speed ents with the he tracking d e been redu ned in pdf fil gical param and direction weather for FIGU ata have bee ced to the zw es. An exam eters needed , and atmosp each hour of RE 33. Nois n reduced to eek file form ple of a daily by an aco heric pressu each day of t e measurem the RAT file at also desc weather file ustic study re. These pa he measurem ent location format desc ribed in Ap is shown in F are the tem rameters are ents. s at PHL. ribed in App pendix E. H igure 34. perature, re clearly labe 5-10 endix ourly lative led in

5.2.5. Ga In data from from mon path and monitors Stansted A T maintaine reflect red at the tim outside te T each mea sound exp slant dista twick Airpo 2001 and 2 takeoff opera itors located between 450 1 and 5 in F irports” (Wh he terrain su d a straight fl uced (cutbac e. It was also mperature 5-3 he measurem surement eve osure level nce) at the t FIGUR rt (LGW) 002, the UK tions at Lon to the west o m and 1500 igure 35, as ite, 2012). rrounding th ight path unt k) engine po noted that t 0 degrees C ent data con nt the acou (SEL), the p ime of data c E 34. Weath ’s Civil Avia don’s Gatwic f the single m m to the sid documented ese monitors il well beyon wer from tak he data were , and outside sisted of aco stic data com osition data apture; and er file from tion Authori k airport duri ain runway. e of the west in “Noise M is relatively d the monito eoff climb se filtered for m relative humi ustical data, prise timest comprise air the meteorolo measuremen ty (CAA) pr ng the years The monito erly flight tr onitor Positi flat, with fe rs. It was no ttings, but th et condition dity 45-90% position data amped A-we craft position gical data w ts at PHL. ovided Volp 1996 to 1999 rs were locat ack of Runw ons at Heath w obstacles; ted that some is was not c s with wind . , and meteor ighted maxi (elevation a ere used to e with monit . These data ed under the ay 26L, simi row, Gatwic aircraft gen of these data onsidered an at less than 1 ological data mum level ( ngle, altitud determine the 5-11 oring were flight lar to k and erally may issue 0 kts, . For Lmax), e and data

met the af data were 5.2.6. Oa O Managem noise and capability ww.bksv. ANOMS. T across the 30. Figur study gro included a the monit ANOMS current st identifyin water. Th histories a included i atmospher orementione used in the d FIGU kland Inter akland Inter ent System ( weather data com/Products aspx. he data from water from r e 36 shows th und-to-groun s the GRU w ors, and we query for the udy, the GR g lateral atte e ANOMS re reduced t n an Excel f ic pressure a d meteorolog evelopment o RE 35. Mea national Air national Airp ANOMS) tha with aircraft can /Environmen OAK includ unway 30, as e locations o d noise prop ill provide a ather. The same perio U data prov nuation when query returne o the zweek ile. It contai t five-minute ical criteria. f SAE-AIR- surement sit port (OAK) ort has inst t combines tracking and tManagemen es spectral so well as a gro f the NMTs agation over dequate sour spectral time d of time ide ided the sp compared w d data in the file format d ns the tempe intervals. These data a 5662 (SAE, 2 es similar to alled Brüel data from no flight plan d be tSolutions/A und pressure und runup u and GRU. T mixed-impe ce event tim history data ntified times ectral data n ith spectral same forma efined in Ap rature, relati re presented 006). those used & Kjær’s A ise monitorin ata. Addition irportEnviron level time h nit (GRU) de he focus of t dance surfac ing, Lmax, SE was collec that aircraft ear the airc data at the N t noted in A pendix E. W ve humidity, in Appendix in the Gatwi irport Nois g terminals al informatio found mentManag istory data fr ployed near t his data colle es; thus, trac L, and durat ted during A departed ru raft to use a MTs on the ppendix D. eather data wind speed C. The proc ck Study. e Monitoring (NMTs) colle n on the AN ement/ om a set of N he start of ru ction effort w king data wa ion of the no ugust 2016. nway 30. Fo s a referenc other side o The spectral from the airp and direction 5-12 essed and cting OMS at MTs nway as to s not ise at An r the e for f the time ort is , and

5.2.7. Lo L (ANOMS and weath data item. and the no this projec A model mi shown ap However, complexit endeavors T 63,093 ev monitors ndon Heath ondon’s Hea ) produced b er. Appendi Using the g ise monitor t from the U note must be xed-impedan peared ideal due to the y of the terra that address he data conta ents at the for arrivals on FIGURE row Airport throw airpo y Brüel & K x C contains eometric var coordinates, K’s CAA do made on the ce surfaces. for this proje fact that the in is outside t undulating te ined for LH noise monito runway 09L 36. Locatio (LHR) rt has the jær that corre an example iables for the the location not have the applicability Originally, th ct due to the bodies of w he current sc rrain. R spans the rs for depar . The data fil ns of Noise Airport Noi lates events of a database point of clos of the aircraf curved flight of this data e monitors d bodies of w ater are elev ope of this pr period from tures off run e format is p Monitors Ar se Monitori at the noise query togeth est approach t can be dete tracks shown set for valida eployed to th ater on eith ated reservoi oject. The d 4 July to 25 way 27L an resented in A ound OAK. ng and Ma monitors wit er with nom (PCA) to ea rmined. The in Figure 37 tion of the a e South of th er side of the rs as shown ata is include September, d 35,781 ev ppendix D. nagement Sy h flight opera enclature for ch of the mo data provide . lgorithms fou e departure t line of mon in Figure 3 d herein for f 2015 and con ents at the 5-13 stem tions each nitors d for nd to racks itors. 8, the uture tains noise

FIGURE 3 FIGURE 38 7. Layout o . Street view f noise moni overl between re tors (numbe aid departur servoirs whe red circles) e tracks. re noise mo west of LHR nitors are de Airport wit ployed at LH 5-14 h R.

5-15 5.2.8. San Francisco International Airport (SFO) The airport has installed Brüel & Kjær’s Airport Noise Monitoring and Management System (ANOMS) that combines data from noise monitoring terminals (NMTs) collecting noise and weather data with aircraft tracking and flight plan data. More information on the ANOMS capability can be found at www.bksv.com/Products/EnvironmentManagementSolutions/AirportEnvironment Management/ANOMS.aspx. Spectral sound pressure level time histories from the monitors shown in Figure 39 have been obtained for August 2016 along with the ANOMS query for the same time. Details of the reduced data for the spectral time histories can be found in Appendix E under the zweek file format. The data fields in the ANOMS query are similar to those described in Appendix D.

FIGURE 39. SFO and surrounding noise monitoring terminals. 5-16

5-17 5.3. Other Data There are other data that can be used to validate the methods and algorithms that result from the sensitivity studies. Measurements in the literature, some of which has been shown or mentioned above, can be used to test the variance of the method with measurements. Some large measurement programs whose data sets are available for use in the project are listed below. 5.3.1. NASA Lateral Attenuation Studies Numerous studies were completed by NASA to measure the lateral attenuation of a variety of aircraft (Wilshire, 1979, 1980, 1981, 1987; Wat, 1999). The data is presented in both raw and analyzed form. The study by Wilshire at Wallops Island, VA in 1978 used a T-38A as the noise source with propagation along a strip of cement and industrial grass. This measurement program attempted to isolate source directivity by presenting the one-third octave band spectrum of sound that was emitted from a single point in the aircraft trajectory and propagated to each of the microphones on the different surfaces. The spectra are listed in tables in the papers accompanying the data archive. The position of the aircraft for the time of emission is described in relative coordinates as are the microphone locations. A detailed sounding of the atmosphere is provided by the raising and lowering of a balloon containing a meteorological station. 5.3.2 ACOU 104 Impedance Measurements A project by Sohlman et al. (2004) measured the impedance of 44 different surfaces using the NORDTEST method (NORDTEST, 1999) that is similar to the ANSI method (ANSI/ASA S1.18, 2010). The data led to the findings in Attenborough et al.’s paper (2011). It may be useful for determining the effects on the A-weighted metrics when using different ground impedance models. The distance between source and receivers is the same for one of the geometries in the ANSI method (1.75 m). 5.4. Data Summary The data described in this report were collected in order to test the algorithms to estimate lateral attenuation of aircraft noise. While some of this data is not appropriate to the immediate task at hand, by collecting it the authors hope to enable future endeavors with an ample database to test against. The data archive can be found on the optical media accompanying this report.