Investigations of environmental effects on freeway acoustics

              Investigations of Environmental
Effects on Freeway Acoustics
Final Report 605(1)
Prepared by:
H.J.S. Fernando
Center for Environmental Fluid Dynamics,
Arizona State University,
Tempe, AZ 85287-9809
N.C. Ovenden
Department of Mathematics,
University College London,
Gower Street, London WC1E 6BT, U.K.
S.R. Shaffer
Center for Environmental Fluid Dynamics,
Arizona State University,
Tempe, AZ85287-9809
March 2010
Prepared for:
Arizona Department of Transportation
in cooperation with
U.S. Department of Transportation
Federal Highway Administration
The contents of this report reflect the views of the authors who are responsible for the
facts and the accuracy of the data presented herein. The contents do not necessarily
reflect the official views or policies of the Arizona Department of Transportation or the
Federal Highway Administration. This report does not constitute a standard,
specification, or regulation. Trade or manufacturers’ names which may appear herein are
cited only because they are considered essential to the objectives of the report. The U.S.
Government and the State of Arizona do not endorse products or manufacturers.
Research Center reports are available on the Arizona Department of Transportation’s
internet site.
Technical Report Documentation Page
1. Report No.
FHWA-AZ-10-605(1)
2. Government Accession No. 3. Recipient's Catalog No.
4. Title and Subtitle
Investigations of Environmental Effects on Freeway Acoustics
5. Report Date
March 2010
6. Performing Organization Code
7. Author
H.J.S. Fernando, N. Ovenden, S. Shaffer
8. Performing Organization Report No.
9. Performing Organization Name and Address
Arizona State University
Office for Research & Sponsored Projects Administration
P.O. Box 873503
Tempe, AZ 85287-3503
10. Work Unit No.
11. Contract or Grant No.
SPR-PL-1(69)ITEM 605
12. Sponsoring Agency Name and Address
Arizona Department of Transportation
206 S. 17th Avenue
Phoenix, Arizona 85007
13.Type of Report & Period Covered
Final Report
05/04/2006 to 06/02/2008
14. Sponsoring Agency Code
15. Supplementary Notes
Prepared in cooperation with the U.S. Department of Transportation, Federal Highway Administration
16. Abstract
The study reported here was designed to examine the impact of background meteorological conditions on the
propagation of noise from urban freeways in the Phoenix area. The aim was to understand and predict how
sound waves emanating from highways respond to the vertical profiles of atmospheric temperature gradients
and velocity shear, so that sound measurements can be interpreted with regard to the environmental variability.
Over the course of four days in late 2006 and two days in early 2007, field experiments were carried out at two
freeway sites, where meteorological data and sound levels were measured and recorded from early morning
until the middle of the day. Such periods span the stable, morning transitional and convective periods of the
atmosphere. From the data collected, three test cases of varying atmospheric density stratification and wind
shear are presented and discussed. These cases represent all measurement periods and were analyzed in
detail.
A parabolic equation model coupled to a Green’s function model close to the source field was developed and
used to compute the refracted sound field for experimental cases up to half a mile from the freeway, permitting
computations of noise exposure of residential areas nearby. The model demonstrates that atmospheric effects
are able to raise sound levels by 10dB–20dB at significant distances from the highway, which at times led to
exceeding acceptable limits imposed by Federal Highway Administration for residential areas. Mitigation
strategies such as barriers and asphalt rubber friction course (ARFC) are also briefly discussed.
17. Key Words
Noise propagation, meteorological effects,
acoustic modeling, field studies, noise exposure,
mitigation strategies, freeway noise
18. Distribution Statement
Document is available to the U.S.
public through the National
Technical Information Service,
Springfield, Virginia, 22161
23. Registrant's Seal
19. Security Classification
Unclassified
20. Security Classification
Unclassified
21. No. of Pages 22. Price
50
SI* (MODERN METRIC) CONVERSION FACTORS
APPROXIMATE CONVERSIONS TO SI UNITS APPROXIMATE CONVERSIONS FROM SI UNITS
Symbol When You Know Multiply By To Find Symbol Symbol When You Know Multiply By To Find Symbol
LENGTH LENGTH
in inches 25.4 millimeters mm mm millimeters 0.039 inches in
ft feet 0.305 meters m m meters 3.28 feet ft
yd yards 0.914 meters m m meters 1.09 yards yd
mi miles 1.61 kilometers km km kilometers 0.621 miles mi
AREA AREA
in2 square inches 645.2 square millimeters mm2 mm2 Square millimeters 0.0016 square inches in2
ft2 square feet 0.093 square meters m2 m2 Square meters 10.764 square feet ft2
yd2 square yards 0.836 square meters m2 m2 Square meters 1.195 square yards yd2
ac acres 0.405 hectares ha ha hectares 2.47 acres ac
mi2 square miles 2.59 square kilometers km2 km2 Square kilometers 0.386 square miles mi2
VOLUME VOLUME
fl oz fluid ounces 29.57 milliliters mL mL milliliters 0.034 fluid ounces fl oz
gal gallons 3.785 liters L L liters 0.264 gallons gal
ft3 cubic feet 0.028 cubic meters m3 m3 Cubic meters 35.315 cubic feet ft3
yd3 cubic yards 0.765 cubic meters m3 m3 Cubic meters 1.308 cubic yards yd3
NOTE: Volumes greater than 1000L shall be shown in m3.
MASS MASS
oz ounces 28.35 grams g g grams 0.035 ounces oz
lb pounds 0.454 kilograms kg kg kilograms 2.205 pounds lb
T short tons (2000lb) 0.907 megagrams
(or “metric ton”)
mg
(or “t”)
mg megagrams
(or “metric ton”)
1.102 short tons (2000lb) T
TEMPERATURE (exact) TEMPERATURE (exact)
ºF Fahrenheit
temperature
5(F-32)/9
or (F-32)/1.8
Celsius temperature ºC ºC Celsius temperature 1.8C + 32 Fahrenheit
temperature
ºF
ILLUMINATION ILLUMINATION
fc foot candles 10.76 lux lx lx lux 0.0929 foot-candles fc
fl foot-Lamberts 3.426 candela/m2 cd/m2 cd/m2 candela/m2 0.2919 foot-Lamberts fl
FORCE AND PRESSURE OR STRESS FORCE AND PRESSURE OR STRESS
lbf poundforce 4.45 newtons N N newtons 0.225 poundforce lbf
lbf/in2 poundforce per
square inch
6.89 kilopascals kPa kPa kilopascals 0.145 poundforce per
square inch
lbf/in2
SI is the symbol for the International System of Units. Appropriate rounding should be made to comply with Section 4 of ASTM E380
TABLE OF CONTENTS
EXECUTIVE SUMMARY .................................................................................................1
I. INTRODUCTION ...........................................................................................................3
II. EXPERIMENTS ............................................................................................................7
III. MODELLING .............................................................................................................15
IV. CHOSEN TEST CASES AND MODELLING PARAMETERS ..............................19
Case A: Nov 7th 2006 (Rt 202) 11am .........................................................................20
Case B: Nov 7th 2006 (Rt 202) 8am ...........................................................................21
Case C: Nov 8th 2006 (Rt 202) 8am ...........................................................................22
V. ANALYSIS OF TRAFFIC SPECTRA TAKEN BY NOISE METERS .....................23
VI. CONSTRUCTION OF LEQ PLOTS ...........................................................................27
VII. CONCLUSIONS/FURTHER COMMENTS ............................................................37
REFERENCES ..................................................................................................................41
BIBLIOGRAPHY ..............................................................................................................44
LIST OF FIGURES
FIGURE 1. MAP OF SITES 3D AND 3E ON THE METROPOLITAN
PHOENIX FREEWAY SYSTEM ...............................................................8
FIGURE 2. LOCATION MAPS .........................................................................................9
FIGURE 3. LOCATION CROSS SECTIONS AND EXPERIMENTAL SETUP 202 ...10
FIGURE 4. LOCATION CROSS SECTIONS AND EXPERIMENTAL SETUP 101 ...11
FIGURE 5. INSTRUMENT IMAGES: SONICS, SODAR-RASS ..................................12
FIGURE 6. INSTRUMENT IMAGES: BALLOON-TETHERSONDE ..........................13
FIGURE 7. SCHEMATIC OF COMPUTATIONAL DOMAIN .....................................16
FIGURE 8. METEOROLOGICAL PROFILES: CASE A ..............................................20
FIGURE 9. METEOROLOGICAL PROFILES: CASE B ...............................................21
FIGURE 10. METEOROLOGICAL PROFILES: CASE C .............................................22
FIGURE 11. TIME DEPENDENCE OF LEQ DIFFERENCE 50 FT–100 FT .................23
FIGURE 12. MODELING SCHEMATIC: LINE SOURCE CONFIGURATION ..........24
FIGURE 13. MODELING SCHEMATIC: SOURCE HEIGHT CALCULATION ........25
FIGURE 14. CALCULATED SOURCE HEIGHT EXAMPLE ......................................25
FIGURE 15. CALCULATED SOURCE STRENGTH EXAMPLE ................................26
FIGURE 16. CASE A LEQ CONTOUR PLOT .................................................................28
FIGURE 17. CASE A LEQRANGE PLOT .......................................................................29
FIGURE 18. CASE A LEQRANGE VS FREQUENCY CONTOUR PLOT ....................30
FIGURE 19. CASE B LEQ CONTOUR PLOT .................................................................31
FIGURE 20. CASE B LEQRANGE PLOT .......................................................................32
FIGURE 21. CASE B LEQRANGE VS FREQUENCY CONTOUR PLOT ....................33
FIGURE 22. CASE C LEQ CONTOUR PLOT .................................................................34
FIGURE 23. CASE C LEQRANGE PLOT .......................................................................35
FIGURE 24. CASE C LEQRANGE VS FREQUENCY CONTOUR PLOT ....................36
FIGURE 25. EXAMPLE OF FLOW DISTORTION OVER BARRIERS ......................38
ACKNOWLEDGMENTS
We are extremely grateful to Arizona Department of Transportation (ADOT), Arizona
State University (ASU), and University College London (UCL) for their support of this
ongoing collaborative research. In particular, we thank Christ Dimitroplos, Fred Garcia,
and Lisa Anderson at ADOT for their interest and encouragement. The consultant
Illingworth & Rodkin’s assistance in the project in measuring and processing the sound
data is also gratefully acknowledged and we also thank Dragan Zajic, Leonard
Montenegro, and Adam Christman for their help with the field experiments and
subsequent data analysis.
1
EXECUTIVE SUMMARY
The study reported here was designed to examine the impact of background
meteorological conditions on the propagation of noise from urban freeways in the
Phoenix area. The aim was to understand and predict how sound waves emanating from
highways respond to the vertical profiles of atmospheric temperature gradients and
velocity shear, so that sound measurements can be interpreted with regard to the
environmental variability. Over the course of four days in late 2006 and two days in
early 2007, field experiments were carried out at two freeway sites, where meteorological
data and sound levels were measured and recorded from early morning until the middle
of the day. Such periods span the stable, morning transitional and convective periods of
the atmosphere. From the data collected, three test cases of varying atmospheric density
stratification and wind shear are presented and discussed. These cases represent all
measurement periods and were analyzed in detail.
A parabolic equation model coupled to a Green’s function model close to the source field
was developed and used to compute the refracted sound field for experimental cases up to
half a mile from the freeway, permitting computations of noise exposure of residential
areas nearby. The model demonstrates that atmospheric effects are able to raise sound
levels by 10dB–20dB at significant distances from the highway, which at times led to
exceeding acceptable limits imposed by the Federal Highway Administration (FHWA)
for residential areas. Mitigation strategies such as barriers and asphalt rubber friction
courses (ARFC) are also briefly discussed.
2
3
I. INTRODUCTION
Noise pollution is a serious and worsening environmental concern in urban areas. Not only
does it diminish the quality of human life,1,2,3 but it also alters wildlife habitats.4 Highway
traffic, airports, heavy industry, railways, and even leisure activities located close to built-up
areas all contribute to the noise menace, and thus urban planners and managers pay
close attention to mitigate it. This report concerns a study on a significant contributor to
noise pollution in urban areas — freeway traffic noise — which varies considerably.
The noise level depends upon a myriad of factors, such as ground conditions; terrain and
the presence of sound barriers; temporal variations in traffic speed, volume, and vehicle
types; and also spatiotemporal variation of meteorological variables such as temperature,
wind velocity, and turbulence.5,6 While some of these factors are accounted for in
operational sound prediction models, available operational models do not take all salient
factors into account.7,8 For example, the latest version of the Federal Highway
Administration's (FHWA) Traffic Noise Model (TNM)9 (Version 2.5 released in 2004)
does not account for the effects of temperature and wind variability: uniform, isothermal
atmospheric conditions are assumed in the calculations. The latter is a reasonable
assumption for shorter (less than 650 ft (198 m) distances from the sound source, but errors
can be substantial when predicting intermediate and far field noise. This drawback is of
particular importance when refraction of sound due to temperature and wind causes
anomalous intensity variations of sound at distance from the source. For example, noise
measurements and analysis conducted in Scottsdale, Arizona, following complaints by
residents living more than ¼ mile (about 400 m) from the eastern portion of Loop 101,
suggest that ground-level inversions (surface stable temperature stratification) can increase
the sound level by as much as 10 decibels to 15 decibels (dB).10
While the noise level under neutral atmospheric conditions is well within the Federal
Highway Administration (FHWA) noise abatement criterion (NAC), an inversion can
cause decibel levels to violate the standard. FHWA-NAC recommends implementing
abatement procedures such as noise walls or modified pavement types (quiet pavements)
when the energy averaged or equivalent sound level (Leq) approaches a value of 67 A-weighted
decibels (dBA). (A-weighting is used to account for typical human sensitivity to
various frequencies of absolute pressure fluctuations following ISO standard IEC 651
(1993-09) by applying a band-pass filter. Note that when referring to a difference in sound
pressure levels, dB are interchangeable with dBA.) Such levels can be observed at some
distance around certain Arizona freeways merely as a result of inversions and wind shear.
The influence of atmospheric factors becomes particularly critical when noise mitigation is
realized via a combination of techniques, for example, noise walls and quiet pavements.
The Arizona Department of Transportation (ADOT) has received approval from the
FHWA for the Quiet Pavement Pilot Program (QPPP) to investigate the usefulness of
pavement-surface type as a noise mitigation strategy, subject to the condition that Arizona
would be a pilot program with specific research objectives and requirements.40 This
research is intended to validate the efficacy of asphalt rubber friction courses (ARFC) as a
noise mitigation method. Over several years ADOT will overlay Portland cement concrete
4
pavement (PCCP) in metropolitan Phoenix with a 1-inch–thick ARFC surface. Where the
ARFC is placed and noise walls are required, the walls may be reduced in height in view
of the extra mitigation offered by ARFC surfacing.40
Sound barrier walls, also known as sound walls, are designed to protect the public and
particularly the nearby residents against noise pollution, which is adverse sound level
exposure that can lead to hearing loss, sleep disturbance, stress, and increased blood
pressure. To decrease noise pollution effects, regulations have been instituted by
governments. Such statutes in the United States include the U.S. National Environmental
Policy Act, the Federal-Aid Highway Act, and the Noise Control Act of 1972. These acts
have promoted decreased noise pollution, quantitative noise analyses, use of sound barrier
walls, and city noise planning. A well-designed noise wall diffracts and reflects sound
waves to optimize the attenuation of far field sound. To determine if sound barrier walls
are effective, the sound is allowed to pass through or over the wall and the transmitted
noise levels are gauged. When a 9dB reduction is reached, a sound barrier wall is
considered adequate. Nine decibels lower in sound is equal to approximately a 90 percent
drop in sound waves traveling past the wall.
Beginning in 2003, ADOT has been monitoring six sites across the Phoenix metropolitan
area for traffic-generated noise over a 10-year period to evaluate the effectiveness of
ARFC. While measurements show that ARFC has reduced freeway noise appreciably
(8dB–10dB) at close-in and community locations, sound refraction due to environmental
conditions can defeat the noise abatement approaches (e.g., the use of walls) at some
distances away. In such instances, noise walls would be of little help as a mitigation tool
and, as noise walls are expensive, the merits of their installation should be carefully
evaluated a priori.
ARFC pavements, sound walls, and environmental factors become dominant only at
certain intrinsic frequency ranges. The relationships between these variables and A-weighted
noise levels in the field thus are intricate and can be delineated only via models
that properly quantify fundamental relationships and their complex interactions. It is
therefore important to develop scientific knowledge and tools to predict atmospheric
effects on freeway noise that help evaluate alternative design options. Such tools will also
help with interpretation of measurements taken at different positions and/or times and in
placing results on a unified scientific basis (i.e., in terms of a certain base or standard
state). The only viable method for predicting sound in complex field situations is the use of
a numerical model that incorporates all governing factors, the straightforward (yet onerous)
method in this context being nesting of an acoustic model with an environmental
forecasting model. Such a modeling system is prohibitively computer intensive and so can
be invoked only under very special circumstances. A simpler method is to use available
representative atmospheric data from the area to feed the acoustic model, assuming local
smaller scale variations are unimportant. The research reported here is of this type and
includes a meteorological measurement component. This study’s aim was to examine how
different meteorological conditions, especially ground-based inversions, can affect freeway
noise under high-pressure, low-synoptic flow conditions prevalent in the desert
Southwest.11,12 The study was particularly motivated by the Quiet Pavement Pilot
5
Program, where the noise reduction capabilities of the rubber friction course are being
measured over a decade. To put results into a consistent framework, the meteorological
effects need to be included in presenting the noise results.
Although physical mechanisms underlying atmospheric sound propagation are well
understood, the lack of both sufficiently detailed atmospheric data and computer models
capable of incorporating them into an integrated formulation have hampered progress on
modeling the impact of environmental effects on sound propagation. A review of literature
suggests that:
for downwind propagation, the magnitude of sound fluctuations increases with the
frequency of the signal and with distance;
for upwind propagation, the fluctuations are greatest near the shadow boundary;
in a stable atmosphere (clear night, weak winds) the range of fluctuations is
typically about 5dB, mainly due to the gravity waves and turbulence, but sound
levels can be enhanced due to refraction at distances beyond ¼ mile (400 m);
in an unstable atmosphere (clear sunny day, strong winds) the range of fluctuation
is typically 15dB–20dB;
the spectrum of fluctuations measured over open ground encompasses a range of
frequencies that humans can hear from 50 Hz to above 3 kHz; and
sound propagation from hilltop to hilltop and from air to ground is frequently
characterized by large low-frequency fluctuations.
A suite of computational approaches is being used for atmospheric sound propagation
studies,5 which include:
Gaussian beam methods — this is a variant of the classical ray tracing technique by
solving the wave equation in the neighborhood of the conventional rays and associating a
Gaussian amplitude profile normal to each ray. An approximate overall solution can then
be constructed as a superposition of these so-called beams.
Fast Field Program Models (FFP) — a semi-analytical method involving computation of
the sound field in a horizontally layered homogeneous atmosphere in horizontal wave
number domain, which is then inverted to the spatial domain using an inverse Fourier
transform.
Parabolic Equation (PE) models — a marching solution based on splitting the governing
wave equations into left- and right-traveling components, originating at the source and
capable of being ―perturbed‖ en route to account for topography, barriers, and turbulence.
Ray theories, although robust for indoor acoustics, rapidly become highly cumbersome to
compute in downward refracting media where many rays are needed and caustics are
problematic. Additional complications, such as diffraction by obstacles, turbulence, and
prediction of acoustic shadow regions, further urge the use of alternative methods. The key
to PE models is the use of an effective sound speed based upon temperature and wind
speed of the actual mean flow field, both of which modify the isotropic adiabatic sound
speed.13,14
6
When assuming a line (or axisymmetric) source, the two-dimensional wave operator is
factored into left- and right-traveling components transverse to the source. The pressure
field due to a source can then be resolved in the domain by marching the solution
numerically away from the source, while discounting any waves that propagate towards the
source. Major disadvantages of this method are that it becomes inaccurate at high elevation
angles and cannot directly account for back scatter unless the more difficult task of
handling propagation in both directions is addressed. It has many advantages, however,
including the ease of incorporating atmospheric absorption and varying boundary
conditions and geometries, along with actual spatially varying meteorological profiles.
Extensions that incorporate turbulence and flow details, such as over a barrier or rough
terrain (e.g., large eddy simulations) have also begun to be incorporated into the scheme.
For these reasons, methodologies based on the PE equation prove highly popular.15-21
The FFPs typically have a faster run time than their PE counterparts and can handle
realistically complicated vertical atmospheric profiles. They can also account for the
vectorial nature of the mean flow without requiring an ―effective‖ sound speed and are
accurate at high elevation angles. However, the required Fourier transformation in the
horizontal direction means that the model is restricted to homogeneous ground surfaces,
with a flat topography containing at most a single and relatively simple topographical
feature.22-27 It is common to use hybrids—models that combine several methods—to
address aspects of the problem at hand in an attempt to circumvent potential drawbacks of
any individual method.5, 28-32
In order to understand and quantify the effects of atmospheric temperature and velocity
profiles on sound propagation, refraction, and diffraction, we have combined a field
measurement campaign with modeling efforts. The field measurements are to provide
realistic vertical profiles of temperature and cross wind velocities to the model and were
performed over six days at two freeway sites in Scottsdale, Arizona, and Mesa, Arizona,
where meteorological and sound data were taken and recorded over roughly a six-hour
period between 6am and 12pm. For the modeling, the sound data is entered into a Green’s
function model to evaluate the near source field generated from the freeway traffic. This
source field, along with the meteorological data, is then input into a parabolic equation
(PE) model to compute the refracted sound field out to a distance of 1968 ft (600 m). The
results are compared to neutral atmospheric conditions; the effect of stratification and wind
shear are separated and quantified in three 20-minute time-averaged cases selected from
the field data.
The outline of this report is as follows. The field experiments and equipment used are
described in detail in Section II. Section III briefly outlines the acoustic propagation
model. The selection of the three test cases to be entered into the model is presented in
Section IV. The procedure of using sound measurements to construct a near-source field
using a Green’s function model and calculated ground impedance is given in Section V.
The evolution of the noise frequency spectra with range and the construction of overall Leq
plots are then presented in Section VI. Finally, conclusions, recommendations to ADOT,
and plans for future work are described in Section VII.
7
II. EXPERIMENTS
To study the influence of meteorological conditions on noise propagation from Phoenix
highways, the Center for Environmental Fluid Dynamics at Arizona State University
(EFD-ASU) conducted a joint field campaign with ADOT and Illingworth & Rodkin, Inc.
The EFD-ASU team provided detailed measurements of atmospheric meteorological
conditions, while Illingworth & Rodkin, Inc. provided sound measurements. ADOT staff
videotaped the traffic and recorded its speed. Field measurements were taken at two
different sites along highways in the Phoenix metropolitan area. The sites are standard
designated sites for ADOT, and the details of these sites are outlined in Saurenman et al.10
The first series of measurements was taken on October 10 and 11, 2006, on the west side
of Loop 101 at milepost 47 (ADOT location site 3E). The second series was carried out
on November 7 and 8, 2006, on the north side of Loop 202 (ADOT location 3D). The
third series was on March 20 and 21, 2007, again at the Phoenix Loop 202 site. Figure 1
shows the location of the sites on the metropolitan Phoenix freeway system. Figure 2
shows maps of both locations with red dots indicating the approximate measurement
sites. Both sites have a relatively flat homogeneous terrain (see cross-sectional profiles in
Figures 3 and 4) with hard sandy soil and sparse bushes. However, away from the sites is
complex topography that may alter the meteorological variables. Measurements were
taken from 7am to 11am, and beginning at 6am for the two days in March, in order to
better understand how noise levels change with atmospheric conditions. The earliest time
for the start of the experiment was determined by the logistical constraints of the
contractor. The goal was to obtain data during periods of temperature inversion, typical
daytime adiabatic lapse conditions, and morning transition, covering representative
periods of the months concerned. It is interesting to note that the temperature conditions
near the surface were found to be unstable even in the early morning hours, and this is
believed to be due to turbulent mixing and heat retention of the freeway surface. Further
work is necessary to investigate such features.
A number of instruments were employed, which included three-dimensional sonic
anemometers, a meteorological balloon with tethersonde system, and a SODAR (SOund
Detection And Ranging) with RASS (Radio Acoustic Sounding System) attachment.
Sound measurement instruments were located at distances of 50 ft and 100 ft from the
center of the nearest lane of the highway at the 3E site and 50 ft, 100 ft, and 250 ft (15.24
m, 30.48 m, and 76.2 m) from the center of the nearest lane at the 3D site. The sonic
anemometers were located on towers at the same distance from the highway as the sound
measurement instruments. Tethersonde and SODAR/RASS systems were located slightly
further away to avoid contamination of sound-level measurements. Schematics of the
cross-sectional area of the sites are given in Figures 3 and 4. Figure 5 has photographs of
the instruments employed.
Figure 1. Map of Sites 3D and 3E on the Metro politan Phoenix Freeway System
8
9
Figure 2. Location maps for Loop 101 (top; location site 3E) and Loop 202 (bottom;
location site 3D). The red dot indicates the approximate location of the measurement
sites.
10
Figure 3.Cross section for Loop 202 location expressed as the elevation above sea level
(ASL) (in feet). Horizontal distance is shown measured in feet from the fence on the
north side. Positions of instruments are shown as squares for microphones, triangles and
stars for sonic anemometers in November (top) and March (bottom), respectively.
Arrows indicate horizontal distances of 50 ft, 100 ft, and 250 ft from the center of the
nearest travel lane on the westbound (WB) side.
11
Figure 4. Location cross section for Loop 101 expressed as the elevation above sea level.
Horizontal distance is shown measured in feet from the fence on the west side. Positions
of instruments are shown as squares for microphones and triangles for sonic anemometers
in the October 2006 field campaign. Arrows indicate horizontal distances of 50 ft and
100 ft from the center of the nearest travel lane on the southbound (SB) side.
12
Figure 5. Photographs of instruments deployed in the experiment. Two sonic
anemometers on a short tripod on right side, and two microphones on the tripod in the left
side of the picture, at Loop 202 in November 2006 (top); and the SODAR-RASS system
at Loop 101 in October (bottom).
13
Figure 6.Photograph of the balloon-tethersonde system used at the Loop 101 site in
October 2006.
The balloon and tethersonde system is an important tool in atmospheric boundary-layer
studies since it provides detailed profiles of wind speed, wind direction, temperature, air
pressure, and humidity in the lower atmosphere. During the two days of field experiments
in October, the balloon system shown in Figure 6 was deployed with a single tethersonde,
thus providing profiles of temperature, wind speed and direction, and atmospheric
pressure, and the data could be obtained up to 164 ft (50 m) above ground level (agl). The
allowable height of balloon flights was determined by the FAA air traffic permit.
However, due to problems with a malfunctioning sonde on the second day, the balloon-tethersonde
system was used only during these two days of measurements. Also, during
the first day, there was a period of stronger winds when the balloon was not used due to
safety reasons. Comparison of sonic anemometer data located on the tower with those
obtained by the balloon system show a good agreement, and hence the data from the
sonic anemometers were used for velocity calculations when the balloon system was
inoperative.
14
The sonic anemometers were operated at a frequency 10 Hz, providing all three velocity
components (one in each dimension Ux, Uy, Uz) and temperature. The large frequency of
measurements provided an opportunity to obtain information on mean flow and
temperature close to the surface, as well as properties of the turbulence. The latter was
used to calculate turbulence statistics, such as the root mean square for velocities and
temperature, as well as for turbulent momentum and heat fluxes. The sonic anemometer
placement is discussed below.
(i) October 10 and 11 — two sonic anemometers were located on a tripod
50 ft (15.24 m) and three on a meteorological tower 100 ft (30.48 m) from
the center line of the closest travel lane. The heights of instruments on the
tripod were 5.9 ft and 9.5 ft (1.8 m and 2.9 m) agl and the heights of those
on the tower were 6.6 ft, 13.2 ft and 19.7 ft (2 m, 4 m and 6 m) agl.
(ii) November 7 and 8 — an additional tripod was also located closest to the
highway 50 ft (15.24 m) where the heights of sonic anemometers were
5.9 ft and 9.5 ft (1.8 m and 2.9 m) agl, while sonic anemometers at the
tower were placed at levels 22.3 ft, 34.1 ft and 45.3 ft (6.8 m, 10.4 m and
13.8 m) agl. On November 8, one more sonic was placed on a tripod at a
location 250 ft (76.20 m) from the center of the near lane at 7.2 ft (2.2 m)
agl in order to measure atmospheric conditions close to the farthest sound
measurement point.
(iii) March 20 and 21 — two towers were set up at distances of 50 ft and 100 ft
(15.24 m and 30.48 m) from the center of the near lane. Two sonic
anemometers were positioned at heights of 10.8 ft and 21.6 ft (3.30 m and
6.60 m) agl on the tower closest to the roadway, while three sonic
anemometers at 12.4 ft, 25.1 ft and 34.1 ft (3.78 m, 7.65 m and 10.39 m)
agl were used on the farthest tower. In some cases, the sonic anemometers
did not work properly, and data from these periods were not included in
the data set and subsequent analysis.
The SODAR/RASS system was utilized to measure wind speed and temperature profiles
between roughly 65 ft to 1968 ft (20 m to 600 m) agl. This system was used in order to
provide more details on the structure of the atmospheric boundary layer at greater
heights, but for the present study the most important were data near the lowest 328 ft
(100 m) or so.
15
III. MODELING
Based on sound data from field experiments provided by Illingworth & Rodkin, a two-dimensional
model can be constructed on acoustic propagation from a single mono-frequency
coherent line source in a vertically layered atmosphere. A rectangular xy
coordinate system is used, with y measuring the vertical height and x measuring the
horizontal range from the center line of the near lane of the highway. All lengths are non-dimensionalized
on a typical source height L0, velocities are non-dimensionalized on the
sound speed measured at the ground level C0, density is non-dimensionalized on the
density of air at 1 atmosphere (ρ0=1.2 kg m-3) and pressure p is non-dimensionalized on
ρ0C0
2. For a given frequency f Hz, we define the Helmholtz number as ω=2π f L0/C0 and
by writing the acoustic pressure perturbation as p(x,y,t)=pc(x,y)e-i ω t, the Helmholtz
equation for a line source at x=x0 of strength S in a vertically layered atmosphere is
obtained as
( ).
~ ( )
~ ( )
~ ( ) 2 0
2
2
2
2
2
2
2
p S x x
y c y
c y p
x c y y
p
c
c c (1)
Here, is the non-dimensional effective sound speed, which includes the effects of both
temperature and crosswind. Given a measured vertical temperature profile T(y) and
crosswind speed profile U0(y), the effective sound speed is defined in a standard manner
to be
where γ is the ratio of specific heats and R is the ideal gas constant. The boundary
conditions imposed are a far-field Sommerfield radiation condition as 2 2 r x y
becomes large, of the form
(2)
and an impedance boundary condition at the surface
Throughout this report, the empirical impedance model of Delany and Bazley33 is used
where, for a ground surface with flow resistivity σ [Pa s m-2], the impedance Z is given by
(4)
16
Two models are used in tandem to compute the far-field sound propagation: (i) a near-field
Green's function method assuming a homogeneous atmosphere and (ii) a parabolic equation
approximation. Figure 7 shows the regions of the xy domain where each model is used.
Figure 7. A schematic of the coupled models used to resolve the far-field propagation of
traffic noise from a freeway corridor. The red dots represent monofrequency coherent
effective line sources positioned above the center of the nearest lane of traffic.
The near-field Green's function method34 is used to obtain the acoustic field in the
vicinity of the line source where the refractive effects of atmospheric factors can be
assumed to be negligible. In other words, the Green's function method assumes a constant
effective sound speed c%=1 and solves equations (1) to (3) with this assumption up to the
edge of the highway, at 22 ft (6.7 m), obtaining the sound field
(5)
where H0
(1) is the zeroth order Hankel function of the first kind, and the term PZ(x,y;y0)
represents the correction to the hard-wall solution for finite z. This correction is derived
by Chandler-Wilde and Hothersall34 and is given in terms of
17
erfc e a
Z
e
s e g s ds
Z
e
i
i a
s
i
/ 4
2
(1 )
0
1/ 2
2 1
( / )
with the result
( , ; ) 0 P x y y Z
where
g(t)=
The first integral expression is calculated using Gauss-Laguerre Quadrature and the
second surface wave term (due to its strong exponential decay away from the ground) is
evaluated using the formula given in Attenborough.35Assume over the near-field
calculation, the ground impedance is typically of porous asphalt with σ = 3x107 Pa s m-2,
which is given in Table 4.9 of Attenborough.6
The near-field Green's function model provides an acoustic field at the edge of the
freeway pini(y)=pc(xedge,y), which is subsequently used as an initial condition for a two-dimensional
Cartesian variant of the standard axisymmetric parabolic equation (PE)
model, first derived by Gilbert and White.13
The PE model used is the parabolic wide-angle approximation of (1) assuming a two-dimensional
line source. The pressure field is rewritten as pc(x,y)=ψ(x,y) eiωx and ψ(x,y) is
obtained by solving the equation
y x c x
c
x c y
~ 1
~ 1
~
1
4
1
2
2
2
2
2 2
~ 1
~ 1
~
1
2 2
2 2
2 y c
c
c y
i . (6)
2 1 / 2 2 1/ 2
2
0 2
0
2
0 2
1 1 1
( ) / ( )
( ) ,
Z
Z
a
y y x y y
x y y
1
1/ 2 2 1 2
/ 4
2 1/ 2
(1 )
( 2 ) 2 (1 / ) ( )
2(1 ) ( )
i
Z it
t i t i Z t Z
e a
Z t ia
18
The equation (6) and the impedance boundary condition (3) are finite-differenced and the
solution is obtained by marching forward in the x direction. Sandy soil is taken to be the
ground surface type beyond the freeway with σ = 4x105 Pa s m-2 and we assume the
ground is completely flat to concentrate strictly on atmospheric effects in this study.
The radiation condition (2) is dealt with numerically by a buffer zone14,36,37 occupying
approximately the upper one third of the grid domain, yatt< y <ymax, where the effective
sound speed in (6) is replaced by
1
3
max
( ) ~( ) 1
att
att
y y
y y
c y c y iA
Here, A is a real parameter that can be optimized for each frequency component. To
ensure the effectiveness of the buffer zone, the initial pressure profile obtained from the
near-field Green's function method pini(y) must also be smoothly reduced to zero within
the buffer zone to prevent spurious reflections from the truncated top of the grid domain.
Thus,
edge
att
att
edge ini i x
y y
B y y
x y p
2
max
2
2
( , ) exp
where1 ≤ B ≤ 4 is another optimized parameter dependent on frequency.
Effects of atmospheric absorption are incorporated following the method outlined in
Salomons5 §B.5 by applying a constant attenuation rate in dB m-1 to each frequency band
at 1 m agl before summing to form the Leq versus range plots. This method follows the
International Standard ISO 9613-1:1993(E). In doing so, it was necessary to approximate
a value for the relative humidity, which was only measured with the balloon-tethersonde
system during the October measurements. We used a value of 20 percent relative
humidity, 20˚C, and atmospheric pressure of 101.325 kPa.
19
IV. CHOSEN TEST CASES AND MODELLING PARAMETERS
The field experiments yielded large amounts of meteorological and sound data. Three
representative cases were selected for sound transmission modeling based on the velocity
profiles and stratification. Temperature and crosswind profiles above 131 ft (40 m) were
obtained from the SODAR/RASS measurements in 32.8 ft (10 m) increments, whereas
data at lower altitudes were gleaned from the sonic anemometer readings (which were
located only at fixed locations). The meteorological profiles are time-averaged over a
period of 20 minutes. To obtain the surface-layer velocity profile for an unstable
convective boundary layer ( <200 ft or 60 m), theoretical curves of the Monin-Obukhov
(MO) similarity theory are then fitted to the sonic data. The MO theory suggests that
near the ground both vertical temperature and velocity gradients have the form
for . (7)
where A and B are parameters fitted to the data.38 Since diverges like 4 / 3 y as y 0,
the chosen temperature profile is made linear near the ground so that T(y)~A y+B and the
velocity takes instead a standard logarithmic form, U0(y)~A log(z/z*), where z* is the
aerodynamic roughness length (which is acceptable below a distance of MO length scale,
as the dominant term therein is shear generated turbulence). Above approximately 200 ft
(60 m), the fitted curve smoothly transitions into the SODAR-RASS data. If the useful
range of data from the SODAR/RASS is less than 984 ft (300 m), the theoretical curve is
held constant at the last entry from the SODAR-RASS. Measurements and theoretical
profiles for the three chosen cases are shown in Figures 8 to 10.
20
Case A: Nov 7, 2006 (Loop 202) 10:40 to 11:00am
This case has wind shear at very high altitudes, but with very little temperature
stratification. Plots of experimental and theoretical profiles for temperature and crosswind
velocity are shown in Figure 8. Note that in Case A, the SODAR-RASS data was usable
up to 820 ft (250 m), compared to 656 ft (200 m) for other cases.
Figure 8. Case A: Temperature and crosswind (to the freeway) data with fitted theoretical
profiles. Open circles are data from the sonic anemometers 50 ft from the center of the near
lane, open squares are from anemometers 100 ft from the near lane, solid circles are from
the SODAR-RASS, and the solid line is the theoretical curve entered into our model.
21
Case B: Nov 7, 2006 (Loop 202) 7:40 to 8:00am
This case is stratified with shear flow. Plots of experimental and theoretical profiles for
temperature and crosswind velocity are shown in Figure 9.
Figure 9. Case B: Temperature and crosswind (to the freeway) data with fitted theoretical
profiles. Open circles are data from the sonic anemometers 50 ft from the center of the
near lane, open squares are anemometers 100 ft from the near lane, solid circles are from
the SODAR-RASS, and the solid line is the theoretical curve entered into our model.
22
Case C: Nov 8, 2006 (Loop 202) 7:40 to 8:00am
This case is strongly stratified with a sharp change in temperature at approximately 394 ft
(120 m) above the ground and a crosswind jet at approximately 164 ft (50 m) above the
ground. Plots of experimental and theoretical profiles for temperature and crosswind
velocity are shown in Figure 10.
Figure 10. Case C: Temperature and crosswind (to the freeway) data with fitted
theoretical profiles. Open circles are data from the sonic anemometers 50 ft from the
center of the near lane, open squares are anemometers 100 ft from the near lane, solid
circles are from the SODAR-RASS, and the solid line is the theoretical curve entered into
our model.
23
V. ANALYSIS OF TRAFFIC SPECTRA TAKEN BY NOISE METERS
The overall acoustic source field we are attempting to replicate consists of a six-lane highway
(three lanes in each direction) with multiple moving sound sources that vary according to
their speed, the traffic density, and the vehicular type. Without knowledge of the exact
acoustic signature of every car and truck, a number of severe but unavoidable assumptions
needed to be made about the nature of the sound sources. We emphasize here that the focus
of this paper is on the meteorological aspect of noise transmission from freeways, as opposed
to understanding the composition of sound sources emitted, and the results are expected to
give useful information on the effects of temperature stratification and wind shear on the
noise propagation.
The sound data consists of five-minute time-averaged one-third octave data from three
sound meters placed close to the highway. We do not have information on sound
generated from separate lanes of traffic and the frequency output of different vehicle
types traveling at different speeds. However, Figure 11 shows the difference between the
five-minute averaged dBA level taken from the sound meter located 50 ft (15.2 m) away
from the center of the nearest travel lane and 5 ft (1.5 m) above the ground and that
located 100 ft (30.5 m) away from the center of the nearest travel lanes and 5 ft (1.5 m)
above the ground on one of the nearest travel lane and 5 ft (1.5 m) above the ground and
that located 100 ft (30.5 m) away from the center of the nearest travel lane and 5 ft (1.5
m) above the ground on one particular day of field experiments. This clearly shows a
geometric attenuation of 3dB as the distance from the source doubles, providing some
justification to the assumption that the freeway can be treated as series of line sources.
Figure 11. The difference in overall A-weighted sound level measured between the sound
meter located 50 ft (15.2 m) from the center of the freeway’s nearest lane at a height of 5 ft
(1.5 m) and the sound meter located 100 ft (30.5 m) from the near lane at a height of 5 ft (1.5
m). The triangles merely display an indication of the traffic conditions at the time (either free
flowing or slow moving). A decrease of 3dB with a doubling of distance corresponds to
what is expected for a line source as Pline~1/r in a neutral atmosphere.
24
We assume that over the ranges in question (up to 1965 ft (600 m)), the sound field in a
neutral atmosphere (i.e., constant temperature with zero mean flow) continues to evolve in a
two-dimensional fashion, with a geometric attenuation of 3 dB as the distance doubles from
the freeway. In our model, the traffic noise is approximated as a series of monofrequency
coherent line sources positioned vertically above the nearest travel lane of the freeway (see
Figure 12).
Figure 12. A schematic showing sound rays emanating from a series of example mono-frequency
coherent sound sources positioned at the center of the near lane of the freeway.
The strength and effective height of these virtual sources are unknowns that must be deter-mined
from the one-third octave data obtained from the three (or four) sound meters. As the
sound meters are positioned relatively close to the source, the influence of meteorological
conditions is regarded as negligible over the range up to the farthest sound meter and a neu-tral
atmosphere is therefore assumed in the near field. This enables the unknown line source
parameters to be determined by using the Green's function model for acoustic propagation
from a line source above an impedance plane as detailed in Section III. As mentioned before,
the impedances chosen are taken from Attenborough6 as σ = 3x107 Pa s m-2 for the asphalt
and σ = 4x105 Pa s m-2 for the sandy soil. Due to a lack of exact knowledge of the highway
and surrounding surface topography, the surface is assumed to be asphalt out to a range of
50 ft from the virtual line sources with sandy soil beyond, as shown in Figure 13. Repeating
the calculation for other impedances suggests that neither the representation of asphalt as a
hard wall (Z = ∞) nor varying sandy soil impedance between 2x105to 6x105 Pa s m-2 change
the results significantly.
For a given one-third octave interval, the height of a representative line source can be
calculated by replicating the differences between the dBA values recorded by the three sound
meters. This is done by varying the source height to minimize a norm based on the sum of
the absolute errors between the differences obtained by the Green's function model and the
recorded differences. The range of possible source heights is restricted to less than 4 m and
the lowest height where the error norm attains a local minimum and takes an absolute value
less than 2dB–3dB is selected. The source heights used for Cases A through C are shown in
Figure 14. Note that there is very good agreement on the source heights obtained in each case
for the higher frequencies (using data taken on different days at different times). Larger
discrepancies for the lower frequencies can be explained as the dBA difference errors do not
vary that much with height due to the large wavelengths. This also means that the accuracy
50ft
100ft
12ft
5ft
600Hz
63Hz
1kH
z
2 kHz
400Hz
25
Figure 13. Schematic demonstrating the calculation of virtual source heights for each one-third
octave frequency component. Shown are two example source heights (blue and red)
along with the direct and once-reflected rays that superimpose upon reaching the sound
meter. Ground impedance boundary conditions change not only amplitude, but also the
phase of the reflected ray (see equation 5).
of the source height for lower frequencies is less crucial as it does not significantly alter the
sound field. Perhaps the most problematic difficulty in selecting source height for 315Hz–
400Hz range, where the norm error at zero height is unacceptably high (possibly 9dB-10dB)
but the error norm approaches zero again at source heights of 11.5 ft–14.8 ft (3.5 m–4.5 m);
thus heights of this order are chosen.
Figure 14. Source heights for Loop 202 cases obtained by minimizing an error norm
based on dB differences between sound meters for Case A (circles), Case B (diamonds),
and Case C (crosses).
26
Following the determination of source heights, it is relatively straightforward to use the
Green's function near-field model to obtain the A-weighted source strengths, and these are
given for Cases A through C in Figure 15. Note the good agreement in the source strength
profile across the frequency ranges 63Hz–2.5kHz for the three cases. The sound signature is
almost identical for Cases B and C, both taken at the same time during rush hour on
consecutive days, whereas Case A has lower sound levels particularly in the 100Hz–200Hz
and 800Hz–2kHz band, possibly due to the lower traffic levels occurring in the late morning.
Figure 15. Source strengths for Loop 202 cases obtained by minimizing an error norm
based on dB differences between the sound meters for Case A (circles), Case B
(diamonds), and Case C (crosses).
27
VI. CONSTRUCTION OF LEQ PLOTS
In each chosen case, the model is run for each frequency component, based on the central
frequency of the one-third octave band, with and without the influence of meteorological
effects for comparison. For efficiency, the frequency range of the computation is reduced
from spanning the entire range of frequency bands from 25Hz–20kHz to only include
those bands between 63Hz and 2.5kHz (17 components in all). Such a restriction
produces an error of less than 0.2 percent in terms of the final overall sound pressure
level when compared to the actual values measured by the sound meters.
The spatial A-weighted sound pressure level distribution for each frequency component
is resolved by the PE model on a grid of size and spacing dependent on the wavelength
(based on a usual 10 grid points per wavelength). These results are subsequently
interpolated onto a grid of 3.28 ft (1 m) spacing with a range of 0 ft–1965 ft (0 m–600 m)
horizontally and 0 ft–984 ft (0 m–300 m) vertically. Then at each grid point the A-weighted
frequency contributions LA(fn) (x,y) are combined to produce the overall Leq
sound pressure level by the formula5
with
[Hz]
Results of the spatial sound pressure levels are presented in Figures 16 to 18 for Case A,
Figures 19 to 21 for Case B, and Figures 22 to 24 for Case C. Each result presents the
equivalent spatial sound field obtained in a neutral atmosphere directly above the
resolved spatial sound field when the temperature and crosswind velocity effects are
included. Note that the downwind side of the freeway is always shown and the vertical
range displayed is only up to 65.6 ft (20 m) agl. It is clear from these figures that the
overall impact of the meteorological effects is significant in all three cases examined.
Indeed, significantly higher noise levels are predicted downwind near ground level for all
cases. For guidance, FHWA’s noise abatement criteria threshold of 67dBA is shown as a
thick contour line on the spatial contour plots of Leq and by a gray area on the sound
pressure level range plots at 3.28 ft (1 m) above the ground. Below, each case is
examined in more detail.
The meteorological effects are weakest for Case A, with very little temperature
stratification and a crosswind on the order of 6.6 ft s-1 (2 m s-1) persisting from about
17
1
( ) /10
10 10log 10
n
L f
eq
L A n
28
98 ft (30 m) to around 492 ft (150 m) in altitude. However, Figure 16 clearly shows how
the crosswind shear flow present up to 98 ft (30 m) above the surface focuses sound into
a thin layer of around 7 ft to 16 ft (2 m to 5 m) in height, where the sound intensity is
raised by roughly 15dB.
Figure 16. Case A: A-weighted sound pressure level contours without meteorological
effects (top) and with meteorological effects (bottom). Each contour line represents a
change of 3dBA.
29
As a result, the sound level close to the ground does not fall below 67dBA until a
horizontal distance of 1637 ft (500 m) from the freeway is reached, as opposed to
approximately 328 ft (100 m) predicted for a neutral atmosphere (see Figure 17).
Figure 17. Case A: Overall A-weighted sound pressure level and frequency components
at 1 m above the ground without meteorological effects (top) and with meteorological
effects (bottom). The gray marks the area over 67dB level. Only a few frequency bands
are shown for clarity.
30
An examination of Figure 18 of the impact of the meteorological effects on individual
frequency components reveals that the frequency band 1kHz–2kHz remains the most intense
out to the far field.
Figure 18. Case A: Contours of A-weighted sound pressure level for each frequency
component for neutral (top) and meteorological (bottom). Each contour line represents a
change of 3dBA.
Case B occurred during the rush-hour traffic on Loop 202 with a stronger wind shear
resulting in speeds of approximately 19fts-1 (6ms-1) at approximately 200 ft (60 m) agl. More
severe temperature gradients are observed, with the temperature falling 5oC with increasing
altitude before rising back to its ground level value at an altitude of approximately 330 ft
31
(100 m). The competition between the near-ground negative temperature gradient and
positive wind shear means that overall near-ground sound levels fall in a similar fashion as in
neutral atmospheric conditions over the first 656 ft (200 m) from the freeway. However, the
refractive effects due to wind shear and the evolution to a temperature inversion at higher
altitudes leads to sound rays being refracted back towards the ground from above and sound
focusing at around 1637 ft (500 m) from the freeway. Indeed, Figures 19 and 20 indicate that
the A-weighted sound pressure level starts to exceed the 67dBA threshold close to the ground
at a range of 1637 ft (500 m), before continuing to exceed 67dBA beyond the calculation
domain.
Figure 19. Case B: A-weighted sound pressure level contours without meteorological effects
(top) and with meteorological effects (bottom). Each contour line represents a change of 3dBA.
32
Figure 20. Case B: Overall A-weighted sound pressure level and frequency components at 1
m above the ground without meteorological effects (top) and with meteorological effects
(bottom). The gray marks the area over 67dBA level.
Similar to the previous case, the frequency range 1kHz–1.25kHz is particularly influenced
and focused most intensely by a combination of wind shear and temperature gradients
(Figure 21), although all frequency ranges appear to be subjected to some degree of focusing
33
at the 1637 ft (500 m) range. The spatial contours in Figure 19 strongly suggest that this case
might be a typical example of excessive sound levels occurring far from the freeway, which
are unlikely to be abated by the use of a sound barrier.
Figure 21. Case B: Contours of A-weighted sound pressure level for each frequency
component. Each contour line represents a change of 3dBA.
Case C is also taken during rush-hour traffic and has the most severely changing meteoro-logical
profiles, being strongly stratified and having a crosswind jet peaking at 4ms-1 at a
height of 50 m above the ground. Figures 22 to 25 show a concentration of sound rays and
34
pockets of constructive and destructive interference between the rays in a roughly 4m-wide
layer close to the ground, particularly beyond the 300 m range. As a result, the combined
effect of wind shear with a mild negative temperature gradient close to the ground leads to
the near-ground sound pressure level persisting in excess of the 67dBA threshold for up to
nearly 600 m (approximately 1/3 mile) away from the freeway. The dominant frequencies
responsible once again appear to be 1.25 kHz and 1.6 kHz with other neighboring
frequencies also being strongly influenced by the meteorological conditions. In addition,
sound in the frequency range 125Hz–160Hz appears to be focused to lesser extent in the
500 m–600 m (1639–1967 ft) range.
Figure 22. Case C: A-weighted sound pressure level contour without meteorological
effects (top) and with meteorological effects (bottom). Each contour line represents a
change of 3dBA.
35
Figure 23. Case C: Overall A-weighted sound pressure level and frequency components
at 1 m above the ground without meteorological effects (top) and with meteorological
effects (bottom). The gray marks the area over 67dB level.
36
Figure 24. Case C: Contours of A-weighted sound pressure level for each frequency
component. Each contour line represents a change of 3dBA.
37
VII. CONCLUSIONS/ FURTHER COMMENTS
This work represents an initial combined experimental and theoretical study into the impact
of meteorological conditions on the propagation of traffic noise from a freeway corridor. The
principal conclusions of the study are as follows:
(i) Traffic noise models used to assess environmental noise impacts on nearby
communities must incorporate expected meteorological conditions that occur in that
location. Noise measurements taken under different meteorological conditions cannot
be directly compared to obtain information on the source, unless corrections are
made for differing meteorology. This is particular important for the case of ARFC
concerned in this study, in that the reduction of the effectiveness of the pavement in
use is deduced via measurements made over certain periods of different years.
Without corrections for the effects of meteorology, the validity of such assessments
is questionable and corrections based on near field data are recommended to
transform the data to some standard conditions.
(ii) The combined Green’s function and PE model developed as a part of this study has
shown its capabilities in taking meteorological data and near-field sound
measurements to generate a spatial map of the predicted noise levels. The model
also enables analyses of individual frequency components (e.g., as in Figures 17 and
18 for Case A), and the model results show that the frequency range 1kHz–2kHz is
the most significantly influenced by meteorological conditions and thus provide the
principal contribution to far-field traffic noise levels; this result, however, awaits
experimental confirmation. If such evidence arises, mitigation strategies targeting
this frequency band would be the most effective in preventing excessive noise levels
at large distances from the freeway corridor.
(iii) This work represents the initiation of further collaborative investigations between
ADOT and EFD-ASU that are focused on assessing the effectiveness of noise
mitigation strategies, in particular comparing the efficacy of different noise barriers
(e.g., absorptive barrier surfaces, land-use control, traffic-management measures,
vegetative barrier, ARFC). Future experiments and development of the theoretical
models will require (a) assessing the effects of turbulence, (b) improved field trials
with more sound meters to give further spatial sound information and meteorological
data, and (c) an examination of terrain effects in the surrounding areas, which is
needed for better accuracy of freeway acoustic work. More accurate knowledge of
the sound field and types of sources is also required.
(iv) Another important issue with physical noise barriers is the flow field distortion
surrounding them. This may greatly influence source characteristics for far-field
noise, in addition to direct influence on near-field characteristics. Inclusion of such
effects needs to be done using complex high-resolution flow models that account for
fine-scale features. These models can be nested in a meso-scale weather prediction
model, as was demonstrated by Fernando et al.12 A model of the genre has been
developed by the EFD group,39 and combining the new acoustic model developed
here with such a nested model will be the most sophisticated methodology that can
be proposed for future highway acoustic evaluation studies. Figure 25 shows an
38
example of a calculation conducted to illustrate the effects of nearby buildings and a
small topographic feature (butte) on an approaching flow field for neutral conditions.
Under convective or stable conditions, the flow field is expected to be more complex
as demonstrated in the figure; for example, it has flow separation, reattachment, and
recirculation regions as well intense turbulence production in these highly sheared
areas. These models, however, will be computer intensive and hence collaboration
with a high-performance computing cluster is recommended. If noise barriers are
present on either side of the freeway, then back scattering of acoustic waves needs to
be taken into account.
6 0
4 0
2 0
0
1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0
6 0
4 0
2 0
0
1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0
6 0
4 0
2 0
0
1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0
6 0
4 0
2 0
0
1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0
Figure 25. The distortion of flow by a set of barriers, as computed by a high-resolution
computational fluid dynamics code that uses Large-Eddy Simulation method. Such a
model can be nested with a meteorological model on a high-performance computational
platform to assess acoustic behavior in built-up areas. Different building heights are
illustrated. Similarly, a noise wall can be included in the simulation.
39
(v) It is expected that ADOT will use the model developed or a further improved
version thereof for a direct comparison of the effectiveness of barriers and
ARFC in the near future. As mentioned, accounting for meteorological
conditions is imperative in further field trials.
(vi) Future field experiments should include far-field monitoring as well as
collecting more high-resolution meteorological data. The former will help
evaluation of the model developed in this study for far-field noise predictions,
and the latter will help to determine the resolution to which the meteorological
data must be collected for making reasonable practical predictions. The present
investigation was limited to near-field sound, as well as general meteorological
conditions surrounding a freeway without any noise walls, which is an
acceptable first step.
40
41
REFERENCES
1. Goines, L. and L. Hagler. ―Noise pollution: A modern plague.‖ Southern Medical
Journal 100(3): 287-294, March 2007.
2. Tapia Granados, J. A. ―Reducing automobile traffic: An urgent policy for health
promotion.‖ Revista Panamericana de Salud Publica 3(4):227-241, 1998.
3. Ward, L. M. and P. Suedfeld. ―Human response to highway noise.‖ Environmental
Research 6(3): 306-326, 1973.
4. Forman, R. T. T. ―Estimate of the area affected ecologically by the road system in the
United States.‖ Conservation Biology 14(1): 31-35, 2000.
5. Salomons, E. M. Computational Atmospheric Acoustics. Kluwer Academic
Publishers, 2001.
6. Attenborough, K., K.M. Li, and K. Horoshenkov. Predicting Outdoor Sound. Taylor
and Francis, 2007.
7. Steele, C. ―A critical review of some traffic noise prediction models.‖ Applied
Acoustics 62(3): 271-287, 2001.
8. Lihoreau, B., B. Gauvreau, M. Berengier, P. Blanc-Benon, and I. Calmet. ―Outdoor
sound propagation modeling in realistic environments: Application of coupled
parabolic and atmospheric models.‖ The Journal of the Acoustical Society of America
120: 110-119, 2006.
9. Wayson, R. L., J. M. MacDonald, R. D. Eaglin, and B. Wendling. ―Simulation
approach to transport noise modeling.‖ Transportation Research Record 1601: 64-70,
1995.
10. Chambers, J., H. Saurenman, R. Bronsdon, L. Sutherland, K. Gilbert, R. Waxler, and
C. Talmadge.―Effects of temperature induced inversion conditions on suburban
highway noise levels.‖ ActaAcustica united with Acustica 92(6): 1060-1070, 2006.
11. Wang, J. X. L. and J. K. Angell. Air Stagnation Climatology for the United States
(1948–1998). NOAA/Air Resources Laboratory ATLAS No. 1. United States.
National Oceanic and Atmospheric Administration, 1999.
12. Fernando, H. J. S., S. M. Lee, J. Anderson, M. Princevac, E. Pardyjak, and S.
Grossman-Clarke. ―Urban fluid mechanics: Air circulation and contaminant
dispersion in cities.‖ Journal of Environmental Fluid Mechanics 1(1): 107-164, 2001.
13. Gilbert, K. E. and M. J. White. ―Application of the parabolic equation to sound
propagation in a refracting atmosphere.‖ The Journal of the Acoustical Society of
America 85: 630-637, 1989.
14. West, M., K. Gilbert, and R. Sack. ―A tutorial on the parabolic equation (PE) model
used for long range sound propagation in the atmosphere.‖ Applied Acoustics 37: 31-
49, 1992.
42
15. Gilbert, K. and X. Di. ―A fast Green’s function method for one-way sound
propagation in the atmosphere.‖ The Journal of the Acoustical Society of America 94:
2343-2352, 1993.
16. Di, X. and K. Gilbert. ―The effect of turbulence and irregular terrain on outdoor
sound propagation.‖ in Proceedings of the 6th International Symposium on Long
Range Sound Propagation, ed. by David I. Havelock, 315-333. Canada National
Research Council, 1994.
17. Sack, R. and M. West. ―A parabolic equation for sound propagation in two
dimensions over any smooth terrain profile: The generalized terrain parabolic
equation (gt-pe).‖ Applied Acoustics 45: 113-129, 1995.
18. West, M. and Y. Lam. ―Prediction of sound fields in the presence of terrain features
which produce a range dependent meteorology using the generalized terrain parabolic
equation (gt-pe) model.‖InterNoise 2000 Proceedings, ed. by Didier Cassereau, 2:
943, Paris: Societe Francaise d’Acousticque, 2000.
19. Wilson, D.K. ―A turbulence spectral model for sound propagation in the atmosphere
that incorporates shear and buoyancy forcing.‖ The Journal of the Acoustical Society
of America 108: 2021-2038, 2000.
20. Wilson, D.K., J.G. Brasseur, and K.E. Gilbert. ―Acoustic scattering and the spectrum
of atmospheric turbulence‖ The Journal of the Acoustical Society of America 105: 30-
34, 1999.
21. Chevret, P., P. Blanc-Benon, and D. Juve. ―A numerical model for sound propagation
through a turbulent atmosphere near the ground.‖ The Journal of the Acoustical
Society of America 100: 3587-3599, 1996.
22. Lee, S., N. Bong, W. Richards, and R. Raspet. ―Impedance formulation of the fast
field program for acoustic wave propagation in the atmosphere.‖ The Journal of the
Acoustical Society of America 79: 628-634, 1986.
23. Franke, S. and G. W. J. Swenson. ―Brief tutorial on the fast field program (ffp) as
applied to sound propagation in the air.‖ Applied Acoustics 27: 203-215, 1989.
24. West, M., R. Sack, and F. Walkden. ―The fast field program (ffp) a second tutorial:
Application to long range sound propagation in the atmosphere.‖ Applied Acoustics
33: 199-228, 1991.
25. Wilson, D.K. ―Sound field computations in a stratified, moving medium.‖ The
Journal of the Acoustical Society of America 94: 400-407, 1993.
26. Nijs, L. and C. Wapenaar. ―The influence of wind and temperature gradients on
sound propagation, calculated with the two-way wave equation.‖ The Journal of the
Acoustical Society of America 87: 1987-1998, 1990.
27. Nijs, L. and C. Wapenaar. ―Reply to: Comments on the influence of wind and
temperature gradients on sound propagation calculated with the two-way wave
equation.‖ The Journal of the Acoustical Society of America 91: 498-500, 1992.
28. Ostashev, V. Acoustics in a Moving Inhomogeneous Media. Spon Press, 1997.
43
29. Ostashev, V., V. Mellert, R. Wandelt, and F. Gerdes. ―Propagation of sound in a
turbulent medium: I. – Plane waves.‖ The Journal of the Acoustical Society of
America 102: 2561-2570, 1997.
30. Ostashev, V., F. Gerdes, V. Mellert, and R. Wandelt. ―Propagation of sound in a
turbulent medium: II. – Spherical waves.‖ The Journal of the Acoustical Society of
America 102: 2571-2578, 1997.
31. Ostashev, V., D. Wilson, L. Liu, D. Aldridge, N. Symons, and D. Marlin. ―Equations
for finite-difference, time-domain simulation of sound propagation in moving
inhomogeneous media and numerical implementation.‖ Journal of the Acoustical
Society of America 117(2): 503-517, 2005.
32. deRoo, F. and I. Noordhoek. ―Harmonoisewp 2-reference sound propagation
model.‖Fortschritte Der Akustik 29: 354-355, 2003.
33. Delany, M. E. and E. N. Bazley. ―Acoustical properties of fibrous absorbent
materials.‖ Applied Acoustics 3: 105-116, 1970.
34. Chandler-Wilde, S. and D. Hothersall. ―Efficient calculation of the Green’s function
for acoustic propagation above a homogeneous impedance plane.‖ Journal of Sound
and Vibration 180: 705-724, 1995.
35. Attenborough, K. ―Sound propagation close to the ground,‖ Annual Review of Fluid
Mechanics 34: 51-82, 2002.
36. Salomons, E. M. ―Improved Green’s function parabolic equation method for
atmospheric sound propagation.‖ The Journal of the Acoustical Society of America
104: 100-111, 1998.
37. Robertson, J. S., W. L. Seigmann, and M. J. Jacobson. ―Low-frequency sound
propagation modeling over a locally reacting boundary with the parabolic
approximation.‖ Journal of the Acoustic Society of America 98: 1130-1137, 1995.
38. Stull, R. B. An Introduction to Boundary Layer Meteorology. Kluwer Academic
Publishers, 1988.
39. Baik, J. J., J. J. Kim, and H. J. S. Fernando. ―A CFD model for simulating urban flow
and dispersion.‖ Journal of Applied Meteorology 42(11): 1636-1648, 2003.
40. Arizona Department of Transportation. Progress Report no. 1, Arizona Department of
Transportation (ADOT) Quiet Pavement Pilot Program (QPPP). Arizona Dept. of
Transportation, 23 December 2004.
44
BIBLIOGRAPHY
1. Harmonoise http://www.imagine-project.org/
2. Counihan, J., J. C. R. Hunt, and P. S. Jackson. ―Wakes behind two-dimensional
surface obstacles in turbulent boundary layers.‖ Journal of Fluid Mechanics 64(3):
529-563, 1974.
3. Saurenman, H., J. Chambers, L. C. Sutherland, R. L. Bronsdon, and H. Forschner.
Atmospheric Effects Associated with Highway Noise Propagation. Final Report 555.
Arizona Dept. of Transportation, 2005.
4. Ovenden, N., S. Shaffer, and H. J. S. Fernando. ―Impact of meteorological conditions
on noise propagation from freeway corridors.‖ Journal of the Acoustical Society of
America 126: 25-35, July 2009.
5. Shaffer, S. Investigations of Environmental Effects on Freeway Acoustics. M.S.
Thesis, Arizona State University, May 2009.