EPA 550/9-82-410
NATIONAL AMBIENT NOISE SURVEY
JANUARY 1982
U.S. Environmental Protection Agency
Office of Noise Abatement and Control
Washington, D.C. 20460
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
National Ambient Noise Survey
5. REPORT DATE
January 1982
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Mark M. Hansen
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME ANO AOORESS
U.S. Environmental Protection Agency
Office of Noise Abatement and Control (ANR 471)
401 M Street, NW.
Washington, D.C. 20007
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME ANO AOORESS
U.S. Environmental Protection Agency
401 M Street, NW.
Washington, D.C. 20460
13. TYPE OF REPORT ANO PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The objectives, methodology, and results of a national survey of outdoor noise
environments in urban residential areas are discussed. The objectives were to
determine overall noise levels, source contributions, and patterns of spatial and
temporal variation in these areas, along with the effect of three locational
factors on these parameters. The survey employed a randomized site selection
procedure, a stratified sampling strategy, and a multifaceted measurement protocol
to meet these objectives. Results of the survey include a simple model which
predicts Ldn in these areas, projections of nationwide noise impact, average
source contributions and temporal noise level histories and average variations
in noise level at different locations around residential units.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lOENTIFIERS/OPEN ENDED TERMS c. COSATI Field/Group
18. DISTRIBUTION STATEMENT
Release to public.
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
167
20. SECURITY CLASS (This page)
22. PRICE
EPA Form 2220-1 (9-73)
-------
-ACKNOWLEDGEMENTS-
The National Ambient Noise Survey employed the talents and efforts of many .
individuals. Lou Sutherland and Daryl May of Wyle Laboratories developed
the survey methodology, and should be commended for having the knowledge and
wisdom to successfully incorporate the many objectives of the study into one
set of site selection and measurement protocols. Special thanks should be given
to the field engineers, Charles Kamerman, Ernie Croughwell, Howard Schechter,
and Dick Blumenthal. Their work took them from San Diego, California to
Portland, Maine. Guided only by their field knowledge and a random numbers
table, they collected valuable data in a wide range of neighborhoods and at all
hours of the day and night.
None of this work would have been performed without the untiring support
of Casey Caccavari of the EPA Noise Office State and Local Programs Division.
His patience, and his attention, with so many other matters competing for them,
are greatly appreciated.
Finally, tribute must be paid to the 124 households which permitted total
strangers to conduct noise measurements around their homes. Their cooperation
in this unique program reaffirms the good will and civic mindedness of the
American people. One must hope that this work will in the long run reward
that spirit with a more pollution free environment.
iii
-------
TABLE OF CONTENTS
Paragraph Page
TITLE PAGE i
ABSTRACT ii
ACKNOWLEDGEMENTS ill
TABLE OF CONTENTS iv
LIST OF FIGURES viii
LIST OF TABLES x
EXECUTIVE SUMMARY ES-1
CHAPTER 1. INTRODUCTION
1-1 BACKGROUND AND PURPOSE 1-1
1-2 ORGANIZATION 1-1
1-3 USING THE REPORT 1-2
CHAPTER 2. SURVEY RATIONALE
2-1 INTRODUCTION 2-1
2-2 ASSESSING NOISE EXPOSURE 2-1
2-3 SOURCE CONTRIBUTIONS 2-3
2-4 TYPOLOGY AND FACTOR ANALYSIS 2-4
2-5 SURVEY EVALUATION . 2-5
CHAPTER 3. METHODOLOGY
3-1 BASIC APPROACH 3-1
, 3-2 CATEGORIES OF RESIDENTIAL UNITS 3-1
3-2-1 Urban Area Size 3-2
3-2-2 Population Density 3-2
3-2-3 Traffic Impact 3-4
3-2-4 Aircraft Impact 3-4
3-2-5 Sampling Cells 3-6
iv
-------
TABLE OF CONTENTS - Continued
Paragraph
3-3
3-4
3-4-1
3-4-2
3-4-3
3-5
3-5-1
3-5-2
4-1
4-2
4-3
4-4
4-4-1
4-4-2
4-4-3
4-5
4-6
4-7
4-8
4-8-1
4-8-2
4-9
-
SITE SELECTION
MEASUREMENT PROTOCOL
Elements of the Protocol
Continuous Measurements
Microsamples
ANALYTICAL PROCEDURES
Data Reduction
Data Analysis
CHAPTER 4. RESULTS
INTRODUCTION
DISTRIBUTION OF URBAN POPULATION OVER LDN
VARIATION IN NOISE LEVELS BY SIDE OF
RESIDENTIAL UNIT
DAILY, HOURLY, AND INSTANTANEOUS VARIATION IN
NOISE LEVELS
Daily Variation
Hourly Variation
Instantaneous Variation
FREQUENCY OF PREDOMINANCE AND AVERAGE NOISE
LEVELS OF SOURCES
VARIATION OF SOURCE CONTRIBUTIONS BY SIDE
OF RESIDENTIAL UNIT
TEMPORAL VARIATION IN SOURCE CONTRIBUTIONS
EFFECT OF URBAN AREA SIZE, POPULATION DENSITY,
AND TRAFFIC IMPACT
Effect on Ldn
Effect on Source Contributions
SUMMARY OF RESULTS
Page
3-9
3-11
3-11
3-11
3-11
3-12
3-12
3-12
4-1
4-1
4-4
4-5
4-5
4-5
4-7
4-9
4-13
4-14
4-27
4-27
4-30
4-49
-------
TABLE OF CONTENTS - Continued
Paragraph - Page
APPENDIX A - METHODOLOGY SUPPLEMENT
A-l INTRODUCTION A-l
A-2 SITE SELECTION A-l
A-2-1 Establishing Site Quotas A-2
A-2-2 Selecting Urbanized Areas and Urban Zones A-4
A-2-3 Selecting Census Tracts and Blocks A-6
A-2-4 Selecting Residential Units A-10
A-3 MEASUREMENT PROTOCOL A-10
A-3-1 Scheduling . A-10
A-3-2 Continuous Monitoring A-ll
A-3-3 Microsamples 'A-13
A-4 ANALYTICAL PROCEDURES A-13
A-4-1 Continuous Monitoring A-13
A-4-2 Microsample Data A-15
A-4-3 Other Data A-15
APPENDIX B - FIELD EXPERIENCE
B-l INTRODUCTION " B-l
B-2 SITE SELECTION B-l
B-3 MEASUREMENT PROTOCOL B-5
B-4 LOGISTICS AND SCHEDULING B-8
B-5 CONCLUSIONS B-9
vi
-------
TABLE OF CONTENTS - Continued
Paragraph Page
APPENDIX C - 24-HOUR CONTINUOUS MEASUREMENT
AND ANALYSIS PROCEDURES
C-l INSTRUMENTATION C-l
C-2 ENCODING PROCEDURE . . . C-2
C-3 DECODING PROCEDURE C-3
C-4 OUTPUT C-4
C-5 MULTIPLE-DAY SITES C-4
C-6 EXTRANEOUS FACTORS , C-5
C-7 EDITING AND CORRECTION PROCESSING -
DIGITAL DATA CASSETTES C-6
APPENDIX D - RESULTS SUPPLEMENT
D-l INTRODUCTION - . D-l
D-2 QUALITY OF DATA OBTAINED D-l
D-2-1 - Continuous Measurements D-l
D-2-2 30-Minute Microsamples D-l
D-3 ANALYTICAL PROCEDURES USED D-3
D-3-1 Distribution of Population over Ldn D-3
D-3-2 Variation'by Side D-5
D-3-3 Temporal Variation D-7
D-3-4 Noise Sources . D-9
D-3-5 Factor Analysis D-10
APPENDIX E - ESTIMATION OF CELL POPULATIONS E-l
APPENDIX F - THE SITES F-l
vii
-------
LIST OF FIGURES
Figure Page
ES-1A Source Contribution Profile, High-Traffic-Impact Areas ES-5
ES-1B Source Contribution Profile, Low-Traffic-Impact Areas ES-6
3-1 Ldn Noise Contours Around a Typical Airport 3-5
4-1 Distribution of Urban Population over Ldn 4-2
4-2 Average Hourly Leq Values, by Traffic Impact 4-6
4-3 Average Hourly Statistical Levels, by Traffic Impact 4-8
4-4 Projected Hourly Decreases in Leq Resulting From
Achievement of Hypothetical Goal for Noise Control Program 4-10
4-5A Source Contribution Profile, High-Traffic-Impact Areas 4-11
4-5B Source Contribution Profile, Low-Traffic-Impact Areas 4-12
4-6A Variations in Automobile Contributions, by Side of Unit 4-15
4-6B Variations in Truck Contributions, by Side of Unit 4-16
4-6C Variations in Motorcycle Contributions, by Side of Unit 4-17
4-6D Variations in Construction Contributions, by Side of Unit 4-18
4-6E Variations in Dog Contributions, by Side of Unit 4-19
4-6F Variations in Bird Contributions, by Side of Unit 4-20
4-7A Variations in Automobile Contributions, by Time of Day 4-21
4-7B Variations in Truck Contributions, by Time of Day 4-22
/
4-7C Variations in Motorcycle Contributions, by Time of Day 4-23
4-7D Variations in Construction Contributions, by Time of Day 4-24
4-7E Variations in Dog Contributions, by Time of Day 4-25
4-7F Variations in Bird Contributions, by Time of Day 4-26
4-8A Source Contribution Profile - Large, High-Density,
High-Traffic-Impact Areas 4-32
4-8B Source Contribution Profile - Large, Medium-High-Density,
High-Traffic-Impact Areas 4-33
viii
-------
LIST OF FIGURES - Continued
Figure Page
4-8C Source Contribution Profile - Medium-Small, High-Density,
High-Traffic-Impact Areas " 4-34
4-8D Source Contribution Profile - Medium-Small, Medium-High-
, Density, High-Traffic-Impact Areas 4-35
4-8E Source Contribution Profile - Medium-Small, Medium-Low-
Density, High-Traffic-Impact Areas 4-36
4-8F Source Contribution Profile - Medium-Small, Low-Density,
High-Traffic-Impact Areas 4-37
4-8C Source Contribution Profile - Large, High-Density,
Low-Traffic-Impact Areas 4-38
4-8H Source Contribution Profile - Large, Medium-High-Density,
Low-Traffic-Impact Areas 4-39
4-81 Source Contribution Profile - Medium-Small, High-Density,
Low-Traffic-Impact Areas 4-40
4-8J Source Contribution Profile -.Medium-Small, Medium-High-
Density, Low-Traffic-Impact Areas . 4-41
4-8K Source Contribution Profile - Medium-Small, Medium-Low-
Density, Low-Traffic-Impact Areas 4-42
4-8L Source Contribution Profile - Medium-Small, Low-Density,
Low-Traffic-Impact Areas 4-43
4-9 Frequency of Identification for Selected Sources -
High- and Low-Traffic-Impact Areas 4-45
4-10 Regression Lines - Leq (id) vs. Log 10 (UZ density) -
For Selected Sources, High- and Low-Traffic-Impact
Areas ' 4-46
A-l Typical Page of Block Statistics A-7
A-2 Typical Census Map A-8
A-3 Measurement Locations Used in the Survey A-12
A-4 Data Sheet Used in Microsamples A-14
E-l Method For Estimating Cell Populations Within Specified
Distance of Roadway Type E-2
IX
-------
LIST OF TABLES
Table Page
ES-1 Summary of Categories Used to Define Sampling Cells ES-2
ES-2 Populations Exposed to Critical Values of Ldn ES-3
ES-3 Differentials in Noise Levels by Side ES-3
3-1 Table From U.S. Summary, Table 20 3-3
3-2 Summary of Categories Used to Define Sampling Cells 3-7
3-3 Populations (Millions) of Sampling Cells Used in
National Ambient Noise Survey (1980) 3-8
3-4 " Site Quotas and Sites Obtained for Each Sampling Cell 3-10
3-5 Source Codes Used in Microsamples 3-13
4-1 Populations Exposed to Critical Values of Ldn . 4-3
4-2 Differentials in Noise Levels by Side 4-4
4-3 Average Ldn by Day of Week - 4-5
4-4 Summary Statistics of Ldn Measurements,
by Sampling Cell 4-28
4-5 Cumulative Populations Distributions over Ldn
by Traffic Impact and Urban Zone Density 4-31
4-6 Source Classes 4-47
4-7 Summary of Regression Results - Source Class Contributions 4-48
A-l Site Quotas and Sites Obtained For Each Sampling Cell A-3
D-l Average Differences in Noise Levels Between Microsample
and Continuous Monitoring Results D-6
D-2 Five-day Site Ldn Values D-8
E-l Distribution of Population by Place Size (Index J) and
Population Density Category (Index ID) E-4
E-2 Roadway Mileages E-5
E-3 Fraction of Roadway Mileages Which Run Through
Occupied Land E-l2
-------
LIST OF TABLES - Continued
Table Page
E-4 Roadway Mileages Through Occupied Land E-13
E-5 Clear Zone Distances (in Feet) by Roadway Type (K),
Population Density Category (ID), and Population
Place Size (J) E-14
E-6 Population (In Thousands) High Traffic Impact by
Roadway Type E-15
E-7 High-Traffic-Impact Population (Millions) by Population
Place Size and Population Density Category E-16
E-8 Population (Millions) and Percentages by Urban Area
Size/Urban Zone Density Category E-16
E-9 Apportionment of NRTNEM Categories to Survey Categories E-17
E-10 Distribution of Urban Population by Urban Area Size,
Urban Zone Population Density, and Traffic Impact E-18
E-ll Population Growth Factors by Place Size (Index J)
For Every Five Years in Time Stream E-19
E-12 Population (Millions) of Sampling Cells Used In National
Ambient Noise Survey (1980) E-20
xi
-------
EXECUTIVE SUMMARY
During 1980 and 1981, the Environmental Protection Agency (EPA) sponsored
a survey of noise environments in urban residential areas. The purpose of the
survey was to generate a statistically valid profile of noise levels and
source contributions. This profile is intended to assist in the evaluation
>
of the need for and effectiveness of noise control measures directed toward
the urban residential environment.
The basic approach of the survey was to perform outdoor noise level
measurements and source identifications at randomly selected residential units
located in urban areas across the United States. The measurements were used
to assess overall noise levels, source contributions, and temporal and posi-
tional variation in these quantities. The residential units were selected by
means of a stratified sampling approach, with the stratifications based oh
urban area population, population density, and proximity to major roadways
(traffic impact). The cell structure is summarized in table ES-1. Residential
units highly impacted by aircraft noise were excluded from consideration.
The results show that 87 percent of the urban population are exposed to
a day-night sound level (Ldn) over 55 dB, the 'threshold of impact for residen-
tial noise based on EPA criteria. The percent exposed to higher levels of
Ldn is given in table ES-2.
Noise levels are usually higher at the front of residential units, with
this tendency most pronounced in areas close to major roadways. Average dif-
ferences in noise levels between the front, rear, and sides of the house are
given in table ES-3.
Daily variation in Ldn is approximately 2 dB. Noise levels are not sig-
nificantly different on weekend days, nor is there any other consistent pat-
tern of daily variation.
ES-1
-------
Table ES-1. Summary of Categories Used to Define Sampling Cells
Parameter
Urban Area
Size
Population
Dens ity
Basis for Defining
. Categories
Urbanized area
population (1970)
Urban zone population
density (1970)
Number of
Categories
2
4
Description of
Categories
Large - _> 2,000,000
Medium/ small - All
others
High - _> 4.500/
square mile
Traffic
Impact
Aircraft
Impact
Distance from major
roadways
Ldn contours around
airports
Medium-high - 3,000-
4,500
Medium-low - 1,500-
3,000
Low - <1,500
High - Within 100
feet of .an arterial
or 300 feet of an
interstate or
freeway
Low - All others
High - Within
Ldn = 65 dB contour
Low - All others
ES-2
-------
Table ES-2. Populations Exposed to Critical Values of Ldn
(Percent of Urban Population)
Ldn > 55 dB
Ldn > 60 dB 53%
Ldn > 65 dB 17%
Ldn > 70 dB 2%
Ldn > 75 dB <1%
Table ES-3. Differentials in Noise Levels by Side
Traffic
Impact Leq (Front)-Leq (Side) Leq (Front)-Leq (Rear)
High 4 dB 9 dB
Low 3 dB 4 dB
Leq - Equivalent sound level. Steady sound level which, if occurring
for a time t, would result in the same amount of sound energy
as the time varying sound level over the same time period.
Ldn - Day-night sound level. The equivalent sound level over a 24-hour
time period, with a 10-dB penalty added for noise levels occurring
between 10 p.m. and 7 a.m.
ES-3
-------
Patterns of hourly variation are roughly the same for high- and low-
traffic-impact areas. Noise levels are lowest at 4 a.m., increase rapidly
until 9 a.m., remain fairly constant through 6 p.m., and decrease rapidly
after that. If, as EPA noise impact criteria suggest, a 10-dB weighting factor
is added to noise levels between the hours of 10 p.m. and 7 a.m., the resulting
levels would be higher than the daytime levels, and highest near the beginning
and end of the nighttime period.
Roadway traffic is the dominant noise source in both high- and low-traffic-
impact areas. The most commonly noted sources in high-traffic areas are autos,
unidentified traffic, trucks, and household sounds. Either autos or unidenti-
fied traffic is heard 75 percent of the time. In low-traffic areas, the most
common sources are unidentified traffic, autos, birds, household sounds,
»
planes, home yard work, trucks, and jets, with autos or unidentified traffic
heard 44 percent of the time. Trucks, buses, motorcycles, automobiles, con-
struction, and aircraft are the loudest sources in both high- and low-traffic-
impact areas. These results are shown in figure ES-1.
Traffic noise is more prominent at the front of the residential unit,
and in the daytime. Most other sources are louder at the front, but are
heard more frequently at other sides where there is less traffic noise. Other
source levels also appear higher in the daytime, partly as a result of the
higher traffic noise levels which the sources must exceed to be identified.
The data were analyzed to determine the effects of traffic impact, urban
area size, and population density on day-night sound levels and on source
contributions. Traffic impact and population density were found to be sig-
nificant. The population density effect is most pronounced when its logarithm
is used as the independent variable. It was found that the day-night sound
ES-4
-------
Frequency - Fraction of
observations in which
source is predominant.
i |i I mi MM I If M i M MM
^*f? iu ,. fc.r«_i 3i 5 -T = 5 z E S
a w
M ^
1 a
MM
Mr
Leq (id) - Equivalent sound level
as averaged over observations in
which source is predominant.
TO.O-
M.O-
q
f »»-
JO.O-
}0.0-
M.O-
10.0-
0.0-
»
f;
-MII
AUTOS
3
s
2
3
2
-i-
T RUCKS
1
RESIDUAL LEVEL
..
;
HOUSEHOLD
i
-
a
3
5
^
MOTORCYCLES
t
5
3
S
|
-1.
FACTORIES
^
"
PLANES
1
RAILROAD
i
JU
5
|
~
MPLIFIED SOUND
*
i
'
ADULT VOICES
|
OME VARD WUKK
z
§~
^
WIND
i
^
UNIDENTItTADI.E
i
i
O1 HER ANIMALS
i
«
IILDREN'S VOICES
C
J-l
a
2
5
GENCV VEHICLES
at
S
s;
"
CONSTRUCTION
S
2
EREO EQUIPMENT
*
("
a
1_
HELICOPTERS
Figure ES-1A. Source Contribution Profile, High-Traffic-Impact Areas
ES-5
-------
o ^
5
I I § I i I i
I I i i i i
v) uisflQMZHU*1'!
SSazS22s3
i " g 1 S § I = |
> i£ I I s I 1 *
| 1
= iS
I I
I a ;
5 ,
5 3
s
i S 2 S !S
5 - I H
! £ z <
J -
3 =
! H I ! i I 58
K 2 g \ \ i S g S S
s 11 i n s
Figure ES-1B. Source Contribution Profile, Low-Traffic-Impact Areas
ES-6
-------
level can be predicted to within a standard error of 3.7 dB by the equation:
Ldn = 42.3 + 4.7 x (log 10 UZ density) + 7.9 (traffic)
where UZ density is the urban zone population density, and,
traffic = 1 for high traffic impact
= 0 for low traffic impact.
This equation predicts day-night levels in excess of the 55-dB threshold of
impact even in low-density, low-traffic-impact areas. It also suggests that
the population exposed in excess of 75 dB, the level at which hearing damage
may result, is restricted to the high-traffic-impact areas.
The frequency of identification of sources was found to depend mostly on
traffic impact, with roadway sources identified more often and other sources
less often in high-traffic-impact areas. Population density was also found
to be somewhat significant in low-traffic-impact areas, with the frequencies
of identification of traffic increasing and those of natural and household
sources decreasing with population density.
Traffic impact was found to significantly affect roadway source noise
levels. Population density was found to significantly affect all source
levels except for those of traffic in high-traffic-impact areas. With this
exception, noise levels associated with virtually all sources increase
with population density.
In conclusion, residential noise is a problem throughout urban America,
and this problem is greatest near major roadways and in areas with high
population densities. Among its solutions, the most effective will focus on
roadway sources, on the especially high noise levels at the front of residen-
tial units, and on the particularly great noise impacts which occur at the
beginning and end of the nighttime period. While roadway sources are by far
ES-7
-------
the greatest contributors, other controllable sources become increasingly
prominent in higher density areas. Thus, as density increases, so do both
noise exposures and the variety of significant causes of them. As noise
increases, so do both the magnitude-and the complexity of the noise problem.
ES-8
-------
CHAPTER 1. INTRODUCTION
1-1 BACKGROUND AND PURPOSE
Section 14(d) of the Noise Control Act of 1972, as amended by the Quiet
Communities Act of 1978, requires the development and implementation of "a
national noise assessment program to identify trends in noise' exposure and
response, ambient levels, and compliance data, and to determine otherwise the
effectiveness of noise abatement actions through the collection of physical,
social, and human response data."
In partial fulfillment of this requirement, a national survey of urban
residential noise environments has been-conducted. The objectives of this sur-
vey were:
a. To assess the residential exposure of the urban population of
the United States to outdoor noise.
b. To determine the relative contributions of various types of
noise sources to this exposure.
c. To assess the influence of various locational factors, such
as population, population density, and proximity to roadways, on
a and b, above.
The methods and results of this survey, called the National Ambient Noise Survey,
are the subject of this report.
1-2 ORGANIZATION
The report is organized so that it may accommodate readers with a variety
of technical backgrounds and interests. The main body of the report outlines
the objectives and methodology and presents the main results. Supplemental
information regarding the methodology and analyses used to obtain the results
is presented in the appendixes. The intention is that the main body of the
1-1
-------
report be accessible and of interest to a wide audience, and that the appendixes
provide documentation for those who have more specialized interests in environ-
mental acoustics, noise control, and statistics.
1-3 USING THE REPORT
The National Ambient Noise Survey generated a large body of data concerning
urban residential noise environments in the United States. This report attempts
to convert this data into information that is useful to those who wish to com-
bat this form of noise pollution.
The information provided will be helpful in addressing two major issues
related to noise control policy development. The first issue is that of problem
definition. In this context, the report can serve as a basis for predicting
noise levels and source contributions in different types of urban residential
settings. Equivalently, the report establishes norms for the noise environ-
ments in these various settings, against which local survey results can be
compared. Such comparison would provide a basis for assessing the relative
severity of a noise climate in a particular area.
The report also provides some input regarding the probable effectiveness
of various types of noise control strategies in different types of urban
settings. In this context, the information provided serves as a basis for
first-order estimates of the acoustical impact of the strategies, especially
as they relate to control of noise emissions from specific types of sources,
or the control of noise levels at specific locations or times of the day. Of
course, the acoustical impact of noise control measures represents just one
dimension against which their desirability must be assessed; political, atti-
tudinal, and economic assessments are also vital in this process. Nonetheless,
noise problems originate with acoustical phenomena, and so it must be with the
solutions to these problems.
1-2
-------
Finally, a note of caution. The National Ambient Noise Survey represents
a first attempt at developing empirically a national profile of residential
noise environments. As such, it was a pilot study, and one which was intended
more as a basis for further investigation than as an end in itself. The results
of the survey must therefore be viewed as tentative, and, one would hope,
stimulative of further investigation. The uses of these results are many, but
it is important to give due consideration to the limited scope and exploratory
character of the study.
1-3
-------
CHAPTER 2. SURVEY RATIONALE
2-1 INTRODUCTION
To develop a study responsive to the objectives identified in chapter 1,
it was necessary to take into account the criteria that the Environmental Pro-
tection Agency (EPA) has adopted for measuring residential noise impact and the
limitations on the types and quantity of acoustical measurements which could be
performed in the field based upon both technical and resource constraints.
This chapter describes how these factors were used to develop a set of informa-
tion objectives, which in turn served as the basis for developing the survey
methodology discussed in chapter 3. .
2-2 ASSESSING NOISE EXPOSURE
Noise exposure experienced in the residential environment is but one of
many modes of noise exposure to which people are subject in their day-to-day
lives.
Such exposure is considered particularly important, however, because
people spend more time home than at any other location. The most widespread
form of residential noise exposure is that which occurs when noise from exterior
sources intrudes upon the interior residential environment. This circumstance
results in interference with speech communication, sleep disruption, and other
types of interference with household activities. This interference, in combina-
tion with various intervening psychological and social factors, can result in
annoyance and adverse community reaction.
EPA has determined that the simplest noise metric which correlates well
with these effects is the day-night noise level (Ldn). This metric is defined
on the basis of a 24-hour day and includes a 10-dB penalty for noise levels
between the hours of 10 p.m. and 7 a.m. It is computed using the equation:
2-1
-------
Ldn = 10 x log ((15/24 x 10Ld/1°) + (9/24 x i0Ln+10/1°) )
Ld - Equivalent sound level from 7 a.m. to 10 p.m.
Ln - Equivalent sound level from 10 p.m. to 7 a.m.
It has been found that, as residential Ldn varies between 55 dB and 75 dB, the
expected adverse reaction varies from little at 55 dB to a high degree of annoy-
ance at 75 dB, with the proportion of individuals highly annoyed assumed to
increase linearly in between. This has led EPA to associate with residential
noise exposure a "fractional impact," which estimates the proportion of a popu-
lation experiencing a high degree of noise-induced annoyance. This impact,
called the Noise Impact Index (Nil), is given by the equation:
75
Nil =.1/20 /L/ P(Ldn)(Ldn-55)
Ldn=55
P(Ldn) - Population exposed to residential noise level Ldn.
P - Total population.
Although the day-night sound level is a useful descriptor for predicting
noise impact, other acoustical factors should also be considered in assessing
residential noise impact. Among these are temporal and spatial variation in
sound levels around the residential unit.
The scenario of noise from exterior sources impinging upon the interior
residential environment suggests that noise levels at locations near the
residential unit facade are the most directly related to residential noise
impact. In most cases, however, these levels change significantly depending
upon which facade (front, rear, or sides) is considered. Although noise levels
at the front of the unit are generally considered the most significant, it is
2-2
-------
expected that any sizable differentials between these levels and those at other
facades will have an effect on the overall reaction experienced within the unit.
These differentials, therefore, require consideration in assessing residential
noise exposure.
The temporal variation of noise levels must also be considered.
Noise levels vary over time, whether measured in seconds, hours, days,
months, or years. These patterns of variation, when conjoined with human
psychological susceptibilities and patterns of activity, represent the actual
conditions under which noise impacts arise. Although it may not be possible
to develop noise impact criteria which take all of these variations into account,
it is desirable to obtain some understanding of these temporal patterns in noise
levels. Of particular interest in the context of this survey were the hourly,
daily, and instantaneous variations.
The study was therefore designed with the intention that it furnish answers
to the following questions:
a. How is the urban population distributed with respect to residen-
tial Ldn?
" b. What is the typical variation in noise levels between the front,
rear, and sides of a residential unit?
c. How do noise levels vary daily, hourly, and instantaneously?
2-3 SOURCE CONTRIBUTIONS
The Noise Impact Index described above is assumed to be source-independent.
Therefore, the contribution of a particular type of noise source to residential
noise exposure can be estimated in terms of the day-night noise level which
results from this type of source. In situations in which a certain type of
source is clearly dominant, this day-night level can be equated with the
overall day-night level. Unfortunately, many noise environments include noise
2-3
-------
from several different types of sources, none of which is consistently
dominant. In these cases, it is not technically feasible to measure the day-
night levels that result from each individual type of source. This necessi-
tates the use of descriptors that are only indirectly related to exposure
contributions, but are directly measurable. Two such descriptors are commonly
employed in this context. One is the frequency .with which a type of source
is predominant. The other is the average noise level when a certain type of
source is predominant. Both of these descriptors rely upon the judgments of
field observers regarding what type of source is predominant at any particular
moment.
As in the case of overall noise levels, spatial and temporal variation
should also be considered in assessing source contributions.
These considerations implied three information objectives regarding soufce
contributions:
a. How frequently are different types of noise sources predominant
in the urban residential noise environment?
b. What are the average noise levels when particular types of
sources are predominant?
c. How do source contributions vary temporally and by side of resi-
dential unit?
2-4 TYPOLOGY AND FACTOR ANALYSIS
In addition to assessing noise exposure and source contributions for
the urban population as a whole, another objective of the survey was to obtain
similar information for the different types of urban environments that compose
this aggregate. This information was desired in order to enhance the quality of
the aggregate information, to allow greater specificity in comparisons with and
predictions of local noise environments, and for factor analysis.
2-4
-------
To define these objectives further, it was necessary to develop a typology
for urban areas that would take into account their most potentially signifi-
cant characteristics, while at the same time limiting the number of categories
so that a reasonable number of measurements could be made, in each one. Four
parameters were selected to be included in this typology. These parameters,
which will be defined more precisely in chapter 3, include metropolitan area
size, area population density, distance from major roadways (traffic impact),
and location with respect to flight paths around major airports (aircraft
impact). These parameters were used because of their significance to noise
environments, the ease with which they can be evaluated for a particular area
or location, and the availability of estimates of the residential populations
of each defined type of urban area. The latter parameter was used to exclude
from consideration areas which are heavily impacted by aircraft noise. The
three other parameters were used to define a sampling cell structure.
Two information objectives concerning these three parameters were defined:
a. How is residential Ldn affected by urban area size, population
density, and proximity to major roadways?
b. How are source contributions affected by these parameters?
2-5 SURVEY EVALUATION
The information needs described in this chapter have two things in common.
First, the information required pertains to the urban population at large, as
opposed to the particular segments of that population with particular noise
exposure problems. Second, the information needs could be met through direct
measurement and observation of acoustical phenomena. Together, these two
attributes defined the set of information objectives that could be reasonably
expected to be met by a study such as the National Ambient Noise Survey.
2-5
-------
However, such reasonable expectations in no way imply certainty of success.
This was especially true in this case, where a relatively small-scale measure-
ment effort was employed to meet a wide variety of information needs pertaining
to a large set of urban locations.. Thus, a final objective was to assess the
utility of direct measurements in obtaining the desired information. In this
respect the survey was not only an investigation of urban noise, but also of
the role of noise monitoring in this investigation.
2-6
-------
CHAPTER 3. METHODOLOGY
3-1 BASIC APPROACH
The .basic approach of the National Ambient Noise Survey was to divide
the set of urban residential units in the United States into subsets, called
sampling cells, select a random sample of units in each sampling cell, and
develop statistically valid profiles of the residential noise environments in
each sampling cell based on measurements taken at the selected units.
To determine the set of urban residential units, the 1970 U.S. Census was
used. The census compiled a list of 248 "urbanized areas." Each area consists
of at least one central city and surrounding closely settled territory. An
urban residential unit is defined as one that is located within one of the
urbanized areas.
With each residential unit is associated a "noise1 environment." This is
defined as the immediate noise field surrounding the exterior facades of the
residential unit. Thus, for the purpose of the study, noise environments are
assumed to be individuated by residential unit.
3-2 CATEGORIES OF RESIDENTIAL UNITS
The set of urban residential units is divided into sampling cells on
the basis of urban area size, population density, distance from major road-
ways ("traffic impact"), and location with respect to aircraft flight paths
("aircraft impact"). A category is specified by four indices, each corre-
sponding to a range of values of one of the parameters.
3-1
-------
3-2-1 Urban Area Size
The urban area size parameter is defined on the basis of the urbanized
area population as determined by the 1970 census. These populations are given
in table 20 of the 1970 U.S. Census of population, U.S. Summary. A portion of
this table is shown in table 3-1.
Two categories of urban area size are defined. "Large" urbanized areas
are those with a population of 2 million or above. All others are classified
as "medium-small" urbanized areas.
Referring to table 3-1, the Chattanooga, Tenn. - Ga. urbanized area is seen
to fall in the medium-small category, while Chicago, 111. - Northwestern Indiana
is in the large urbanized area category.
3-2-2 Population Density
As depicted in table 3-1, each urbanized area is divided into two or
more components. For example,' the Chattanooga urbanized area consists of
Chattanooga City and outside Chattanooga City, and the Chicago area consists
of Chicago, East Chicago, Gary, Hammond, and outside central cities. These
components will be called "urban zones."
The population density metric used in the survey is the urban zone popu-
lation density based on the 1970 census. These densities, based on gross land
area, are also given in table 20.
Four categories of population density (persons per square mile) are defined.
Densities of 4,500 or over are classified as "high." Densities between 2,500
and 4,499 are "medium-high". Those between 1,500 and 2,499 are "medium-low".
Densities below 1,500 are classified as "low."
3-2
-------
Table 3-1. Excerpt from U.S. Summary, Table 20
OJ
LO
Table 20 Population and Land Area of Urbanized Areas: 1970 and 1960Continued
(For meaning ol symbols, see lexlj
Areas
Champaign-Urhana, 111 100,417 1OO.O
Inside central cities 89,332 89.0
Champaign 56,532 56.3
Urbana 32,800 32.7
Outside central cities 11,085 11.0
Charleston, S.C 228,399 1OO.O
Charleston city 66,945 29.3
Outside central city 161,454 70.7
Charleston, U. Va 157,662 1OO.O
Charleston city 71,SOS 45.4
Outside central city 86,157 54.6
Charlotte, N.C 279,S3O 1OO.O
Charlotte city 241,178 86.3
Outside central city 38,352 13.7
Chattanooga, Tenn.-Ga 223,580 1OO.O
Chattanooga city 119,082 53.3
Outside central city 104,498 46.7
Chicago, 111.-Northwestern Ind 6,714,578 1OO.O
Inside central cities 3,697,144 55.1
Chicago 3,366,957 50.1.
East Chicago 46,982 0.7
Gary 175,415 2.6
Hammond 107,790 1.6
Outside central cities 3,017,434 44.9
Cincinnati, Ohlo-Ky 1,110,514" ' 1OO.O
Cincinnati city 452,524 40.7
Outside central city 657,990 59.3
Cleveland, Ohio 1,959,B8Q 1OO.O
Cleveland city 7SO,9O3 38.3
Outside central city 1,208,977 61.7
Colorado Springs Colo 204,766 1OO.O
Colorado Springs city 135.O6O 66.0
Outside central city 69,706 34.0
1970
Population
Number
Percent
distribution
Land area
in square
miles
Population
per square
mile ol
land area
18.3
13.4
8.3
5.1
4.9
99.2
17.2
82.0
61.8
27.2
34.6
105.7
76.0
29.7
5,487
6,667
6,811
6,431
2,262
2,3O2
3,892
1,969
2,551
2,629'
2,49O
2,645
3,173
1,291
116.7
52.5
64.2
1,916
2,268
1.628
1,277.2
301.0
222.6
12.3
42.0
24.1
976.2
5,257
12,283
15,126
3,820
4,177
4,473
3.O91
335.1
78.1
257.0
646.1
75.9
57O.2
90.0
60.8
29.2
3,314
5,794
2,560
3,033
9,893
2,120
2,275
2,221
2.387
I960
Population
Number
Percent
distribution
78,O14
76,877
49,583
27,294
1,137
160,113
65,925
94,188
169,500
85,796
83,704
209,551
201,564
7.987
1OO.O
98.5
63.6
35.0
1.5
1OO.O
41.2
58.8
100.0
50.6
49.4
1OO.O
96.2
3.8
205,143
130,009
75,134
100.0
63.4
36.6
'5,961,634
3,898,091
3,550,404
57,669
178,320
111,698
'2,063,543
1OO.O
65.4
S9.fi
1.0
3.0
1.9
34.6
993,568
502,550
491,018
1,783,436
876,050
907,386
1OO,22O
70,194
30,026
1OO.O
50.6
49.4
1OO.O
49.1
50.9
100.0
7O.O
30.0
Land aiea
in square
miles
Population
per square
mile ol
land area
11.7
10.7
e.s
4.2
l.O
31.2
5.5
25.7
53.6
26.1
27.5
72.1
63.0
9.1
89.0
36.6
52.4
955.7
296.7
221.7
11.8
40.5
22.7
659.0
241.5
76.5
165.0
581.4
75.9
SOS. 5
28.3
15.7
12.6
6,668
7,185
7,628
6,499
1,137
5,132
11.986
3,665
3,162
3,287
3,O44
2,906
3,199
878
2,305
3,552
1.434
6,238
13,138
16,014
4,887
4,4O3
4,921
3.131
4,114
6,569
2,976
3,067
11,542
1,793
3,541
4,471
2,383
Percent
change in
population,
I960 to 1970
28.7
16.2
14.0
20.2
874.9
42.6
1.5
71.4
-7.0
-16.7
2.9
33.4
19.7
380.2
9.0
-8.4
12.6
-5.2
-5.2
-18.5
-1.6
-3.5
46.2
11.8
-10. 0
34.0
9.9
-14.3
33.2
1O4.3
92.4
132.2
-------
Again referring to table 3-1, Chattanooga City is seen to fall in the
medium-low-density category. In the Chicago urbanized area, Chicago is in
the high-density category, while all other urban zones are in the medium-high
/
category.
3-2-3 Traffic Impact
The traffic impact parameter, which represents the proximity of
the residential unit to major roadways, is based upon the Federal Highway
Administration's classification scheme for urban roadways. This scheme includes
six functional classes: interstates, freeways and expressways other than inter-
states, principal arterials, minor arterials, collectors, and local streets.
Two categories are defined. Residential units that are within 300 feet
of an interstate or urban freeway or within 100 feet of a principal or minor
arterial are put in the "high-traffic-impact" category.* All other residential
units are in the "low-traffic-impact" category.
These distances are necessarily somewhat arbitrary. The intent is to
include in the high-traffic-impact category only those residential units that
are chiefly impacted by major roadways. In general, only residential units
that face on arterial streets or are less than three houses distant from
interstates and freeways fall into the high-traffic-impact category.
3-2-4 Aircraft Impact
To define the aircraft impact categories, day-night sound level contours
around airports were used. These contours delineated areas within which
specified ranges of day-night noise levels result from aircraft operations.
Examples of these contours are shown in figure 3-1.
*These distances are measured from the center of the nearest lane of
the roadway to the center of the residential unit.
3-4
-------
OJ
I
Ul
Figure 3-1. Ldn Noise Contours Around a Typical Airport
-------
All residential units located within the Ldn = 65-dB contour are placed
in the "high-aircraft-impact" category. All other residential units are
in the "low-aircraft-impact" category.
Residential units in the high-aircraft-impact category were not included
in the survey. This exclusion was justified by the ability of computer models
to predict noise levels in these areas and by the daily variability of these
noise environments, resulting from day-to-day changes in aircraft flight paths
and air traffic volumes.
3-2-5 Sampling Cells
Table 3-2 summarizes the categories of area size, population density,
traffic impact, and aircraft impact described in the previous sections. To
/
define the sampling cells, these four sets of categories are simply combined.
Thus, a sampling cell consists of all residential units that fall within the
same urban area size, population density, traffic-impact, and aircraft-impact
categories.
One important feature of this cell structure is that data are available
that permit a calculation of the human population corresponding to each combi-
nation of area size, population density, and traffic-impact categories. The
calculations rely on 1970 census data and data from the National Roadway
Traffic Noise Exposure Model data base. These populations are given in
table 3-3 and derived in appendix E.
No data are available that permit the apportionment of these populations
into high- and low-aircraft-impact categories. It is estimated that a total of
5.22 million people reside within Ldn = 65-dB contours. Because this repre-
sents only about 4 percent of the urban population, little uncertainty is
introduced from assuming that this population is distributed among the other
categories in the same way as the total population. This assumption was made,
3-6
-------
Table 3-2. Summary of Categories Used to Define Sampling Cells
Parameter
Urban Area
Size
Basis for Defining
Categories
Urbanized area
population (1970)
Number of
Categories
2
Description of
Categories
Large - _> 2,000,000
Population
Density
Traffic
Impact
Aircraft
Impact
Urban zone population
density (1970) ,
Distance from major
roadways
Ldn contours around
airports
Medium/small - All
others
High - _> 4,5007
square mile
Medium-high - 3,000-
4,500
Medium-low - 1,500-
3,000
Low - <1,500
High - Within 100
feet of an arterial
or 300 feet of an
interstate or
freeway
Low - All others
High - Within
Ldn = 65 dB contour
Low - All others
3-7
-------
Table 3-3. Population (Millions) of Sampling Cells Used in
National Ambient Noise Survey (1980 Estimates)
Large Medium/Small
Urban Areas Urban Areas
HigK Low High Low
High 4.29 25.00 1.47 14.96
Medium-High 3.82 19.49 2.20 19.92
Medium-Low 0 0 3.07 31.7
Low 00 0.53 5.78
3-8
-------
and the populations given in table 3-3 were assumed in all subsequent calcula-
tions. The survey may thus be conceptualized as a survey of the total urban
population, but one that does not consider the increment-of noise that results
from a residential unit being located within an Ldn = 65-dB contour.
The total number of sampling cells defined by the four categories of urban
area size, urban zone population density, traffic impact, and aircraft impact
is 32. Excluding the high-aircraft-impact cells reduces this number to 16.
Of these, four have zero population. This leaves a total of 12 sampling cells
to be included in the survey.
3-3 SITE SELECTION
Each of the sampling cells included in the survey represents a subpopula-
tion of urban residential units. Units in each subpopulation were then
randomly selected to be used as measurement sites.
The selection process involved five steps. First, the sizes of the samples
desired from each sampling cell were determined. Second, urbanized areas were
randomly selected. Third, census .tracts within each selected urbanized area
were randomly selected. Fourth, blocks were randomly selected from within each
census tract. Finally, residential units, either high- or low-traffic impact'
as required, were selected from each selected block. A detailed description of
the procedures followed in determining sample sizes and in making these random
selections is described in appendix A.
A profile of the sites obtained in the survey is given in table 3-4. A
complete listing of the sites is given in appendix F.
3-9
-------
Table 3-4. Site Quotas and Sites Obtained for Each Sampling Cell
1980
1981
Total
Urban Area
Size
Large
Large
Large
u) Large
Medium- Small
Medium-Small
Medium- Small
Medium-Small
Medium- Small
Medium-Small
Medium- Small
Medium-Small
Urban Zone
De ns i ty
High
High
Medium-High
Medium-High
High
High
Medium-High
Medium-High
Medium- Low
Medium- Low
Low
Low
Traffic
Impact
High
Low
High
Low
High
Low
High
Low
High
Low
High
Low
Quota Obtained
4
4
4
4
4
4
4
4"
4
4
4
4
4
4
4
4
4
4
3
3
3
4
3
3
Quota
1.
16
1
11
1
11
1
21
1
26
1
6
Obtained
1
15
0
7
1
9
1
18
1
22
1
5
Quota
5
20
5
15
5
15
5
25
5
30
5
10
Obtained
5
19
4
11
5
13
4
21
4
26
4
8
Average
Average Urban Zone
Urban Area Population
Population Density
4,844,
6,678,
5,049,
8,938,
585,
765,
184,
588,
477,
739,
187,
208,
195
032
977
089
593
491
195
187
083
398
740
261
12,524
15,773
3,786
3,557
8,317
7,279
4,111
3,610
1,964
2,327
1,122
1,248
Total
48
43
97
81
145
124
-------
3-4 MEASUREMENT PROTOCOL
3-4-1 Elements of the Protocol
Two types of sound-level measurements were performed at each selected
residential unit. Continuous sound-level measurements employing an automated
noise-level recorder equipment were performed at the front of each selected
unit. Manual sound-level measurements and source identifications, employing
a technique called microsampling, were made at accessible sides of each
selected unit.
3-4-2 Continuous Measurements
The purposes of the continuous measurements were (1) to accurately measure
the Ldn and (2) to provide a detailed time history of sound levels over a
24-hour or longer period at one location near the residential unit. This loca-
tion was the architectural front of the residential unit and was further speci-
fied by rules that took into account such factors as driveway location, window
location, and security. The intent of these rules was to obtain a location
that represented as closely as possible the external-source-generated noise
field at the front side of the unit.
The continuous measurements were generally conducted for a 24-hour period.
At a few sites, 5-day continuous measurements were performed to assess daily
variation in sound levels.
A discussion of the equipment and analytical procedures used in continuous
monitoring is provided in appendix C.
3-4-3 Microsamples
Microsamples were collected to assess source contributions and differentials
in noise levels between the front and other sides of the residential unit.
3-11
-------
Microsamples consisted of 120 sound-level measurements, employing a Type 2
or better sound-level meter set to "slow" response, and source identifications
taken over a 30-minute time period. The microsamples were usually gathered
in sets of two, one at the unit near the continuous-measurement location and
one at another side. Whenever possible, three such sets were obtained at each
site, one each in the daytime, evening, and nighttime periods.* The exact
locations at which these measurements were taken depended upon window placement
and other architectural factors. As in the case of the continuous measurement,
the objective was to accurately represent.the external-source-generated noise
field at the given side of the unit.
The sound-level measurements were taken every 15 seconds. At the time
of each measurement, the type of noise source judged to be predominant was
also recorded. Table 3-5 shows the list of source types used in the survey.
3-5 ANALYTICAL PROCEDURES
3-5-1 Data Reduction
The continuous noise-level data were encoded on digital cassette tapes.
These tapes were analyzed using a digital translator and computer. The micro-
sample data were keypunched and computer reduced. The reduced data were
assembled into a set of Statistical Analysis System data sets for further
analysis.
3-5-2 Data Analysis
The data were analyzed to develop information relevant to the objectives
discussed in chapter 2. A discussion of some of the more important analytical
procedures used is included in appendix D. The results of this analysis are
presented in chapter 4.
*Daytime is 0700-1900; evening is 1900-2200; nighttime is 2200-0700.
3-12
-------
Table 3-5. Source Codes Used in Microsamples
A Auto
B Bus
C Construction Equip, (not Y or S)
D Dog
E Emergency Vehicle
F Factory Equip.
G Unamplified Adult Voice
H Helicopter
I Person Using Nonpowered Equip, (not Y or S)
J Jet
K Unamplified Child Voice
L Amplified Sound (not E)
M Motorcycle
N Other Animal (not D or 0)
0 Bird
P Prop. Plane
Q Wind
R Railroad
S Household (not G, Y, or K)
T Truck
U Unidentified Road Traffic
V Off-Road Vehicle
W Water Vehicle
X Unidentifiable Source
Y Home Yard Work (not G or K)
Z Residual Level
3-13
-------
CHAPTER 4. RESULTS
4-1 INTRODUCTION
In this chapter, the results of the survey are described. They are pre-
sented as answers to the questions posed in chapter 2. The answers presented
are based solely upon the data obtained in the survey. Therefore they should
be viewed cautiously: 124 sites are being used to represent a population many
orders of magnitude greater.
With each answer are included a few remarks regarding its policy implica-
tions. These remarks are far from exhaustive, but are intended to illustrate
the connections which exist between the information presented and noise control
policy development. It is expected that other such connections will be made
by the interested reader in light of the local situation which he faces.
The emphasis in this chapter is on presenting information, not a detailed
account of how this information was derived from the measurement results. Such
an account may be of value to some readers, and is included, along with an
error analysis, in appendix D.
It must be emphasized that these results apply to the urban population not
residing within Ldn = 65 dB contours around airports. Equivalently, the
results may be considered to apply to the entire urban population, but without
including the increment of noise which results from living within that contour.
4-2 DISTRIBUTION OF THE URBAN POPULATION OVER Ldn.
Figure 4-1 shows two versions of the distribution. The bars show the "raw"
distribution, in which the results of the individual measurements were simply
weighted according to the populations of the sampling cells. The curve is the
normal curve derived from the individual measurements. The mean Ldn value is
60.4 dB, and the standard deviation is 4.8 dB.
4-1
-------
13
12
II
IA
O
^
i
""" 9
o
4>
a>
o
x 8
UJ
c
0 7
** o
1 -=
1x3 a.
£ 6
c 5
4)
'35
a>
n* A
3
2
1
/
/
/
./ *
_x I
y
/
/ V
t
j
I
'
'
,'
1
i
i
1
1
i
.
*
I
1
t
1
*
1
'
:
'
i
i
X
DISTRIBUTION BASED ON RAW DATA
DISTRIBUTION BASED ON NORMAL
CURVE FITTED TO RAW DATA
N
\
s
X
\
X
\
\
X
.
I
V
\
\
\
\
' \
J
\
\
\v
\
\
\
\
y
\
!
\
\
\
\
\
' \
!'
\
\₯ ;'
*
"^ I
45
50
55
65
7O
6O
Ldn
Figure 4-1. Distribution of Urban Population Over Ldn
75
-------
The irregularity of the raw distribution reflects the small number of data
points upon which it is based. The normal curve smooths out the irregularities,
and is therefore the preferred approximation of the true distribution.
Table 4-1 shows the percentage of the urban population exposed to residen-
tial Ldn over 55, 60, 65, 70, and 75 dB respectively. Earlier EPA 'estimates,
as based on the 100-site study*, are also included.
Table 4-1. Populations Exposed to Critical Values of Ldn
(Percent of Urban Population)
Ldn >
Ldn >
Ldn >
Ldn >
Ldn >
55
60
65
70
75
dB
dB
dB
dB
dB
National Ambient
Noise Survey
87%
53%
17%
2%
<1%
100-Site
Study
70%
44%
18%
5%
1%
These estimates indicate that the vast majority of the urban population is
exposed to residential noise sufficient to create some adverse impact, but that
only a tiny fraction experience the full impact associated with Ldn over
75 dB. Eighty-seven percent of the urban population are sufficiently exposed
to experience some benefit from a reduction in noise levels.
The Noise Impact Index, defined in chapter 2, can be computed from these
results, and is found to be .28. Thus, based on EPA criteria, 28 percent of
the urban population are expected to be highly annoyed as a result of noise in
the residential environment.
*The 100-site study is the previous nationwide study of urban residential
noise exposure sponsored by EPA. The study established a relation between Ldn
and census tract population density. This relation, in combination with the
distribution of the urban population over census tract density, was used to
develop the cited estimates.
4-3
-------
These results consider noise levels at the front of residential units only.
The next section discusses the differentials between these levels and those at
other sides.
4-3 VARIATION IN NOISE LEVELS BY SIDE OF RESIDENTIAL UNIT.
Table 4-2 shows the average differentials between noise levels at the
front and those at the rear and sides. The information is presented for both
high and low traffic areas.
Table 4-2. Differentials in Noise Levels by Side
Traffic
Impact Leq(Front)-Leq(Side) Leq(Front)-Leq(Rear)
High 4 dB 9 dB
Low . 3 dB 4 dB
!
These differentials result primarily from screening effects and the gen-
erally greater distances between roadways and the rear and side locations.
They therefore reflect the predominance of roadway traffic in the noise environ-
ment in low aircraft impact areas. Whether and how these differentials affect
noise impact are interesting and important questions yet to be answered.
These results nonetheless have useful implications for the design and
retrofitting of urban housing. Housing layouts which locate the most noise-
sensitive areas, such as bedrooms, in the rear will afford such areas maximal
protection from exterior noise. Noiseproofing measures, such as double
glazing of windows, will realize their greatest benefits when applied to the
front of residential units. These considerations are especially important in
high traffic impact areas.
4-4
-------
4-4 DAILY, HOURLY. AND INSTANTANEOUS VARIATION IN NOISE LEVELS
4-4-1 Daily Variation
Table 4-3 shows the average day-by-day variation in Ldn by day of the week.
It is based on the results of the eight sites which were monitored for 5 days
continuously.
Table 4-3. Average Ldn by Day of Week
Day ' M T W TH F S S
Ldn (dB) 59 59 60 61 61 59 60
The small variations obtained indicate that the day of the week does not
significantly affect the Ldn. This conclusion is especially surprising in the
case of weekend days, when traffic patterns are significantly different from
those on weekdays.
The standard deviation in Ldn for the eight 5-day sites ranged from 1.0
to 2.9 dB, with an average of 2.0 dB. This means that one day of monitoring
is usually sufficient to obtain a .reliable estimate of the Ldn in low aircraft
impact areas.
4-4-2 Hourly Variation
Figure 4-2 shows the variation in hourly Leq values over a 24-hour period
for high- and low-traffic-impact areas. The patterns of variation are seen to
be remarkably similar in both types of areas, especially during the evening and
nighttime hours. Starting at midnight, the levels decrease steadily until about
4 a.m., and increase rapidly between 4 a.m. and 7 a.m. The levels begin to
decrease steadily after 6 p.m. During the hours between 7 a.m. and 6 p.m.,
noise levels remain fairly constant in high-traffic-impact areas, while increas-
ing slightly in low-traffic-irapact areas.
4-5
-------
LEQ
64-
58-
66-
54-
46-
LEO
72-
70-
68-
62-
60-
54-
HOUR
135 7 9 11 13 15 17 19 21 23
HOUR
IS
23
High-Traffic Impact
Figure 4-2.
Low-Traffic Impact
Average Hourly Leq Values, by Traffic Impact
(Dashed line includes 10-dB penalty for nighttime levels.)-
-------
The dashed lines in figure 4-2 indicate Leq values adjusted to reflect
the 10-dB weighting factor for noise levels during the evening and nighttime
hours used in the calculation of Ldn. When this factor is considered, noise
during the nighttime appears to have a substantially greater impact than noise
during the daytime, with by far the greatest impacts occurring at the beginning
and end of the nighttime hours.
4-4-3 Instantaneous Variation
Instantaneous variation is best described by statistical levels. These
descriptors indicate what noise level is exceeded for a given percentage of
the time. LI, for example, is the level exceeded 1 percent of the time, and
thus describes peak and near-peak levels. L99 is the level exceeded 99 percent
of the time, thus defining the "noise floor" of the environment.
«
Figure 4-3 shows the variation in the statistical levels L99, L90, L50,
L10, and LI over a 24-hour period for high- and low-traffic-impact areas.
Instantaneous noise levels are seen to vary more greatly in the daytime and
in high-traffic-impact areas. Although all five statistical levels follow
the same basic pattern of variation, this pattern is most pronounced in the LI
and L10 levels. The variation in these upper statistical levels is the primary
source of the hourly Leq variation described in paragraph 4-4-2.
Of particular interest is the increment between LI and L10. This incre-
ment represents the noise levels resulting from the loudest noise sources. As
such it would be most affected by an abatement program targeted at such sources.
A plausible goal of such a program would be to reduce all noise levels above
the pre-program L10 to levels at or below this value.
4-7
-------
I
00
uj SO |
i
HOUR
I I I I I I I I I I I I I I I. I I I I I I I I
1 3 5 7 9 11 13 IS 17 19 21 23
60-|
3H
1 3 5 7 9 II 13 15 17 19 2t 23
HOUR
High-Traffic-Impact Areas Low-Traffic-Impact Areas
Figure 4-3. Average Hourly Statistical Levels, by Traffic Impact
-------
Figure 4-4 shows the hourly decreases in Leq that would result if this
goal were to be' achieved, based on the statistical levels shown in figures 4-3A
and 4-3B.* In low-traffic areas, these decreases are generally about 3 dB,
with a substantial decrease for the late night hours and some increase for rush
hours. The expected decreases are generally about 2 dB in high-traffic areas,
but substantially greater during the late night and rush hours.
These observations suggest that a noise abatement program targeted at the
noisiest 10 percent of sources would decrease noise levels in low- traffic-
impact areas by about 3 dB, which in most cases would result in an Ldn at or
below 55 dB, the assumed threshold of impact in the EPA criteria. Such a
decrease would also be equivalent to reducing traffic volumes by 50 percent.
In high-traffic-impact areas, a somewhat smaller overall decrease would result.
However, an abatement program targeting noise levels in the late night hours
(perhaps employing curfews) would be especially effective in such areas.
4-5 FREQUENCY OF PREDOMINANCE AND AVERAGE NOISE LEVELS OF SOURCES
Source contributions are described by means of two descriptors, called
frequency and Leq (id). Frequency is the proportion of time in which the source
was identified as predominant. Leq (id) is the equivalent sound level as aver-
aged over the instances in which the source was so identified. Figures 4-5A
and 4-5B show these values for high- and low-traffic areas, based on micro-
samples taken at the front measurement site only.
*The average noise level during the noisiest 10 percent of the time is
assumed to be:
10 x Log 10((10L1/1° + 10L10/10)/2)
4-9
-------
4.8-
4.5-
4.2-
3.9-
3.6-1
3.3-
3.0-
2.7-
2.4-
2.1 -
1.8-
1.5-
1.2-
1 HIGH TRAFFIC IMPACT
2 LOW TRAFFIC IMPACT
1 3 5 7 9 11 13 15 17 19 21 23
HOUR
Figure 4-4. Projected Hourly Decreases in Leq Resulting From Achievement
of Hypothetical Goal for Noise Control Program
4-10
-------
S S
9 .500-
s
£
.«0-
.100-
200-
.100 -
1
i
3
= ,
i S S »
6 3 a s
| d 9 o
- 3 i 3 3
£ u > o *
" " i I
2 1 |
^ *
Z at
a
Z
S S
S o
*"^ZC^5
3 5
rt 3
i
o
i
- i I
a a =
x * 0
O < at
5 " I
ik *
= 5 Z £ 8 I
1 1 § 1 i i
5 1 i I f ?
i : 1 s
s = S 5
5 < S i
2 0 =
< r
OOOMJI
MALS 1
R ANI
0
S
s
ilN'i V<
q
C
3 S
I 3
i "
z
Jj
>
s
a
Z
000176)
|
|
1
6(OIOO'0
i
i
H
at
LU
1
o
s d
s I
~ C
«
X
0
6
S
.
5
»
.2
» 50.0-
! 1
S is 3 g § S 3
= »
III
Figure 4-5A. Source Contribution Profile, High-Traffic-Impact Areas
4-11
-------
M.O-
50.0-
.10.0-
30.0-
20.0
10.0
=
£
| I ,
5 3 S
r~H
s
^
I S) 4-WX.
-lisa
-
I i
i
i
: 1
I ?
5
a
s
i
§ i ^ c
*
f:
£
*
If i 1 1 1 \ I \ - i I i i i I i f = ! I I s 1 1 1
i i 3i s ! ^ i I i M I H Hi
Figure 4-5B. Source Contribution Profile, Low-Traffic-Impact Areas
4-12
-------
Unidentified traffic and automobiles are by far the most frequently
mentioned sources. Other frequently identified sources are trucks and house-
hold sounds in high-traffic areas, and birds, household sounds, voices, air-
craft, home yard work, and trucks in low-traffic areas. Residual sound, that
which is heard when no particular source is sufficiently loud to be identified,
is identified about 3 percent of the time in high-traffic-impact areas and
10 percent of the time in low-traffic-impact areas.
The loudest sources in both traffic impact areas are autos, buses,'trucks,
motorcycles, aircraft, and construction, all with Leq (id) values over 65 dB in
high-traffic impact areas and over 55 dB in low-traffic-impact areas. The
sources are the usual targets of noise control programs. Another frequent tar-
get which is'not among this group is dogs, with Leq (id) values of 56 dB in
high-traffic-impact areas and 52 dB in low-traffic-impact areas. This suggests
that dog-barking is a noise problem primarily as a result of its expressive
content.
Leq (id) values are consistently higher in high-traffic-impact areas.
This indicates a more "competitive" noise environment, in which the higher
traffic-noise levels' must be exceeded by another source in order for that
source to be identified. This creates the illusion that even sources unrelated
to traffic (birds, for example) are louder in such areas.
4-6 VARIATION OF SOURCE CONTRIBUTIONS BY SIDE OF RESIDENTIAL UNIT
Six sources were selected for analysis of locational and temporal varia-
tion of frequency and Leq (id). Three of the sources, autos, trucks, and
motorcycles, are roadway traffic sources frequently targeted in noise control
programs. Two sources, construction and dogs, are non-transportation sources
which are also targeted in many programs. The other source, birds, was selected
4-13
-------
because of the commonly made association between bird sounds and a pollution-
free sound environment.
Figures 4-6A through 4-6F show the frequencies of identification and Leg (id)
values of these sources by side for high- and low-traffic-impact areas. They
are based on data taken in the daytime hours. The roadway traffic sources are
clearly most prominent at the front, with this tendency most strongly pronounced
in the high-traffic-impact areas. Construction and dogs are more frequently
identified at the rear and sides, although the associated noise levels are
.higher when they are heard at the front. This reflects the higher competing
levels which these sources must exceed in order to be heard at the front.
These effects are again more pronounced in high-traffic-impact areas. Birds
follow the same basic pattern as dogs and construction with regard to frequency,
with little locational variation in Leq (id) values. These observations sug-
gest that noise control programs oriented toward roadway traffic will have their
greatest impact on noise levels at the front of residential units. Non-roadway
traffic oriented measures will have a more uniform, though smaller, impact.
4-7 TEMPORAL VARIATION IN SOURCE CONTRIBUTIONS
Figures 4-7A through 4-7F show the frequencies of identification and Leq
(id) sources of the six sources identified in paragraph 4-6 by time of day,
based on data taken at the front only. Daytime refers to hours between 7 a.m.
and 7 p.m. Evening refers to the time between 7 p.m. and 10 p.m. Nighttime
includes the period between 10 p.m. and 7 a.m.
The roadway traffic sources are predictably less prominent in the night-
time than in the daytime, both in terms of frequency and of Leq (id). In the
evening period, autos and motorcycles are as prominent as in the daytime, while
trucks are much less prominent. Noise levels associated with roadway vehicles
are consistently lower in the nighttime.
4-14
-------
FREQUENCY
.900
.800
.700
.600
.500
.400
.p-
1
H-*
01 .300
.200
.100 -
«? 8 C>j
"
0
(0
v^f'
^ "! ""
; V .
;;H
}-; /,
'" - ; -\
/'%.?
'", ^'
' ';-//'
^ ^
/);/
j;|;'
Kg :*
O)
in
T';^
'#-;-'
k f
^^ *"
\ --^^
<
^'A ' v '
f \f '
/ '.
:,wf
fl
^t~!'
in
S
f,.
S$ 'V
f y fj
V
X' '
/
'""//'.
''-
"
v ", 5. .
l$|
/ J''^
: :
S
Oj
in
t-yjOC Hc/>CC Ht/5
HQ]< Z 'UJ < Z HI
O Q ill O Q LU O Q
DC OC CCrXQC OC (/)
^
-
oc
UJ
oc
IT)
in
' , * -
'\; \
£ f
s f
f ' ^
,
» ° '" !
4 '"
,- '
%"/
_,
V?
/
CO -
1 ;,i
^'^ ^^
, jl ;
ff
' ''''.
.".,'
;/ ,v
/ '
*' '.
' , '
/v';>
in
i^"*
^
°' - '
" ^
'' -^
s f f
£
S
Jf >'
LEQ(ID)
90
80
70
60
50
40
30
20
10
H- co a:
Z LU rf
O Q uj
.f C5 a:
FREQUENCY LEQ(ID)
HIGH TRAFFIC IMPACT
FREQUENCY LEQ(ID)
LOW TRAFFIC IMPACT
Figure 4-6A. Variations in Automobile Contributions, by Side of Unit
-------
FREQUENCY
.090
.080
.070
.060
.050
.040
.030
.020
.010
q
0
cc
n
q
to
UJ
Q
to
1
DC
UJ
CC
O
oi '
CD
J,
\'-i*
$'',','>
,v
ft'
k"l
*f
% '
_ *
' '
'
*,,
\'"C'
'*. $r '.
'''?.'/.
',''''',
V?;
H
z
O
DC
to
',
jf f
f
}''
f
* '
,
f
Jf
'" - J
ff ' ;
V' «,i
, ' ';
t'-'V'!
to
UJ
O
co
CO
£8
^ ' ;
' \
'>
' x '/':
" j
' :
/< '
f*
?'-'
''''-'''
'''-. '
x. _.
'' '"
DC
UJ
DC
0
s
1-
0
DC
CM
s
to
UJ
Q
CO
n
CM
o
cc
UJ
DC
5
' ^ J
^V'
"""
->;!$
^ %?(i-'
" ,
" * , "s
, ''
''*'l
^'fl.
r;<^
^ *'-'
'V *"'
''-'t"'
''* %'
1-
o
DC
in
S
*'* /'
'/"? f '':
,%''*
,v »
*'«,'
^ ' *
v/% '> "*
>{''
'*$*'
' , '-v
{&
to
UJ
O
to
to
s
,^
' ' ^'-> J
* '^'^'
fr ^^ %
y"'' :
- ^ "":
<
V
~^fff
;'<\>;
?,'xll,
"/''''%
^ ''.
'^V:
cc
UJ
DC
LEQ(IO)
90
80
70
60
50
40
30
20
10
o
FREQUENCY LEQ(ID)
HIGH TRAFFIC IMPACT
FREQUENCY LEQ(ID)
LOW TRAFFIC IMPACT
Figure 4-6B. Variations in Truck Contributions, by Side of Unit
-------
FREQUENCY
.018
.016
.014
nirt
.U I f.
.010
.008
j>
1
^ .006
.004
.002
r.
p
H
O
tc.
u.
CO
p
CO
UJ
2
co
in
p
cc
UJ
cc
^
8
U$
''V'-,*?
4/r '''
"y '''',
/-'' ./<
,"> ''^'
'*' -\'
A. ff
~ff ~t
' f" ^
1 '<*' *
^ fff fif
''t/l''',
'ft-Sft '
'%?''£
'£'?">,'-'
'^frff
Hk
O
1C
u.
n
CD
&$&
"ffi";
%~1
' ,','?'
'';'&'
'"< JJC'
:-M
^
CO
Ul
O
CO
iri
U)
i''^'"-'
' »/'
" X <'^ t
,,,?;
; J'
' ^ i
'" r.
' '.V ^ '
')'£:
^' '-V
' A"
11 ?; ''
'H''', I
'":'"'';
tc.
UJ
oc
CM
p
1-
O
oc
u.
§
p
UJ
0
CO
^
p
1C
UJ
cc
p
r-'
'*?':
'% ' 'f
't'f ^ :
'^
i-
o
tc
u.
o
,s
v-ry;-:
'f^t
I/* ?
'"'\:
' ff
' f fff *f
' ]^f
\%>
\'i*;
" _, f
-.' ,
f* f *f
»"'*','
If/;
CO
UJ
0
CO
o>
rj
IT)
/'/,-.-'
'?' *
"^/'\
' '''* \
' 4 &
'( * *<"'
'^
*":'
-------
I
I1
oo
FREQUENCY
.036
.032
.028
.024
.020
.016
.012
nno
.Uuo
.004
CO
g
H
0
cc
CM
O
O
in
LU
0
in
°
oc
LU
OC
in
to
?-!!
$ , ^
's'*-.
?'&f.
'*? <' '
&
^;v
f x
''i;
,'',' f
f, '
'" ' ,
" fff
' "S
,/-,,''
1-
z
o
QC
in
'''*''*"'
V/^;;
'{>$!?.
;£"'"
;-i?
'''' , *
" ff 5 '
^ '',
' ''' j
^ * '
' '* ' ,
ff' f'ff
?'$' ''*'
'f',t "''
CO
LU
a
CO
CD
r-
>^
-;;J
'"&'
f f ;%
"' '' ',
'; ' ^
f, "
* j
'?;,''/'
''" t
* ,*
j " t /
'*"'.
cc
LU
OC
q
H
0
QC
CM
q
CO
LU
O
in
at
CM
q
oc
LU
CC
U)
in
' '/f '
'~ , -, !.,
* f ''-.
''';
' ;- ' ;
",^
? /'
,i'
fft f
'*
, "f
'
f . '
>>-,
t-
O
DC
in
S'
:'C '
p't
< ""
^ ^
/, Vi
' '':-;:
;AI
.- ..* .
*" '; *' f "
* '
11 ' '**.-
"* _,
> * ^ ^
/^'
CO
LU
D
CO
o
s
't'f^
/ ^^ ' >.
*>v<
V-''
'*', ''
f'ff'
' O'
^ /
' '*
" "'
:^ /
' f'°-t
QC
LU
OC
LEQ(ID)
90
80
'70
60
40
30
20
10
FREQUENCY LEQ(ID) FREQUENCY LEQ(ID)
HIGH TRAFFIC IMPACT LOW TRAFFIC IMPACT
Figure 4-6D. Variations in Construction Contributions, by Side of Unit
-------
FREQUENCY
.027
.024
.021
.018
.015
.009
nnfi
,ULJD ^^""
.003
« « «
CM
8
H
o
CC
(O
8
in
LU
Q
in
CM
q
CC
LU
CC
en
in
f?%f '''*}'
:Vi.^ f!'
',J'""/
> ^
";"f %
'-<
/T^
t~',"
*" .
/','y
? f'fiff
K
o
CC
CO
*
,-^y,-
'>,'<",
''*£/;
\f~"-'f'>
'friy.
Crfi Z **,
t '"
'tfi
'"V
^1
'**"':-
' *
in
LU
0
CO
in
tn
i4 ??
"':''
%'v<^^
*> r "
S',
"&jl.
*f f "
v/. ^
'H'' f
f'f" '^
r*/f -'
'A
% >
N'X V
'?A;
"/»',
',/'''
f \ 'v
....?.f :
t-
0
CC
p
if)
.*<,' '
''A
', , '»
J '''-<''
"> "^ -.
'"",,? '
'^ *
~" ""' s
/"^ /
P" s''V
"^\_
,- ^
\J''
^ 'x
CO
LU
g.
CO
O!
LO
^ '"j
, '^i
,'"''^
* ''
'ft
\^. ^
''-i s
' -V n'
_f f " f ,
XJ
'sc.
i '-,''
\ s
. v
CC
LU
CC
LEQ(ID)
90
80
70
60
/i A
4O
30
10
FREQUENCY LEQ(ID)
HIGH TRAFFIC IMPACT
FREQUENCY LEQ(ID)
LOW TRAFFIC IMPACT
Figure 4-6E. Variations in Dog Contributions, by Side of Unit
-------
I
N3
O
FREQUENCY
.108
.096
.084
072
.060
.048
.036
.024
.012
000
p
O
cc
i
in
UJ
O
CO
in
r*
p
cc
UJ
oc
CN
in
- '.'?'
./c<-
, 'W£,
*''< ?
'?/
,^'-'1
"" J\S'ff *:
f Sf'
->'-£ *';
' fff s *
^/i/fs
U'/V"
f& fff
O
CC
in
in
f ,.f "'£
fjf "''
' >' ,
\ ^
""' t
'"£}'
'-;£
/''V
',';'
: '»'
'%"'':
f£tfr
to
UJ
Q
CO
,_
in
\ f^ /
'' '" 'll
-',' ^
f *"'
'"' '',;<
\ ':';.?.',.;
l'r;"J"'
" $ '-.
, "f ,
«*'* ,
f's/ ''
'' ''"'.
'- ',''",
% ' ' s>.
cc
Ul
cc
B
*
CM
O
H
O
CC
°
in
UJ
0
in
%
p
cc
UJ
cc
CO
$
"\""";s$
> t *\
''~A\ '
'^'f I
'-"''
"\' ' ''
'*'''':
'' ', '
7''-
/ ^-"
' -.-.^
' V '
ff
o
cc
^1
*
"/''''',
>-,V
f " f
,"' ' ,
* , ' '
v ;'' '
/ f '
v-% ^ '
'x
f '
Ul
O
to
CO
O3
;%*
/-^'
\r *
c-^'
<". *
V
' ' 'V
^ 'f -."'
%
X
ff-
CC
Ul
cc
LEQ(IO)
90
80
70
60
50
40
30
20
10
FREQUENCY LEQ(ID)
HIGH TRAFFIC IMPACT
FREQUENCY LEQ(ID)
LOW TRAFFIC IMPACT
Figure 4-6F. Variations in Bird Contributions, by Side of Unit
-------
I
N>
FREQUENCY
.900
.800
.700
.600
.500
.300
f\f\n
.100
in rxj
s
CO
CO "~
g
CO
't^'d
' ,.i, "':
, ' - v* I
'/ *
f ff ff
y
j"f'' ( f*
f f f
f * f
/ "/
'*' f "*f*
f
*',<,;-
' ^% "*
*/,''' -
f\--"
T;
,,v
\,..,,
'*\r
"""/;*
< fJ^f '
yf
* "f ftf>
*'$ ,
' '''','
s \ ff "'
j'', '/
-.A
^ '?'
' *.
S
<- *^
*
l-f'^
^ -'^
''t -I
s x ":
:^' *,
fffff
"^ ?<
'', ^
; ^'
>- 0 I- >- 0 f-
$ - S < Z I
Q ^ & O ^ O
S * % * -
in ...
"
s
^
>
0
o
CM
o
z
z
UJ
2
l-
o
Z
id ^-
in f«
\''"f,f.
?<&
P
>£:;
s ff
^i"'':
,3 f>
" "<"'
' ^. .,,'
&£
',}$
^ f.
',* Vy
" ' * V* '
"V^>'J
:?>-'
fJJ;
W"-
^ x
-?V"'''
\f f f
'"' -''
'j /'
' '''-
^ % ^J
' < f
'> /'
^ .V /*
(O
CD
in
:?/&'
*0,
', '/>;-
' -y x '
''"''s ^"*
* />'!
% -^ '^-
i 'iv
_. ^AJ-
;,; ^
'*?' *
''
*
& >
LEQ(ID)
90
80
70
60
50
At\
30
10
>- O 1-
oil
UJ Z
>
FREQUENCY LEQ(ID)
HIGH TRAFFIC IMPACT
FREQUENCY LEQ(ID)
LOW TRAFFIC IMPACT
Figure 4-7A. Variations in Automobile Contributions, by Time of Day
-------
FREQUENCY
.090
.080
.070
ncri
.UDU ""
.050
.040
.030
.020
.010
.000 L
s
q
O)
£ S «>.
o
Q
tj
in
q
"V«V,"
*p5y
'^'.P f.^1,
irji
_? _." x
Y"- ^
^ ' ' :
'
: ^
/y
;/>%
. -.4^ .
X t *
rf'S
".' '
, ,; f
'f'
--'\
?$>v.
^ ''
fV ^f
''*!'",
''..t. f-.
W-
* f%f ""
;'"-"
' ''"
~-
'V/'
5J;
/*-",
'* '^
^ ^^;--'.
'"/*""
"'{
f ' '
'' / '
'£ ",
(O
'f;^
/".. "^
f ;
"* f '' f
/
" j
\f '
'
_.
ss "*
A''
^
// f f
* ^ ,
'' -',!
/?^"
o
s
-
in
o
^" C3 h >-CDl *>- rn
* 1 I z z z i 1
uj -^ ~ Z
> LU 2 UJ
UJ -Jj >
(0
5
H-
O
Z
" r?
Li ^*
.T:"",
^
:l",'''
o-s
v ^ '"
^ f /
, 'r
^ *
" /'
%'^ '
',
"',
*' ff ,' '
j* ''''
f->C:
/ ' ^* '
&V"
^ X- ^ '
',%
V"
;J-
x\'
'v ^
''V
""^1- '"'
"
. _,
;- ;
o
Si
c^'
"*X -, "^v.
/ '\ ""'
/
* '
J> W
%%;
:''vh
%
fff
^ ff
f
..,-'
'»
LEQ(ID)
90
80
70
60
50
40
30
20
10
1 . 1 '. F 1 1 0
>- 0 h-
< Z I
Q 2 u
UJ Z
>
FREQUENCY LEQIIDI
HIGH TRAFFIC IMPACT
FREQUENCY LEQ(ID)
LOW TRAFFIC IMPACT
Figure 4-7B. Variations in Truck Contributions, by Time of
Day
-------
4>
I
FREQUENCY
.016
.014
/\* f\
.008
.006
.004
.002
.000
00
p
p
-
s
a- co
CO (O
co to
5
"v!x.
A:';;,
L, ..j*^'
'4^' '
41i
"?'"
'",
V ' '
<. '
';''
''/""
-
i. '' '
;xf
- % t
f * f
'''/''*<
?%. ? e
'"'j/
f ;
/ --*
/'
x '
/"'
i
/ '
a (N
S °
'*> ;
- O i- >- o l-
< Z * < 2 I < 2 I
' a 2 _ Q^O Q^O
UJ 2 UJ 2 UJ 2
UJ LU UJ
0
Cj ^.' iri
"v? *;
*'f '$
»1"
'* X %
4, '
'* ff
''" '
-«-"
^ >'
'"V
, ' J,
' ''«'* "
'"??
\
f f ?
_,*"
.,
'
, ' '
' s
in
*-/;
»w
-;V;
^ '
, V
'^
% ^
' '
LEQ(IO)
90
80
70
60
50
40
30
20
10
S s S
1 1
UJ
FREQUENCY LEQ(ID)
HIGH TRAFFIC IMPACT
FREQUENCY LEQ(ID)
LOW TRAFFIC IMPACT
Figure 4-7C. Variations in Motorcycle Contributions, by Time of Day
-------
FREQUENCY
.018
.016
.014
.012
.010
.008
.006
.004
.002
o
q
£
. Q
5
in
(£>
i §
S'l
*-"^*'
'<<'", "A-
V '; *£
'">', '"'\
:l?£,.
''.&''
'*" ^ < '"'V''
^?
'/<;. 1^
^S
' '''X
,
-
in
§ 1
3ft
&!'!*#
IB
?/{i
j.'^;'
\&*
'K
/;££"'
n
"^ f ''£
LEQ(ID)
90
80
70
60
50
40
30
20
10
Ol-' >- ' ID ' h- ' >- 0 H- ' >- O H "
ZI < Z I < Z I
-------
N)
Ul
FREQUENCY
.036
.032
.028
.020
.016
.012
.008
CN
O
O
>-
Q
§
C3
Z
UJ
UJ
1
Z
FREQUENCY
tn
en
in
vf ?
f,f
' '"' \
'*'*'-,' ''
' jf % " |
(0
in
''>:
*
,>:
''!"
oi
in
?..!
"/
,/'"
ff ff
s
Csl
q
s
p
O)
p
CM -.:
tn u,
^ *
:^'»
V
'*' f '
< s;;
'V"
," "
'X
P
rf"'
:,;.-
* ^ %
' V
i, '"
LEQ(ID)
90
80
70
50
40
30
20
10
> O J- >- C3 t ~> O t
< Z I
-------
I
ho
FREQUENCY
.090
.080
.070
.060
050
.040
.020
.010
f.
o
>-
<
0
CO
q
O
Z
LU
>
o>
CM
q
O
Z
S
'.,'-.' ?;.'.
^f"!
"_.
':t;£
,V^
~>ff ':
','., '}
"&/.
f ''*"/'';
'?'*'..%
i^]-.
f ffftAf
"> '' ^f''
i'fy*^
ff"$fy
*'$£
'j 'f''>'-s
>
0
in
1
'',,-,
fff Y'r
f ' ^ ' '
: ',
'\' '' *
"""
,
\ >"'
"i'-t '
/ *J/,
y
**?&%
"" *"""&'
f f fff f
ff ff /
'y/15 "
fH'
r^i/ 'V
o
z
LU
en
r-'
"*
?-!'' s/
^ ^ ' -1
-' ^
:',' '^
_,
S\'^
* ' /
>" i
U?'' ''
iL3?
^f ^ "~tt f
\ f S-S
'* ' \
£.' "\
t, '/'','
&J /':
I
O
z
CO
q
>-
Q
CM
CM
q
O
z
LU
CM
o
I
O
Z
00
»
' '"?',
A* '_,
,'-'^
'<'^ '
^ '/'
*X '> '
^> %
^ ''
( ^'^-
f
f> ,""
ff *
* > s'
jC "
',
0
CO
od
''->' ,'
'^ " '
! ' "[
, /*
'^
: ' «* ^
: >/
; ',",*
':, ^*
'&>'
f f
f "
; ' v ^
'<
-';',''-
'
O
z
2;
LU
in
5
.- ^
r, ,^
f ,
f^f ^
;/
fff^-'
' ^ '?'''
? ','
/^ % *"
f ** '%
^ '
*
-------
As expected, construction noise is usually present only during the daytime.
Dogs are very rarely heard in high-traffic-impact areas. In low-traffic-
impact areas, they are heard equally frequently throughout the day, though their
associated noise levels decrease during the nighttime.
Birds are somewhat louder in the daytime hours. Their frequency of
identification in high-traffic-impact areas is greatest in the nighttime, while
in low-traffic areas it is greatest in the daytime. This difference results
from the stronger tendency of traffic in high-traffic-impact areas to drown out
bird sounds during the daytime. Thus, the prominence of birds in high-traffic-
impact areas is basically a function of the prominence of other competing
sources, while in low-traffic-impact areas the temporal patterns of the birds
themselves are more important.
From a policy perspective, these results suggest that roadway traffic
abatement measures will have their greatest acoustical effects in the daytime
and evening. This consideration must be balanced against the greater noise
impacts which occur in the nighttime as predicted by the 10-dB weighting
factor used in EPA criteria. With this factor included, roadway traffic during
the nighttime would appear to be the most important contributor to adverse
noise impacts. The penalty for nighttime sources is thus of critical importance
in assessing the relative benefits of roadway traffic noise abatement measures
over the temporal domain.
4-8 EFFECT OF URBAN AREA SIZE, POPULATION DENSITY, AND TRAFFIC IMPACT
4-8-1 Effect on Ldn
Table 4-4 shows summary statistics describing the Ldn measurement results
in the 12 sampling cells described in chapter 3. The results indicate that
average Ldn values are above the 55-dB threshold of impact in all 12 cells.
Standard deviations of these levels are approximately 4 dB, suggesting that
4-27
-------
Table 4-4. Summary Statistics of Ldn Measurement Results,
by Sampling Cell
Traffic
Impact
High
High
High
High
High
High
Low
Low
Low
Low
Low
Low
UA
Size
Large
Large
Medium-
Small
Medium-
Small
Medium- .
Small
Medium-
Small
Large
Large
Medium-
Small
Medium-
Small
Medium-
Small
Medium-
Small
UZ
Density
High
Medium-
High
High
Medium-
High
Medium-
Low
Low
High
Medium-
High
High
Medium-
High
Medium-
Low
Low
n
5
4
5
5
4
4
19
11
13
21
26
8
Mean
Ldn
69.2
66.8
67.6
66.9
65.2
66.0
62.4
59.0
60.6
58.3
58.2
57.4
Std.
Dev.
4.4
4.3
2.1
3.0
2.5
3.3
4.0
3.3
4.5
4.0
4.3
2.3
Error of
Mean
2.0
2.1
0.9
1.3
1.2
1.7
0.9
0.9
1.2
0.9
0.8
0.9
4-28
-------
66 percent of the population within each cell are exposed to a residential Ldn
within 4 dB of the average. Ninety percent are expected to be exposed to within
8 dB of the average Ldn.
The.data reveal that the average residential noise exposure is sufficient
to generate some adverse impact in all types of urban areas. While pockets of
relative quiet certainly exist on the urban landscape, such areas must be viewed
as exceptional even in low-density areas.
Inspection of table 4-4 indicates that traffic impact has a pronounced
effect on residential noise levels, with the average Ldn in high-traffic-impact
areas between 6 dB and 9 dB greater than in low-traffic-impact areas.
Analysis of the variation in Ldn over urban area size and urban zone
density reveals that density is, the other significant factor. Its effect is
best represented when the logarithm of the actual density is used as the inde-
pendent variable instead of the density category used to define the sampling
cell structure. Urban area size is a significant factor only insofar as it
correlates with urban zone density:' noise levels in large urban areas are
higher as a result of the greater population densities of these areas.
A regression analysis of the results indicates that a simple and reasonably
accurate prediction of Ldn is given by:
Ldn = 42.3 + 4.7 x (log 10 UZ density) + 7.9 x (Traffic)
where UZ density is the urban zone population density, and,
Traffic = 1 for high-traffic-impact
= 0 for low-traffic-impact
The standard deviation of the data from the regression line is 3.7 dB. Varia-
tion in the independent variables of traffic impact and Ldn accounts for approxi-
mately half of the variation in Ldn found in the survey.
4-29
-------
The regression formula and its standard deviation may be used to predict
population distributions over Ldn in areas of varying traffic impact and popu- -
lation density. Table 4-5 indicates the predicted percentages of population
exposed over certain values of Ldn as a function of traffic impact and popula-
tion density. Examples of urban zones of various densities are also included.
Table 4-5 shows that virtually the entire population exposed to Ldn over
75 dB in their residential environment resides in high-traffic-impact areas
(except, of course, for those so exposed as a result of aircraft noise). This
is important because 75 dB represents "full" noise impact, and represents the
threshold at which auditory damage may begin to occur as a result of residen-
tial exposure. Attempts to protect this highly impacted population should thus
focus almost entirely on areas within 100 feet of arterials or 300 feet of
freeways.
In conclusion, variation in Ldn'as a function of traffic impact, urban
area size, and urban zone density is significant, but rarely extends below the
55-dB threshold of impact. Sources of the variation include traffic impact and
urban zone density, with the former having a more pronounced effect. The small
population that is fully impacted according to EPA criteria resides almost
exclusively in high-traffic-irapact areas.
4-8-2 Effect on Source Contributions
Figures 4-8A through 4-8L show average frequency and Leq (id) values
obtained for each source in each of the 12 sampling cells. These values are
based on measurements taken at the front only.
Source contributions were found to vary greatly from site to site. The
average values indicated in figures 4-8A through 4-8L must, therefore, not be
4-30
-------
Ldn
Table 4-5. Cumulative Population Distributions Over Ldn by Traffic Impact and
Urban Zone Density
Traffic
Impact
Percent of Population Exposed
55
60
65
70
75
High
Low
High
Low
High
Low
High
Low
High
Low
99%
65
87
17
42
1
6
0
0
0
100%
78
94
28
58
3
12
0
1
0
100%
86
97
39
69
5
20
0
1
0
100%
93
98
56
83
12
34
1
4
0
100%
97
100
72
91
22
50
2
9
0
100%
98
100
83
96
35
65
4
17
0
-p-
Population Density
(thousands per
square mile)
Example Urban
Zones
'
t
L ;
, i
> 4
> '
J
, t
J 1
> . t
6
, t
32
1 »-
1
*
Leominister,
MA
Newport
News, VA
Outside
Central
City
Washington,
DC
Lancaster,
PA
Chicago,
IL
New York
City
-------
3 .500.
?
£
.400-
300-
.200-
.100-
1
,
«,
|
6
'
- 1 § S
1 I 3 3
0 ° 0 6
1 1 1
1 1 1
a 2 s i a
S o « > §
< s s 3 i
at m
S I <
UI w S
1 3
J_§
""^- 1
s 2
U -i
3 w
s S
p.
1 1 1
o o o
000
r... .. r . r- ,
1 1 E
o
at
<
I "
a 6
2 a
I f
i i
£
s
d
a
0.
p
3
6
tf>
U
o
V VEH
3
«
|
1U
00
<
i
P
z
Ul
R S
I !
o a
s §
at o
_i y>
5" Q
S
1
1
ei
at
(U
Z
1-
c
i I
3
? ">oH
s_s
s
a-SSss
is
aaas-S
=
Figure 4-8A. Source Contribution Profile - Large, High-Density,
High-Traffic-Irapact Areas
4-32
-------
1 I
100.0-
90.0-
80.0-
70.0-
60.0-
_
a
| 50.0-
^
40.0-
30.0-
20.0-
10.0
'
a
^
S'
f
I
a
s
5
1
UNIDENTIFIED TR
^
i
o
VI
2
£
uu
z
w»
R
a
at
i
O
3
I
i
a
1
I
g
|
Mt
9
o
MOTORC
j"
5
3
EMERGENCY VEH
s
s
1
3
UNIDENTIF
3
s
1
§
5
i
^
a
a
"
VI
at
Ul
t-
a.
HELICOI
3
*i
S
5
ADULT V
§
'" ~~
i
RESIDUAL 1
§
2
WORK
HOME YARD
a
§
3
Q
Z
3
3
AMPLIFIED S
1 8
1
1
CHILDREN'S V
9
6
tt
Ul
O
Figure 4-8B. Source Contribution Profile - Large, Medium-High-Density,
High-Traffic-Irapact Areas
4-33
-------
.700- °
.600-
3 .500-
?
U,
.400-
.300-
.200-
.100-
§
AUTOS 1
_.
O
§ 5 °
6 3 S " I s S! s s 3 V 3 8 \
~" 'i 3 e » r- 3 S $ £ 5 * 3 5 '
= ^__| 555sllli38i
(/I Q {_> J ' f *" MQOV IU (fl
n 1 ! i 5 § i ! i !
" 1 1 i n I'll
H a s * « ° z
z S i = j i
s < =
. 1
^ u
i s 8 s a 8
1 I i s § I
n n 5 f
1 ' 3
u >
Q uj
u 3
I l
80.0 -
70.0 -
60.0
a
T 30.0 -
40.0 -
JO.O -
200-
1 0.0 -
s §
S *»
3 1
1
O $
s
«
3
5
3
i 1
3
~
^^
--,,
i
** ^
i
--
e
~
|
3 ^
2
9 °
-i ii
X
^^^
f
1 8 § M 3 8 i § I I S I | § 5 i i s 3
3 1 I 1 2 S « S § > i 1 ° 1 2 ! 5 §
aoJo-i B«zz * o 2
1 1 i I ! = If i I «
'I s 1 1 § 2 1
O U U J
I SI
Figure 4-8C. Source Contribution Profile - Medium-Small, High-Density,
High-Traffic-Impact Areas
4-34 .
-------
.700
.«00
I I
111
caps?
I I i I I
;
* s E
S 8 8
u M .4 o « o
z S
2
z
3
A i
| 1
I i
s
tu =
o S
?
i
60.0-
§
=f 30.0-
.J
40.0 -
30.0-
20.0-
100-
X
3
1 "
i
?
-
5
^
J
g
>^
^~
is i
ss a
1
i
s
i ^
5
| ! g
1 *-
1
£
1
s
a .
3
i
1
6
SUci^Qtno wiviMviviujviHzav1 d
?MUI^jUJ< QuJUJUJtU^UZOZ^" ^
i| n i |i s s » 1 3 ; 1 1 § s ?
1 ill' i i § i f i
t 2 ^ 3* zHiuS
Figure 4-8D.
Source Contribution Profile - Medium-Small, Medium-High-Density,
High-Traffic-Irapact Areas
4-35
-------
-------
§ J"H
E s *
o 6 6
ooo
V) (ft
-------
.400- £
§
6
.300-
.200-
100-
NTIFIED TRAFFIC |
8
Z
|
3
S <*
i i §
AUTOS
JETS
ADULT VOICES
RESIDUAL LEVEL
9. f
S E S S 5 3 .
I i 5 5 a 1 3
S ° § S S § a
8 j < I I 1 §
I 2 1 a s * I
- 1
1
IILDREN-S VOICES y
S
CONSTRUCTION 11
LMPLIFIED SOUND N
.DOCS (]
^
Z s
I i
0 A
MOTORCYCLES 1]
UNIDENTIFIABLE |'
§
OME YARD WORK 1'
Z
§ I
ERED EQUIPMENT I'
DENCY VEHICLES 1 '
i i
7 Ul
§
FACTORIES |<
|
HELICOPTERS l<
i f
-ROAD VEHICLES '
OTHER <
S.
60.0-
3
| so.o-
40.0-
30.0-
20.0-
10.0-
I
i
a
a
41.044)
|
S
f*
a
i
|
s
§.
s
s
-o
JIM «|
1
i
s
si
i
s
a
i
i.
o
a g E e s a i 1 s § i § a i I jj a * i i g § s g i
2 3 - i a ! 2 i a a * ! s § s s g 2 ! 2 i i § I i
i- ,-H'-< .g
-------
a 2
I 5
S 8
.
- I
g g
I 6
OK
UJ
S
50.0 -
40.0 -
JO.O -
20.0 -
10.0 -
5
s
DENTIFIEU TRAFFIC
Z
S
AUTOS
_
Z
RESIDUAL LEVEL
5
S
Q
DC
a
3
ADULT VOICES
i
s
a
I
S
S
FACTORIES
a
s
HOME YARD WORK
2
5
CHILDREN'S VOICES
3
|-Z-.
5
§
PLANES
X
RAILROAD
s
?
D
Z
5
«
Si
AMPLIFIED SOUND
S
s
1
5
§
a
~
TRUCKS
S
«
-^
CONSTRUCTION
9
Z
&
ERGENCV VEHICLES
S
s
"*.
3
MOTORCYCLES
g
s
UJ
S
Z
BUSES
|
s
WERED EQUIPMENT
i
^
~
a
V)
at
UJ
U
Z
f.
j
«
OTHER ANIMALS
g
6
OTHER
Figure 4-8H. Source Contribution Profile
Low-Traffic-Impact Areas
- Large, Medium-High-Density,
4-39
-------
s
e
3 .sa>4
e
.200-
2 g
.100- 1 §
.000 1 1 1 '
u a en *
E § 2 3
2 < s I
i I
C ui
i i
1 x
z
O.OW927
0.0410)1
1 004411)
o o J
i 1 §
i i
£
Ul
at
0.0)829)
i
<
0.0)1)21
0026807
W1 VI
(U H
U uj
ADULT
002)610
LROAD 1
0.0172)]
0012184
\ I
POWERED EQU
5 i
1 1
UCTION 1
CYCLES
CONSTR
MOTOR
0.009820
0.008940
0005 |
NIMALS
OTHER A
S
1
VOICES
CHILDREN'S
§ §
1 §
f IAB1.E
BUSES 1
UNIOENT
»
i
SOUND 1
AMPLIFIED
III
I 3 i
1 H s
y >
i s
UJ
O
ec
UJ
2
s I
s 1
I
'
I
s
Figure 4-81. Source Contribution Profile - Medium-Small, High-Density,
Low-Traffic-Impact Areas
4-40
-------
s
£
53
"II I I l l i
s
is
ii
! !
1
S
I
IDENTIFIED TRAFFIC
""
~^^
AUTOS
S
RESIDUAL LEVEL
J3
S
i |
1
11
a
z
5
"
TRUCKS
S
OTHER ANIMALS
Z
J
*
ADULT VOICES
*
"
PLANES
S
j
^
HOME YARD WORK
J
r
S
5
*,
i^^
MOTORCYCLES
-t
J
*
I
5
aWERED EQUIPMENT
3
^
p i
CHILDREN'S VOICES
y
^
HOUSEHOLD
.
^
RAILROAD
J^
OFF-ROAD VEHICLES
T
AMPLIFIED SOUND
a
^
FACTORIES
.
-
i
3
*
'
UNIDENTIFIABLE
i
4ERCENCY VEHICLES
S
CONSTRUCTION
^
5
HELICOPTERS
WATERCRAFT
Figure 4-8J. Source Contribution Profile - Medium-Small, Medium-High-Density,
Low-Traffic-Impact Areas
4-41
-------
5K £ 2
?* r- 5
o d p____ o' ei b 6
! I 5
at S Z U U
= i 5 § =
Q a. > t-
5
1
VO
s 2
100.0 .
90.0-
10.0-
70.0-
60.0-
f
? 50.0-
40.0 .
300-
20.0-
10.0-
i
s
s' g ' . a 1 .1
45.6244
~
46.6705
3
S
**
1
I I
s
P 1
s s
s
irt
S
I
7
=
S a 1
5 =
§
1 "
i
i
a
^2OO'2<'T^ i5^i0°§!°
!T 3 << < « i- CaQ"; mis
§ i 1 ' 5 ! . 1 ! Ms
I s 8 | i i | s g
Figure 4-8K. Source Contribution Profile - Medium-Small, Medium-Low-Density,
Low-Traffic-Impact Areas
4-42
-------
| .500-
?
u.
.400-
.MO.
.200-
.100
.
|
b
DENTIFIED TRAFFIC 1
Z
S
S
2
6 2 «
. m f « *
| 1 | |
|
g i 3 3 1
| 2 z
i S
5
f» «
i 1 1
~1~T~
RESIDUAL LEVEL
BIRDS
RAILROAD
1 iM i
§ S o s
I
PLANES
HOME YARD WORK
IX HiS
ADULT VOICES
S 5 I 1
§ § i 1
JETS
UNIDENTIFIABLE
MOTORCYCLES
AMPLIFIED SOUND
5
ERGENCY VEHICLES 1
£
LU
..
1 I *'
bob
BUSES
CHILDREN'S VOICES
CONSTRUCTION
|
b
IWERED EQUIPMENT
£
|
b
OTHER 1
60.0-
50.0-
40.0 -
10.0.
200-
10.0-
s ;
5 . S|5S | a .
|
*
K
I s
;
i
a
r*
O
S
M
1 49. 1524
S
s
1
§
s
f
5
r-=-
5 I
7
S
gt
Lfl 3842
!
?
s
_ -
1 '
6
||s|||l|y||||||||||o||||||.
Figure 4-8L.
Source Contribution Profile - Medium-Small, Low-Density,
Low-Traffic-Impact Areas
4-43
-------
taken as predictors of these contributions at any particular site. Local
factors not considered in the survey sampling cell structure would have to be
considered in the development of such predictors. Nonetheless, .these results
justify some tentative conclusions regarding the variation in source contri-
butions that does result from variation in the cell structure parameters.
The dominance of roadway traffic sources is evident throughout these
results. As expected, this dominance is less pronounced in low-traffic-impact
areas. Such areas are seen to have a generally more diverse noise environment.
Autos, trucks, motorcycles, construction, dogs, and birds were selected
for analysis of variation with respect to traffic impact, urban area size, and
population density. It was found that frequencies of identification are
strongly affected by traffic impact, but not by population density. Leq (id)
values are generally affected both by traffic impact and by population density.
As expected, autos, trucks, and motorcycles are identified more frequently
in high-traffic areas. Construction, dogs, and birds are identified less fre-
quently. This is shown in figure 4-9.
Figure 4-10 shows the regression lines obtained when, for each of the six
sources, the dependent variable Leq (id) is plotted against the log of the
population density. Density is seen to have more of an effect on roadway source
levels in low-traffic-impact areas. In the case of construction, density has a
pronounced effect regardless of traffic impact. For dogs and birds, the effect
is less pronounced, but again independent of traffic impact.
A similar analysis was also performed in which the dependent variables were
the aggregate contributions of four classes of sources: roadway traffic, air-
craft, other abatable sources, and nonabatable sources. The sources included
in these classes are shown in table 4-6, and the results of the regression
analysis in table 4-7. It was found that both traffic impact and urban zone
4-44
-------
>
o
z
01
a
UJ
cc
.600-
.500
.000
TRAFFIC IMPACT
SOURCE
Hi(|h Low High Low High Low High Low High Low High Low
Figure 4-9. Frequency of Identification for Selected Sources -
High- and Low-Traffic-Impact Areas
4-45
-------
High Traffic
70 Low Traffic
60
ffl
8" 50
40
1 24 8 16 32
Density (thousands/sq. mile) Autos
High Traffic ^
70-, Low Traffic ___ "
60
50
40
2 4 8 16
Motorcycles
32
High Traffic
70 -, Low Traffic
60
50
40
2 4
Dogs
8 16 32
70
60
50
40
High Traffic
Low Traffic
1248
Trucks
High Traffic
70 ^ Low Traffic
60
S
40
12 4 8
Constuction
High Traffic
70 Low Traffic
60-
50
40
16 32
16 32
248
Birds
16 32
Figure 4-10. Regression Lines - Leq (id) vs. Log 10 (UZ density) -
For Selected Sources, High- and Low-Traffic-Irapact Areas
4-46
-------
Table 4-6. Source Classes
Roadway Traffic
Aircraft
Other
Abatable
Sources
Nonabatable
Sources
Autos
Trucks
Buses
Motorcycles
Emergency Vehicles
Unidentified Traffic
Jets
Planes
Helicopters
Construction
Factories
Off-road Vehicles
Water Vehicles
Amplified Sound
Home Yard Work .
Railroads
Household
Nonpowered Equipment
Residual
Dogs
Birds
Other Animals
Wind
Child Voice
Adult Voice
Unidentifiable
4-47
-------
Table 4-7. Summary of Regression Results - Source Class Contributions
High Traffic Low Traffic
Source
Type
Roadway
Traffic
5" Aircraft
a
<1)
g. Other
2 Abatable
^ Sources
Non-
js abatable
js Sources
00
Density
Intercept Coefficient
.57 .07
.013 .00005
.21 -0.04
.20 -0.02
Probability
of
Significance
0.32
0.04
0.82
0.29
Density
R Intercept Coefficient
.039 0.16 0.10
.000085 -0.06 0.03
.067721 .21 -0.03
.005 0.68 -0.10
Probability
of
Significance
.89
.93
.70
.96
R2
.027
.036
.012
.031
Roadway
Traffic
^ Aircraft
o
11
^ Other
of Abatable
1-3 Sources
Non-
abatable
Sources
61.0 .72
14.7 12.6
27.7 7.9
18.8 8.9
0.22
0.99
0.84
0.97
.003 33.9 5.44
.26 38.0 4.7
.11 26.3 7.0
.16 25.6 6.9
.99
.99
.99
.99
.13
.075
.14
.200
Predicted value = Intercept + (Density Coefficient) x log 10 (urban zone density)
Probability of Significance - The probability, based on the observed data, that urban
zone density is statistically significant.
2
R - The amount of variation in the dependent variable (Frequency or Leq (id)) which is
accounted for by variation in the independent variable (log 10 urban zone density).
-------
density are significant factors to frequencies of identification of these
classes, especially in low-traffic-impact areas. Roadway traffic and aircraft
source identifications increase with frequency, while other abatable sources
and nonabatable sources identifications decrease. Leq (id) values of all
classes increase with density in both high- and low-traffic-impact areas, with
the exception of roadway traffic in high-traffic-impact areas, where no sig-
nificant relationship is found.
In conclusion, traffic impact and population density significantly affect
the source composition of the residential noise environment. Traffic impact is
the most important factor in determining the frequencies of identification of
noise sources. Population density is also somewhat significant to these
frequencies when classes of sources are considered. Both density and traffic
impact are significant in determining noise levels associated with either
individual sources or classes of sources; these levels increase with popula-
tion density.
4-9 SUMMARY OF RESULTS
Noise pollution is a problem to which the vast majority of the urban popu-
lation is subject. It is most of all a problem in high density areas, and in
areas located near major roadways, but remains considerable in areas with
neither of these characteristics.
Noise levels vary significantly both temporally and locationally. The
highest levels occur in the daytime but, if the EPA 10-dB weighting factors for
nighttime noise is considered, the greatest impacts occur at the beginning and
end of the nighttime. Higher levels also occur at the front of residential
units. To the extent that noise control programs can be focused spatially and
temporally, these considerations should be used as a guide to such focusing.
4-49
-------
Roadway traffic is the most significant noise source, and should be the
primary target of noise control programs. In areas located near major roadways,
roadway sources should be targeted almost exclusively. In other areas, noise
from other sources is also significant, and becomes more so with greater popu-
lation density. In areas not near major roadways, the severity of noise is
associated with the number of significant noise sources, and thus the degree of
the noise problem with the complexity of its solution.
4-50
-------
APPENDIX A
METHODOLOGY SUPPLEMENT
A-l INTRODUCTION
The site selection procedures, measurement protocol, and analytical
procedures used in the National Ambient Noise Survey were summarized in
chapter 3. More detailed information on these topics is presented here.
A-2 SITE SELECTION
The urban population, defined for the purpose of the survey as the popula-
*
tion residing within urbanized areas, was divided into 12 subpopulations on
the basis of urban area size, urban zone population density, and traffic
impact. In the selection of measurement sites, the objective was to obtain
random samples of residential units of these subpopulations. This was a five-
step, process. First, a quota of sites was established for each cell. Second,
appropriate numbers of large and medium-small urban areas, and urban zones
within them, were randomly selected. Third, census tracts were randomly
selected within each selected urban zone. Fourth, blocks were randomly
selected within each census tract. Finally, residential units, either high- or
low-traffic impact units as required, were randomly selected with'in each block.
Within this basic site selection framework, a fair amount of procedural
flexibility was considered appropriate to the pilot-like nature of the study.
/
As a result, a number of changes in procedural detail were made between the
first and second years of the study. In recognition of the fact that the site
selection framework, rather than the particular procedures followed within that
framework, is more important, the site selection process will be described on a
step-by-step basis with year-to-year differences noted for each step.
A-l
-------
A-2-1 Establishing Site Quotas
The use of measurement resources allows considerable flexibility in the
allocation of measurement resources. (This ability had already made it pos-
sible to systematically exclude high-aircraft-impact sites.) Two objectives
might be considered in exercising this flexibility. The first is to optimize
the sample distribution from the point of view of assessing the significance of
various parameters used in the cell structure to the variables of interest (e.g.,
determining the effect of urban area size on Ldn). The second is to optimize
the sample distribution from the standpoint of establishing a maximally accurate
f
statistical profile of the population as a whole (e.g., determining as accurately
as possible the average Ldn value to which residents are exposed). The first
objective is best served by distributing measurement sites evenly among the
sampling cells. The second objective suggests a distribution which is a func-
tion of the cell populations.
In the first year of the study, the focus was upon the first objective,
and the sites were therefore distributed evenly. Four sites were allocated to
each of the 12 sampling cells.
In the second year emphasis was given to achieving maximally accurate
national estimates. Thus, in the second'year, sites were apportioned according
to cell populations. The small populations of the high-traffic cells resulted
in a second-year allocation that emphasized much more strongly the low traffic
cells. The cell quotas, along with the numbers of sites actually obtained, are
given in table A-l.
A-2
-------
Table A-l. Site Quotas and Sites Obtained for Each Sampling Cell
1980
1981
Total
Urban Area
Size
Large
Large
Large
Large
Medium- Small
Medium-Small
Medium-Small
Medium-Small
Medium- Small
Medium-Small
Medium-Small
Medium- Small
Urban Zone
Density
High
High
Medium-High
Medium-High
High
High
Medium-High
Medium-High
Medium-Low
Medium- Low
Low
Low
Traffic
Impact
High
Low
High
Low
High
Low
High
Low
High
Low
High
Low
Quota
4
4
4
4
4
4
4
4
4
4
4
4
Obtained
4
4
4
4
4
4
3
3
3
4
3
3
Quota
1
16
1
11
1
11
1
21
1
26
1
6
Obtained
1
15
0
7
1
9
1
18
1
22
1
5
Quota
5
20
5
15
5
15
5
25
5
30
5
10
Obtained
5
19
4
11
5
13
4
21
4
26
4
8 '
Total
48
43
97
81
145
124
-------
A-2-2 Selecting Urbanized Areas and Urban Zones
The next objective was to select residential units to fill the cell quotas
established. To limit travel costs, the residential units were clustered in
certain randomly selected residential areas. This clustering was balanced
against the need to obtain nationally representative results.
In the first year of the study, the equal apportionment of sites among the
sampling cells suggested a straightforward approach to accomplishing the desired
clustering. This was to select initially eight urbanized areas, called primary
areas, four in large urbanized areas and four in medium-small urbanized areas.
0
One site in each of the four large urban area cells was selected in each primary
large urbanized area. When possible, the analogous procedure, this time involv-
ing the four medium-small urbanized area cells, was performed for the primary
medium-small urbanized areas. In fact, most medium-small urbanized areas do
not contain urban zones in each of the four categories of population density.
This necessitated two procedural refinements. First, urbanized areas close to
the primary areas, and having populations on the same side of 200,000 as the
primary areas, were selected to provide the sites not available in the primary
areas. Second, a 10-percent leeway was allowed when necessary: An urban zone
with a density as high as 3,300 persons per square mile or as low as 1,350 per-
sons per square mile could be counted as medium-low density and likewise for
the other density categories.
To further insure that the sample of urbanized areas was representative of
such areas in the United States, two additional constraints were imposed upon
the urbanized area selection. First, it was decided that one medium-small and
one large urbanized area be located in the west, south, northcentral, and
A-4
-------
northeast geographical regions of the United States. Second, for the medium-
small urbanized areas, it was determined that two of the selected areas would
have a population over 200,000 and two have a population under 200,000. With
these requirements taken into account, the urbanized areas were selected at
random from the list compiled by the 1970 census.
In the second year, urbanized area and urban zone selection procedures
were modified as the result of a number of considerations. These included the
unequal allocation of sites to sampling cells used in the second year; a belief
that further clustering was possible in urbanized areas and urban zones of
large population; and a desire to insure that the larger population centers in
each urbanized area size category, along with those regions of the country with
a greater urbanized population, have a more proportionate representation in the
sample. To accommodate these factors, the following changes were made:
a. An urban zone was allowed to contain at least two measurement
sites, with an additional two, sites for each million of popula-
tion (an urban zone with a population between one and two million
could have four sites).
b. The random selection process was conducted on a population-
weighted basis, so that the probability of a certain urbanized
area being selected was proportional to its population.
c. No geographical quotas, and no quotas concerning urbanized areas
with populations greater or less than 200,000, were set.
Large and medium-small urbanized areas were selected on a one-by-one basis fol-
lowing the random selection process described in (2), above. Each selected
area was made the locus of as many sites as was consistent with (1), above,
A-5
-------
subject to the limitation of the cell quotas specified in table A-l. This
process continued until all site quotas were achieved. At this point sites in
urban zones in which more than one site was located were transferred, when
possible, to urbanized zones.in the same density category in other urbanized
areas which had been selected after the quota for that category had been filled.
The refined urbanized area selection procedure used in the second year is
preferable from the standpoints of conceptual simplicity and the representive-
ness of the generated sample. There is no evidence, however, that the slight
biases inherent in the first-year procedure have any acoustical significance.
Thus, it is assumed that on a cell-by-cell basis, the data acquired in the first
year is completely comparable with the second.
A-2-3 Selecting Census Tracts and Blocks
According to the U.S. Census, "census tracts are small areas into which
large cities and metropolitan areas are divided for statistical purposes."
These tracts are in turn divided into blocks. A block is usually a well-
defined rectangular piece of land bounded by streets or roads. A typical tract
has a population of between 1,000 and 10,000; a typical block has a population
between 50 and 500. These divisions provided a convenient basis for carrying
the site-selection process from the urban zone level to a higher level of
specificity.
The United States Census Bureau publishes a set of census tract and block
statistics, along with a set of maps that, contain these divisions, for each
urbanized area. Figure A-l shows a typical page of block statistics. The outer-
most numbers correspond to census tracts, the numbers beneath them to blocks
within those tracts. Figure A-2 shows the census map of the area described in
figure A-l. The larger outlined'areas are census tracts; the smaller areas
within them are blocks. Urban zone boundaries are also identified on the maps.
A-6
-------
Table 2. Characteristics of Housing Units and Population, by Blocks: 1970-Con.
Fresno County, Calif.
(Oato exclude vacant seasonal and vacant migratory housing units, for minimum base for derived figures (percent, average, etc.) ond meaning of symbols, see lexll
ocks
filhin
ensus
acts
903
903
904
905
906
907
909
910
911
912
VI3
914
915
916
917
918
919
921
922
923
924
925
926
927
926
929
930
931
932
934
935
101
102
103
104
105
106
107
108
109
lolal
popu-
la-
tion
1584
22
8
25
51
47
26
24
21
42
26
106
e
76
64
58
45
64
76
156
72
96
13
9
19
20
55
38
59
36
32
65
49
74
7274
30
46
79
53
94
19
13
II
3
Percent of total population
In Un- 62
group der years
Ne- quar- 18 ond
gro ters years over
1 - 39 7
41 9
56 8
4 - 45 4
43 4
39
42 8
26 13
29 12
50
45 3
50 25
43 3
52 6
35 7
47 4
42 3
34 9
40 13
27 12
54
II
26 5
25 5
II S3 9
34 16
46 2
a 33 6
34 3
46 2
41 14
34 3
3 32 14
33 17
41 9
30 6
34 2
37 2
16 21
15 46
9 46
Year-round housing units
tack-
ing
some
or oil
plumb-
ing
facili-
Total ties
468 14
6
6
14
12
8 1
6
7
14
5
25
2
22 1
15
19 1
II
20 1
22
53 1
22
42 2
3
4
6 2
6
13 2
II
13
13
8
19 2
13
24 '
2461 29
9
II 1
28
14
24
10 1
7
4
1
Units in -
Struc-
tures
One- of
unil 10 or
struc- more
lures units
449
6
6
14
12
8
6
7
13
5
25
2J -
15
17
II
19
20
' 51
20
35
5
6
13
II
13
13
a
19
13
24
1720 120
9
II
28
14
24
8
7
Occupied housing units
Owner
lock
ing
some Aver-
or oil oge Aver-
plumb- num- oge
ing ber volue Per-
facili- of. (dol- cent
Tola) lies rooms lors) Negro
122 5 58 27100 1
5 - 60
5 - ,56
9 - 61 25000
10 - 54 21500
6 - 57 27100
6 - 57 ...
3
19 - 59 25500
18 i 56 26500
13 - 55 21600
15 1 64 23900
8 - 61 ...
15 - '54 21000
21 - 63 31500
38 1 56 27400
15 - 52 17100
20 1 53 18900
I
3
10 1 61 31900 10
8 68
13 68 44200
10 - 56 23400 10
5 - 60 ...
14 54 28000
9 - 6 1 31400
12 - 64 27300
1607 7 51 15700
5 54 22800
5 - 66 15000
13 - 49 11600
10 52 13500
16 54 13000
6 - ' 50 15300
5 68 18000
Renter
lock-
i.ig Aver-
some Aver- oge
or all oge con-
plumb- num- tract
ing ber rent Per-
focili- . of (dot- cent
Total lies rooms lars) Negro
111 2 45 73
1
1
3
2
1
1
a - si ...
5 - 50 ...
6 - 43 ...
1
2
2
2
3
4 ...
16 3 5 62
3
3
2
2
3
3 ...
4 ...
II - 4.7 103
609 13 43 96
4
5 - 52 79 -
15 - 38 63
4
8 50 99
3
2
1.01 or more
persons
per room
With
oil
plumb-
ing
tocili-
Tolol ties
47 45
1 1
1 1
1 1
1 1
2 2
2 2
6 6
3 2
3 3
2 2
1 1
4 4
2 2
2 1
3 3
1
1
1
3 3
168 168
1 1
4 4
3 3
1 1
4 4
With
room-
ers.
With boord-
One- female ers,
person head or
house- of lodg-
holds family ers
17 18 II
1 i I
i
i - i
i
i
1 3
2 1
9 3
1
1 - 1
1 1
1
2 1
1
3 1
1
3 1 1
416 176 52
1
2 1
4 2
1
1 3
1 1
1
Figure A-l. Typical Page of Block Statistics
-------
00
r Y*l
Fresno pj s|*
Div. Ill; ClovisDi>
Cloyjs D
,-EDI
ED 44S 1!
ED ll 79 111
RESNO
ED "181
54.02
ED 21
Part
ED 182
sfresno
E022.
Part'"
. oiJiBUL
71/ . L .^.."»!.»" "*'" *'
Clovi
Figure A-2. Typical Census Map
-------
These materials were used to randomly select appropriate numbers of census
tracts and blocks within each selected urbanized area.
In the first year, this was done on a zone-by-zone basis. For each urban
zone in which sites were to be located, tracts were from the set of tracts con-
tained in that zone. Next, for each selected tract, a block was randomly
selected from the set of blocks which compose that tract.
In the second year, tracts were selected from the urbanized area as a
whole, with the random selection process weighted for population in the same
manner as the second year urbanized area selection process. Tracts were
selected until each urban zone had a sufficient number of tracts. Extra tracts
which were obtained in this process were used as alternates. Blocks were then
selected from each tract, again on a population-weighted basis.
Alternate census tracts and blocks were selected in both years for use in
the event that certain blocks were unacceptable for monitoring. This would 'be
the case if: (1) the block fell within an Ldn 65 contour around an airport;
(2) the block contained no inhabited residential units, or was otherwise sig-
nificantly altered during the period between 1970 and the time of the study;
(3) the block was located in an area judged too unsafe for measurement activity;
or (4) no residential units on the block were made available for monitoring, by
their occupants.
Once the census tracts and blocks were selected, they were randomly desig-
nated as high- or low-traffic impact according to the number of sites needed in
the particular urban zone in the particular traffic impact category. These
designations were adjusted in the cases where the block had no available
residential units in its designated category.
A-9
-------
A-2-4 Selecting Residential Units
The selection of residential units from the selected blocks was made by
the monitoring field team, usually just before initiation of measurement activ-
ity at the selected site.
The first step was to identify the residential units on the selected block .
which fell into the traffic impact category for which the block was designated.
This was done by counting mailboxes or doorbells.
Next a residential unit was randomly selected from the set of identified
units by use of a random number table. Permission to conduct noise measurements
at the selected unit was then sought. If permission was not obtained, permis-
sion to monitor an adjacent unit was requested. This process continued until a
measurement site was obtained.
The selection of residential units for measurement sites was one of the
more difficult activities in the survey. Field personnel had to make judgments
regarding the safety of the area and its conformity to the 1970 census maps,
as well as be prepared to encounter the diverse reactions which a stranger
asking permission to measure noise levels on the front yard has every reason
to expect. A discussion of the field experience of selecting residential units
is given in appendix B.
A-3 MEASUREMENT PROTOCOL
The measurement protocol used in the survey was typically carried out over
a 24-hour period. It consisted of two elements, continuous noise monitoring
and microsamples.
A-3-1 Scheduling
The measurement protocol was usually begun immediately after the measure-
ment site had been obtained. No formal scheduling procedures were followed
in selecting a day on which to obtain and monitor a particular site. Rather,
A-10
-------
this was determined by the travel schedules of the field measurement teams and
the geographical layout of selected census tracts and blocks. It was antici-
pated that this would result in a set of measurement days which was aggregately
representative with respect to such variables as weather, holidays, and day of
the week. Of course, the results at any particular site are likely to be some-
what less representative.
A-3-2 Continuous Monitoring
The continuous monitoring was usually conducted for a 24-hour period.
The measurement apparatus consisted of a microphone and an automated sound
level recorder. The apparatus are described in detail in appendix C.
The microphone was located at the architectural front of the residential
unit, and the recorder placed in a secure location near it. In the case of
single family homes, the microphone position was specified at 1.2 m above the
ground, 2 m out from the front of the building, and 2 m from the corner of the
house farthest from the driveway. In the case of houses with no driveways or
with driveways on both sides, the microphone was placed equidistant between
the sides of the house. This location is illustrated in figure A-3.
For residential units in apartment buildings, the microphone was placed
at the same height as the selected residential unit, 2 m from the front of the
unit, on a convenient balcony or other outside location.
After the continuous-measurement system was set up at the specified loca-
tion, it was calibrated and the recorder system started. The system was
inspected during each subsequent visit to the site, and was recalibrated at
the end of the measurement period.
A-ll
-------
!-*-
-»<>*-
-w
2m
S<
2m
g
LU
E
Q
BACK
SELECTED RESIDENCE
FRONT
2m
< »
2m
O
M Sr
C.ll.l.
<
g
© =MICROPHONE LOCATION
M = CONTINUOUS MEASUREMENTS
SF SS ,SR , = FRONT, SIDE, AND REAR SATELLITE MEASUREMENTS
Figure A-3. Measurement Locations Used in the Survey
A-12
-------
The continuous monitoring was extended to 5 days at eight of the measure-
ment sites. This required no changes in the measurement procedure.
A-3-3 Microsamples
Microsamples were collected manually over 30-minute periods at accessible
sides of the residential units. The microsamples were usually obtained in
sets of two, one at the front of the unit and one at another side. Three such
sets were usually collected, one each in the daytime (0700-1900), evening
(1900-2200), and nighttime (2200-0700). Scheduling within this framework was
based primarily on logistical considerations.
The microphone location for these measurements was 2 m outside the midpoint
of each exterior wall, and 1.2 m above the ground. Figure A-3 illustrates
these locations. Satellites were omitted when the distance between a wall and
another building was less than 3 m.
Each microsample consisted of about 120 sound-level measurements and
source identifications, made once every 15 seconds. The sound-level meter was
set to "slow response," and was calibrated before and after each microsample.
A sample microsample data sheet is shown in figure A-4.
A-4 ANALYTICAL PROCEDURES
A-4-1 Continuous Monitoring
Sound levels measured in the continuous-monitoring procedure were encoded
on digital tape. The tapes were analyzed by means of a digital translator
interfaced with a computer,, as described in appendix C. Computer output con-
sisted of hourly equivalent sound levels and statistical levels; the daytime,
\
nighttime, and 24-hour equivalent sound levels; and the day/night sound level
(Ldn). These data were stored in a Statistical Analysis System (SAS) data base.
A-13
-------
J
o
S
O
I
('
4N
J
.?
£
COMMUNITY NOISE MEASUREMENT DATA SHEET
a
k^
Uv*l
ID
NumUr
20
t. isoc*
MtlwCUdi
k»»yt O.K.
CclltnllMi
SIN
NdTH
Mhc.ll.MMM f AftM 0*1 C«N*«tl«li
VA
UhwMi 9lMtl««i v WohmiH*
TW
W«MhM
T-,.
^
HM.
WM
Db.
S»«|«
total
SoiiylM
Figure A-4. Data Sheet Used in Microsamples
JM
HHU«pi«i
W«l«Cnl>
Oft-M. V*h.
AC CxJl
e_
v
MacMr^r
C«mliwll«» I
Yvrf M«M. b«»lf.
O.U.
0_
0_
l_
X
Oo,
UhUmtlfbU*
-------
A-4-2 Microsample Data
Each inicrosainple was reduced to find:
a. the equivalent sound level as averaged over the entire set
of measurements taken in the microsample;
b. the frequency of identification of each source type; and
c. the equivalent sound level of each source type as averaged over
the set of measurements in which it was identified.
These data, along with the time of day and the designation of the side of
the unit at which the microsaraple was collected, were stored in a SAS data
base.
A-4-3 Other Data
In addition to the acoustical data, the urban area size, urban zone
density, and traffic impact categories were stored in a SAS data base. Also
included were the actual urbanized area population and urban zone density of
each site, based on the 1970 census. These data were used in developing cell-
by-cell profiles and for factor analysis.
A-15
-------
APPENDIX B
FIELD EXPERIENCE AND SURVEY CRITIQUE
B-l INTRODUCTION
The site selection and measurement protocols described in chapter 3 and
appendix A are based upon the Environmental Protection Agency (EPA) and
contractor experience in community noise assessment. Essentially, these
protocols are scaled-up versions of similar procedures which were developed
for local application. Although such adaptations are quite straightforward,
their practicability and utility can be ascertained only through implementation.
To provide useful input for future studies, as well as to allow a more informed
assessment of the results of this study, some remarks on this experience are
required.
B-2 SITE SELECTION
The most difficult part of the methodology was the site selection. Several
features of the study contributed to this difficulty:
a. The field engineers were usually in unfamiliar territory, making
it more difficult to find preselected census tracts and blocks.
b. The limited travel budget and tight schedule usually allowed only
one visit to the selected block for the purposes of site selection.
c. The random block selection process resulted in a number of unsafe
or otherwise unusable blocks being selected.
d. The survey required permission from residents before a measurement
site could be established.
The above circumstances resulted in the loss of some measurement sites,
and in the relocation of others. The lost sites were the direct result of
the limited travel budget and the tight measurement schedule. The length of
B-l
-------
stay in each urban area was predetermined by this schedule, which was generally
adhered to even if the desired number of sites had not been obtained. Reloca-
tion of measurement sites occurred at both the block and residential unit level.
Block relocation occurred 19 times. In nine of these cases, the original block
was judged too unsafe to enter or leave measurement equipment in. In other
cases, the entire block fell within an Ldn = 65 dB contour around an airport
(three times), the block had been changed into an exclusively nonresidential
area (three times), or permission to measure could not be obtained at any
residential unit on the block (three times). Finally, in the Buffalo urbanized
area, a selected block was condemned because of the Love Canal toxic waste dump
site.
When blocks had to be reselected, two different procedures were used
depending upon the circumstances. In cases where the problems causing reselec-
tion were confined to one particular block, a block adjacent to the initially
chosen one was selected randomly. In cases where reselection problems con-
cerned a larger area, a new block was selected randomly from a set of blocks
not included in the area. In some cases, selection of a new census tract was
required also.
Residential unit relocation was much more common than block relocation.
By far the most prevalent cause for this factor was failure to obtain permis-
sion to conduct measurements at the residential unit that had been selected
initially. The lack of permission was generally due to no one being at home
when the unit was called upon. In other cases, permission was refused or
could not be granted as a result of a neighborhood association or other
cooperative living arrangement. In all, about 40 percent of the initially
selected residential units could not be used for these reasons. Of the cases
B-2
-------
described, 40 percent resulted from no one being at home, 50 percent involved
a refusal or inability to grant permission on the part of the occupants,
and the remainder were the result of safety considerations.
Selection of alternate residential units was accomplished deterministically
on the basis of the initial residential unit selection. The residential units
on the selected block had been assigned numbers in the initial selection
process. If the nth unit had been selected initially and could not be used,
the unit numbered n+1 was then selected, followed by n-1, n+2, n-2, etc.
Thus, the unit used for the measurement site was always the one nearest to
the initial selection, which was located on the same block and at which permis-
sion to monitor could be obtained. In only nine cases was this unit more than
two units distant from the initially selected one.
Apartment units and other multiunit complexes posed the greatest problems
in the residential unit selection process. For example, measurements made
above the first story required entrance into the residential unit by the field
engineers, and this made obtainraent of permission from the occupants substan-
tially more difficult. Another contributing factor in some cases was the lack
of authority of occupants to grant permission. No time had been allocated
for contacting landlords or meeting with residents' associations to obtain
this permission, and only a limited attempt could be made to do so in these
cases.
Of the 11 instances in which the randomly selected unit was part of a large
apartment building or multiunit complex, the selected unit was successfully
used three times. In two cases, the unit was relocated to a first-floor
residential unit in the same building. In six cases, an alternate site
outside of the apartment building or multiunit complex was used.
B-3
-------
In three of the six alternate sites, the site used was a public building
adjacent to or across the street from the original selection. The use of such
a building allowed a measurement location at the same height as the initially
selected one, and was considered preferable to relocating to a less similar
residential unit.
The necessity for selecting alternate residential units and blocks during
the survey introduces the possibility of biased results. The three most likely
sources of bias are:
a. Relocation of sites out of dangerous neighborhoods.
b. Relocation of sites to residential units where the occupants were
at home at the time of the site visit.
c. Relocation of sites to residential units from multiunit complexes,
and especially from upper story units in these complexes, to units
outside of the complexes.
The relocation procedures described in a through c above result in a sample
of residential units which:
a. Underrepresents areas that appear unsafe.
b. Overrepresents residential units that are occupied during the
daytime hours (during which site selection generally took place).
c. Underrepresents residential units located in multiunit complexes,
and especially those located above the first story in these
complexes.
While it is desirable to avoid such biases, their expected effect on
the survey results is minimal. None of the circumstances that gave rise
to the biases occurred very often, and it is doubtful that any substantial
correlations exist between noise environments and the affected variables.
B-4
-------
One possible exception to this is the underrepresentation of non-first-story
residential units. A systematic approach to avoiding this bias is recommended
for future studies of this kind.
B-3 MEASUREMENT PROTOCOL
The basic measurement procedures that were included in the measurement
protocol are 24-hour continuous noise monitoring and 30-minute raicrosamples.
While both of these procedures have been used extensively in local surveys,
some problems arose in their implementation for the National Ambient Noise
Survey. These problems resulted from:
a. The statistical nature of the survey.
b. The tight measurement schedule.
c. The diversity of residential units being used as measurement sites.
d. The application of noise assessment techniques to noise-free
environments.
The above features of the survey led to three areas of difficulty:
(1) choosing measurement positions, (2) minimizing the impact of the field
personnel on the noise environments being studied, and (3).identifying sources
observed during the survey.
/ The measurement positions were specified in the measurement protocol,
and were defined relative to the outer facades of the residential units.
These specifications were necessary to obtain comparable data for each site
without performing time-consuming preliminary measurements. The specified
positions were found to be satisfactory about 90 percent of the time. Of the
10 percent of the exceptional cases, the most frequent problem resulted from
obstructions such as trees and bushes or from the presence of a walkway that
could not be obstructed. In such cases, the measurement positions were simply
moved laterally to the nearest suitable location.
B-5
-------
In three cases, the location was found to be unsuitable as a result of
acoustical considerations. This occurred when the specified position was
extremely close to an operating air-conditioner or a window that was :trans-
mitting a high volume of noise from the residential interior. . Because the
objective of the survey was to study noise from exterior sources, the use of
these specified positions would have produced distorted results, and a lateral
adjustment of the measurement position was made to avoid this outcome.
While the infrequency of this circumstance minimizes its significance,
the importance of the field engineer's judgment in making these adjustments
must be realized. There is a basic conflict between the need to perform
these measurements where they will most accurately reflect the noise impinging
upon the building facade and the need to avoid noise emanating from that facade.
Future protocols should address this conflict, and allow field personnel some
flexibility in dealing with it.
The problem of survey personnel impacting the noise environments being
studied occurred when either the survey personnel generated the noise or when
they acted as a stimulus for others to do so. In some areas, the noise of
the field team's motor vehicle coming to and leaving the site was a noticeable
acoustical event. This was especially a problem during the late night measure-
ments. In other cases, the presence of the field team aroused either human or
canine curiosity and, with that, either talking or barking.
In all of these cases, only the 24-hour measurement was impacted. From the
point of view of this single measurement, the visits to the site to collect
microsaraples might have been better avoided. In any case, a cost-benefit
analysis of these visits should be undertaken in the development of future
protocols.
B-6
-------
The source identifications used in the microsample procedure were the third
source of difficulty. Problems arose when predominant sources could not be
seen, or when there was no obvious predominant source.
Cases in which predominant sources could not be seen occurred most fre-
quently when a solitary vehicle on an unseen roadway was the predominant source.
In some cases, the type of vehicle could be surmised on the basis of hearing
alone, but more often the source was recorded as "unidentified traffic." Thus,
both aggregate traffic noise and unseen solitary vehicle noise are included
under this designation.
Cases in which there was no obvious predominant source occurred when
natural sounds, such as wind, birds, and crickets, and distant man-made sources
(usually distant traffic) were both present in the acoustic environment. If
the natural sound was distinct (i.e., produced a deflection of 3 dB or more
on the sound level meter) the source was identified as the natural source.
If the natural sound was indistinct but had .a masking effect on the man-made
noise (i.e., produced a deflection of 1 to 2 dB) the source was identified as
residual. . If the natural sound was detectable but produced no deflection in
the meter, the man-made noise was identified as predominant.
These rules of thumb evolved over the course of the survey, and arose from
the attempt to use microsampling in situations for which the procedure was not
really intended. In the future, careful consideration should be given to
the utility of performing microsampling in environments that include only
low-level sources. Microsampling is best considered as a diagnostic procedure
that is most useful when one or more high-level sources are present. When
only low-level sources are present, the reliability of the procedure is greatly
diminished.
B-7
-------
B-4 LOGISTICS AND SCHEDULING
Reference has been made to the severe time constraints under which the
survey was performed. The efficient use of time and the careful planning of
travel were critical to the success of the survey. It was found that such
efficiencies allowed the field team to make two complete sets of measurements
per day.
The most time-consuming element of the survey was travel between cities.
The large volume of measurement equipment made driving the only practical mode
of transportation, although flying would obviously have been faster. The survey
was scheduled to minimize the necessary intracity travel, and this efficiency
was increased somewhat during the second year by the greater number of urbanized
areas and sites being visited.
Intercity travel time to sites varied greatly but generally was between
45 and 90 minutes. Attempts were made to minimize the travel time between
two sites that were being measured at the same time or consecutively. The
larger urbanized areas include substantial amounts of land area, making such
planning especially important.
Implementing the protocol itself was much less time-consuming. Securing
a measurement site usually took about 15 minutes, and each visit to the site
lasted about 45 minutes.
The field team found that performing two sets of measurements daily was
possible yet extremely taxing. Certainly this represents the upper limit of
measurement activity which can be reasonably expected of a two-person team.
Moreover, occasional slippages in schedule are bound to occur when the two-
site-per-day regimen is followed.
B-8
-------
B-5 CONCLUSIONS
As a pilot study intended to provide a statistically valid profile of
residential noise environments across the United States, the National Ambient
Noise Survey was an overall success. The survey methodology can serve as
a basic framework for future surveys at either the national or subnational
level. The survey results represent the first statistically valid national
profile of urban residential noise environments ever assembled.
The problems with methodology that are most in need of consideration
are (1) the treatment of multiunit residential complexes, (2) the method of
specifying microphone locations, (3) the tendency of field perso.nnel to impact
the acoustical environment being measured, and (4) the use of microsampling
when only low-level sources are present. Further consideration may lead to
solutions of these problems, or may reveal that, given resource, constraints,
the approaches used in the survey are the best available. In either case, '
a survey such as this one must always represent compromises between what is
desired and what is possible. So long as the terms of these compromises are
understood, the resulting data can be interpreted in a correct and meaningful
way.
Although the methodology used was developed especially for this national
survey, its applicability extends to surveys on any other scale: community,
regional, or statewide. Such application is certainly desirable, for it is
at these other jurisdictional levels that most noise control policy is made.
It is recommended, however, that these applications be carefully considered
as to their purpose and the obvious logistical advantages that localized
survey programs have over national ones. While the benefits, on an informa-
tional level, of conducting local surveys analagous to the National Ambient
Noise Survey are obvious, no such benefits accrue from rigorous adherence to
B-9
-------
all aspects of that survey's methodology. Many of the problems encountered
in the National survey may be avoided altogether in local surveys if this
point is remembered.
B-10
-------
APPENDIX C
24-HOUR CONTINUOUS MEASUREMENT AND ANALYSIS PROCEDURES
C-l INSTRUMENTATION
Acoustic instrumentation utilized for community noise surveys which
employ continuous statistical monitoring is available in a number of different
formats and with varying capabilities. Three main categories exist and these
include:
a. Equipment designed to hold the data internally in software
storage in the field, and relinquish same upon command at the
end of the 24-hour sampling.
b. Equipment that must be monitored periodically during the 24-
hour period to determine hourly, or shorter-duration, statis-
tical information.
c. Equipment designed to encode the data on tape (either magneti-
cally or on paper) in the field for later decoding in' the
laboratory.
Each of the above equipment types has its inherent advantages and dis-
advantages. In general it has been found useful to back up 24-hour samplings
with periodic manually sampled checks of the "real time" noise values. Port-
able sound-level meters placed in proximity to the stationary 24-hour micro-
phone are used for this type of monitoring. This procedure was part of the
National Ambient Noise Survey measurement protocol and was followed rigorously
throughout the survey. Three half-hour samples, one each during daytime,
evening, and late night, were performed adjacent to the stationary continuous
24-hour monitoring equipment. It was determined from comparisons of these
C-l
-------
satellite half-hour samplings with the same digitally instrumented sampled data,
that the statistical parameters were within 2 to 3 dBA.
Twenty-.four-hour statistical sampling instrumentation utilized for the
survey consisted of a Bruel and Kjaer (B&K) Model 181 digital data logger,
Model 182 digital data translator, Dec-interface, and Digital Dynamics model
PDF 11/45 computer with Dec printer and 9-track tape reader. The general for-
mat of this system is in Fortran IV with 20- to 30-minute turnover time per
24-hour survey.
The B&K 4161 1-inch microphone and 2619 preamplifier units at the front
end of the 181 system meet American National Standards Institute (ANSI)
SI.7-1971 Type I specifications and the placement of the microphone/tripod
combination was in substantial conformity to ANSI SI.13-1970, Field Method
techniques. The 181 data logger digitization rate was preset to A-weighting
and a 0.5-second sampling period for all but the 5-day surveys, in which a
1.5-second sampling was achieved.
The field implementation of this system fo'r real-time gathering of 24-
hour data is given below.
C-2 ENCODING PROCEDURE
a. An Information Terminal Certified Digital cassette* was mounted
into the drive system of the 181 system. The tape was advanced
forward beyond the clear leader to the beginning of the ferric
oxide backing.
*A11 cassettes were bulk-erased prior to the survey to insure a clean
encoded word stream.
C-2
-------
b. The 181 system was energized and allowed to stabilize for 15
minutes prior to calibration. A B&K Model 4230 1000-Hz cali-
brator was used to calibrate the system. Prior to calibration
internal system noise was measured by isolating the microphone
with a Model 4220 Pistonphone coupler and recording the output
onto a B&K Model 2306 graphic level recorder.
c. Calibration commenced by engaging the 181 drive system and loading
forward the tape to the point of digitization. The calibrator
was placed over the microphone and held in the ON position for
30 seconds. The microphone/tripod system was already in-situ for
the site analysis.
d. The calibrator was carefully removed and replaced with a 1-inch
windscreen. The 181 system was sealed, covered with a weather-
proof cover, and carefully placed behind bushes or shrubs in
proximity to the source house.
e. At completion of each 24-hour survey the calibration procedure
was repeated and the tape removed. The tape was either mailed
or handcarried back to the computer laboratory for decoding and
processing.
C-3 DECODING PROCEDURE
a. The 181 encoded digital tape was placed in the 182 digital trans-
later unit located in the PDF 11/45 main frame room. An input name
was assigned to each cassette. This name appeared as output file
titles at the top of the statistical printout sheets for each site.
Total decoding time for each digital tape was 18 minutes.
b. The digital tape output was transferred to disk storage for pro-
cessing. The processing time for the output printing was 6 minutes.
C-3
-------
c. The output files were retained on 9-track tape storage, with
retrieval time approximately 3 minutes, including printing.
C-4 OUTPUT
The output of statistical data sheets provide the following information:
a. Number of samples taken per 15-minute, hourly, 24-hour, day-
time, and.nighttime sampling periods.
b. The percentage of time A-Weighted noise levels were exceeded
for 1 percent, 10 percent, 50 percent, 90 percent and 99 per-
cent (LI, L10, L50, L90, L99) of the measurement period (15-
minute, hourly, 24-hour, etc.).
c. The standard deviation (tf) for each measurement period.
d. The mean A-Ueighted level for each measurement period.
e. The equivalent A-Weighted. sound pressure level (Leq) for each
measurement period.
The above parameters are formatted for each 15-minute as well as hourly
period. Four sheets of data output were provided per site. The fourth sheet
included summary information for the following statistical data:
a. LI, L10, L50, L90, L99 percentile units for daytime (7 a.m. -
10 p.m.), nighttime (10 p.m. - 7 a.m.), and 24 hours.
b. Equivalent sound pressure level for daytime, nighttime, 24
hours.
c. Mean standard deviation over 24 hours.
d. The day-night sound level (Ldn) for the survey.
C-5 MULTIPLE-DAY SITES
The encoding/decoding procedures.were utilized for the multiple-day sur-
veys, with the only difference being the use of a 1.5-second sampling rate
C-4
-------
for five of the eight sites. This required hand calculation of the daytime,
nighttime, and 24-hour statistical parameters in each instance. The sampling
rate was increased so as to disturb the environment as little as possible by
supplying the 181 system with a tape and sample rate configuration requiring
only one recalibration per 5 to 6 days of sampling. In retrospect a change
of tapes once a day (allowing for the normal 0.5-second sampling) would have
been preferable.
C-6 EXTRANEOUS FACTORS
Ninety-four of the 128 survey tapes encoded in the field were decoded
with no extraneous errors or modification. In 30 of the remaining 34 surveys,
minor editing corrections were required for a variety of reasons, but in no
way were the results affected to an extent beyond the inherent inaccuracy of
the measurement system (±0.5 dB fast response). Some of the reasons for modi-
fication of digital tape results included the following phenomena:
a. Dropout errors in the digital tape.
b. Rain or extraneous wind or moisture conditions occurring ran-
domly during the 24 hours of sampling.
c. Events precipitated by the presence of the satellite data
sampling at the survey site. These could include dog barking,
curious onlookers, children playing around the microphone,
extra horn honks, accidental bumping of the 24-hour microphone,
noise caused by setup of satellite measuring equipment, etc.
d. Minor encoding errors caused by harsh environmental conditions
(overheated or excessively cold) and general wear and tear of
instrumentation utilized 7 days a week, 24 hours a day for
nearly 5 months of continuous monitoring. The major problem in
C-5
-------
this instance was the 181 unit tape drive malfunction due to
the need for internal lubrication. Four of the remaining 34
surveys required modifications due to this type of error. This
included extensive digital review of the original cassette tape.
Detailed procedures for the modifications discussed above are
presented below.
C-7 ' EDITING'AND CORRECTION PROCESSING - DIGITAL DATA CASSETTES
As might be expected from any statistical data gathering instrumentation,
hardware-generated encoding errors often occur in the field and most systems
provide for correction either at the front end of the system or during decod-
ing. The field measurement methodology provided for absolute calibration at
the front end of each tape, encoded with the corresponding "word code" for
94 dB. Utilizing this number as a reference, all other sound pressure level
values, as encoded every 0.5 second, were assigned corresponding word values.
Encoding errors always appeared as words with values that could not possibly
represent actual noise levels. This is to say that the octal* word error
would be encoded as 10020 or 4230 which translated with the corrected base
o o
value (usually 31-34 dBA) would come out 543 dBA or 327 dBA , an obvious
data error. The reasons for encoding errors such as the above could include:
a. Dropouts in the digital cassette.
b. Someone hitting the measurement microphone.
c. Someone shouting into the microphone.
d. Moisture condensation on the microphone.
e. Mechanical drive system discontinuities in the 181 system unit.
f. Arcing in the preamplifier.
*The PDF 11/45 System uses octal number system.
C-6
-------
In cases where the drive system actually backed up on itself for a momen-
tary half-second sample, two words could go down on tape where one word be-
longed, thus effectively doubling the total numerical value. What would be
translated as 64 dBA could actually be encoded as 128 dBA. Such is one pit-
fall of word encoding of digital values. In instances where very hot or
humid conditions prevailed the drive system could actually speed up as much
as 25 percent due to crystal oscillator control malfunctions.
Error correction for these phenomena took place in both the encoding and
decoding instrumentation process.
In the field unit, the following error correction techniques were
utilized.
a. Minute/hour markers.
'b. Drop-out compensator.
c. Anti-aliaser.
' The decoding error correction capabilities are more extensive since the
computer software capability is essentially limited only by word space. For
the PDP-11/45, over 130,000 storage blocks (256 words per block) are available
per disk storage unit. For digital cassettes recorded in the field on the
National Noise Assessment, decoding error correction consisted of the
following:
a. The B&K decoding unit contains an 80-dB dynamic range cutoff
switch that, when activated, will automatically truncate data
words above an 80-dB base level input. Thus, no more than a
nominal 120-dB peak level will be passed to the digital sur-
face, eliminating real-time information above such level.
C-7
-------
b. In instances in which dropout errors or word-over-word encoding
errors occur, a special software routine was written to search
for such discrepancies. In this instance any word (noise level)
encoded above a certain preset decibel level is replaced with
the previous word value. Thus, if the level is set at 100 dB,
any word written in excess of 100 dB will be replaced with the'
previous value. This is actually a software version of the
error correction technique previously described, with the added
\
advantage of smoothing the time domain contour with previously
i
/
encoded real-time data. This technique allows the removal of
such extraneous events as children yelling into the micro-
phone, dog barking caused by the satellite analysis team's
presence in the environment, arcing in the microphone due to
excess moisture or dewpoint conditions, and sundry disturbances
of the 24-hour setup. It should be noted that such extraneous
data appears as abnormally encoded data on the decoded computer
printout and is readily identifiable as such.
c. The previously mentioned decoding errors due to digitization
failure or mechanical drive failure show up in decoding as
extraneous word codes in the output. These are automatically
printed as sample errors under the sampling rate column in the
decoded output files for each cassette survey tape. These num-
bers are easily corrected by the edit routine by simply running
a .special software program that accepts only numbers less than
110 dB and prints out the real-time sampled levels as if the
doubling or duplication of words never occurred.
Co
o
-------
d. The last editing program available was utilized for surveys on
which the drive "system and attendant control-crystal malfunc-
tion caused less than 24 hours of data to be encoded on the
input cassette tape. It should be noted that this type of
error occurred in less than 10 percent of the survey. The
correction consisted of spreading the less-than-full-day
sample out to 24 hours, realizing a full-day sampling, and
comparing satellite-sampled 1/2-hour readings with the finished
edited 24-hour sample to determine correlations and differences.
The theory behind this editing procedure is that random mal-
functions throughout a particular 24-hour day occur in an even
distribution, making the corrected sample correspond evenly to
the real-time noise events presented to the measuring micro-
phone. Data -reduced utilizing this procedure indicated that
the theory, as applied to the cassettes so edited, fit the
normal distribution of given values within the sample error of
the survey format.
A formal breakdown of the data encoded for 24-hour surveys indicated
the following:
a. Four surveys edited for less than 24 hours of data.
b. Twenty-six surveys edited for minor encoding errors caused by
yelling into microphones, moisture, and other extraneous
phenomena.
c. Ninety-seven surveys decoded with no error correction required.
C-9
-------
APPENDIX D
RESULTS SUPPLEMENT
D-l INTRODUCTION
The results presented in chapter 4 are based upon statistical analyses of
"raw" acoustical data obtained in the field. To appraise the results criti-
cally, it is necessary to consider both the quality of the data obtained, and
the analytical procedures which were performed on that data.
D-2 QUALITY OF DATA OBTAINED
D-2-1 Continuous Measurements
The accuracy of continuous sound level measurement is largely determined
by the sampling rate employed. This rate was 2 samples per second (sps) for
the 24-hour monitoring and 0.66 sps for the 5-day monitoring.
EPA has found that sampling rates in excess of 0.4 sps can be expected
to generate Leq measurements within 0.5 dB of the true value. Sampling
errors in the continuous measurements are thus not significant sources of
error to the survey results.
D-2-2 30-Minute Microsamples.
EPA has found the 30-minute microsamples to yield estimates of the Leq
accurate to within 3 dB when roadway traffic is the predominant noise source.
Further discussion of this error is included in paragraph D-3-2.
Estimates of the errors in source contribution data developed from these
microsamples are difficult. Among the sources of error are:
o errors in source identifications
o sampling errors
o contributions of non-predominant sources to noise levels
D-l
-------
It is not possible to quantify the errors resulting from mistaken source
identifications. Consideration of this phenomena would require a study in-
volving the simultaneous collection of microsamples by different observers.
A small scale study of this sort is highly recommended to those who make
extensive use of this procedure.
Sampling errors affect both frequency and Leq(id) estimates. In the case
of frequency, the situation is best approximated by a binomial distribution,
where a trial is considered a single source identification, and a success the
identification of a particular source. Thus each- satellite consists of 120
trials, subsets of which are successful trials for each source identified in
the microsample.
Using this model, the expected error in frequency is given by
AFi = N 120 Fi(l-Fi)/120
where Fi is the frequency of source type i as obtained from the microsample.
Thus a frequency of .01 percent has an estimated error of .01,.a frequency
of .1 has an error of .03, and a frequency of .5 has an error of about .05.
The error associated with Leq(id) values also depends on the number of
times the source was identified, as this also determines the number of measure-
ments. The error is given by the equation
ALeq(id) = a±l \120xFi
where
-------
Leq(id) values for these latter sources are also strongly effected by
the presence of competing sources. Such competing sources tend to drown out
low level sources. Therefore, a source is most likely to be heard when it is
emitting high noise levels, thus biasing upward the sample of sound level
measurements upon which the Leq(id) value is based.
Contributions from unidentified sources also result in an upward bias of
this sample. If these sources combine to account for half the sound energy
at the time of a measurement, then the noise level associated with the identi-
fied source is 3 dB more than the actual level of that source. A 2-dB error
would result if such sources accounted for 40 percent of the energy, and 1-dB
error if they accounted for 20 percent of the energy. While it is reasonable
to expect that most situations encountered fell somewhere within this range,
a more precise treatment of this source of error is not possible.
D-3 ANALYTICAL PROCEDURES USED
D-3-1 Distribution of Population over Ldn
The basic problem in obtaining this distribution was to translate
measurement results from 12 subpopulations (the sampling cells) into results
pertaining to the aggregated urban population.
To obtain the raw distribution, the distributions obtained for each
sampling cell were combined by means of a weighting scheme which took into
account the human populations of the individual cells. Thus, if 20 percent
of the Ldn measurements obtained in a particular cell were between 60 and
61 dB, and if that cell had a population of 10 million, then 2 million of that
population was assumed to be exposed to Ldn between 60 and 61 dB. Summing of
these individual cell distributions produced the overall distribution.
To obtain the normal distribution, the individual cell results were used
to generate estimates of the mean Ldn and the standard deviation in Ldn. The
mean was obtained using the equation
D-3
-------
12
M = 2-J Pi x Xi,
where Pi is the fraction of the population in the ith sampling cell, and Xi
is the mean of the Ldn measurements obtained in that cell. The standard
deviation was obtained from the equation
712 \ 1/2
nr 2 - - 2
-\/ vPifai +(X±-v) ) I
\i-l /
where i is the-standard deviation of the measurements obtained in the ith
cell, and the other terms are defined as above.. A n of 60.4 dB and a o of
4.8 dB were obtained from these calculations. Together, these values define
a normal curve, according to '
. 1 -(X-M)2/2o2
where P(X) is the fraction of the urban population exposed to an Ldn of X.
In all practical applications, this equation must be transformed to
P(X -AX
-------
D-3-2 Variation by Side
To obtain estimates on the variation in noise levels by side, microsample
data was compared with data obtained at the continuous measurement site. Each
30 minute microsample consisted of 120 sound level measurements, and these
measurements were used to obtain an estimate of the Leq over the 30 minute
period. This Leq value was then compared with the Leq value obtained at the
continuous measurement site for the hour in which the microsample was obtained.
(Note that only half of this hour was covered by the microsample.) The dif-
ferentials in Leq values were then averaged by side at which the microsample
was obtained and by sampling cell. The results are shown in table D-l.
Clearly the weakest link in this procedure is the Leq obtained from the
microsample. This is because of the sampling errors inherent in the micro-
sample procedure, and because the microsample was obtained over only half of
the time period considered in the continuous monitoring procedure. Some
estimate of the resulting errors can be obtained by means of the differentials
between the Leq values based on continuous monitoring and those obtained from
microsamples at the front of the residential units. These microsamples were
obtained at locations very near to the continuous monitoring site, thus, the
differentials obtained represent sampling errors instead of noise level
variations. Inspection of table D-l shows an average error of about 1.5 dB.
Note that, in 10 of the 12 sampling cells, the average Leq based on the
front microsample is less than that obtained from continuous monitoring. This
is the result not of faulty measurement equipment but of the nature of the
averaging procedure. Leq is a logarithmic measure of sound energy, and it is
the average energy obtained in the two measurement procedures which should be
equal. Table D-l, on the other hand, reflects arithmetic averaging of the
differences in sound level. If Leq(c) is the Leq obtained on the basis of
D-5
-------
Table D-l. Average Differences in Noise Levels Between
Microsaraple and Continuous Monitoring Results
Traffic
Impact
High
High
High
High
High
High
Low
Low
Low
Low
Low
Low
Urban Area
Size
Large
Large
Medium-
Small
Medium-
Small
Medium-
Small
Medium-
Small
Large
Large
Medium-
Small
Medium-
Small
Medium-
Small
Medium-
Small
Urban Zone
Density
High
Medium-
High
High
Medium-
High
Medium-
Low
Low
"High
Medium-
High
High
Medium-
High
Medium-
Low
Low
Front
Delta
-1.2
1.6
3.1
0.8
0.5
-0.4
1.9
1.1
1-7
3.1
2.3
-0.2
Side
Delta
3.7
3.6
3.4
3.5.
5.9
5.4
1.2
3.9
3.1
4.7
4.0
4.0
Rear
Delta
8.2
8.5
6.9
8.9
11.4
14.7
1.9
3.0
6.6
6.0
4.9
1.4
Delta - Average value of the quantity Leq(c) - Leq(m), where
Leq(c) is the equivalent sound level as measured at the front
of the residential unit by means of continuous monitoring, and
Leq(m) is the equivalent sound level as measured at the front,
rear, or side of the unit by means of a microsample.
D-6
-------
continuous monitoring, and Leq(m) is that obtained on the basis of micro-
samples, then,table D-l reflects the value:
or, equivalently,
(leq(c) - Leq(m)y
,Leq(c) - Leq(m)
The energy average, on the other hand, is given by
____ 2
Leq = Leq + k ac ,
implying that Leq(c) > Leq(m) in order for the equality to be maintained.
D-3-3 Temporal Variation
Table D-2 shows the results of the eight raultiday measurements upon which
the assessment of daily variation was based. Overall daily averages were
obtained by averaging the differences between Ldn values obtained for a par-
ticular day and the average Ldn for each multiday site. These average dif-
ferences were then added to the average Ldn obtained over all sites, 60 dB, to
obtain the average daily Ldn values.
The daily variation in Ldn shown in table D-2 reflects both true varia-
tion in the noise environment and sampling errors inherent in the continuous
monitoring procedure. Based upon the discussion of sampling errors in section
D-2, the variation may be assumed to be almost entirely true variation.
D-7
-------
Table D-2. Five-day Site Ldn Values
Site
San Francisco
TR 3870
BL 205
San Diego
TR 170.01
BL 112
Tampa
TR 113
BL 211
o Philadelphia
oo TR 148
BL 105
New York
TR 5202
BL 108
San Jose
TR 5065.03
BL 210
Cleveland
TR 1192
BL 104
Chicago
TR 6305
BL 108
Dates
of M T . W TH F S S Mean
Measurement
3/9/81
to 54.8 53.3 54.6 58.0 55.2
3/13/81
3/13/81
to 53.2 56.8 60.7 54.6 52.9 54.5 55.5
3/19/81
4/10/81
to 60.5 56.9 62.7 61.7 60.4
4/15/81
5/3/81
to 63.4 62.5 61.0 66.3 67.4 64.1
5/8/81
5/27/81
to 59.4 61.8 59.5 61.3 59.8 60.4
5/31/81
3/6/81
to 51.2 51.1 56.8 49.5 50.7 51.8
3/11/81
6/4/81
to 70.2 69.7 68.5 59.3 71.3 69.8
6/10/81
6/16/81
to 62.0 63.3 64.3 63.2 61.7 62.9
6/22/81
Standard
Deviation
2.0
2.9
1.6
2.7
1.6
2.8
1.0
1.1
Mean
60.0
2.0
-------
The hourly noise level histories were obtained by averaging the levels
both by hour and by sampling cell, and then taking weighted averages of these
results by traffic impact.
D-3-4 Noise Sources
The results concerning source contributions are based upon 700 micro-
samples and 84,000 individual source identifications and sound level measure-
ments. A two-stage process was used to reduce these data. First, the data
obtained in each microsample was reduced to obtain the overall Leq, according
to the equation
/120 , \
Leq = 10 x log 101 £ 10 ' /120J
\i=l /
Li - Level obtained in ith measurement of microsample,
the frequency of identification of each source, n/120, where n is the number
of times the source was identified, and the Leq(id) for each source,
Leq (id) = 10 x log 10 ( £ 10Lj/1°/n)
\j = l /
Lj - Level measured at time of jth identification.
The acoustical data for a microsample thus contained one Leq value, 26 fre-
quency values (one for each source considered in the survey), and one Leq(id)
value for each source whose frequency was greater than zero. If a source was
not identified in a microsample, its frequency was.zero and its Leq(id) was
considered a missing value.
The results for each microsample 'taken at a particular site were then
reduced to obtain a source profile of each site. To control for the various
sources of variation, different subsets of the microsamples were used accord-
ing to the phenomena being considered. Cell by cell variation and factor
analysis was based upon microsamples taken at the front only. The frequency
D-9
-------
used was computed from the equation
F = 12/24 Fd + 3/24 Fe + 9/24 Fn .
Fd - Frequency'in front, daytime, microsample
Fe - Frequency in front, evening, microsample
Fn - Frequency in front, nighttime, microsample,
where the fractional weighting factors reflect the number of hours associated
with these three time periods. Likewise, Leq(id) was computed according to:
Leq(id) = 10 log 10 (ioLe^id>d x 12/24 + !0Leq(id)e x 3/24
+10Leq(id)n x g/24)
Leq(id)d - Leq(id) in front, daytime, microsample
Leq(id)e - Leq(id) in front, evening, microsample
Leq(id)n - Leq(id) in front, nighttime, microsample
The same Fd,e,n and Leq(id). descriptors were used to assess variation
d,e,n
in source contributions by time of day. To assess variation by side of unit,
only microsamples collected during the daytime were considered. Thus, loca-
tional variation was eliminated in the consideration of temporal variation,
N
and temporal variation was minimized in the assessment of locational variation,
and both locational and temporal variation were eliminated in considering
variation by urban area size, population, density, and traffic impact. In-
sufficient data were available to evaluate the interactive effects of these
sources of variation.
i
D-3-5 Factor Analysis
Analysis of variation with respect to urban area size, population density,
and traffic impact began with the definition of the dependent variables.
These were the Ldn, and the frequency and Leq(id), based on temporally averaged
data collected at the front as described in paragraph D-3-4 of the six
selected individual sources and four source classes considered in the analysis.
D-10
-------
In cases where a source was not identified at a particular site, its frequency
was zero and its Leq(id) considered a missing value.
The independent variables were defined initially as dummy variables which
reflected the categories of traffic impact, urban area size, and urban zone
population density used to define the sampling cell structure. An initial
regression analysis was run using these three variables as well as their cross
terms. It was found that the significant factors were traffic impact, density,
and the urban area size-density product. This lead to the hypothesis that
urban area size is significant only because larger urban areas contain urban
zones of extremely high density, so that a "high" density area in a large
urban area tends to be significantly more dense than one in a smaller urban
area. To test this hypothesis, the density variable was redefined as the
logarithm of the actual density. Results of regression analysis performed on
this modified set of variables, confirmed the-hypothesis, with only traffic
impact and log 10 (UZ density) found to be statistically significant.
D-ll
-------
APPENDIX E
. ESTIMATION OF CELL POPULATIONS
As described in the text, the cell structure used in the National Ambient
Noise Survey incorporates three parameters: urban area (UA) size, urban zone
(UZ) density, and traffic impact. The first two of these are based on data
available in table 20 of the 1970 Census of the Population - Number of In-
habitants. The traffic parameter is defined in terms of distances from free-
ways and arterials.
To plan the survey efficiently and to draw national conclusions from it,
it was necessary to estimate the 1980 populations corresponding to each cell.
This was a straightforward matter with respect to the UA size and UZ density
parameters. The difficulty came in apportioning the population according to
the traffic impact categories.
This apportionment was accomplished by means of the National Roadway
Traffic Noise Exposure Model (NRTNEM) data base. The model employs a cell
structure very similar to that used in the National Ambient Noise Survey and
was designed specifically to produce the type of population distribution over
distance from roadway information which was desired.
Figure E-l shows the method used to determine these distributions. It
represents the total roadway mileage of a particular roadway classification
which runs through the total occupied area of a particular population place
size, population density category. The distance d. + d~ is the distance from
the roadway center of the nearest lane (CNL) at which the clear zone ends and
the populated area begins. Thus the populated area within a particular dis-
tance d_ of the roadway is defined by M, the roadway mileage, multiplied by
E-l
-------
I
NJ
OCCUPIED AREA
CLEAR ZONE
M MILES
ROADWAY
T
8
S
2 DO O <2 O
- r-\ T3
>0
2>
m o
O m ON
mo mo
Figure E-l. Method for Estimating Populations Within Specified Distance of Roadway Type
-------
the quantity 2 [d--(cL+d2)]. This area, when combined with its population
density, defines a population.
. Table E-l shows the population place size and population density cate-
gories used in the NRTNEM data base. Population place size is equivalent
(though with a few minor deviations) to the urban area size. The population
density parameter is the urban zone population density.
Despite these equivalences with respect to the variables employed, the
categories used in NRTNEM and the cells used in the National Ambient Noise
Survey differ with, respect to the intervals of these variables which they
represent. In the case of population place size, NRTNEM has nine categories,
eight urban and one rural, while the survey cell structure has only two urban
categories. Both NRTNEM and the survey employ four categories of population
density, but NRTNEM uses logarithmic intervals for its categories, whereas the
survey uses linear intervals.
In addition to the population and land area of each cell, table E-l con-
tains a parameter called P*. This parameter represents a population density
adjusted to exclude the land area in each cell which is unoccupied. This
parameter thus defines the population density of the occupied area shown in
figure E-l.
As the survey considers only urbanized areas, which by definition must
include a central city with a population over 50,000, only population place
size categories 1-6 of table E-l are considered in the subsequent calculations
and tables.
Tables E-2A through E-2E show the mileage distribution of the six roadway
classifications used by FHWA, broken down by average travel speed and the
E-3
-------
Table E-l. Distribution of Population and Land Area by Place Size
(Index J) and Population Density Category (Index ID)
Based on National Roadway Traffic Noise Exposure Model Data Base
POPULATION PLACE SIZEINDEX J
7
G
X
0)
o
g
t-H
flj
Oi
^
£
H
1
"
c
0
to
,-H
D
D.
£
TOTAL
TOTAL
PARAMETER
1 Population
Area
P*
2 Population
Area
P*
3 Population
Area
P*
4 Population
Area
P*
POPIIUTION
ARKA
1
>2M
5.61
134.2
64,711
22.28
3576
12,638
21.59
8358
6, 107
0.0
0.0
49.48
12064.2
2
1M
-2M
2.10
272
13,451
4.08
775
9,092
11.13
5080
5,014
5.35
4089
5,505
22.66
10216.0
3
500k
-1M
0.36
63
9,368
2.04
488
6,967
8.40
4426
3,842
5.30
4584
2,336
16.09
9561.0 ,
4
200k
-500k
1.61
215
9,368
10.43
4558
3697.0
6.75
5790
2,264
0.0
0.0
18.78
10563.0
5
100k
-200k
1.16
279
5,831
2.93
1305
3,384
6.84
5266
2.O11
0.0
0.0
10.93
6850.0
6
SOk
-100k
1.07 .
329
4,186
2.12
1115
2.863
4.53
4195
1.612.0
0.0
0.0
7.71
5639.0
7
25k
-SOk
0.47
58
13,091
2.98
896
8,506
3.51
2230
4,698
1.92
2769
2.147
8.88
5953.0
8
5k
-25k
1.85
220
16,988
4.97
1261
10,681
8.46
4527
6,271
2.70
5820
1,673
17.98
11828.0
URBAN
TOTAL
14.23
1570.2
51.83
13970.0
71.20
39872.0
15.27
17262.0
152.52
72674.2
9
RURAL
64.18
3,476,938
18.0
0.0
0.0
0.0
0.0
0.0
0.0
64.18
3476938
Total Population = 216.70 million
Total Und Area = 3,549,612.2 square miles
p* = Population/ (Area) (Area Factor), Adjusted
Population Density In People per Square Mile
Population Place Size = Urban Area Size
Population Density = Urban Zone Density
-------
Table E-2A. Roadway Mileages
J » Population Place Size
ID - Population Density
K Roadway Type, where:
1 - Interstates 4 - Minor Arterials
2 - Urban Freeways 5 - Collectors
3 - Principal Arterials 6 - Locals
AVERAGE TRAVEL SPEED 20 MPH
ID = 1
HIGH POPULATION DENSITY AREAS
V K> 1 2 3 4 5
K>
K>
K >
0
0
0
0
0
0
3
7
t
3
5
5
16
21
4
17
24
29
41
71
11
45
58
67
37
71
12
42
61
69
94
172
31
119
149
171
ID - 2
MEDIUM TO HIGH POPULATION DENSITY AREAS
6
1
1
7
2
1
78
19
6
69
23
18
438
59
31
360
110
99
1085
201
84
963
273
229
989
203
95
386
283
233
2494
491
242
2514
699
579
ID - 3
MEDIUM TO LOW POPULATION DENSITY AREAS
14
7
7
9
7
4
182
125
51
88
92
67
1025
384
280
458
444
372
2540
1321
761
1223
1100
860
2314
1333
866
1125
1142
877
5837
3216
2197
3193
2821
2178
ID - 4
LOU POPULATION DENSITY AREAS
0
6
7
0
0
0
0
101
53
0
0
0
0
309
290
0
0
0
0
1063
788
0
0
0
0
1073
897
0
0
0
0
2589
2276
0
0
0
E-5
-------
Table E-2B. Roadway Mileages
J « Population Place Size
ID - Population Density
K = Roadway Type, where:
1 - Interstates
2 - Urban Freeways
3 - Principal Arterials
4 - Minor Arterials
5 - Collectors
6 - Locals
K>
K>
K>
AVERAGE TRAVEL SPEED 30 MPH
ID - 1
HIGH POPULATION DENSITY AREAS
1
t
0
1
1
1
3
16
2
9
13
14
43
54
10
44
61
76
83
144
22
92
119
137
76
145
25
85
123
140
422
775
141
534
673
769
ID = 2
MEDIUM TO HIGH POPULATION DENSITY AREAS
13
4
2
22
5
3
202
44
. 15
182
60
46
1138
152
80
937
286
257
2213
411
171
1958
556
466
2017
' 415
195
1805
577
475
11225
2208
1090
11311
3146
2606
ID = 3
MEDIUM TO LOW POPULATION DENSITY AREAS
42
28
20
29
20
11
472
291
133
231
241
174
2662
999
728
1191
1154
967
5179
2693
1551
2487
2242
1753
4720
2717
1765
2293
2328
1787
26264
14473
9888
14368
12695
9803
ID 4
LOW POPULATION DENSITY AREAS
K>
0
22
21
0
0
0
0
234
138
0
0
0
0
805
754
0
0
0
0
2167
1606
0
0
0
0
2187
1828
0
0
0
0
U649
10241
0
0
0
E-6
-------
Table E-2C. Roadway Mileages
J » Population Place Size
ID = Population Density
K » Roadway Type, where:
1 - Interstatas
2 - Urban Freeways
3 - Principal Arterials
4 - Minor Arterials
5 - Collectors
6 - Locals
K>
K>
AVERAGE TRAVEL SPEED 40 MPH
ID = 1
HIGH POPULATION DENSITY AREAS
I
2
0
1
1
1
5
10
1
6
9
9
29
36
7
29
41
51
24
41
6
26
34
39
21
41
7
24
35
40
422
775
141
534
673
769
ID » 2
MEDIUM TO HIGH POPULATION DENSITY AREAS
24
6
3
30
6
4
134
30
10
121
40
31
759
102
54
624
191
172
626
116
43
555
157
132
571
117
55
511
163
134
11225
2208
1090
11311
3146
2606
ID « 3
MEDIUM TO LOW POPULATION DENSITY AREAS
K>
K>
55
37
27
38
26
16
315
194
89
154
161
116
1776
667
486
793
769
646
1465
762
439
705
635
495
1336
769
499
650
659
506
26264
14473
9888
14368
12695
9803
ID - 4
LOW POPULATION DENSITY AREAS
0
30
28
0
0
0
0
156
92
0
0
0
0
537
503
0
0
0
0
614
455
0
0
0
0
619
517
0
0
0
0
11649
10421
0
0
0
E-7
-------
Table E-2D. Roadway Mileages
J » Population Place Size
ID = Population Density
K » Roadway Type, where:
I - Interstates 4 - Minor Arterials
2 - Urban Freeways 5 - Collectors
3 - Principal Arterials 6 - Locals
AVERAGE TRAVEL SPEED 50 MPH
ID - 1
HIGH POPULATION DENSITY AREAS
K>
K>
K>
8 .
17
3
12
12
10
3
6
I
3
5
5
16
21
4
17
24
29
8
14
2
9
11
13
7
14
2
8
12
13
0
0
0
0
0
0
ID - 2
MEDIUM TO HIGH POPULATION DENSITY AREAS
201
48
26
256
55
34
78
17
6
69
23
18
438
59
31
360
110
99
209
39
16
185
52
44
191
39
18
170
54
45
0
0
0
0
0
0
ID - 3
MEDIUM TO LOU POPULATION DENSITY AREAS
K> 1 2 3 4 5,
470 182 1025 488 446
317
233
325
224
129
112
51
38
92
67
384
280
458
444
372
254
146
235
211
165
256
166
217
220
169
0
0
0
0
0
ID = 4
LOW POPULATION DENSITY AREAS
0
255
241
0
0
0
0
90
53
0
0
0
0
309
290
0
0
0
0
205
152
0
0
0
0
206
172
0
0
0
0
0
0
0
0
0
E-8
-------
Table E-2E. Roadway Mileages
J = Population Place Size
10 » Population Density
K - Roadway Type, where:
1 - Interstates 4 - Minor Arterials
2 - Urban Freeways 5 - Collectors
3 - Principal Arterials 6 - Locals
AVERAGE TRAVEL SPEED 60 MPH
ID = 1
HIGH POPULATION DENSITY AREAS
K>
K>
K>
13
29
6
21
20
17
1
2
0
1
2
2
6
7
1
6
8
10
2
3
0
2
2
3
1
3
0
2
2
3
0
0
0
0
0
0
ID » 2
MEDIUM TO HIGH POPULATION DENSITY AREAS
343
83
44
437
94
59
26
6
2
23
7
6
147
20
10
120
37
33
42
8
3
37
10
9
38
8
4
35
11
9
0
0
0
0
0
0
ID - 3
MEDIUM TO LOW POPULATION DENSITY AREAS
K> 1 2 3.4 5
802 60 343 98 90
541
397
555
381
222 .
" 37
17
29
30
22
128
94
152
148
123
51
29
' 47
42
33
51 .
33
44
44
34
0
0
0
t
0
ID - 4
LOW POPULATION DENSITY AREAS
0
435
411
0
0
0
0
30
18
0
0
0
0
103
97
0
0
0
0
41
30
0
0
0
0 .
41
34
0
0
0
0
0
0
0
0
0
E-9
-------
NRTNEM population place size and population density categories. Table E-3
shows the fraction of these mileages which run through occupied land. To-
gether, tables E-2 and E-3 can be used to construct table E-4, which gives
mileages of roadway running through occupied land by roadway classification,
population place size category, and population density category. This data
is given only for the roadway classifications used in defining the high-
traffic-impact-category used in the survey: interstates, urban freeways,
principal arterials, and minor arterials.
The data in table E-5 specify values of M as defined in figure E-l for
*
each combination of roadway classification, population place size category,
and population density category. There remains the need to define values of
d and d«. The value d , the lane halfwidth of the roadway, is estimated as
7.5 feet for interstates and 6 feet for all other roadways. The value d_ for
each combination of roadway type, urban area size category, and population
density category is shown in table E-5.
The high-traffic-impact areas, as defined by the NRTNEM, include all
areas within either 300 feet of an interstate or urban freeway, or 100 feet
of a principal or minor arterial. The data given in tables E-l through E-5
can be used to estimate the populations of each of these four areas for each
population place size, population density category. These estimates are given
in table E-6. This data is then used to construct table E-7, which gives the
total high traffic impact population of each category.
To obtain the high-traffic-impact populations in the urban area size,
urban zone density categories used in the National Ambient Noise Survey, it is
necessary to apportion the NRTNEM categories to the survey categories. Table
E-8 shows the 1970 urban population distribution over urban area size and
urban zone density, the latter distribution based on the category definitions
E-10
-------
used in the survey. Tables E-l and E-8 can be used to perform the necessary
.apportionment within each urban area size category. The results are shown in
tables E-9A through E-9F.
Applying these tables to table E-7 yields the populations of each urban
area size, urban zone density, and traffic impact category used in the survey.
The resulting estimates are shown in table E-10.
Table E-l and the subsequent tables based upon it reflect 1974 data. To
update these estimates to 1980, the population growth factors shown in table
E-ll are used. These growth factors are defined at the level of population
place size and are assumed to apply to all population density/traffic impact
categories within each category. They are assumed to reflect migration of
urban areas within categories as well as net population growth. The applica-
tion of table E-ll to table E-10 results in table E-l2, in which the urban
area size categories have been collapsed into the two used for the National
Ambient Noise Survey. Table E-12 thus gives the population estimates of the
sampling cells used in the survey.
E-ll
-------
Table E-3. Fractions of Roadway Mileages Which Run Through Occupied Land
Population Place Size, Index
K
1
2
3
4
5
' 6
0
0
0
0
0
0
1
.764
.738
.866
.845
.852
.852
2
0.764
0.738
0.866
0.845
0.852
0.852
0
0
,0
0
0
0
3
.764
.738
.866
.845
.852
.852
4
0.764
0.738
0.866
0.845
0.852
0.852
J
5
0.764
0.738
0.866
0.845
0.852
0.852
6
0.764
0.738
0.866
0.845
0.852
0.852
_,
7
0.656
0.679
0.843
0.849
0.867
0.867
8
0.656
0.679
0.843
0.849
0.867
0.867
9
1.000
1.000
1.000
1.000
1.000
1.000
J is Population Place Size Index
K is Roadway Type Index
E-12
-------
Table E-4. Roadway Mileage Through Occupied Land
J=l J=2 J=3 J=4 J=5 J=6
PI
I
ID=1
ID=2
ID=3
ID=4
ID=1
ID=2
ID=3
ID=4
17.6 37.4 6.9 26.7
452 109 58 575
1057 711 523 730
0 571 541 0
Interstates
(K=l)
J=l J=2 J=3 J=4
95 114 22.5 79
2529 339 178 2079
5916 2219 1618 2643
0 1787 1617 0
26.0 22.2
124 77
503 292
0 0
J=5 J=6
137 169
636 572
2562 2148
0 0
Principal Arterials
(K=3)
J=l J=2 J=3 J=4 J=5 J=6
ID=1 14.8 30.3 3.7 16.2 25.1 25.8
ID=2 382
ID=4
ID=4
86 36.9 342 113
0 451 261
0 3456 2561
Minor Arterials
(K=4)
88
ID=3 849 560 252 435 455 329
Urban Freeways
(K=2)
J=l J=2 J=3 J=4 J=5 J=6
ID=1 134 288 34.6 147 189 219
ID=2 3528 655 272 3125 886 744
ID=3 8256 4293 2472 3969 3574 2794
-------
Table E-5. Clear Zone Distances (In Feet) by Roadway Type (K),
Population Density Category (ID), and Population
Place Size (J)
K
1
2
3
4
5
6
ID
ALL
All
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
Population
1
50.
30.
10.
15.
20.
30.
10.
15.
20.
30.
5.
10.
15.
20.
5.
10.
15.
20.
2
50.
30.
10.
15.
20.
30.
10.
15.
20.
30.
5.
10.
15.
20.
5.
10.
15.
20.
3
50.
30.
10.
15.
20.
30.
10.
15.
20.
30.
5.
10.
15.
20
5.
10.
15.
20.
Place
4
50.
40.
10.
20.
30.
40.
10.
20.
30.
40.
10.
20.
30.
40.
10.
20.
30.
40.
Size,
5
50.
40.
10.
20.
30.
40.
10.
20.
30.
40.
10.
20.
30.
40.
10.
20.
30.
40.
Index
6
50.
40.
10.
20.
30.
40.
10."
20.
30.
40.
10.
20.
30.
40.
10.
20.
30.
40.
J
7
50.
40.
10.
20.
30.
40.
10.
20.
30'.
40.
10.
20.
30.
40.
10.
20.
30.
40.
8
50.
40.
10.
20.
30.
-40.
10.
20.
30.
40.
10.
20.
30.
40.
10.
20.
30.
40,
9
50.
50.
40.
40.
40.
40.
40.
40.
40.
40.
40.
40.
40.
40.
40.
40.
40.
40.
Index K denotes highway type; Index ID denotes population density category
E-14
-------
Table E-6. Population (in Thousands) High Traffic Impact by Roadway Type
J=l J=2 J=3 J-4 J=5 J=6
J=l J=2 J=3 J=4 J=5 J=6
ID=1 105 46
ID=2 525 91
ID=3 593 327
ID=4 0 131
6
37
185
116
23 14 9
195 39 20
152 93 43
0 0 0
Interstates
d3=300"
i
Ln
J=l J=2
ID=1 196 49
ID=2 956 92
ID=3 1013 312
ID=4 109
J=3
7
37
174
92
J=4 J=5 J=6
24 25 23
215 60 46
145 125 84
ID=1 96 41
ID=2 483 78
ID=3 518 281
ID=4 113
J=l J=2
ID=1 276 123
ID=2 1334 178
ID=3 1413 603
ID=4 210
3
26
97
61
Urban
d3
J=3
10
57
266
145
15 14
122 37
95 88
Freeways
=300"
J=4 J=5
44 35
324 84
218 174
10
24
51
J=6
29
60
109
Principal Arterials
d3=100"
Minor Arterials
d3=100"
-------
Table E-7. High-Traffic-Impact Population (Millions) by Population
Place Size and Population Density Category .
ID=1
ID=2
ID=3
ID=4
.J-l
0.67
3.30
3.54
0
J=2
0.26
0.44
1.52 .
0.56
J=3
0.03
0.16
0.72
0.41
J=4
0.11
0.86
0.61
0
J=5
0.09
0.22
0.48
0
J=6
0.07
0.15
0.29
0
Table E-8. Population (Millions) and Percentages by Urban Area Size/
Urban Zone Density Category
Source: 1970 United States Census
UA Population
UZ
H
MH
ML
L
Density2
(per mile )
(>_4500)
3000-
4499
1500-
2999
<1499
>2M
26.3
56.6%
20.2
43.4%
0
0
46.5
100%
1M-2M
6.02
26.3%
7.50
32.7%
7.40
32.2%
2.01
8.8%
23.0
100%
500K-1M
3.49
24.1%
3.41
23.5%
7.06
48.7%
.53
3.7%
14.5
100%
200K-
500K
2.40
14.7%
5.00
30.6%
7.94
48.5%
1.01
6.2%
16.4
100%
100K-
200K
1.74
18.1%
2.27
23.6%
4.75
49.5%
.85
8.8%
9.6
100%
50K-
100K
1.04
13.6%
20.4
20.4%
3.65
47.7%
1.40
18.3%
7.6
100%
E-16
-------
Tables E-9A - E-9F. Apportionment of NRTNEM Categories to Survey Categories
J=l J=2 J=3
ID=1 ID=2 ID=3 ID=4 ID=1 ID=2 ID=3 ID=4 ID=1 ID=2 ID=3 ID=4
H 1.00 1.00 H 1.00 0.94 H 1.00 1.00 0.18
MH 1.00 MH 0.06 0.65 MH 0.45
ML ML 0.35 0.63 ML 0.37 0.89
0.37 L 0.11
1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
(A) (B) (C)
J=4 J=5 J=6
ID=1 ID=2 ID=3 ID=4 ID=1 ID=2 ID=3 ID=4 ID=1 ID=2 ID=3 ID=4
H 1.00 0.11 H 1.00 0.28 H 0.98
MH 0.55 MH 0.72 0.07 MH 0.02 0.73
ML 0.34 0.83 ML 0.79 ML 0.27 0.69
0.17 L - 0.14 L 0.31
1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
(D) (E) (F)
-------
Table E-10. Distribution of Urban Population by Urban
Area Size, Urban Zone Population Density,
and Traffic Impact
Population
Density
H _>4500
MH 3000-
4499
ML 1500-
2999
L <_1499
High
Low
High
Low
High
Low
High
Low
Large
UA's
J-l
3.97
23.92
3.54
18.05
0
0
0
0
Medium Small UA's
J=2
0.67
5.27
1.01
6.47
0.88
6.39
0.21
1.77
J=3
0.32
3.59
0.32
3.46
0.63
7.20 '
0.05
0.53
J=4
0.20
2.56
0.47
5.27
0.80
8.35
0.10
1.05
J=5
0.15
1.83
0.19
2.40
0.38
5.02
0.07
0.89
J=6
0.07
0.98
0.11
1.44
0.24
3.46
0.09
1.31
E-18
-------
Table E-ll. Population Growth Factors by Place Size (Index J)
For Every Five Years in the Time Stream
AREA TYPE, J
PLACE SIZE,
THOUSANDS
YEAR VARIABLE
1974
5 198°
1986
1988
1990
1995
2000
2005
2010
1
OVER
2000
2
1000-
2000
3
500-
1000
4
200^
500
5
100-
200
6
50-
100
7
25-
50
8
5- -
25
9
ALL J
RURAL
POP (YEAR) / POP (BASELINE)
1
1
1
1
1
1
1
1
1
.00
.08
.17
.19
,22
.29
.36
.43
.50
1.00
1.07
1.16
1.19
1.22
1.29
1.36
1.44
1.51
1.00
1.07
1.16
1.19
1.22
1.29
1.36
1.44
1.51
1.00
1.02
1.04
1.05
1.05
1.07
1.08
1.10
1.12
1.00
1.02
1.04
1*.05
1.05
1.07
1.08
1.10
1.12
1.00
1.02
1.04
1.05
1.05
1.07
1.09
1.10
1.12
1.00
1.02
1.04
1.05
1.05
1.07
1.09
1.10
1.12
1.00
1.02
1.04
1.05
1.05
1.07
1.09
1.10
1.12
1.
1.
1.
1.
1.
1.
1.
1.
1.
00
12
23
27
31
39
48
57
65
-------
Table E-12. Populations (Millions) of Sampling Cells Used
in National Ambient Noise Survey (1980)
Large Medium/Small
UA's UA's
High Low High . Low
High 4.29 25.00 1.47 14.96
Medium-High 3.82 19.49 2.20 19.92
Medium-Low 0 0 3.07 31.7
Low 0 0 0.53 5.78
E-20
-------
APPENDIX F
THE SITES
LARGE URBAN AREA SITES
Urban Area
Boston, MA
Boston, MA
Boston, MA
Urban Zone
Boston, MA
Outside Central
City
Community Tract Block
Brighton, MA 4 102
Outside Central Quincy, MA
City
4180 717
West Newton, 3746
MA
107
Dates
Measured
Sun-Mon
Jul 6-7,
1980
Sun-Mon
Jul 6-7,
1980
Mon-Tue
Jul 7-8,
1980
Chicago, IL -
Northwestern IN
Outside Central
Cities
LaGrange Park, 8189
IL
320
Fri-Sat
May 2-3,
1980
Chicago, IL -
Northwestern IN
Chicago, IL -
Northwestern IN
Chicago, IL -
Northwestern IN
Chicago, IL -
Northwestern IN
Chicago, IL -
Northwestern IN
Chicago, IL -
Northwestern IN
Chicago, IL -
Northwestern IN
Outside Central Chicago 8293
Cities Heights, IL
Chicago, IL
Chicago, IL
Chicago, IL
Chicago, IL
Chicago, IL
Chicago, IL
Chicago, IL 2434
Chicago, IL 4402
Chicago, IL 631
221
104
305
102
Chicago, IL 315 302
Chicago, IL 5801 406
Chicago, IL 6109 109
Fri-Sat
May 2-3,
1980
Sat-Sun
May 3-4,
1980
Sun-Mon
May 4-5,
1980
Tue-Wed
Jun 16-17,
1981
Wed-Thur
Jun 17-18,
1981
Thu-Fri
Jun 18-19,
1981
Fri-Sat
Jun 19-20,
1981
F-l
-------
LARGE URBAN AREA SITES - Continued
Urban Area
Chicago, IL -
Northwestern IN
Chicago, IL -
Northwestern IN
Chicago, IL -
Northwestern IN
Chicago, IL -
Northwestern, IN
Chicago, IL -
Northwestern IN
Los Angeles -
Long Beach, CA
Los Angeles -
Long Beach, CA
Los Angeles -
Long Beach, CA
Los Angeles -
Long Beach, CA
New York, NY -
Northeastern NJ
New York, NY -
Northeastern NJ
Urban Zone
Chicago, IL
Chicago, IL
Chicago, IL
Chicago, IL
Chicago, IL
Los Angeles, CA
Los Angeles, CA
Outside Central
Cities
Outside Central
Cities
Outside Central
Cities
Outside Central
Cities
Community
Chicago, IL
Chicago, IL
Chicago, IL
Chicago, IL
Chicago, IL
Los Angeles,
CA
Los Angeles,
CA
Burbank, CA
Hacienda
Heights, CA
Garden City
Park, NY
Deer Park, NY
Tract
6501
6305
201
2316
609
1926
2352.
01
3118
4086.
01
303.
202
1227.
02
Dates
Block Measured
107 Sat-Sun
Jun 20-21,
1981
108 Mon-Tue
Jun 22-23,
1981
101 Tue-Wed
Jun 23-24,
1981
102 Thu-Fri
Jul 9-10,
1981
201 Tue-Wed
Jul 16-17,
1981
205 Thu-Fri
Mar 20-21,
1980
102 Thu-Fri
Mar 27-28,
1980
412 Mon-Tue
Mar 31-
Apr 1,
1980
116 Mon-Tue
Mar 31-
Apr 1,
1980
104 Thu-Fri
May 28-29,
1981
510 Fri-Sat
May 29-30,
1981
F-2
-------
LARGE URBAN AREA SITES - Continued
Urban Area
New York, NY -
Northeastern NJ .
New York, NY -
Northeastern NJ
New York, NY -
Northeastern NJ
New York, NY -
Northeastern NJ
New York, NY -
Northeastern NJ
Philadelphia,
PA - NJ
Philadelphia,
PA - NJ
Philadelphia,
PA - NJ
Philadelphia,
PA - NJ
Philadelphia,
PA - NJ
Philadelphia,
PA - NJ
Philadelphia,
PA - NJ
Urban Zone
Outside Central
Cities . .
Outside Central
Cities
New York, NY
New York, NY
New York, NY
Philadelphia,
PA
Outside Central
Cities
Outside Central
Cities
Outside Central
Cities
Philadelphia,
PA
Philadelphia,
PA
Philadelphia,
PA
Community
Levitown, NY
Sayville, NY
Brooklyn, NY
Brooklyn, NY
Bronx, NY
Philadelphia,
PA
*
Warainster,
PA
Narbeth,
PA
Collingdale,
PA
Philadelphia,
PA
Philadelphia,
PA
Philadelphia,
PA
Tract
5202
1473.
02
198
374
394
148
1016.
04
2056
4031.
03
238
259
348
Block
108
311
201
504
104
105
33
201
106
201
401
206
Dates
Measured
Fri-Sat
May 29-30,
1981
Sat-Wed
May 30-
Jun 3, 1981
Tue-Wed
Jun 2-3,
1981
Tue-Wed
Jun 2-3,
1981
Wed-Thu
Jun 4-5,
1981
Sun-Thu
May 3-7,
1981
Mon-Tue
May 4-5,
1981
Tue-Wed
May 5-6,
1981
Wed-Thu
May 6-7,.
1981
Thu-Fri
May 7-8,
1981
Fri-Sat
May 8-9,
1981
Fri-Sat
May 8-9,
1981
F-3
-------
LARGE URBAN AREA SITES - Continued
Urban Area
San Francisco -
Oakland, CA
San Francisco -
Oakland, CA
Urban Zone
Community
Tract
Outside Central El Cerrito, CA 3870
Cities
Oakland, CA
Oakland, CA 4080
Dates
Block Measured
205
101
Mon-Tue
Mar 9-10,
1981
Tue-Wed
Mar 10-11,
1981
San Francisco -
Oakland, CA
San Francisco,
CA
San Francisco, 307
CA
215
Wed-Thu
Mar 11-12,
1981
Washington, DC -
MD - VA
Washington, DC -
MD - VA
Outside Central Alexandria, 2003.
City VA 02
203
Washington, DC Washington, DC 78.04 501
Sun-Mon
Jul 6-7,
1980
Thu-Fri
Jul 10-11,
1980
Washington, DC -
MD - VA
Washington, DC -
MD - VA
Outside Central Mt. Rainer, MD 8047
City
204
Washington, DC Washington, DC 52.01 107
Thu-Fri
Jul 10-11,
1980
Fri-Sat
Jul 11-12,
1980
F-4
-------
SMALL URBAN AREA SITES
Urban Area
Allentown -
Bethlehem -
Easton, PA, NJ
Allentown -
Bethlehem -
Easton, PA, NJ
Augusta, GA
SC
Urban Zone
Community
Tract Block
Bethlehem, PA Bethlehem, PA 101 109
Bethlehem, PA Bethlehem, PA 176 715
Outside Central Augusta, GA
City
103 208
Dates
Measured
Sun-Mon
May 10-11,
1981
Sun-Mon
May 10-11,
1981
Sun
Aug 10,
1980
Binghamton, NY
Outside Central Owego, NY
City
203 905
Tue-Wed
May 12-13,
1981
Binghamton, NY
Buffalo, NY
Buffalo, NY
Cleveland, OH
Outside Central Union, NY
City
133.01 318
Binghamton, NY Bingharaton, NY Binghamton, NY 18 202
Outside Central Niagara Falls, 224 411
City NY
Outside Central Niagara Falls, 201 101
City NY
Cleveland, OH Cleveland, OH 1192 104
Tue-Wed
May 12-13,
1981
Wed-Thu
May 13-14,
1981
Sun-Mon
May 17-18,
1981
Sun-Mon
May 17-18,
1981
Thu-Mon
Jun 4-8,
1981
Cleveland, OH
Cleveland, OH
Cleveland, OH
Outside Central
City
Outside Central
City
Wycliff, OH
2009
Cuyahoga Hts., 1659
OH
204
102
Outside Central Middleburg 1731 509
City Heights, OH
Fri-Sat
Jun 5-6,
1981
Sat-Sun
Jun 6-7,
1981
Sun-Mon
Jun 7-8,
1981
F-5
-------
SMALL URBAN AREA SITES - Continued
Urban -Area
Duluth -
Superior,
MN - WI
Duluth -
Superior,
MN - WI
Duluth -
Superior,
MN - WI
Duluth -
Superior,
MN - WI
Fargo-Moorhead ,
ND - MN
Fargo-Moo'rhead ,
ND - MN
Fargo-Moorhead ,
ND - MN
Fargo-Moorhead,
ND - MN
Fresno, CA
Madison, WI
Milwaukee, WI
Milwaukee , WI
Urban Zone
Duluth, MN
Duluth, MN
Superior, WI
Superior, WI
Fargo, ND
Moorhead, MN
Moorhead, MN
Fargo, ND
Fresno, CA
Outside Central
City
Milwaukee, WI
Outside Central
City
Community
Duluth, MN
Duluth, MN
Superior, WI
Superior, WI
Fargo, ND
Moorhead, MN
Moorhead, MN
Fargo, ND
Fresno, CA
Madison, WI
Milwaukee, WI
Menomonee
Falls, WI
Dates
Tract Block Measured
33 408 Mon-Tue
May 12-13,
1980
14 106 Mon-Tue
May 12-13,
1980
204 124 Tue-Wed
May 13-14,
1980
210 109 Tue-Wed
May 13-14,
1980
9 422 Thu-Fri
May 15-16,
1980
206 405 Thu-Fri
May 15-16,
1980
203 220 Thu-Fri
May 15-16,
1980
5 506 Fri-Sat
May 16-17,
1980
11 120 Sat-Sun
Feb 28-
Mar 1,
1981
16.02 104 Wed-Thu
Jul 1-2,
1981
170 106 Fri-Sat
Jun 26-27,
1981
2001 425 Fri-Sat
Jun 26-27,
1981
F-6
-------
SMALL URBAN AREA SITES - Continued
Urban Area
Minneapolis -
St. Paul, MN
Urban Zone
Community
Tract Block
Outside Central Plymouth, MN 266.02 112
Cities
Dates
Measured
Sun-Mon
Jun 28-29,
1981
Minneapolis -
St. Paul, MN
Outside Central Bloomington, 252 209
Cities MN
Mon-Tue
Jun 29-30,
1981
Minneapolis -
St. Paul, MN
Minneapolis,
MN
Minneapolis,
MN
109
413
Tue-Wed
Jun 30-
Jul 1,
1981
Minneapolis -
St. Paul, MN
Minneapolis,
MN
Muskegon - Muskegon, MI
Muskegon Hgts. ,
MI
Minneapolis, 59 103
MN
Muskegon, MI 10 113
Tue-Wed
Jun 30-
Jul 1,
1981
Thu-Fri
' Jun 11-12,
1981
Muskegon - Muskegon, MI
Muskegon Hgts.,
MI
Muskegon, MI
04 501
Thu-Fri
Jun 11-12,
1981
Muskegon - Muskegon Heights, Muskegon 14.02
Muskegon Hgts., MI Heights, MI
MI
606
Fri-Sat
Jun 12-13,
1981
Muskegon -
Outside Central Norton Shores, 26.01 201.
Muskegon Hgts., Cities
MI
MI
Sat-Sun
Jun 13-14,
1981
Oxnard -
Ventura-
Thousand Oaks, CA
Thousand Oaks,
CA
Oxnard - Thousand Oaks,
Ventura - CA
Thousand Oaks, CA
Thousand Oaks, 72.01
CA
Thousand Oaks, 63
CA
Oxnard - Outside Central El Rio, CA
Ventura - Cities
Thousand Oaks, CA
50
307
102
216
Tue-Wed
Apr 8-9,
1980
Thu-Fri
Apr 10-11,
1980
Thu-Fri
Apr 10-11,
1980
F-7
-------
SMALL URBAN AREA SITES - Continued
Urban Area
Urban Zone
Oxnard - Outside Central
Ventura - Cities
Thousand Oaks, CA
Portland, ME
Portland, ME
Community
Tract Block
Port Hueneme, 42 213
CA
Portland, ME
13 104
Dates
Measured
Tue-Wed
Apr 15-16,
1980
Thu-Fri
May 21-22,
1981
Portland, ME
Portland, ME
Portland, ME 01 301
Thu-Fri
May 21-22,
1981
Portland, OR -
WA
Portland, OR -
WA
Providence -
Pawtucket -
Warwick, RI - MA
Portland, OR
WA
Portland, OR
WA
Pawtucket, Rl
Providence - Outside Central
Pawtucket - Cities
Warwick, RI, MA
Reno, NV
Reno, NV
Reno, NV
Reno, NV
Portland, OR 17.01 516
Portland, OR 66.02 113
Pawtucket, RI 157 413
E. Greenwich, 209.02 .108
CT
Reno, NV
Reno, NV
05 121
03 115
Mon-Tue
Apr 21-22,
1980
Mon-Tue
Apr 21-22,
1980
Fri-Sat
May 22-23,
1981
Sat-Sun
May 23-24,
1981
Mon-Tue
Mar 2-3,
1981
Mon-Tue
Mar 2-3,
1981
Reno, NV
Rochester, NY
Outside Central Sparks, NV
City
Rochester, NY
19
106
Rochester, NY 82 209
Tues-Wed
Mar 3-4,
1981
Mon-Tue
Jun 30-
Jul 1,
1980
Rochester, NY
Rochester, NY Rochester, NY 78 601
Tue
Jul 1,
1980
F-8
-------
SMALL URBAN AREA SITES - Continued
Urban Area
Rochester, NY
Urban Zone
Community
Tract
Outside:Central E. Rochester, 120
City NY
Block
403
Dates
Measured
Wed-Thu
Jul 2-3,
1980
Rochester, NY
Rochester, NY
Rochester, NY 78 512
Wed-Thu
Jul 2-3,
1980
Saginaw, MI
Saginaw, MI
Saginaw, MI
103
907
Wed-Thu
Jun 10-11,
1981
Saginaw, MI
San Antonio, TX
San Antonio, TX
San Antonio, TX
San Antonio, TX
San Bernardino -
Riverside, CA
San Bernardino -
Riverside, CA
San Bernardino -
Richmond, CA
San Bernardino -
Riverside, CA
Outside Central
City
Outside Central
City
Carrolton, MI 107
416
Lackland, TX 1719 -218
San Antonio, TX San Antonio, 1702
TX
San Antonio, TX San Antonio, 1212
TX
San Antonio, TX San Antonio
TX
1410
San Bernardino, San Bernardino, 58
CA CA
San Bernardino,
CA
Outside Central
Cities
San Bernardino, 62
CA
Loma Linda, CA 73
Outside Central Bloomington, 36
Cities CA
310
207
308
203
205
606
401
Wed-Thu
Jun 10-11,
1981
Mon-Tue
Apr 6-7,
1981
Mon-Tues
Apr 6r7,
1981
Tue-Wed
Apr 7-8,
1981
Tue-Wed
Apr 7-8,
1981
Wed-Thu
Feb 25-26,
1981
Wed-Thu
Feb 25-26,
1981
Thu-Fri
Feb 26-27,
1981
Thu-Fri
Feb 26-27,
1981
F-9
-------
SMALL URBAN AREA SITES - Continued
Urban Area
San Diego, CA
Urban Zone
Community
Tract Block
Outside Central Rancho 170.01 112
City Bernardo, CA
San Diego, CA San Diego, CA San Diego, CA 80.02 122
San Diego, CA San Diego, CA San Diego, CA 25.02 109
Dates
Measured
Fri-Thu
Feb 13-19,
1981
Sat-Sun
Feb 21-22,
1981
Sat-Sun
Feb 21-22,
1981
San Diego, CA
San Jose, CA
San Jose, CA
Savannah, GA
Savannah, GA
Savannah, GA
Savannah, GA
Savannah, GA
Scranton, PA
Outside Central
City
Outside Central
City
San Jose, CA
Savannah, GA
Savannah, GA
Savannah, GA
Savannah, GA
Savannah, GA
Scranton, PA
Oceanside, CA 185.04 104
Campbell, CA 5065. 210
03
San Jose, CA 5079. - 101
02
Savannah, GA
28 412
Savannah, GA 30 122
Savannah,
Savannah ,
Savannah ,
Scranton,
GA
GA
GA
PA
24
38
104
10
209
312
112
208
Mon-Tue
Feb 23-24,
1981
Fri-Wed
Mar 6-11,
1981
Fri-Sat
Mar 6-7,
1981
Wed-Thu
Jul 16-17,
1980
Wed-Thu
Jul 16-17,
1980
Fri-Sat
Jul 18-19,
1980
Sat-Sun
Jul 19-20,
1980
Mon-Tue
Jul 21-22,
1980
Tue-Wed
Jun 24-25,
1980
F-10
-------
SMALL URBAN AREA SITES - Continued
Urban Area
Scranton, PA
Scranton, PA
Scranton, PA
Springfield -
Chicopee -
Holyoke, MA - CT
Springfield - .
Chicopee -
Holyoke, MA - CT
Springfield -
Chicopee -
Holyoke, MA - CT
Syracuse, NY
Syracuse, NY
Tampa , FL
Tampa, FL
Tampa, FL
Tampa , FL
Urban Zone
Scranton, PA
Outside Central
City
Outside Central
City
Outside Central
Cities
Chicopee Falls,
MA
Chicopee Falls,
MA
Syracuse, NY
Syracuse, NY
Outside Central
City
Outside Central
City
Tampa , FL
Tampa, FL
Community
Scranton, PA
Dickson City,
PA
Blakely, PA
Enfield, CT
Chicopee
Falls, MA
Chicopee
Falls, 'MA
Syracuse, NY
Syracuse, NY
Lake Carroll,
FL
University,
FL
Tampa, FL
Tampa, FL
Tract Block
16 311
115 107
112 225
4805 121
8111 115
8113 307
108 107
125 215
113 211
108 502
33 405
18 703
Dates
Measured
Tue-Wed
Jun 24-25,
1980
Wed-Thu
Jun 25-26,
1980
Thu-Fri
Jun 26-27,
1980
Sun-Mon
May 24-25,
1981
Mon-Tue
May 25-26,
1981
Mon-Tue s
May 25-26,
1981
Thu-Fri
May 14-15,
1981
Thu-Fri
May 14-15,
1981
Fri-Wed
May 10-15,
1981
Fri-Sat
Apr 10-11,
1981
Mon-Tue
Apr 13-14,
1981
Tue-Wed
Apr 14-15,
1981
F-ll
-------
SMALL URBAN AREA SITES - Continued
Dates .
Urban Area Urban Zone Community Tract Block Measured
Waterbury, CT Waterbury, CT Waterbury, CT 3523 105 Tue-Wed
May 26-27,
1981
West Palm Beach, W. Palm Beach, W. Palm Beach, 28 114 Thu-Fri
FL FL. FL Apr 16-17,
1981
F-12
------- |