EPA-450/3-75-077
SELECTING SITES
FOR CARBON MONOXIDE
MONITORING
by
F.L. Ludwig and J.H.S, Kealoha
Stanford Research Institute
Menlo Park, California 94025
Contract No. 68-02-1471
EPA Project Officer: NeilJ. Berg, Jr.
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Caroolin 27711
September 1975
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors
and grantees, and nonprofit organizations - as supplies permit - from
the Air Pollution Technical Information Center, Environmental Protection
Agency, Research Triangle Park, North Carolina 27711; or, for a
fee, from the National Technical Information Service, 5285 Port Royal
Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency
by Stanford Research Institute, Menlo Park, California, in fulfillment
of Contract No. 68-02-1471. The contents of this report are repro-
duced herein as received from the Stanford Research Institute. The
opinions, findings, and conclusions expressed are those of the author
and not necessarily those of the Environmental Protection Agency.
Mention of company or product names is not to be considered as an
endorsement by the Environmental Protection Agency.
Publication No. EPA-450/3-75-077
11
-------
ABSTRACT
This report presents procedures and criteria for selecting appro-
priate locations for carbon monoxide (CO) monitoring stations. The
purposes for which CO concentrations are measured are reviewed and
classified according to a system based on special scales of representa-
tiveness. Different purposes require measurements representative of
areas of differing size. The first step of the site selection process
is to decide the purpose of the measurements. Then this primary purpose
must be related to the kind of area that should be represented by the
measurements. A matrix of purposes and spatial scales is included to
assist in this determination.
Procedures are given for selecting locations that will provide CO
measurements representative of downtown street canyon areas, along major
traffic corridors, urban neighborhoods, and larger interurban regions.
Specific recommendations are included for inlet heights, distances from
major and minor roadways and placement relative to urban areas. Less
detailed discussions of monitoring around indirect sources such as
shopping centers and stadia are included. The rationale behind the
specific recommendations is given. In general, the objective has been
to place the monitor so that it is not disproportionately influenced by
any one source within the area to be represented.
Appendices discuss sources of information useful to the site
selection process, such as climatological data, land use information,
and traffic data. A bibliography is also included. It is classified
according to monitoring purposes and scales of representativeness. A
computer program designed to identify "worst-case" conditions and the
relative contributions of sources at different distances is presented.
111
-------
CONTENTS
ABSTRACT iii
LIST OF ILLUSTRATIONS vii
LIST OF TABLES x
ACKNOWLEDGMENTS xi
SUMMARY xii
I INTRODUCTION 1
A. Monitoring Site Standards 1
B. Philosophy of Approach 1
C. Special Characteristics of Carbon Monoxide that
Affect Monitoring Site Selection 2
D. Organization of this Report 4
II DECIDING THE TYPE OF MEASUREMENTS THAT ARE TO BE MADE . 7
A. Uses of Carbon Monoxide Measurements 7
B. Scale of Representativeness ..... 9
C. Relative Importance of the Different Scales of
Measurement 12
D. Selecting the Required Scale of Representativeness. 14
iv
-------
CONTENTS (Continued)
III SELECTING STATION LOCATION 21
A. Background 21
B. Regional Stations 21
C. Neighborhood Stations . , 30
D. Middle Scale Stations 35
1. General 35
2. Street Canyon Sites 38
3. Roadway Sites 42
4. Indirect Source Sites 46
E. Other Types of Stations 49
1. Microscale 49
2. Urban and National Scales 51
3. Global Scale 53
IV RATIONALE FOR SITE SELECTION CRITERIA 55
A. Background 55
B. Inlet Locations 56
C. Minimum Separation Between Monitoring
Sites and Sources 60
D. The Importance of Sources at Various Distances
from the Monitor 64
-------
CONTENTS (Concluded)
APPENDICES
APPENDIX A A SIMPLE MODEL OF CONCENTRATION/EMISSION
RELATIONSHIPS 75
APPENDIX B SOURCES OF TRAFFIC INFORMATION 105
APPENDIX C SOURCES OF CLIMATOLOGICAL AND METEOROLOGICAL
INFORMATION 115
APPENDIX D SOURCES OF LAND USE INFORMATION 121
APPENDIX E BIBLIOGRAPHY 129
REFERENCES 145
VI
-------
ILLUSTRATIONS
1 Measured Hour-Average CO Patterns at Two Heights for a
San Jose, California Intersection During Late Afternoon
on 11 December 1970 5
2 Schematic Diagram of a Procedure for Locating Regional
Stations 22
3 Example of Earth Resources Technology Satellite (ERTS)
Photograph and a Mosiac of the U.S 23
4 Aerial Photograph of a Rural Area 24
5 Example of Wind Roses 26
6 Schematic Diagram of Appropriate Areas for a Regional
Monitoring Site 28
7 Schematic Diagrams of Appropriate Siting Areas for
Regional Monitors When Two Sites are Planned 29
8 Schematic Diagram Illustrating the Unsuitability of Areas
Aligned with Major Roads 30
9 Example of a Site with Surroundings Appropriate to
Regional Monitoring of CO 31
10 Schematic Diagram of a Procedure for Locating Neighborhood
Monitoring Stations . 32
11 Portion of a Typical Traffic Map 34
12 Aerial Photographs of Urban Residential Neighborhoods . 36
13 A Typical Urban Neighborhood Depicted on a Topographical
Map 37
14 Schematic Diagram of a Procedure for Locating Street
Canyon Stations 39
VII
-------
ILLUSTRATIONS (Continued)
15 Example of a Pair of One-Way Streets with Peak
Traffic on One in the Morning and in-the Afternoon
on the Other 41
16 Sample Inlet in a Downtown Street Canyon . 43
17 Schematic Diagram of a Procedure for Locating Traffic
Corridor Stations 44
18 An Example of a Bag Sampling Array Near a Highway. ... 47
19 An Example of the Distribution of Normalized CO
Concentration Downwind of Elevated Roadways 48
20 Diagram of Sensor and Air Inlet Location for an
Experiment to Study Distributions of CO in Street
Canyons 50
21 Diagram of Sensor and Air Inlet Location for an
Experiment to Study Diffusion Near a Highway 52
22 The Vertical Distribution of CO Concentration in
a Street Canyon with Traffic Volume of 1,500
Vehicles/Hour 57
23 Average Distribution of CO Concentration in a Street
Canyon for Different Wind Conditions ..... 58
24 Normalized Concentrations Computed with a Gaussian
Dispersion Model 65
25 Averages of the 1969-1970 Annual Maximum Hourly CO
Concentrations Versus Slant Distances to the Street . . 67
26 Daily Traffic Cycle (Los Angeles) 68
27 Frequency Distributions of the Ratios of Street
Contributions to City Contributions for 8-hour Average
CO Concentrations . 70
viii
-------
ILLUSTRATIONS (Concluded)
28 Frequency Distributions of the Ratios of Contributions to
8-hour Average CO Concentrations from Sources Nearer and
Farther than 2-km 71
29 Frequency Distribution of the Ratios of Street Contributions
to Citywide Contributions for the Highest 8-hour Average
CO Concentrations 72
A-l Diagram of Segments Used for Spatial Partitioning of
Emissions 78
A-2 Vertical Diffusion as a Function of Travel Distance and
Stability Category for Rural Areas 79
A-3 Vertical Diffusion as a Function of Travel Distance and
Stability Category for Urban Areas 80
A-4 Simplified Flowchart of Computer Program 89
B-l Sample Traffic Map for a Downtown Area 110
B-2 Sample Traffic Map for an Intercity Area, Virginia . . . Ill
B-3 Sample Traffic Map for an Intercity Area, Colorado . . . 112
B-4 Sample of Gridded Traffic Data 113
C-l Example of Meteorological Records Available from the
National Climatic Center ...... 117
D-l An Example of the Informational Content of the Large-,
Medium-, and Small-Scale Topographic Map 125
D-2 A Portion of the Orthophotoquad Index Showing the Legend
and a Portion of the State of Florida 126
D-3 A Sanborn Map for a Section of Portland, Oregon .... 128
IX
-------
TABLES
1 Nationwide Estimates of Carbon Monoxide Emissions ... 3
2 Scales of Measurement Applicable to Various Purposes . . 16
3 Example of a Tabulated Wind Summary 25
4 Values of Constants Used to Represent Vertical
Dispersion as a Function of Downwind Travel Distance . . 62
5 Minimum Distances (km) Between a Large Roadway and
a Neighborhood Monitoring Site . 62
A-l Mixing Depth Classes . 83
A-2 Contributions to the Concentration at a Point from
Emissions of Unit Strength at Various Upwind Distance
Intervals, Rural 84
A-3 Contributions to the Concentration at a Point from
Emissions of Unit Strength at Various Upwind Distance
Intervals, Urban 85
A-4 FORTRAN Variables Used in SMOCER 91
A-5 Inputs for a Simple Model of Concentration/Emission
Relationships 93
B-l Example of Tabulated Traffic Volumes 109
C-l Published Volumes of Worldwide Airfield Summaries . . . 119
D-l National Topographic Maps 124
D-2 Picture Products Available from ERTS 127
E-l Bibliography Index, Arranged According to Measurement
Purposes and Scales 131
-------
ACKNOWLEDGMENTS
The authors have been greatly assisted by the comments and tech-
nical advice of many people, particularly Dr. Wayne Ott and Mr. Neil Berg
of the Environmental Protection Agency and Messrs. Eugene Shelar, Hisao
Shigeishi, and Ms. Sarita Skidmore of Stanford Research Institute. We
are also indebted to Ms. Kathey Mabrey, Mr. Gary Parsons, and Mr. Roger
Bass for their contributions to the preparation of this report.
XI
-------
SUMMARY
It does not require an extensive review of the literature to
conclude that many measurements of carbon monoxide (CO) concentrations
are being made and used for a wide variety of purposes. In fact, it
sometimes appears that the number of purposes exceeds the number of
stations. It is very seldom that an attempt is made to show that the
physical characteristics of a given sampling location are appropriate to
the problem being addressed with the data that are collected there.
This may be because a coherent scheme has not been devised for classi-
fying sites and relating their characteristics to intended data usage.
There is a great need for a site classification system since most
monitoring stations will be expected to operate for many years, tens of
thousands of dollars will be spent for equipment and installation, and
further tens of thousands will be expended for maintenance and data pro-
cessing during the period of operation. It would be foolish to locate a
station where its data would not be appropriate to the intended uses.
The costs of poor siting procedures can extend well beyond the actual
costs of establishing and operating the stations. Whenever the data are
used as a rationale for large-scale programs such as air quality improve-
ment projects, there are likely to be great economic and social impacts,
which means that the siting of data collection stations becomes even
more important.
The uses of CO data can be broadly categorized as being related to
the following:
• Enforcement of air quality regulations
• Development and evaluation of control measures
• Public health
• Scientific research
• Miscellaneous purposes.
Each of these broad categories contains several narrower subcategories
of use, but at no level of classification is this system directly re-
lated to physical factors. In trying to devise a site classification
system that can be used to define an appropriate set of physical char-
acteristics for each site type, it seems reasonable to examine the uses
of the data in terms of the physical factors that influence the data.
For example, each of the above categories of use has a level of spatial
smoothing that is appropriate for that use. Sometimes, the requirement
is for many closely spaced measurements that will reveal small—scale
features of the distribution of CO in space; at other times the
xii
-------
requirement is for measurements to typify a whole city or perhaps even a
whole region of the country.
Those uses of data that are related to the development of air
quality control measures illustrate the variety of spatial scales that
can be of concern within a single, broad category of use. Measurements
that are typical of a street canyon or a freeway corridor will be
suitable for formulating control plans to reduce emissions along the
specific roadway or along a limited number of similarly congested
streets, but they will not necessarily be useful for the formulation of
plans that are city-wide in scope. Plans of the latter type require
data that represent much larger areas.
The concept of spatial representativeness arises from examples like
those given above and provides a useful basis for classifying stations
and the uses to which their data are put. Furthermore, it has a
physical basis that can serve to define station characteristics. In
general, the measurement scales that are of greatest importance are:
• Microscale, to define concentrations in volumes with
dimensions of the order of meters to tens of meters.
• Middle scale, generally defining concentrations typical of
areas with dimensions of tens to hundreds of meters. This
category includes measurements to define concentrations along
streets and roads and typical areas can be elongated,
measuring tens of meters by hundreds of meters or even
kilometers.
• Neighborhood scale, defining concentrations within some
extended area of the city that has relatively uniform land
use; dimensions are of the order of kilometers.
• Urban scale, to define the overall, citywide conditions on
a scale of tens of kilometers; in general, more than one
site will be required for such definition.
• Regional scale, to provide a measure of CO concentrations
typical of a large, usually rural area of reasonably homogeneous
geography and extending for tens to hundreds of kilometers.
Of course, the concept of scale can be extended upward to the
global scale, but for this discussion, those scales enumerated above are
sufficient. In fact, the discussion is limited to the middle, neighbor-
hood and regional scales. Urban scale conditions, as noted, cannot be
specified with measurements at a single location, but can be synthesized
from measurements representing middle and neighborhood scales. When the
latter two scales are well represented, it will not be necessary to find
the ideal urban scale site. Microscale measurements are often made for
very specialized research purposes that make a generalized set of siting
Xlll
-------
requirements nearly impossible to devise. Factors other than scale of
representativeness are important to the definition of types of moni-
toring site and their required physical characteristics. One of these
other factors arises because virtually all routine CO monitoring is
related directly, or indirectly, to health or public exposure and
therefore samples taken at breathing level are most appropriate. Inlets
at breathing level will constitute an obstacle in many locations and
will also be subject to vandalism. Ott (1975) has suggested 3 m i 0.5 m
as a compromise and that same height range has been adopted here.
The adoption of the same inlet height as recommended by Ott pro-
vides consistency. Consistency, in turn, allows data from different
locations to be compared with the assurance that any differences in the
data sets will reflect something other than anomolous site character-
istics. The site standards that Ott (1975) proposed were derived from
a combination of empirical evidence and practical requirements. The
representativeness concept, not surprisingly, leads to similar results
so that consistency between the recommendations of this study and those
of Ott is maintained.
In proceeding from the concept of spatial representativeness to the
concrete requirements of siting, some arbitrary decisions are required;
these decisions are made easier if they can be translated into very
simple quantitative terms. The most pervasive of the translations that
were used provided equivalence between the contribution that a specific
source makes to observed concentrations at a monitor and the distance to
that source. From this relationship, maximum allowable contributions
(in mg/m3 ) could be converted to minimum separations (in m) between the
monitor and certain kinds of sources. For instance, regional sites
should be far enough removed from cities so that the concentrations
arising from emissions within the city will be less than typical rural
background concentrations of 0.2 mg/m3. Similarly, neighborhood sites
should not be so close to large roadways that the emissions from the
roadway increase measured concentrations more than about 1 mg/m3.
Because smaller roadways are less important contributors to the overall
neighborhood concentrations, they should be permitted even less
influence on the observed concentrations.
The relationships between minimum separation distance, source to
monitor, and the maximum concentration impact of the source on the
monitor can be derived from common diffusion equations. The physical
requirements of the major site types are summarized below. These
requirements have been derived according to the principles outlined
above and are described in much greater detail in the body of this
report.
xiv
-------
Street Canyon Monitor
Ott type:
Scale of
representativeness:
Inlet location:
Other requirements:
Traffic Corridor Monitor
Ott type:
Scale of
representativeness:
Inlet location:
Other requirements:
"A" - downtown pedestrian exposure
station.
Middle
3 - 0.5m, over center of sidewalk,
but at least 2 m from building front
and 10 m from intersection.
Street canyon width and depth
should be typical of others in the
area. Traffic should be either average
or maximum for the area, depending on
intended data use. Avoid bus stops,
loading zones and other unusual source
areas. Traffic counts would be
valuable. Orientation relative to wind
directions, use of multiple inlets, and
the importance of one-way streets and
daily traffic cycles are discussed in
the text (Section III).
None comparable
Middle
3 - 0.5 m, at edge of right-of-way, or
at typical distance to nearest
residences.
Roadway should be at grade-level and
as typical of its type as possible.
Site should be well removed from
intersections and overpasses or other
large obstructions. Traffic counts and
wind measurements would be valuable.
Orientation relative to wind directions
and the use of multiple inlets are
discussed in the text, as are nongrade-
level roadways (Section III).
xv
-------
Neighborhood Monitor
Ott type:
Scale of
representativeness:
Inlet location:
Other requirements:
Regional Monitor
Ott type:
Scale of
representativeness:
Inlet location:
Other requirements:
"C" - Residential population exposure
station.
Neighborhood
3 - 0.5m. 35 m from nearest traffic.*
2.5 km from nearest major roadway,
50,000 vehicles per day or more.
Reasonably homogeneous land use within
1 or 2 km of the site. Large cities,
of diverse land use, should probably
have enough stations of this type to
characterize the variety of neighbor-
hoods in the town. Traffic counts on
nearby streets and wind measurements
would be valuable.
'E" - Nonurban background stations.
Regional
3 3 0.5 m. Small vertical gradients
make heights of up to 10 m acceptable
in such locations if absolutely
necessary. 35 km from nearest city, in
the direction that is least frequently
downwind; if more than one such station
is planned, see Section III for
preferred directions. 5 km from
nearest major roadway — 50,000 or more
vehicles per day. 400 m from nearest
traffic.
Should not be aligned with any long
straight section of a major roadway.
Low-lying areas, subject to cold air
drainage and stagnation should be
avoided. Wind measurements are
desirable.
Ott (1975) specifies a minimum distance of 100 m to nearest
street with more than 500 vehicles per day, but this seems unduly
stringent.
xvi
-------
The above summaries of the physical characteristics of the most
important types of monitors make mention of traffic on certain roadways,
of homogeneity of land use, and of other considerations that imply that
it is not sufficient to simply place an inlet at 3 m height at some
arbitrary location. In general, some areas are preferable to others,
and within the preferred areas there are some specific sites that are
better than others. Step by step procedures for site selection are
described in Section III. The first step is usually to gather traffic
data, land use data, topographic maps, and climatological information.
Sources of these kinds of information are discussed in appendices to
this report. The information is used to identify generally desirable
areas, and reduce their numbers and their size until a final selection
is carefully made. When this has been done, the data should be
reasonably representative of the desired conditions.
The original premise that served as a basis for categorizing the
sites and defining their physical requirements was that the purposes for
data collection could be classified according to appropriate scales of
representativeness. It was suggested, for example, that air quality
improvement measures should be consistent in scale with the measurements
that inspired them. This premise has been tested by simple calculations
and it appears to be valid.
The simple diffusion model described in Appendix A was applied with
a wide variety of meteorological data and some reasonable, idealized,
urban distributions of CO emissions. The results show that a street
canyon monitor on a typical, heavily traveled street will measure
concentrations that are strongly influenced by the local traffic. In
fact, it appears that more than half of the time the adjacent street
will contribute over one-third of the observed CO. Such results suggest
that observations are substantially influenced by emissions within a few
blocks of the monitor (and that control plans should be designed
accordingly).
As is to be expected, neighborhood stations are not so dra-
matically influenced by sources in their immediate vicinity. In fact,
the siting criteria specifically seek to exclude such influences.
However, the influences of sources within a few kilometers should be
important if the scale of representativeness of the measurements is to be
used to suggest appropriate scales for practical applications. The
model from Appendix A was again used to estimate the region of influence.
It was found that in most instances, the contribution of sources within
2 km of the monitor was more than one-third of the total, again demon-
strating a reasonable relationship between the scale of representativity
of the measurement and the appropriate scales for interpretation and
application of the data.
xvii
-------
In summary, the guidelines presented here should serve as a good
basis for selecting sites that can be classified into a limited number
of comparable types. The standardization of physical characteristics
will ensure that comparison among sites of the same type will not be
clouded by anomolies in the data, arising from peculiarities in the
siting. Furthermore, the guidelines are sufficiently consistent with
those already proposed by Ott (1975) that there s-hould be little
difficulty in combining the two schemes.
Use of the classification scheme does more than ensure compati-
bility of dataj it also provides a physical basis for the interpretation
and application of those data. This should help to prevent mismatches
between what the data actually represent and what they are interpreted
to mean. If carefully considered selection of monitoring sites could
prevent one instance where a large-scale control plan is designed to
cure a small-scale problem, then that alone would probably justify the
effort required for proper selection of monitoring sites.
xviii
-------
I INTRODUCTION
A. Monitoring Site Standards
As the importance of air pollution increases, so does the necessity
to measure the concentrations of important pollutants in the air. Many
reasons for making the measurements are discussed later in this report;
they include attempts to understand pollutant behavior in the
atmosphere, assessment of the effectiveness of control measures, and
assessment of public health effects. Each purpose is best served by
some combination of monitoring site characteristics. This report is an
attempt to codify the monitoring site characteristics suitable for each
of a variety of monitoring purposes and to provide procedures for
selecting sites with characteristics optimum to the intended purposes.
The need for such codification and selection procedures can be
illustrated with a few examples. If we wish to use measurements to
estimate the public health effects of carbon monoxide (CO), the
measurements must be made in areas where concentrations are representa-
tive of those to which the public is exposed. In many instances, the
air has been sampled for this purpose at heights where the public has no
access. Any vertical gradients of concentration will greatly reduce the
usefulness of such data. Intercity comparisons of air quality must be
based on data from comparable stations, which means that some station
classification scheme will be required to judge the degree of compara-
bility of the stations. Ott (1975) has shown that the range of average
CO concentrations from four different locations in a single city, New
York, is greater than the range of the measurements for 79 other sites
in locations throughout the rest of the country. In fact, the range
within New York was almost twice that of all the other sites. New York
is not unique in this respect. In any large city there will be loca-
tions with widely differing concentrations, many of which are not repre-
sentative of the city's general air quality. In fact, the diversity of
measured concentrations and the diversity of land use suggest that there
may be no one station that is representative of the entire city.
Therefore, stations should probably be chosen to represent various
aspects of the city's CO concentration distribution.
The preceding examples were presented only to illustrate a need for
logical, consistent procedures that can be used to locate and categorize
CO monitoring stations. The remainder of the report discusses various
aspects of that topic.
B. Philosophy of Approach
An inspection of the interpretations given to concentration data
reveals that the measurements are assumed to represent areas and volumes
-------
that extend well beyond the small volume that is actually sampled. The
area presumed to be represented by a measurement may be relatively
small, such as one side of a downtown street canyon; intermediate in
size, like a neighborhood; or much larger, like a city, or even a whole
region of the country.
The methods for locating stations that have evolved during this
study have been based on the premise that the major purposes of
establishing a certain monitoring site can be identified and then paired
with a scale of representativeness that is most suitable for those
purposes. Thus, the site selection process begins by identifying the
purpose of the monitoring. Then, this purpose provides a basis for
selecting a station type, based on the area that the measurements should
represent. Finally, procedures are followed that lead to sites that
represent areas of the appropriate size.
Although the site selection procedures are simple, they require
labor. It is inconsistent to locate a monitoring station that will be
expected to operate for many years, and will cost tens of thousands of
dollars for instrumentation and facilities without commensurate
expenditure of effort toward site selection. Furthermore, the
measurements made at the monitoring site may serve as the basis for
large-scale air quality improvement plans with enormous economic or
social impact, again dictating a careful site selection process.
All costs and potential costs associated with monitoring stations
warrant the expenditure of considerable effort in the site selection
process. Thus, we have not hesitated to recommend the acquisition and
interpretation of a diverse body of background information. In some
instances, special measurements are warranted before making a final
decision. The importance of the task demands that it be given time and
thought.
C. Special Characteristics of Carbon Monoxide
That Affect Monitoring Site Selection
Some of the special characteristics of CO as an air pollutant will
limit the things that can be done in the siting of monitoring stations.
Other CO characteristics will allow approaches that might not be
possible with other pollutants. Fortunately, the limitations and
special requirements imposed by the nature of CO sources can, in many
instances, be met through exploitation of the relatively inert
properties of the gas.
Table 1 summarizes the nationwide emissions for 1970. On a
nationwide basis, transportation sources accounted for nearly three-
fourths of the total and motor vehicles for nearly two-thirds. In urban
areas the relative contribution is probably greater since several of the
other major source categories (e.g., agricultural burning) take place
almost entirely outside of the urban areas).
-------
Table 1
NATIONWIDE ESTIMATES OF CARBON MONOXIDE EMISSIONS (1970)
Source: Cavender, Kircher and Hoffman, 1973
Emissions,
Source Category io6 tons/year
Transportation
Motor Vehicles
Gasoline
Diesel
Aircraft
Railroads
Vessels
Other nonhighway use of motor fuels
Fuel combustion in stationary sources
Coal
Fuel oil
Natural gas
Wood
Industrial process losses
Solid waste disposal
Agricultural burning
Miscellaneous
Forest fires*
Structural fires
Coal refuse burning
111.0
96.6
95.8
0.8
3.0
0.1
1.7
9.5
0.8
0.5
0.1
0.1
0.1
11.4
7.2
13.8
4.5
4.0
0.2
0.3
Percent of
Total
74.5
64.8
64.3
0.5
2.0
0.1
1.2
6.4
0.6
0.3
0.1
0.1
0.1
7.7
4.9
9.3
3.0
2.7
0.1
0.2
Total 149.0
Includes prescribed burning.
-------
The fact that motor vehicles constitute the largest urban source of
CO means that most CO emissions take place near ground level along
roadways. This, in turn, means that the sources and the public, or the
monitoring station, can be very close to each other; therefore, the
public or the monitor can easily be exposed in areas where little
dilution has taken place. This contrasts with some other pollutant that
may be emitted well above ground level, or may form slowly from chemical
reactions in the atmosphere; such pollutants may not appear at locations
accessible to the public until considerable mixing and dilution has
taken place.
The nature of the CO source means that very large concentration
gradients can be found near ground level, especially by streets or
roads. Figure 1 illustrates this. It shows that concentrations at 3 m
above a downtown street can change by several parts per million (or a
factor of nearly 2) over distances of only a few tens of meters. If we
were trying to represent the exposures of pedestrians, measurements made
on either side of the street would be misleading if they were
interpreted as applicable to all pedestrians on both sides of the
street. The problem is to devise methods for taking measurements that
are more representative, whatever purposes are being pursued.
Fortunately, the relative inertness of CO allows us to recommend
procedures and techniques that can increase representativity but which
would not be possible to use for many other pollutants. Much longer
lengths of inlet tubing can be used for CO sampling than would otherwise
be the case, because of it low reactivity. This, in turn, means that it
is feasible to draw air to a single sampler from several separate
points. If care is taken to equalize the flow from the different
inlets, such samples can give an average concentration from opposite
sides of the street and provide a number that is more representative of
the exposure experienced by the total pedestrian population in the area.
The use of multiple inlets to average concentrations from different
parts of an area will overcome some of the problems posed by the large
gradients in the vicinity of strong ground-level source areas.
D. Organization of This Report
The purpose of this document is to provide procedures for locating
CO monitoring sites. As noted before, it is our opinion that the first
step in such procedures must always be the determination of the primary
purposes for which the monitoring is to be conducted so that the site
can be classified and its physical requirements can be identified.
Section II of this report discusses monitoring objectives and a site
classification system that can be used to relate the objectives to the
physical characteristics of the site.
Section III gives step-by-step procedures for selecting sites in
the most important measurement categories. Section III is intended as
the working part of the report.
-------
in
o
o
EC
D
O
U)
DC
O
<
O
CO
O
<
to
<
O 05
LL T-
tO DC
I m
1 LU
I- H
< O
sl
o <
D
a
x
O
Q I-
Lll O
or
D
-------
The site selection procedures would not be valid if there were no
rational quantitative reasons for them. Section IV is included to
provide the reasoning that was used to arrive at the suggested site
selection methods, particularly the limiting values that are suggested
for such things as inlet height, separation distances between sources
and inlets, and so forth.
In some of the recommended procedures, the person making the site
selection decision will have to use unfamiliar data. The appendices
provide some discussion of these different kinds, of data and possible
sources from which they might be obtained. A simple model that can
assess the relative contributions of sources at different distances is
also discussed and a program listing and user instructions are given.
Finally, a bibliography is included to show the diverse uses of CO
(and in some cases, other pollutants) data. The bibliography also
includes a sampling of citations dealing with monitoring and siting
philosophies and procedures.
-------
II DECIDING THE TYPE OF CO MEASUREMENTS THAT ARE TO BE MADE
A. Uses of Carbon Monoxide Measurements
The objective of this section is to present a unified theme for
classifying monitoring stations. Our approach to this objective has
been to start with a simple list of the uses of ambient CO concentration
data, which was compiled by searching the literature for reports and
papers that made use of such data (see Appendix E). The ways in which
the data had been used were assembled and listed without regard to their
relative importance or to the possible overlapping of objectives. In
the list that follows we have tried to group the purposes into general
categories:
1. Enforcement of Air Quality Regulations
• Determine compliance with air quality standards:
- Federal primary
- Federal secondary
- State or local.
• Provide information for preparation of environmental impact
statements:
- Highway projects
- Transportation plans
- Large developments
- Indirect sources.
2. Research
• Determine relations between concentrations and sources
(or sinks):
- Specific emitters
- Emissions related to land-use patterns, etc.
- Indirect sources, shopping centers, stadiums, etc.
- Classes of emitters, e.g., vehicular, stationary, etc.
• Provide information for better understanding of the
processes affecting CO concentrations:
- Subjective understanding
-------
- Objective modeling: development, evaluation, and
refinement.
• Describe the "microdistribution" of CO in special areas:
- Intersections
- Street canyons
- Tunnels
- Below-grade highways.
• Provide measurements of the magnitude of sources and sinks:
- Anthropogenic
- Biological.
• Test monitoring equipment.
• Evaluate interstate and international transport processes.
• Provide measurements of natural, worldwide background CO
concentrations.
3. Development and Evaluation of Controls
• Evaluate results of control measures
- "Hot spots"
- Overall urban conditions
- State, regional, or nationwide conditions.
, Determine long-term trends in CO concentrations:
- In an urban area
- In a rural area
- Worldwide background.
• Provide information for city and regional planners and
decision-makers.
4. Public Health
• Alert authorities to existing or impending critical
situations.
• Evaluate the effects of human exposure to ambient CO.
• Determine the relationship of ambient outdoor CO levels and
those inside buildings.
-------
5. Miscellaneous
Provide information for comparisons among locations and
areas of the same general class:
- City streets
- Neighborhoods
- Urban regions
- Larger regions.
Evaluate representativeness of existing and proposed
monitoring sites.
The above list is quite diverse and it is not immediately obvious
that any coherent, physically sound, applicable taxonomy exists. We are
indebted to a paper by Dr. Wayne Ott (1975) that provided much of the
stimulation for the following analysis.
Since the goal is to devise a set of physical characteristics and
procedures for selecting monitoring sites, it seems reasonable that the
objectives should be related to a physical classification scheme. Spa-
tial scale of representativeness has been chosen as the basis for the
classification system. As noted before, any useful measurement of CO
concentration is supposed to represent some volume and some period of
time. Actually, the measurements always involve averaging over some
period of time because the analyzed volume is collected over a finite
period; at the same time, spatial averaging occurs over the same volume.
In general, this volume does not correspond to the objectives of the
monitoring. The objectives may require representation of volumes
considerably different from those from which the samples are drawn.
Site selection, when viewed from this perspective, consists of selecting
a location or locations where concentrations in the volume that we
sample can be related to concentrations representative of the volume
that is required to meet our objectives.
B. Spatial Scale of Representativeness
The categories for classifying the various objectives are listed
below. They are presented in order, from the smallest scale to the
largest scale of measurement. Their relative importance will be
discussed later.
1. Microscale
This refers to volumes with dimensions of meters to a few tens
of meters, smaller than downtown street canyons or below-grade highways.
Those special studies attempting to determine CO distributions within
parking lots, within street canyons, over highways, and the like require
microscale measurements. Similarly, the development, testing, and
revision of models that seek to describe the processes that produce
-------
these concentration distributions require data of this scale. This type
of measurement might also be used to define health effects for certain
individuals, such policemen, who remain near a fixed location for
extended periods.
2. Middle Scale
This category covers dimensions from tens of meters to
hundreds of meters and more. In certain cases discussed below we apply
it to regions that may have a total length of kilometers. In many cases
of interest, sources and land use may be reasonably homogeneous for long
distances along a street, but very inhomogeneous normal to the street.
This is the case, with strip development, freeway corridors, and downtown
street canyons. We have included in this category measurements to
characterize the CO concentrations along the urban features just
enumerated.
When a site is chosen to represent conditions in a single
street canyon or a block of strip development, then the characteristic
dimensions of this scale are tens of meters by hundreds of meters. If
we try to characterize street canyon conditions throughout the downtown
area, or along an extended stretch of freeway, the dimensions may be
tens of meters by kilometers.
Public health effects and air quality standards related to
public health effects would use measurements of this scale for their
assessment. People moving through city streets tend to be exposed to CO
concentrations consistent with a scale like this. This is also a scale
of measurements that would provide valuable information for alerts, for
devising and evaluating "hot spot" control measures, for comparing
central business districts in different cities, and for providing the
outdoor measurement necessary to the study the relationship between
outdoor and indoor concentrations in large public buildings.
This important class would also include the characteristic
concentrations for other areas, with dimensions of a few hundred meters,
such as the parking lot and feeder streets associated with indirect
sources—that is, complexes that are not themselves pollutant emitters,
but which attract significant numbers of pollutant emitters,
particularly autos. Shopping centers, stadia, and office buildings are
examples of indirect sources.
3. Neighborhood
Measurements in this category would represent conditions
throughout some reasonably homogeneous urban subregion, with dimensions
of a few kilometers and generally more regularly shaped than the middle-
scale. Homogeneity refers to CO concentration, but it probably also
applies to land use. In some cases, a site carefully chosen to provide
neighborhood-scale data, might represent not only the immediate
neighborhood, but also neighborhoods of the same type in other parts of
the city. These kinds of stations would provide information relating to
10
-------
health effects and compliance with regulations because they would
represent conditions in areas where people live and work.
Neighborhood-scale data would provide valuable information for
developing, testing, and revising concepts and models that describe the
larger-scale concentration patterns, especially those models relying on
spatially smoothed emission fields for inputs. These types of
measurements could also be used for interneighborhood comparisons
within, or between, cities. This is also the most likely scale of
measurement that would meet most of the objective's of city and regional
planners and decision-makers.
4. Urban
This class of measurement would be made to typify CO
concentration over an entire metropolitan area. Such measurements would
be useful for assessing trends in city-wide air quality and, hence, the
effectiveness of larger-scale air pollution control strategies.
Measurements that represent a city-wide area would also serve as a valid
basis for comparisons among different cities.
5. Regional
These measurements would characterize conditions over areas
with dimensions of as much as hundreds of kilometers. As noted earlier,
representative conditions in an area imply some degree of homogeneity in
the area. For this reason, the class of regional measurements would be
most applicable to sparsely populated areas, without major settlements.
Data characteristic of this scale would provide information about
interstate transport processes.
6. National
Measurements that defined concentrations on this scale would
characterize the CO level of the nation as a whole. Characterizing
national CO levels would provide trend data and allow assessment of
national policies. Such data would also be useful for studying
international and global transport processes.
7. Global
Such measurements would provide information useful to the
identification of world-wide trends in CO concentration.
The above list categorizes measurements according to the
spatial scale to be represented. Virtually all uses of CO measurements
involve the characterization of CO concentration on one or more of these
scales. Some scales would be difficult, and perhaps impossible, to
represent with a measurement at a single site. Thus, there need not
necessarily be a site category that corresponds to each of the above
scales of measurement; some scales will have to be represented as a
composite of measurements characterizing smaller areas.
11
-------
C. Relative Importance of the Different Scales of Measurement
Some of the scales of measurement will be more important than
others for at least two reasons. One of these is the number of separate
purposes that can be served by a particular scale. The other reason is
based on a judgment of the relative importance of the purposes that are
served. Most continuing measurements of pollutant concentrations are
used to define air quality, particularly as it relates to the national
primary ambient air quality standards. Accordingly, we have chosen to
emphasize those scales of measurement that are most closely related to
those standards.
The national primary ambient air quality standards are "those
which, ... based on the air quality criteria and allowing for an ade-
quate margin of safety are requisite to protect the public health."
(Federal Register, 1971). The emphasis on public health gives the air
quality standards great importance. The emphasis on public exposure is
further shown by the definition of ambient air that is used in connec-
tion with the air quality standards: "Ambient air is that portion of
the atmosphere, external to buildings, to which the general public has
access." If we accept ambient air quality standards as the major moti-
vation for monitoring, then the ranking of scales of measurement can be
carried out so that the most important scales are those which are most
characteristic of the kinds of exposures that the general population
encounters. Although the air quality standards must be met on all
scales, the emphasis on public health and public exposure is apparent in
the definitions quoted above, and hence it seems reasonable that site
selection should be subject to the same emphasis. Finally, experience
has shown that areas of greater public exposure are more likely to
experience standards violations than are areas that contain few people.
With such factors in mind, the following ranking was developed:
(1) Neighborhood
(2) Middle
(3) Urban
(4) Regional
(5) Micro
(6) National
(7) Global
As noted in the discussion of the various scales of measurement, most
people are exposed to CO concentrations on the neighborhood or the
middle scale. Thus, these two scales should occupy the most important
positions in the list. Many people, over a period of a day or week,
will be exposed to neighborhood and middle scale concentrations of CO.
A mixture of such exposures can be considered as representative of the
urban environment as a whole, and hence the relatively high ranking of
urban scale measurements.
Since it would be virtually impossible to represent the hodge-podge
of neighborhood and middle-scale concentrations with measurements at a
single site, no "urban-scale" measuring station is described in spite of
12
-------
its acknowledged importance. We feel that urban areas must be
characterized by networks of stations that describe the range of
conditions within the area.
Another large segment of the population dwells in areas that could
be characterized by regional-scale measurements; therefore, this scale
also deserved a high place on the list. It ranks lower than the urban
scale because "regional" areas will be rural and have lower
concentrations of both CO and people. Therefore, they are likely to be
less critical.
Since the microscale measurements are, as noted before, of
importance in determining the health effects on only a very limited
class of citizens, it will usually be adequate to rely on middle-scale
measurements for the protection of this group. Other reasons for
relegating this category to fifth place are discussed later. The final
two categories, national and global, are representative of such large
areas that they do not relate to the exposure of the public. There are
also other grounds, which are discussed later, for placing these
categories at the end of the list.
The ranking of measurement scales given above, which is based
specifically on the requirements of national air quality standards, is
the same ranking that would have resulted if the primary rationale had
been public health, because public health is the rationale for the
primary air quality standards.
The needs of urban diffusion modelers are best met by the first two
scales listed (neighborhood and middle) because the models generally use
emissions inventories of these scales to predict concentrations on the
same scales. Regional measurements will provide the data necessary for
evaluating the performance of models that are designed to describe
larger-scale transport processes.
Preparation of environmental impact statements is becoming an
increasingly important activity that requires CO concentration data.
Since one of the motivations for the preparation of these statements is
to define present air quality, and the potential changes in that air
quality as they relate to air quality standards, it is not surprising
that the most important scales of measurement for evaluating environ-
mental impact are much the same as shown in the above list. The CO
impacts of indirect sources are felt most on the middle- and
neighborhood-scales. Conceivably, very large projects could have urban-
wide or regional CO impacts but CO is not likely to be the worst
pollutant in such instances.
Our survey of the literature suggests that those making microscale
measurements of CO concentration have specialized requirements that
would be difficult to generalize. The users are usually research
oriented and develop their own criteria, carefully matched to their own
specific aims. The usual requirements associated with microscale
measurements tend to be beyond what we consider to be the primary scope
13
-------
of this report—to provide guidelines for siting the more routine, long-
term monitoring stations.
Finally, the last two categories on the list can be relegated to
the end because they are largely redundant. To some extent this is also
true of the urban scale. If the middle- and neighborhood-scale
measurements have adequately characterized the city's downtown streets,
shopping centers, highway corridors, and neighborhoods, then all the
necessary information is available for characterizing the city as a
whole. Similarly, if the cities and regions are adequately described,
their descriptions can be synthesized into a national description and on
to the global scale.
The conclusion of the above discussion is that the most important
specific site types can be associated with three scales of measurement:
. Middle
• Neighborhood
• Regional.
The site selection process begins with an identification of the appro-
priate scale of measurement, which is discussed in the next section.
D. Selecting the Required Scale of Representativeness
Table 2 summarizes the most important of the purposes for which
measurements may be made and the relationships between the scales of
measurement. If the relationships were unambiguous and there were only
one scale that were suitable for a given purpose, it would only be
necessary to find the measurement purpose in the first column, see which
one of the other columns contains a check mark, and proceed to the
appropriate part: of Section III of this report.
In some instances, the ideal case described above is the one that
prevails, but most often, some subjective decisions are required.
Before proceeding to a discussion of the many less than ideal cases
represented in the table, note that the middle-scale class of
measurements has now been subdivided into three special subtypes.
(1) Downtown street canyon—Middle-scale measurements to define
the CO concentrations along streets in areas of dense,
multistory buildings.
(2) Indirect source—Middle-scale measurements to define the CO
concentrations prevailing outdoors along walkways, in parking
lots, and on streets surrounding so-called indirect sources
that do not emit pollutants themselves but they do attract
pollutant emitters, particularly automobiles. Examples
include, shopping centers, theaters, and stadia.
14
-------
(3) Corridor—Middle-scale measurements to define the CO
concentrations along major highway corridors, another kind
of indirect source.
In Table 2, the kinds of measurement that will be required for a
given purpose are indicated by an "X". An "X" 'enclosed in parentheses
indicates that the needs might be satisfied, in particular cases, by
temporary or mobile units. Blank spaces have been left where
measurements of the designated scale are not usually very useful,
although there will be exceptions. The purposes for which the
measurement data are to be used are discussed on the following pages.
1. Determine Compliance with Ambient Air Quality Standards
It can be inferred from the air quality standards that the
most important locations for such measurements will be those that com-
bine the highest concentrations with the greatest exposures of
population. Among the more likely locations for such a combination is
the downtown street canyon, which has high traffic densities, confined
spaces between buildings, and large numbers of people present. In the
neighborhoods, concentrations may be less, but people spend greater
periods of time exposed to them. Thus, high density residential
neighborhoods are also likely candidates for stations devoted to this
purpose. Indirect source sites are less likely to be required for this
purpose, because of the lower frequency of instances where high
concentrations are reached. Mobile monitors could be used to measure
concentrations when scheduled events are expected to produce high
traffic densities. Corridor measurements may be appropriate where
residences are near heavily traveled highways.
2. Alert Authorities to Existing or Impending Critical
Situations,
Traffic jams are most likely to produce critically high
pollutant concentrations in already congested downtown areas, around
indirect sources, or along stretches of freeway. However, the latter
two are poor candidates for permanent measuring sites devoted to this
purpose because of the infrequent or erratic occurrence of such
conditions at locations of those types. As noted before, indirect
source monitoring for purposes of detecting high concentrations could
probably be done best on an "as needed" basis with mobile equipment.
Similarly, when severely congested conditions occur on freeways, mobile
equipment might be brought to the scene via surface streets.
3. Evaluate Results of Control Measures
There are many possible control measures that could be
evaluated. Some are essentially city-wide, such as transportation plan
revisions or exhaust emission controls. For these, sites characterizing
neighborhood concentrations are suitable because they will measure the
results of the controls without being unduly influenced by large,
unrelated fluctuations in local emissions. For specific, smaller scale
15
-------
Table 2
SCALES OF MEASUREMENT APPLICABLE TO VARIOUS PURPOSES
Purpose
1. Determine compliance
with ambient air
quality standards
2. Alert authorities to
existing or impending
critical situations
3. Evaluate results of
control measures
• Hot spots
• City-wide
4. Determine long-term
trends
• Urban
• Rural
5 . Provide information
for developing eval-
uating and refining
air pollution models
6. Provide information
for comparisons
among locations of
the same general
class, e. g. ,
• Street canyons
• Highways
• Neighborhoods
• Rural areas
Applicable Scales of Measurement
Middle
Downtown
Street
Canyon
X
X
(X)
X
(X)
X
Indirect
Source
(X)
(X)
(X)
(X)
Traffic
Corridor
(X)
(X)
(X)
X
(X)
X
Neigh-
borhood
X
(X)
X
X
X
X
Regional
X
X
X
16
-------
Table 2 (Continued)
SCALES OF MEASUREMENT APPLICABLE TO VARIOUS PURPOSES
Purpose
7. Serve as data
base for city and
regional planners
and decision makers
8. Serve as data base
for environmental
impact statements,
e.g.,
• Highway projects
• Transportation
plans
• Large develop-
ments
• Indirect sources
9. Provide measures of
the magnitude of
sources and sinks
• Anthropogenic
• Biological
10. Provide measures
of indoor/outdoor
concentration
relationships
Applicable Scales of Measurement
Middle
Downtown
Street
Canyon
X
(X)
Indirect
Source
(X)
(X)
Traffic
Corridor
(X)
X
(X)
(X)
Neigh-
borhood
X
X
X
X
X
X
Regional
X
X
Note: X - Fixed site
(X) - Mobile or temporary site
17
-------
control measures, monitors that will characterize the middle scale are
more appropriate. The evaluation of some control measures on this scale
will only require monitoring for a limited period of time, perhaps a
week or two in different seasons. Among such control measures are
traffic engineering changes that would improve traffic flow in the
downtown area. Such control measures might require any of the three
types of middle-scale measurement, depending on the facility to which
they were applied.
4. Determine Long-Term Trends
Middle-scale measurements will not be as useful for this
purpose as neighborhood-scale measurements because they are highly
influenced by large, and often erratic, fluctuations in emissions over a
relatively small area. When the emissions influencing a location are
averaged over a larger area, then the fluctuations in concentration will
more realistically reflect meteorological factors and changes in
emissions that are more widespread than those that affect the middle
scale. Therefore, the neighborhood- and regional-scale measurements are
likely to be more suitable for evaluating long-term trends in CO
concentration.
5. Provide Information for Developing, Evaluating and Refining
Air Pollution Models
Inasmuch as air pollution models have been developed or
proposed to describe phenomena on almost any scale, the measurements
required by those models are apt to be of any of the scales. Usually,
those models of smaller scale phenomena—for example, street canyon
effects or distributions of CO in parking lots or along the edges of
freeways—will use microscale or middle-scale measurements collected
during programs devoted especially to the purpose.
Models describing the distributions of pollutants throughout
the urban area will often make use of available data bases from stations
that are tacitly assumed to represent conditions on the neighborhood
scale. Generally, the appropriate scale of measurements to be used for
model validation studies is approximately the same level of detail as is
used for the emissions inventory that serves as an input to the model.
6. Provide Information for Comparisons Among Locations
of the Same General Class
Often it is desirable to compare CO concentrations in
different neighborhoods in the same city, or similar neighborhoods in
different cities. In such instances, valid comparisons can only be
achieved between measurements of comparable scales of representa-
tiveness. Similarly, conditions in the downtown street canyons of two
different cities may be compared if appropriate data, representative of
the street canyons, are available. Almost any kind of comparisons can
be made, but they will have more meaning if the types of area repre-
sented by the measurements are known.
18
-------
7. Serve as a Data Base for Planners and Decision Makers
The activities of regional planners and other officials making
long-range decisions are usually on the neighborhood scale or larger.
There are exceptions, such as traffic engineers who may be concerned
with smaller scale phenomena, e.g., traffic flow in the central business
district or congestion at specific intersections, but most long-range
decisions are not concerned with small areas, short segments of
individual streets, or localized activities, but with the distribution
of business, industry, residences, and transportation throughout an
area. If air quality measurements are incorporated into this level of
the decision-making process, it will usually be on a large-scale,
through the analysis of long-term trends or the use of urban air quality
models.
8. Serve as a Data Base for Environmental Impact Statements
Very often, projects for which impact statements are required
will affect air quality on the middle scale. If the area in which the
project is to be located is reasonably homogeneous, then neighborhood-
scale measurements can be used to characterize conditions prevailing
before the project is begun. Sometimes it will be necessary to make
special middle-scale measurements for the purpose. For example, if the
project is a highway widening, then corridor-type measurements will be
required in the area where the widening is planned. Similarly, expan-
sion of a shopping area might require measurements of the concentrations
around that indirect source in its unaltered configuration. Preparation
of environmental impact statements will require measures of preproject
conditions and data that will serve as bases for estimating the changes
caused by the project.
9. Provide Measures of the Magnitude of Sources and Sinks
The magnitude of sources and sinks can be estimated from
concentration and wind measurements on the upwind and downwind sides of
the area in question. Comparisons of the fluxes will provide estimates
of the magnitude of an intervening source or sink. Regional-scale
measurements on the windward and leeward sides of a city might provide
some estimates of its larger-scale impact as a source of CO. Emission
rates from smaller sources, such as highways, can also be estimated from
appropriate micro- or middle-scale measurements, but these usually
involve special, nonpermanent facilities as indicated in Table 2.
10. Provide Measures of Indoor/Outdoor Concentration
Relationships
Applications of this type will require special, nonstandard
measurements; in particular, indoor observations will be required. The
outdoor measurements may also include nonstandard locations to collect
data representing CO levels near air conditioning inlets. Standard
neighborhood scale measurements will probably be useful for studies of
indoor/outdoor concentrations in residential buildings, but special
19
-------
middle scale measurements will be the rule for such studies with
commercial buildings.
An evaluation of Table 2 suggests that the most important
types of permanent stations are those that characterize neighborhood and
regional scales and the type used for middle-scale concentrations in
downtown areas and along traffic corridors. The following sections
therefore emphasize sites appropriate to the representation of regional,
neighborhood, city center and traffic corridor conditions.
20
-------
Ill SELECTING STATION LOCATION
A. Background
The purpose of this section is to provide, as nearly as possible,
step-by-step procedures for locating monitoring stations that will
represent the three most important measurement scales discussed in the
preceding section. The form and major steps of the site selection
process are presented in flowcharts with accompanying discussion. To
avoid obscuring the steps of the procedure, this section includes little
justification for specific recommendations; this has been deferred to
the next section of the report. Similarly, sources of the special data
or analytical tools that may be required during site selection processes
are not discussed extensively in this section; the appendices can be
consulted for such information.
The different procedures to be used in selecting sites are
presented below in an order that proceeds from large- to small-scale
representation. This is the same order that would arise if the
procedures were organized according to increasing complexity.
B. Regional Stations
Figure 2 is a schematic diagram of the site selection process for a
regional-type station. The process reflects the fact that such stations
may be used to estimate the CO concentrations entering an urban area.
The selection process also reveals that siting criteria for the
establishment of a single station will differ from those used for more
than one station.
The site selection process begins with acquisition of the necessary
background material. This material is to be used as a basis for the
judgmental decisions that are required during the selection process.
Two basic kinds of information are required: geographical and clima-
tological. The geographical material is used to determine the distri-
bution of natural features — forests, rivers, lakes — and the works of
man. Useful sources of such information include: road and topograph-
ical maps, aerial photographs, and satellite photographs, particularly
those from the Earth Resources Technology Satellite (ERTS). Naturally,
the site selection process is one that winnows out unsuitable regions
and proceeds to fewer and smaller areas. Thus, the photographs and maps
will generally be used in an order that proceeds from those that depict
large areas, such as shown in an ERTS photograph like Figure 3, to those
that show smaller areas in greater detail, like the aerial photograph in
Figure 4.
21
-------
Acquire necessary background material
1. Regional maps and aerial
photographs showing
a. Topography
b. Settlements
c. Major highways
2. Climatological information
a. Wind roses
b. Dilution climate
(Holzworth, 1970)
Is station to define regional concen-
trations or background concentrations
for a specific city?
( Background for specific city J
1
Will there be more than one
monitoring station?
c
One station
Two (or more)
stations
I
Select tentative siting
area about 35 km or more
downwind for the least
frequent direction. Sites
should be at least 5 km
from any major intercity
roadways in a region of
reasonably uniform
topography.
I
Select tentative siting
areas that are about 35
km or more upwind for the
two most frequent wind
directions. Add other
sites for other cardinal
directions. Sites should
be at least 5 km from any
major intercity roadways
in a region of reasonably
uniform topography.
Regional concentrations
Select tentative siting
area, about 35 km or more
from major metropolitan
areas; in direction of
low wind frequency from
nearest city, at least 5
km from major intercity
roadways in a region of
reasonably uniform
topography.
Eliminate specific sites in low lying
areas, within about 5 degrees of
alignment with major highways, and
within about 400 m of any roadways
with traffic greater than a few
hundred vehicles per day.
Locate specific sites. Inlets at 3-10 m.
Supplemental monitoring: wind.
SA-3515-27
FIGURE 2 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING
REGIONAL STATIONS
22
-------
in
i
in
u>
n
CJ
to
O
<
Q
Z
<
I
o_
DC
O
O
O
I
a.
tr
LLJ
to
O
O
o
LLJ
I-
CO
UJ
u
ir
D
O
CO
LU CO
a: LU
x ^
01 O
PO
LLJ
QC
23
-------
SOURCE: Dabberdt and Davis, 1972.
FIGURE 4 AERIAL PHOTOGRAPH OF A RURAL AREA
24
-------
The climatological summaries of greatest use are the frequency
distributions of wind speed and direction. This information will
usually come in one of two forms. One of these is a tabulated joint
frequency distribution like that shown in Table 3, an example of the
material that is available from the National Climatic Center. The wind
rose is an easily interpreted graphical presentation of the directional
frequencies. Examples of wind roses are shown in Figure 5, from the
National Climatic Atlas (National Oceanographic and Atmospheric
Administration, 1968). Other types of useful climatological data are
also available but, generally, are not as directly applicable to the
site selection process as are the wind statistics. These summaries are
discussed in Appendix C.
Table 3
EXAMPLE OF A TABULATED WIND SUMMARY
PERCENTAGE FREQUENCIES
OF WIND DIRECTION AND SPEED:
DIRECTION
N
NNE
NE
ENE
E
ESE
SE
SSE
S
ssw
sw
wsw
w
WNW
NW
NNW
CALM
TOTAL
HOURLY OBSERVATIONS OF WIND SPEED
,IN MILES PER HOUR)
0.3
+
+
+
A
•f
•f
•f
•f
1
i
47 8 12 13 - 18
1 2.1
1
1
1
1
1
1
1
2
+ 1
+ 1
2
3
2
1
1
2
2
2
1
1
1
2
1
+
+
1
1
3
3
3
+ 13 4
+ 135
+ 12
+
•f
1
2
1 1
+
4 I 13
30
3
2
1
30
19 J4 , 35 31
+
•f
•f
•f
+
*
^
2
2
2
4
3
1
1
•f
17
+
+
+
+
+
1
1
1
32 38 39 44
47
OVEt
|
I
+
•f
+
1 *
+
+
+
+
5
+
+
+
+
+
+
1 + +
TOTAL
4
4
6
4
3
2
4
4
10
8
8
14
12
7
6
4
+
100
AV
SPEED
11.4
10.5
11.7
11.4
9.0
8.6
8.9
11.0
13.3
14.4
15.5
17.3
15.3
14.6
13.1
12.0
13.5
Source: National Climatic Center
Asheville, N. C.
25
-------
CO
LLJ
CO
o
QC
<
X
IT)
LLJ
oc
26
-------
After the background material has been assembled, a decision must
be made regarding whether the monitor is to be used to supply informa-
tion on CO concentrations that enter a specific city. For selecting a
purely regional site, maps or satellite photographs should be used to
identify general areas that are suitable for locating such stations.
Uniform topography is desirable. The climatological information should
be incorporated into the selection process to determine those areas
which will be least frequently downwind of the closest major urban
areas, especially during periods of low wind speed. If possible, the
prospective siting areas should be at least 35 km from that urban area
as illustrated in Figure 6. It also illustrates the fact that the
siting area should not be within about 5 km of the nearest major
intercity arterial.
The requirements for regional stations that will provide data on
the background CO concentrations entering an urban area are much the
same as for the purely regional site described above. In fact, if only
one background station is being established, the requirements are
identical. If more than one upwind urban background site is planned,
then the potential siting areas should be chosen to be upwind for the
two most frequent wind directions. Figure 7 (a) illustrates this
schematically. If the two most frequent wind directions are within 90
degrees of each other [Figure 7 (b) ] , then the most frequent direction
will be used, along with the most common direction from among all those
that are more than 90 degrees different from it, as Figure 7(b)
illustrates. Any prospective siting area should be more than 5 km from
the nearest major intercity arterial road.
Once the general siting area has been selected, a winnowing process
begins. Low-lying locations are unsuitable, as are sites that are
located within about 5 degrees of alignment with extended straight
highway segments, as shown schematically in Figure 8. The monitoring
site should not be too close to any road that has more than a few cars
per day traffic; specifically, no road with a few hundred cars per day
should be within about 400 m of the site.
monitoring site should be in an area, open sufficiently that
the air is not likely to stagnate. In farmland or other open areas,
this will be no problem. In a forested area, a clearing would be desir
able; if none is available, then the inlet could be raised a few meters
above the tops of the surrounding trees. In most instances, the inlet
height at the site finally selected can be between about 3 and 10 m
above the surface although 3 m is the most desirable. Figure 9 illus-
trates a typical site that might be used for regional CO monitoring.
A CO monitoring site can also be used for measuring other environ-
mental factors that are related to the air pollution problem. In
particular, winds are quite important, and they should be measured at a
height of about 10 m above the general level of the surrounding surface.
27
-------
LL1
H
to
QC
O
g
u
LU
cn
cc
O
u.
to
CC
LU
I-
<
cr
o.
O
cc
o_
D.
DC
CJ
I-
LU
I
O
to
to
LU
CC
28
-------
NEXT MOST FREQUENT
WIND DIRECTION
MOST FREQUENT
WIND DIRECTION
SECOND
MOST FREQUENT
WIND DIRECTION
SECOND WIND DIRECTION
FOR LOCATING SITING AREA
PROSPECTIVE
SITING AREA
(b)
SA-3515-8
FIGURE 7 SCHEMATIC DIAGRAMS OF APPROPRIATE SITING AREAS FOR REGIONAL
MONITORS WHEN TWO SITES ARE PLANNED
29
-------
AVOID
THIS
AREA
\
PROSPECTIVE
SITING
AREA
MAJOR
ROAD
SA-3515-1
FIGURE 8 SCHEMATIC DIAGRAM ILLUSTRATING
THE UNSUITABILITY OF AREAS ALIGNED
WITH MAJOR ROADS
In open fields this means 10 m above the ground; in a heavily forested
area, the anemometer should be that distance above the general forest
canopy.
C. Neighborhood Stations
A schematic diagram of the procedures used to select a neighborhood
iiK-.itoring station is shown in Figure 10. As with a regional station,
the first step of the selection process for neighborhood stations is the
acquisition of necessary background materials. Climatological informa-
tion will be required, especially a joint frequency distribution of
winds and atmospheric stability. This is provided by the output of the
National Climatic Center's "STAR" program, which is described in more
detail in Appendix C.
Aerial photographs and topographic maps will also be useful. Maps
that describe the land use more specifically than conventional maps will
provide valuable input to the selection process. Often such maps are
only available for study in the offices of the planning agency.
Finally, maps, or other sources, showing the distribution of traffic on
30
-------
o
o
o
z
g
o
LLJ
IT
LU
tr
a.
O
cc
0.
CL
CO
O
oc
QC
CO
I
I-
co
<
U-
O
LU
_J
0.
X
til
o>
at
oc
31
-------
ASSEMBLE BACKGROUND
INFORMATION
* Land Use Maps
* Aerial Photos
* Traffic Maps
* Climatology
Identify neighborhood type to be repre-
sented. If commercial or warehousing,
decide if street canyon site selection
process would be better.
/street canyon process^
\jiot appropriate J
\
Identify neighborhoods of the selected
type on maps.
Use simple climatological-type model
to obtain relative distribution of CO
concentrations. Plot of traffic or
emissions field could be substituted.
[Street canyon processV
^ould be better )
•
Use street
canyon process
Superimpose CO distribution map on map
of candidate neighborhoods.
_n
'Conduct optional Bag Sampling to better •
I define conditions in the area of interest.!
1 5. site to
example of
represent
selec ted
"typ ical1
type of nt
or "worst"
iehborhood7
Typical^)
Identify neighborhood
with mode led relative
concentration nearest
the average for all
the neighborhoods of
the selected type.
Identify neighborhood
with highest modeled
relative concentra-
tions .
Identifv all high volume roadways
(daily traffic- greater than 5x10 )
in the selected area.
Mark off area at least 2500 m from
nearest high volume road-way preferably
near center of selected neighborhood.
Survey area for inlet locations at least
35 m from any street. (Optional: use
bag sampling technique to better define
most representative locations.) Use
inlets at 3 m height.
SA-3515-10
FIGURE 10 SCHEMATIC DIAGRAM OF A PROCEDURE
FOR LOCATING NEIGHBORHOOD
MONITORING STATIONS
32
-------
arterials are necessary. An example of this kind of information is
shown in Figure 11.
If the type of neighborhood to be represented is characterized by
relatively tall, closely spaced buildings—that is, the typical streets,
or other open areas between buildings being less than twice the building
heights—then the procedures used to locate a middle-scale monitoring
station, such as a street canyon site would be more appropriate to the
problem than those given below.
After acquiring the necessary background materials, the next step
is to use the maps and aerial photographs to'identify neighborhoods of
the type that are of concern to the particular monitoring objective.
Aerial photographs of two different kinds of residential neighborhood
are shown in Figure 12. Figure 12(a) shows an area that was built more
recently than that shown in Figure 12(b), with its more mature trees and
shrubbery. A section of a topographic map corresponding to Figure 12(b)
is shown in Figure 13.
After potentially suitable CO monitoring areas have been identified
on a map, a simple numerical simulation model can be applied to provide
an estimate of average CO concentration over the city. Several suitable
models are available, but probably the climatological dispersion model
(CDM; Busse and Zimmerman, 1973) would be easiest to apply; its required
inputs are the output of the STAR program referred to earlier and an
inventory of traffic data converted to CO emission rates in grid squares
throughout the city. If computational facilities are not available,
then the emissions inventory itself can serve as an approximation to
average CO distributions, because of the close relationship between
concentration and nearby emissions; see Section IV, D.
If time and resources are available, a program of bag sampling
would be highly desirable to provide better data on the relative
concentrations in the various candidate neighborhoods. Ott and Mage's
(1974) statistics suggest that about 25 samples (24-hour average) at
each location would be sufficient. The samples should be collected on
random days over a period that will include the annual climatic
extremes. When the estimated spatial distribution of average annual
concentration—either from modeling or measurements—has been completed,
it should be compared with the map showing candidate neighborhoods. If
the objective of the monitoring is to obtain data for the particular
neighborhood that is likely to have the highest average CO concentra-
tions among all the neighborhoods of the chosen type, the measured or
modeled average CO distribution will serve to identify the appropriate
area. Similarly, the distribution of average CO concentration can be
used to judge which of the candidate neighborhoods is nearest the
average for all of them.
33
-------
\ m\^-—-^r ~\\ •"'0(iE
>oU%fJT i
sk^pf730\\Wfe2^^ >l
l°1 / Jy r^^—«»e SAzo'oS^Srr-L
) ©/%-S!!f5^
/ SpX7720 K66'° ^^^1 ,J^~~~--T^~M&?^^
J ^^-\ ir i/^^-JSri^^
i/ Oo;^^ / ./ ",40-y ^BoTr^ss^^^4
Jjgoooo ««o/ | | «C2lFg?*°Z'>
1970
TRAFFIC MAP
OF
ST. LOUIS
ST. LOUIS COUNTY
KlOTE: Figures represent estimated average annual two-way weekday traffic
volumes.
SOURCE: Prepared by the Missouri State Highway Department Division of
Planning in cooperation with the U.S. Department of Transporta-
tion Federal Highway Administration.
FIGURE 11 PORTION OF A TYPICAL TRAFFIC MAP
TA-8563-128
34
-------
The above discussion represents rather minimal preparatory efforts
in the site selection process. Obviously, modeling is subject to
considerable uncertainty, which can be reduced by a limited sampling
program. Even the limited sampling program only provides estimates of
average CO concentration and averages do not allow as complete a
comparison among sites as would be desirable. A more extensive sampling
program could determine more reliable frequency distributions which
would be better bases for comparison.
Once the neighborhoods that most nearly meet the monitoring objec-
tives have been identified, then more specific locations must be
selected. These locations must be at least 2,500 m from the nearest
highway carrying 50,000 or more vehicles per day. This spacing will
limit the contribution of the roadway to less than 1 ppm. Lower traffic
volumes will allow for closer spacing. An example of the calculation of
a roadway's contribution is presented in Section IV. The inlet must be
at least 35 m from any local street having peak traffic of about 800
vehicles per hour. Here again, a limited bag sampling program could be
undertaken to define gradients (especially horizontal) in the area. If
the gradients are very strong, then a location more distant from the
nearest street might be required. Conversely, weak gradients would
allow closer placement. The inlet should be at a height of 2.5 to
3.5m.
Some auxilliary measurements are desirable for a neighborhood
station, such as wind. As for the regional sites, the anemometer should
be placed about 10 m above the general level of the surroundings; it
should be well away from any structures of comparable height. Traffic
counts on one or more of the nearby streets would provide valuable
supplemental information for interpreting anomalous CO readings.
D. Middle-Scale Stations
1. General
As noted earlier, this report considers three specific types
of monitoring site for measuring middle-scale CO concentrations. These
are:
• Street canyon
• Roadway or traffic corridor
• Indirect source.
Only the first two of these are considered in detail because indirect
source monitors are most likely to be needed for special studies of
limited duration. Thus, this section provides a detailed procedure for
selecting a street canyon and traffic corridor monitoring locations, but
provides only guidelines for locating monitors that will determine
concentrations around the parking areas of indirect sources.
35
-------
(•) RECENT RESIDENTIAL
(b) OLDER NEIGHBORHOOD WITH MATURE TREES
SOURCE: Dabberdt and Davis, 1972.
FIGURE 12 AERIAL PHOTOGRAPHS OF URBAN
RESIDENTIAL NEIGHBORHOODS
36
-------
APPROXIMATE AREA SHOWN
IN FIGURE 12(b)
FIGURE 13 A TYPICAL URBAN NEIGHBORHOOD DEPICTED ON A TOPOGRAPHICAL MAP
37
-------
2. Street Canyon Sites
The first step of the selection process (see Figure 14) is to
acquire the necessary background information. This includes the average
daily traffic on all the streets in the area, wind roses for different
hours of the day, and maps showing one-way streets, street widths, and
building heights. These can be obtained from land use maps or by having
personnel survey the area and estimate a typical height for each side of
each block. Hourly weather observations for a year from a nearby U.S.
weather service station will be necessary in some cases.
As indicated in Figure 14, there are two basic monitoring
objectives for this type of station. One is to typify the worst
conditions to which the public is regularly exposed in the area, and the
second is to typify conditions throughout the area.
If the station is to typify the area with highest
concentrations, the streets with the greatest daily traffic should be
identified. If some of streets are one-way and if local traffic
authorities say that they have an asymmetric distribution of daily
traffic (i.e., one of the two rush hour periods has much greater traffic
volumes than the other), those streets that have their greatest traffic
during the afternoon and evening hours should be selected as tentative
sites, because the periods of high traffic volume are usually of
greatest duration through the evening hours.
When several blocks have been selected as candidates, then the
simple computer model described in Appendix A can be applied to each of
the tentative locations to determine which is likely to have the highest
eight-hour average CO concentrations. If the necessary computational
facilities are not available to apply the model, the equations in
Appendix A, describing street canyon CO concentration, can be applied,
using the most frequent wind direction for the rush-hour period (or
periods) of the day. As with the computer model, the object is to
determine the block and the side of the street in that block where
highest CO concentrations are likely to occur.
If the monitoring site is supposed to typify the entire
downtown area, the average daily traffic should be calculated based on
all the important street segments in the area. Select those blocks where
the daily traffic is nearest the average for the whole area. If
possible, avoid street segments that have radically different widths or
building heights from those typical of the area.
Attempt to select only two-way streets or one-way streets with
similar traffic volumes in the morning and afternoon. If there are no
such streets with near-average traffic, attempt to find a pair of
38
-------
COLLECT BACKGROUND DATA
* Traffic
* Climatology
* Maps
* Plans
Is station to be representative of
entire area, or of that part of the
area with highest concentrations?
f Typify entire area J
Determine average daily traffic on
streets within the area of concern.
Select streets with daily traffic near
average. Eliminate one-way streets.
If all streets are one-way, select a
pair with peak traffic at opposite
ends of the work day; see text.
Are the frequencies of wind direction
and speed distributed more or less
symmetrically?
Symmetrlcaljj
UnsymmetricalJ
If possible, select
street approximately
parallel to an axis of
wind symmetry
ossiblAd
Possible
(Typify area with A
highest concentrationsJ
\
Select streets with greatest daily
traffic. If some streets are one way,
select those with peak traffic period
of greatest duration... probably the
afternoon; see text.
Choose street and side of street with
highest expec ted concentrations. Use
model described in Appendix A, or on
basis of wind rose, select most fre-
quent direction for light winds during
rush hours.
Tentatively select inlet
location on most conve-
nient side of street.
Tentatively select inlet
locations on opposite
sides of street.
J
I Optional bag sampling program to verify .
I or modify conclusions. I
Hake final inlet site(s) selection ...
3 m high, over the middle of the side-
walk, but at least 2 m from the
building; near midblock; away from bus
stops, loading zones, lines of
vehicles, or other unusual CO sources.
SA-3515-11
FIGURE 14 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING STREET
CANYON STATIONS
39
-------
streets forming adjacent sides of the same block, as shown in Figure 15
(a), and having near-average traffic. The peak traffic should occur
during the morning on one of the two streets and during the afternoon on
the other. Figure 15(b) illustrates such a case, with the two streets
having peak traffic at opposite ends of the working day.
The next step is to study the frequencies of the various wind
directions. If they are distributed with reasonable symmetry about some
axis, select a street nearly parallel to that axis. For example, in
Figure 5, Salt Lake City, Utah, has a reasonably symmetric wind rose
with an axis of symmetry running approproximately north-northwest to
south-southeast. North-south streets would be sufficiently parallel to
this axis to satisfy the requirement. Most, but not all, of the wind
roses in Figure 5 are nearly symmetric; Albuquerque, New Mexico, and
Kansas City, Missouri, are examples of asymmetric wind roses.
The Weather Service station for which wind frequencies are
available may not always be typical of conditions in the part of the
city where the site is to be located, especially in regions with complex
topography. Therefore, it will often be wise to seek advice from local
meteorologists and make adjustments where necessary.
If data are available showing directional frequencies as
function of wind speed, as in Table 3, then the frequencies at the lower
wind speeds should be more heavily weighted in the site selection
process than those at higher speeds. This is because the lowt_r wind
speeds will usually be associated with higher CO concentrations*
If typLcal concentrations are sought, some consideration must
be given to the strong gradients that occur in street canyons. To
minimize the effects of these gradients, it is desirable to have a
street where the wind blows with equal frequency from the opposite sides
of the street, that is the street should be aligned with the axis of
symmetry of the wind rose. In such a case, the side of the street on
which the inlet is located is of less importance. Another alternative
is to use more than one inlet, placing them on opposite sides of the
street, or in extreme cases, on more than one street, as shown in Figure
15, to compensate for asymmetries in the daily traffic cycles on one-way
streets. If multiple inlets are used, the air flows must be the same
through each.
A limited program of bag sampling can be used to check whether
the tentatively selected locations are indeed representative of
conditions in the downtown area. Such a program could also be used to
find other representative locations that were more convenient for one
reason or another; for example, space might be more readily available
and less expensive, security might be better, and so forth.
40
-------
TRAFFIC DIRECTION
TRAFFIC
DIRECTION
POSSIBLE
INLET LOCATIONS
CANDIDATE
STREETS
(a)
DIURNAL
TRAFFIC
DISTRIBUTION
(b)
SA-3515-2
FIGURE 15 EXAMPLE OF A PAIR OF ONE-WAY STREETS WITH PEAK TRAFFIC
ON ONE IN THE MORNING, AND IN THE AFTERNOON ON THE OTHER
41
-------
The actual location of the inlets should be at a height
between 2.5 and 3.5 m and over the center of the sidewalk, but not
closer than about 2 m to the buildings. The locations should not be
closer than 10 m to a cross street to better typify conditions in the
larger spaces between intersections. Before finally selecting a
location, at least several days should be spent observing activity
around that location to make sure that vehicles do not commonly stop and
spend extended periods of time idling. Bus stops, loading zones, and
areas where lines of cars form regularly should be avoided. Figure 16
illustrates how a downtown street canyon sample inlet might look. The
anemometer is shown because this inlet was used as part of a special
study in which winds in the canyon were also measured.
A vehicle counter would be a valuable instrument to use in
conjunction with the CO monitor. It would provide information that
could be used to interpret the records of CO concentration and to
identify unusually heavy traffic conditions. Wind measurements in a
downtown area are very difficult to obtain without some bias introduced
by surrounding buildings; therefore, wind measurements would not have
high priority.
3. Roadway or Traffic Corridor Sites
a. Long-term monitors
The site selection procedure for traffic corridor sites
is shown schematically in Figure 17. The procedure is very similar to
that suggested for selecting street canyon sites. It begins, as should
all site selection, with the acquisition of background information,
e.g., traffic data, street maps, and climatological information. A
decision must also be made whether the purpose of the monitor is to
determine the highest carbon monoxide concentrations in the vicinity of
roadways, or to monitor more typical conditions. If the purpose is to
characterize a high concentration road segment, then the monitor should
be located near an area of maximum traffic volume, and on the side of
the roadway that is most frequently downwind, especially during periods
of light winds.
If the monitor is supposed to characterize concentrations
typical of those found near roadways in the area, then average traffic
volumes will be sought. The roadway should parallel the axis of
symmetry of the wind rose, if possible. (The concept of symmetry of
wind roses was discussed in connection with the selection of street
canyon sites.) If it does, then the monitor can be placed on either
side of the roadway. The basis for selecting one side over the other
might be convenience, or greater population exposure.
If it is not possible to locate an appropriately aligned
and traveled roadway, then it would be desirable to place inlets in both
42
-------
-3v v'SkSvssfe.-^S'
;; *
l.-o" »'
.,;, -?;•''"*.•"!; ~"'fTS~ '-'-.ft'*4? SA-351S-T2
FIGURE 16 SAMPLE INLET IN A DOWNTOWN STREET CANYON
43
-------
COLLECT BACKGROUND DATA
* Traffic
* Climatology
* Maps
Is station to be representative of
typical exposures along the roadway,
or of the highest exposures
experienced by the public?
Typify more generally
representative exposures
Typify highest
exposures
Select at-grade sections of highway
with traffic volumes that are
approximately average for roadways of
the same type (arterial, expressway,
limited access freeway, etc.) in the
area of interest
Select at-grade section of highway
with greatest traffic volumes for the
highway type
c
Are the frequencies
distributed more or
•
•
of wind direction
less symmetrically?
i
f
Choose side of roadway that is most
frequently downwind
Symmetrical
U n s y mm e t r i c a 1
Select a section of roadway
parallel to axis of wind
symmetry
Select locations on
opposite sides of roadway
Select location on most
convenient or most populated
side of roadway
Optional bag sampling to verify that site is typical
of the type of conditions being sought
r
Make final selection of inlet site(s);
3 m high outside highway right-of-way,
at least 2 m from buildings away from
anomolous sources, e.g. metered
on-ramps, toll gates, etc.
SA-3515-28
FIGURE 17 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING TRAFFIC
CORRIDOR STATIONS
44
-------
sides of the roadway to obtain average concentrations. This may be more
difficult than for street canyons, but it should be possible to use
existing supports for road-signs, or to use utility poles so that the
inlet tubing can pass over the highway at a height where it will not
obstruct traffic.
To this point, the discussion has dealt only with siting
in the vicinity of roadways that are at, or below grade. Elevated
roadways present some special problems and their differing heights
introduce effects that make site selection for the long-term monitoring
of highest concentrations or of typical concentrations nearly
impossible. However, special surveys can help to describe CO
concentrations in the vicinity of specific sections of elevated roadway
and the relationships of these concentrations to meteorological and
traffic factors. Siting for such special sites is discussed in the next
section. Special, short-term surveys can also provide valuable
information for the selection of sites for the long-term monitors.
Once a general, location for a monitor has been selected,
the inlet should be placed at a height of 2.5 to 3.5 m, outside the right-
of-way. If the purpose is to characterize the typical population
exposure, then the site should be at a distance from the roadway that is
about equal to the average building setback from the roadway. If a
worst example is sought, then the monitor should be as near the edge of
the right-of-way as possible.
Finally, monitors should not be placed in the vicinity of
possibly anomolous source areas. Examples of such anomolous areas
include toll gates on turnpikes, metered freeway ramps, and drawbridge
approaches. Traffic counters near the monitoring site will provide
valuable data for interpreting the observed CO concentrations. An
anemometer will also provide valuable corollary information that can be
used to determine the extent to which traffic emission might have been
carried toward the monitor.
b. Special Surveys
The preparation of environmental impact statements (EIS)
for highway projects often requires fairly detailed descriptions of CO
concentrations in their vicinity. The California Department of
Transportation (formerly the Division of Highways) has developed methods
for assessing the air quality impacts of highway projects (Beaton, et
al., 1972). These methods are generally quite comprehensive and the
documents should be consulted for solutions to monitoring problems of
this sort.
In general, the effects of highways at or below grade are
most pronounced within about 100 m of the edge of the highway.
Measurements made within this distance will usually suffice to define
the important impacts. Bag sampling is efficient because it allows
45
-------
samples to be collected at several points while requiring only one CO
analyzer. Figure 18 shows an elaborate array of bag samplers used to
measure the emission, transport, and diffusion of CO in the vicinity of
a divided highway (Dabberdt, 1975). Each of the barrel-like containers
shown in the figure holds several bags, which are sequentially filled.
In this experiment, CO concentrations and winds were measured at several
elevations in the immediate vicinity of the road.
Bags filled over a period of an hour would provide
samples compatible with the short-term air quality standard. Of course,
most of the samples should be collected on the downwind side of the
highway, but at least one sample should be collected upwind of it.
Several samples should be taken on each side of the road when winds are
nearly parallel to it. Inlets should normally be 2.5 to 3.5 m above
ground unless the special, three-dimensional characteristics of the
diffusion processes are being studied.
For elevated roadways, samples should be collected in the
vicinity of the maximum concentrations that are likely to occur. The
expected locations of the maximum values can be most easily estimated
from graphs, based on numerical modeling, prepared by the California
Division of Highways (Beaton et al., 1972e). An example of one of these
graphs is given in Figure 19. The sampling network would be arranged to
correspond to the atmospheric stability and wind prevailing during any
given test.
4. Indirect Source Sites
The monitoring of CO concentrations in the vicinity of an
indirect source will most often be for a special purpose and of limited
duration. To establish prevailing conditions before the indirect source
is built, a site of the neighborhood or regional type would be required
and the siting requirements would be the same as for any other sites of
those types.
To determine the CO levels at an existing indirect source, a
combined program of modeling and sampling would be desirable. The
modeling would provide information on the areas where the greatest
concentrations and the most typical concentrations are to be expected.
Models have been developed to describe CO concentrations around indirect
sources. (See, for example, Sandys et al., 1975; Patterson and Record,
1974.) A bag sampling program would provide similar information from
measurements of the actual concentration distributions that occur in the
indirect source region. Such information would be required if a more
permanent monitor was planned. Patterson and Record's (1975)
measurement program around a Chicago shopping center will provide the
reader with some insight about what is required to define CO
concentrations around an indirect source.
46
-------
I
I
<
CC
111
QC
QC
O
2
_l
CL
CO
00
UL
o
X
111
<
CO
LU
cc
47
-------
ELEVATED SECTION
STABILITY CLASS B
Q
UJ
UJ
UJ
CO
CO
UJ
o
tr
co
0=67.5° ANGLE OF INTERSECTION BETWEEN
WIND DIRECTION AND HIGHWAY
ALIGNMENT IN DEGREES
H = Height of'povement above
surrounding terrain in feet
0
50
00
=150
-- 200
= 250
OOOl
0 600 1200 1800 2400 3000
NORMAL DISTANCE FROM DOWNWIND EDGE OF SHOULDER -FEET
3600
SOURCE: Beaton et al , 1972.
SA-3515-14
FIGURE 19 AN EXAMPLE OF THE DISTRIBUTION OF NORMALIZED CO
CONCENTRATION DOWNWIND OF ELEVATED ROADWAYS
48
-------
As is true for the selection of sites for other kinds of
monitors, a decision must be made whether to monitor the highest CO
concentrations or to monitor concentrations that are more nearly
representative. Once that decision is made, the modeling and bag
sampling results can be used as a basis for selecting specific sites.
The procedures will be similar to those described for other kinds of
sites. If a representative measurement is desired, the use of multiple
inlets should be considered seriously. An inlet height of 2.5 to 3.5 m
is recommended.
E. Other Types of Stations
The preceding sections have dealt with the location of stations to
monitor CO on the regional-, neighborhood-, and middle-scales. Micro-,
urban-, national-, and global-scale measurements were not discussed.
The following sections present brief discussions of these other scales
and some of the factors that bear on the representation of their CO
concentrations.
1. Microscale
Every experiment that requires microscale measurements is
likely to differ from every other such experiment, and so it is not
possible to establish a single set of criteria for this kind of
measurement. However, there are some areas of commonality. Microscale
measurements are often made in connection with the study of some
particular physical phenomenon. The phenomenon will often have
definable dimensions and the measurements must be made at locations
spaced closely enough to define its significant features. Perhaps, the
best way of understanding this is through the use of specific examples.
If one is concerned with the behavior of CO emissions within a
street canyon, it is easy to define the overall dimensions of that
problem, because the canyon is confined by the street and the buildings
on either side of the street. Measurements must be made at enough
points within these bounds to define the processes that are taking
place. Figure 20 shows one array of inlets that has been used to study
street canyon processes. This array was established to determine the
transport in the street canyon of CO emitted at ground level. It was
hypothesized that during times when winds blew across the street,
relatively clean air flowed down one side, across the street (becoming
more contaminiated by the emissions), and then up the other side of the
street canyon. Since the evidence indicated that there was a helical
circulation around the street canyon, the inlets were placed around the
edges to the extent possible—obviously inlets cannot be placed at
street level where there is traffic, so they were located as low as
feasible.
49
-------
o
CJ
y,
o
to
in
m
t-
D
CD
CC
CO
Q
Q
D
CO
O
H
Z
LU
DC
UJ
Q.
O
LL
z
o
o
o
cc
1»
CM
01
UJ
0
cn
<
Q ?
o
CM
cc
50
-------
Figure 21 illustrates another array of microscale samplers.
In this instance, the objective was to study the behavior of pollutants
in the vicinity of a major divided highway. A photograph of part of
this array is shown in Figure 18. The initial rapid mixing that is
caused by vehicular turbulence was of particular interest. An estimate
had to be made concerning the volume that might be expected to be
affected by such turbulent mixing. The array shown in Figure 21 is a
reflection of that estimate; the several towers defined the expected
vertical and horizontal extent. The placement of the towers and the
samplers is the final compromise achieved by the experimenter between
the desire to measure at as many places as possible and the limitations
imposed by traffic and experimental resources.
The two examples given above illustrate the importance of
defining the scale to be studied through microscale measurements. In
the above cases, the total area was rather small. If the distributions
of CO in a large parking lot were to be studied, the dimension of
interest would expand to those of the parking lot and the measurements
would probably be spaced at intervals comparable to the zones into which
most parking lots are divided.
Note that microscale measurement programs will generally
require a considerable number of corollary measurements to provide all
the information required for proper interpretation of the CO concentra-
tions. In the examples discussed earlier, this corollary information
included winds and turbulence, traffic, and vertical temperature
profiles. The physical relationships that are studied will define the
required parameters for any particular case.
2. Urban and National Scales
It is probable that there will be no single site in a large
urban area that will represent conditions throughout the city. The
obvious approach to this problem is to combine measurements from a
number of individual stations through simple averaging, but this can
provide misleading results. Better results could be achieved through
weighted averaging of the individual measurements, with the methods used
for this weighting determined by the objectives that are to be met by
the derived measure of urban CO concentration. Although a weighted
average might not be appropriate to the purpose of determining
compliance with air quality standards, it could be quite useful for
comparing CO concentrations in cities of different types.
To define an average CO concentration over the area, each
individual measurement would be weighted according to the total area it
represents. For instance, measurements from a neighborhood site would
be weighted proportionally to the area of the neighborhood(s) that it
represents. Similarly, if health effects are important, the weighting
of measurements from a given station might be based on the population
exposed to concentrations like those represented by that station.
51
-------
fn
U
ft
W
0)
•P
0)
I
01
c
•P
c
fi
0
ft
co
0)
fi
!H
0)
Q)
ft
O
JH
Q)
ft
0)
EH
O
01
fl
^ H
ft
+->
fn
0)
tH
•H
Q
3
+->
cj
Q)
ft
Ofl ^ CJ CO
^p CXI ,-H
0 000
i
& & &
0 O O O
1
•^1 ^ ^ ^
3 33
* •* «*
EH EH EH
<3 •<]
O O O O
T
^ ^ ;£ ^
& > !> >
3 3 3
0 000
EH
3 33
•S •» •*
£i <3 H
O 0 O O
1
ft ft ft
1
in
in
n
2
L
t.
!
j
0
o
~
m
CO
T5
c
3
0 0
M rt ^
01 cj >>
* iJ «
CO -D
cS
O
C ft *
oi -H 01
•H JH >
•O -P -H
01 M -P
S 0
SH
"O O
rt
3 ^
O b£
J2 01 -O
•P d) P4
01 c
W J
CO ^ x
u
•>
co
in
r~
m °>
to
N
V
13
a
Q
111
0
DC
D
o
t/)
g
CO
=>
Q
I
o
LLJ
tr
LU
a.
X
cc
o
LL.
CJ
o
2
QC
Q
Z
<
QC
O
CO v
LU <
CO g
si
l<
fn QC
CN
LU
CE
D
52
-------
Hence, a downtown street canyon site would be weighted according to the
number of people on the downtown streets at the time of the measurement;
data from neighborhood sites would be weighted by the numbers of people
living in the neighborhood.
The above examples have been presented to illustrate factors
that might be important in defining an "urban scale" CO concentration.
The importance of objectives had already been emphasized in connection
with the selection of individual sampling sites. The examples given
above indicate that the objectives are also important in the
interpretation of the data and in the synthesis of measures of total
urban area CO concentration.
Similar arguments can be applied to measures of national
levels of CO as were just applied to the urban scale. The synthesis of
a national measure from regional and urban measures could be
accomplished through weighted averaging, with the weighting methods
dependent on the objectives.
3. Global Scale
To a large extent, the atmosphere itself will accomplish the
averaging that is necessary for assessing global scale CO concentra-
tions. If the station is sufficiently remote, the contributions of
sources and sinks will be "averaged" by thorough mixing in the air.
This report does not address the definition of "sufficiently remote" as
it applies to global sites. Note, however, that the site should
probably be remote from sinks as well as from sources, the problem being
complicated by a lack of understanding and identification of the sinks
for CO in the atmosphere.
53
-------
IV RATIONALE FOR SITE SELECTION CRITERIA
A. Background
The site selection procedures contain some very specific recommenda-
tions concerning such things as spacing between monitors and sources or
the heights of inlets. The recommendations have been derived through a
variety of methods. In most cases, an a_ priori judgment is required
during some phase of the process. The objective of this work has been
to make those j. priori judgments as recognizable and consistent as
possible. This section presents the reasoning and judgments that were
used to arrive at the recommendations.
In some cases, such as the recommended heights of inlets, the
choices are straightforward. The importance of population exposure to
CO concentrations demands that the air be sampled at average breathing
heights. However, practical factors, like prevention of vandalism and
the potential obstruction to pedestrians, require that the air samples
be higher—hence, the recommended 3 m which is an admitted a priori
compromise between these two requirements. The recommendation of a
range of heights about 3 m requires some analysis to translate the
specified range of heights to corresponding measures of the representa-
tiveness and the spatial variability of CO concentration. In those
latter terms, the a priori judgments are much easier.
Similarly, the recommended spacing between sites and specific
sources is clearly understandable if it is restated in terms of the
expected maximum contributions of the source to the measured CO concen-
trations at the site. Thus, we have decided on acceptable levels of
interference by a specific source and proceeded to find the minimum
spacing between the source and the monitoring site where that level is
not likely to be exceeded.
Some of the procedures require other kinds of justification. For
instance, potential street canyon sites are identified on the basis of
traffic on the streets in the area. The recommended procedure tacitly
uses traffic volumes as a surrogate for CO emissions and CO concen-
trations.
By spelling out the reasoning and the assumptions behind the
choices that have been made, it should help others to make rational
decisions in cases where their requirements are not met by the
recommendations of this report.
Finally, a word about the organization of the following material;
it does not parallel the organization of the preceding section.
55
-------
Instead, it is divided according to the important questions that must be
answered for every station: inlet height, distance from specific
sources, and so forth. In general, this allows us to present a specific
line of reasoning and then precede to show how that line of reasoning
applies to each of the different station types.
B. Inlet Locations
It was recommended that inlets for most kinds of sampling should be
at a height between 2.5 and 3.5 m. The choice of 3 m for the median
height has already been explained as as compromise between
representation of breathing height and prevention of vandalism.
The recommended 1-m range of heights is also a compromise to some
extent. For consistency and comparability, it would be desirable to
have all inlets at exactly the same height, but practical considerations
will often prevent this. Therefore, some reasonable range must be speci-
fied and 1 m should provide adequate leeway to meet most requirements.
The variability of CO concentration with height in a street canyon
is sufficiently large that the representativeness of the measurements
will be strongly affected by variability of the inlet height. Figure 22
based on observations by Georgii et al (1967) shows vertical gradients
of CO concentration from about 0.3 ppm/m to 0.5 ppm/m in the lower
levels of a street canyon. Similar gradients are evident in Figure 23
from Ludwig and Dabberdt (1972). The gradients will depend on traffic
emissions and on street canyon dimensions, but available observations
and the empirical model presented in Appendix A suggest that hour-
average vertical gradients of 1 ppm/m are quite possible. Aim range
of inlet height then corresponds to a range in concentrations of about 1
ppm or less. This seems a reasonable value for measurements in this
kind of environment. The reasonableness of the 1 ppm range can be
subjectively judged by comparing it to air quality standards—it is
about 2 or 3 percent of the one-hour standard and about 10 percent of
the eight-hour standard.
The inlet heights for neighborhood and regional monitoring
stations were also specified to be between 2.5 and 3.5 m, although
regional monitoring inlets up to 10 m above the surface were specified
as acceptable. The objective is to obtain measurements that are
consistent with the street canyon type measurements and are taken as
near as practical to average breathing heights. As is the case with the
street canyon monitors, we would like to minimize differences from the 3
m concentrations. Since most CO is emitted near the surface from line
sources (roads and streets), it seems reasonable to use equations for
ground-level line sources to evaluate the variability of concentrations
56
-------
40
30
20
LU
I
10
LEEWARD SIDE
OF STREET
WINDWARD SIDE
OF STREET
10
CO — ppm
15
SOURCE: Georgii et al., 1967
20
SA-3515-4
FIGURE 22 THE VERTICAL DISTRIBUTION OF CO CONCENTRATION
IN A STREET CANYON WITH TRAFFIC VOLUME OF
1500 VEHICLES/HOUR
57
-------
o
00
I p>
I ID
I in
I- °?
Z
LU
OC
LU
oc
O
O
>
z
<
o
E
CO
E
(0
Ad0931VD Q33dS QNIM
10
A
E
.Q
(Q
Q
UJ
O
DC
D
O
DC
CO
Z
z
o
QC
I-
CJ
z
o
o
o
o
o
I-
QQ
cc
00
Q CO
QC Q
LLJ 2
> O
< O
CO
CN
LLJ
DC
58
-------
in the vertical. The commonly used Gaussian formulation for such
sources is as follows (e.g., see Turner, 1969):
The relative change of concentration, C, with height is given by:
1 bC z
C Sz a,
2
Where
C = ^ exp -0,5 (—)2
VJFTucrsin 9 a
z
C = concentration (g/m3)
Q = source strength (g/m s)
u = wind speed (m/s)
az = standard deviation of a Gaussian plume in the vertical,
a function of downwind distance, x (m)
9 = angle between wind and line source; validity of
equation degrades for cp < 45 degrees
z = height to instrument inlet (m),
We have specified minimum separations between monitors and line
sources, the bases for the recommended separation are discussed later,
but for now it is sufficient to note that minimum distances from
neighborhood monitors to major roadways should be about 2 to 3 km and to
smaller streets about 30 to 40 m. In general, roadway sites will be 15
m or more from the roadway center. At such distances, az will be about
5 m or more in urban areas (see for example, Johnson et al., 1972;
Ludwig, et al., 1975); thus, at z = 3 m the variation of concentration
with height should be about 10 percent/m or less. This is the
variability caused by the closest sources^ CO from more remote sources
will be more uniformly mixed, thus we can be assured that concentration
variations associated with aim range of inlet height should be limited
to a few percent of the overall concentration.
Minimum separations between regional sites and major line sources
should be about 5 km. Rural diffusion proceeds so that o~z will be about
30 m or greater at these distances (e.g., see Turner, 1969). Over the
59
-------
acceptable range from 3 to 10 m, the variation in observed concentra-
tions should be limited to a few percent, as in the city. Nevertheless,
the preferred height is still 3 m, and should be retained as nearly as
possible.
It has been recommended that street canyon sensor inlets be
located, at least 10 m from the intersection. The choice of midblock
locations over intersection locations was made because intersections
represent a much smaller portion of downtown space than do the streets
between them. The pedestrian exposure times are probably also greater
in street canyons than at intersections. Finally, the practical diffi-
culties of positioning inlets are less at midblock locations than at an
intersection. Of course, some special studies might want to place the
sensor inlet so as to detect the effects of queuing at the intersection,
but most purposes will be better served by more representative locations
back from the intersection. Some observations made in St. Louis street
canyons (Ludwig and Dabberdt, 1972) suggest that there is a reasonable
uniformity of CO concentration throughout the part of the block that is
more than 10 m from the intersection. This figure has been used as a
guide.
C. Minimum Separation Between Monitoring Sites and Sources
In determining minimum separation between a monitoring site and a
specific source, the presumption has been that the neighborhood or re-
gional measurements should not be unduly influenced by any one source. A
subjective judgment is required as to what constitutes undue influence.
There are at least two bases for making that judgment. A maximum concen-
tration value can be assigned and the separation between the monitoring
site and all sources should be such that the contribution from any one
source does not exceed the assigned maximum. The other approach is to
examine the disturbance in the concentration gradients caused by sources
and assign a maximum allowable gradient from any given source. Both
approaches were tried in connection with roadway sources and are dis-
cussed below. In the examples that follow, typical emission rates have
been assigned. These are based on typical CO emission values for
vehicles during the early- to mid-1970s and on perusal of traffic maps
and other documents to obtain reasonable values for average daily
traffic.
The line source equation that was presented earlier and served as a
basis for estimating vertical variability can also serve as a basis for
determining the minimum separations that are required between sources
and monitors. Only a very small error is introduced if we assume that
the sources and the monitor inlets are at the same height. The equation
then simplifies to:
2Q
C =
V2rr ua sin cp
z
60
-------
As noted before, if a monitoring site is to be representative of a
fairly large surrounding area, the contributions from any one source
should either be small or the gradients of those contributions should be
small. That is, one of the following inequalities should be satisfied:
2Q
uazsin cp
< C
max
or
1 Be
C d(xsin cp)
1 SC
a sin cp 3x
Z
max
where Gmax is the desired maximum of relative gradient of concentration
away from the roadway and Cmax is the maximum allowable contribution to
the concentration. Johnson et al. (1971) have suggested that in urban
areas a can be related to x by an equation of the following form:
z
Thus
= ax
b
dx
and the two inequalities become
2Q
uax 2n sin cp
max
xsin cp
max
At this point, some values must be assigned to the terms so that some
reasonable values can be chosen for the minimum distance (x sin cp ) from
the roadway, at which at least one of the criteria can be satisfied.
The following values have been chosen as reasonable for major urban
roadways : _ _ nr7 ,
Q = 0.07 g/m s
Note, that this corresponds to about 10,000 vehicles/hr,* emitting about
40 g/vehicle-mile.
cp = 40 degrees
u = 1 m/s
max
max
= 0.001 gm/ m3 (approximately 1 ppm)
= 0.0002/m (20 percent/km)
* Peak hour traffic is usually about 10 percent or less of the total
daily traffic; hence, the corresponding total daily traffic is around
100,000 vehicles.
61
-------
Johnson et al. (1971) have proposed the values of a and b given in Table
4 for use with different atmospheric stabilities in urban areas.
Table 4
VALUES OF CONSTANTS USED TO REPRESENT
VERTICAL DISPERSION AS A FUNCTION OF
DOWNWIND TRAVEL DISTANCE
Stability Type
Very unstable
Unstable
Slightly unstable
Neutral
Slightly stable
Values
a
0.07
0.12
0.23
0.50
1.35
b
1.28
1.14
0.97
0.77
0.51
After making the substitutions, minimum distances between monitoring
sites and the hypothesized roadway can be estimated for each atmospheric
stability class according to the two different criteria. The results
are summarized in Table 5. Since only one criterion (not both) has to
be satisfied, it appears that neighborhood monitoring sites more than
about 2 to 3 km from major highway sources can be expected to be
generally free of undue influence from those sources.
Table 5
MINIMUM DISTANCES (KM) BETWEEN A LARGE
ROADWAY AND A NEIGHBORHOOD MONITORING SITE
Stability Class Gradient Criterion Concentration Criterion
Very unstable
Unstable
Slightly unstable
Neutral
Slightly stable
6
6
5
4
2.5
0.3
0.3
0.4
0.81
3
62
-------
For a source strength of 0.007 g/m s, corresponding to a smaller
city street with peak traffic about 500 vehicles per hour and emissions
of about 80 gm/vehicle-mile, the 1 ppm concentration criterion indicates
minimum distances of about 30 to 40 m between the monitor and the
source. Of course, it can be argued that the contribution from any
individual smaller street ought to be kept at even lower levels than the
contributions from the larger thoroughfares if the monitored values are
to be representative of the contributions from all sources, large and
small. If we specify the maximum acceptable contribution from a source
of 0.007 g/m s to be 0.5 mg/ m3 , then the minimum allowable separation
would be about 100 m from a middle size street source such as that speci-
fied. For smaller streets the permissible separation can be less. It
is not too difficult to find inlet sites that are 30 or 40 m from the
nearest street of any appreciable size, so we have chosen 35 m as the
minimum setback from any street for a neighborhood site.
Street canyon and traffic corridor sites are chosen to provide a
measure of the influence of the immediate source. For such purposes, no
minimum separation distance is required. This is consistent with the
smaller scale area to be represented by street canyon or other traffic
oriented sites as compared to the scale being represented at a
neighborhood site.
At the other extreme, for regional monitoring in rural areas, a
very small limiting concentration will be desirable; we have chosen
0.2 mg/m3, which is comparable to world-wide background concentrations
(Robinson and Robbins, 1967). Away from cities, major roadways may
carry about 3,000 vehicles/hr at peak times. Speeds will generally be
faster than at peak hours in the city, so emission rates of about 15
g/mile would be reasonable. This leads to peak line-source emissions of
about 0.008 g/m s. Rural diffusion takes place at somewhat slower rates
than in the city. Turner (1969) has given curves defining o~z in rural
areas as a function of x. Using those curves and the above figures for
source strengths and concentration limits gives a minimum separation of
about 4 or 5 km between the monitor and any large intercity highway.
If the wind blows parallel to an extended straight section of
highway, there can be substantial concentrations along the direction of
the section, even at substantial distances downwind of the point where
the road has turned so that it is no longer parallel to the wind. To
avoid exposure to such concentrations, we have recommended that the site
not be chosen within about 5 degrees of the extension of the road align-
ment (see Figure 8). Under stable atmospheric conditions, a Gaussian
plume from a point source will have a horizontal width that falls within
this subtended angle at distances of more than a few kilometers. In
this instance, "width" of the plume is defined as two standard devia-
tions on either side of the center line. If the line source is
considered as a series of point sources, then keeping outside the i 5
degree angle will generally avoid very strong influences during periods
when the wind parallels the roadway.
63
-------
The above approach does not consider climatological factors. If
winds are only infrequently from certain directions, a site could be
placed so that it would be influenced only infrequently. However, such
an approach might increase chances for subsequent misinterpretation of
the data by introducing correlations with wind direction that were not
related to general, larger-scale factors.
Our approach to defining the distance that a regional station
should be outside of an urban area has been based on reasoning similar
to that used above in connection with the determination of minimum
separations between neighborhood sites and roadways. That is, an
attempt has been made to limit the influence of the city on the nearby
regional monitors. We have taken advantage of similar calculations made
for other purposes. Figure 24 (from Stanford Research Institute, 1972)
shows calculated relative concentrations downwind of a circular area
source, 70 km in diameter. Within the area, source emission rates are
highest at the center, decreasing in Gaussian fashion to 24 percent of
the central value at the edge. The figure illustrates that where the
atmosphere is fairly stable, mixing proceeds slowly and ground level
concentrations remain high for fairly great distances.
Estimates based on data published by EPA (1973) for Washington,
D.C., and on the Los Angeles traffic densities of Roberts et al. (1971)
indicate that maximum emission rates, Q0, of around 1.3 x 10~4g/ m2s
are typical. If we assume the lightest winds to be about 1 m/s and if,
as before, we try to limit the effects on regional monitoring to about
0.2 mg/m3 then, as Figure 24 shows, the regional site should be 35 km
or farther outside the city.
D. The Importance of Sources at Various Distances from the Monitor
It is probably evident to the reader that the selection processes
that have been recommended emphasize nearby sources, or lack thereof.
In selecting a site for typifying some aspect of downtown street canyon
CO concentrations, the process is largely limited to finding a typical
block, or a worst block. Little attention is given to the interrelation-
ships with all the other sources in the urban area. For neighborhood
sites, the area of concern expands. In a "neighborhood" that is
reasonably homogeneous (i.e., without individual major sources) the area
of concern expands to a few kilometers. The scale expands still further
to tens of kilometers for regional sites. The reader can infer from all
this that concentrations at a site are highly influenced by sources
close to that site. Thus, the monitor readings are expected to be
characteristic of the immediate surroundings. This expectation has
prompted many of the recommendations that have been made.
The influence of local sources has been shown often. For example,
Ott (1971) observed that in San Jose, California, stations more than
"several hundred feet" from traffic should read approximately the same
CO concentration. Perkins (1973) concluded that a monitoring site in
the southwest part of the Los Angeles basin was probably strongly
64
-------
Q
O
CO
cc.
LU
Q.
CO
CO
CO
D
<
I
Q
LJJ
O
CJ
CO
z
O
QC
h-
_
z
O
Q
LJJ
N
QC
O
CM
LJJ
QC
65
-------
influenced by local traffic. Kinosian and Simeroth (1973) found that
the annual, maximum hour-average CO concentration varied strongly with
distance from the nearest traffic for nine Los Angeles area monitoring
sites; their results are rather startling, as shown in Figure 25. This
figure is based on the average of the maxima for two different years.
The simple model presented in Appendix A of this report was
developed so that the relative importance of nearby, and more remote,
sources could be evaluated. Application of the model can provide useful
information concerning the degree to which a site represents the area
that it is supposed to represent.
We have used the model with three sets of hourly meteorological
data representing three different regions of the country. Conventional
airport observations for the year 1960 were available from San Diego,
California, and St. Louis, Missouri. Special observations made at
Millstone, Connecticut were available for the period from late summer
1973 to late summer 1974. Thus, we have used meteorological data from
the two coasts and the central United States.
Each set of meteorological data was used in combination with a
hypothetical source distribution such as might be characteristic of an
idealized, radially symmetric city. The particular distribution that
was used for this problem decreased linearly from 1.3 x 10~4g/ m2 at
the center to zero at 32 km from the center.
First, we investigated the effects associated with a street canyon
monitor at five locations in the hypothetical city: at the center and
at 17.5 km from the center in each of the cardinal directions. For the
test, the street was chosen to be oriented in a northwest-southeast
direction and to have traffic of 200,000 vehicles per day, emitting an
average of 40 g of CO per mile. The street canyon was taken to be 20-m
wide and 30—m deep. The inlet was specified to be 3-m high and 4-m from
the traffic. The hypothesized daily traffic cycle (based on Roberts et
al., 1971, nonfreeway data for Los Angeles) is shown in Figure 26. It
shows the morning peak, sustained high midday values, a strong afternoon
peak and a gradual decline to the very low post-midnight levels. No
differentiation was made for weekends.
Figure 27 shows the frequency distributions of the contribution (to
8-hr average concentrations) of the local street relative to the
contributions from the rest of the city. The figure shows that there
are differences in the effect of the street that arise from differences
in the climatology of the area, but these differences are not so great
that generalizations cannot be made. For instance, the graphs show that
at the center of the hypothetical city, the street contribution exceeds
that from the rest of the city 20 to 30 percent of the time. The
dominance of the local street increases as the monitor is moved toward
the edge of the city. At 17.5 km from the center, the local street
contribution exceeds that from elsewhere in the city 50 to 80 percent of
the time, depending on the climatology and the direction from the city
center to the site. Some differences in the importance of the local
66
-------
in
in
X
o
c
c
o>
11)
.c
o
c
O
CO
O
O
UJ
O
00
O
<£>
CO
t-
en
O
OJ
h-
Z
<
CO
CO
CO
cc
LU
CO
z
o
cc
h-
z
01
o
z
o
o
o
o
a:
D
O
I
X
<
<
o
ill
LLJ
CC
CO
u- O
O t-
CO CO
LU LLJ
O O
< g
EC <
LLJ l~
> ¥
< D
LO
CM
IO
O
10
IT)
o
st-
ro
o
ro
in
CM
01
o
a:
uudd —
NOliVaiN30NOD
CC
13
67
-------
I
0> CO Is-
o
y
LL
<
QC
CD
CM
QC
13
(D
68
-------
street source are apparent from one side of the street to the other as
can be seen in Figures 27c and 27d, but these differences are not as
pronounced as those that are attributable to differences in location
relative to the city center.
If the monitor is sufficiently removed from strong local sources,
the city's emissions can be treated as area sources by the model
described in Appendix A. When the model is applied in this way, it
approximates the contributions to the CO concentrations monitored at a
neighborhood or regional site. Figure 28 presents the frequency
distribution of the ratios of the contributions from sources within 2 km
to those at greater distances. As before, the model was applied to the
hypothetical city emission distribution for locations at the city center
and at radial distances of 17.5 km; the same three sets of hourly
meteorological data were used.
The graphs in Figure 28 suggest that a substantial fraction of the
CO concentrations observed at neighborhood sites will arise from sources
within 2 km. There is considerable variablity, but sources within that
range will contribute half, or more, of the observed CO on from 5 to 40
percent of the occasions. Differences in the importance of the nearer
sources arise from meteorological factors and from the location relative
to the center of the city. The number of instances in which the nearer
sources contribute more than a third of the CO (i.e., the ratio of
contributions from within 2km to those from greater distances exceeds
0.5) almost always exceeds, the number of instances when the nearer
contributions are less than one third of the total.
The model confirms that nearby sources generally contribute a major
portion of the observed CO concentrations at a point. Often the
greatest concern is with those instances in which CO concentrations are
the highest and a question arises concerning whether the contributions
of the sources are as important during the periods of high concentra-
tions as they are at other times. The model given in Appendix A
identifies the 10 periods during the year when the calculated CO
concentrations are highest. The frequency distribution of the ratios of
street contributions to city-wide contributions for these "worst" 8-
hour average cases is shown in Figure 29. In this figure the results
from all the meteorogical data sets and from the different locations
relative to the city center have been combined. Comparison of Figure 27
with Figure 29 suggests that during the periods of highest CO
concentration, the contribution from the nearby street canyon is usually
less than for other cases. In only a third of these "worst" cases do
the street canyon emissions account for more than about a third of the
total concentration. However, about two-thirds of the worst cases have
contributions from the local street canyon that are more than 25 percent
of the totalo
For neighborhood type sites, the contributions from the nearest 2km
were found to account for 15 to 40 percent of the total CO in virtually
all cases. While this tends to be somewhat less than was true for the
total population, it still constitutes a substantial fraction.
69
-------
z
o
m
oc
o
o
tr
UJ
i
H
O
ffi
tr
O
O
t-
Ul
UJ
IT
O
o
4.0
2.0
1.0
.8
.6
.4
.2
4.0
2.0
1.0
.8
.6
.4
.2
T I I I I I
a. San Diego data, SW side of
^street
i i i
l I I I I I l
b. Millstone data, SWside of
street
T [ r i i i
St. Louis data, SW side of
street
I
d. St. Louis data, NE side of
street
10 30 50 70 90 10 30 50 70
PERCENTAGE GREATER THAN ORDINATE VALUE
90
D Center of city
• 17,5 km west of center
O 175 km south of center
A 17.5 km east of center
A 17.5 km north of center
SA-35 15-32
FIGURE 27 FREQUENCY DISTRIBUTIONS OF THE RATIOS OF STREET CONTRIBUTIONS
TO CITYWIDE CONTRIBUTIONS FOR 8-HOUR AVERAGE CO CONCENTRATIONS
70
-------
2.0
CO
2 i.o
i
CO
O
or
00
or
o
o
o
o
H
<
rr
•8
.6
.4
m
or
\-
z
o
o
or
LJ
o .2
P 2.0
• 0
.8
.6
.4
C/) O
z
2 2.0
1.0
.8
.6
.4
.2
1 I r i r
a. Son Diego data
I
I I i
J
D Center of city
• 17.5 km west of center
O 175 km south of center
A 17.5 km east of center
A 17.5 km north of center
T I i i i r
b. Millstone data
J I
I
J I
1 I
10 30 50 70 90
PERCENTAGE GREATER THAN ORDINATE VALUE
SA-3515-29
FIGURE 28 FREQUENCY DISTRIBUTIONS OF THE RATIOS OF CONTRIBUTIONS
TO 8-HOUR CO CONCENTRATIONS FROM SOURCES NEARER AND
FARTHER THAN 2 KM
71
-------
z
o
CD
£
z
z 2
5* O
0 O
L_ .y,
u u
UJ 5:
^ *
^~ O
fe2
O
5
QC
2.0
.0
.7
.5
I
I
I
I I
I
10 30 50 70 90
PERCENTAGE GREATER THAN ORDINATE VALUE
SA-3f> 15-30
FIGURE 29 FREQUENCY DISTRIBUTION OF THE RATIOS OF STREET
CONTRIBUTIONS TO CITYWIDE CONTRIBUTIONS FOR THE
HIGHEST 8-HOUR AVERAGE CO CONCENTRATIONS
The results of the calculations,based on a hypothetical but
reasonable distribution of emissions, confirm the importance of nearby
sources in determining observed CO concentrations, in general, and
during the occurrence of the highest concentrations. The variations
from case to case indicate that calculations based on a particular
distribution of emissions would be in order when there was concern about
the effects observed at a specific site. Such calculations might be
warranted at times other than during the initial siting process;
calculations of the sort presented above would provide valuable guidance
in the application of existing data, particularly in devising emission
control strategies.
Emission control strategies would be much more efficient if they
focused on those sources most responsible for violations of air quality
standards. The results of the simulations presented earlier suggest
72
-------
that violations of the 8-hour air quality standard at a neighborhood
station do not necessarily warrant the imposition of control measures
throughout the urban region. Emission reductions in an area that is
within a few kilometers of the site might be all that is needed to meet
standards.
Violations of air quality standards at street canyon monitors might
be the result of even more localized causes than those responsible at
neighborhood sites. The control strategies required to meet standards
at street canyon locations may only have to deal with streets that are
within a few blocks of the monitor.
The preceding discussion of the importance of the sources at
various distances from the monitor support the intuititve notion that
siting criteria based largely on the distribution of sources near the
monitor will be adequate to define, at least qualitatively, the spatial
representativeness of the station. The argument comes full circle when
we start deducing the influences of sources around a monitor from its
type, as specified in terms of scales of representativeness. Finally,
from the degree of influence exerted by various sources, we can deduce
those areas that are the most important in reducing CO concentrations at
the monitor. An important principle emerges—the application of the
monitoring data should be consistent in spatial scale with the area
represented by data from the site. Thus, if the concept of spatial
representativeness is carefully applied during site selection, then the
same concept can also be invoked for guidance when the data are later
interpreted and applied as bases for finding solutions to practical
problems.
73
-------
-------
Appendix A
A SIMPLE MODEL OF CONCENTRATION/EMISSION RELATIONSHIPS
75
-------
Appendix A
A SIMPLE MODEL OF CONCENTRATION/EMISSION RELATIONSHIPS
A. Introduction
This appendix describes a simple pollutant concentration simulation
model for inert, ground-level emissions. It can be used to process many
hours of meteorological data economically and answer two practical
questions:
What sequence of meteorological conditions, from
historical records, would have led to the highest multi-
hour average concentrations at a given location?
What are the relative contributions of sources at various
distances to the observed multi-hour average concentrations?
The answer to the first question identifies critical periods so that
they can be examined in greater detail. The problem of identifying
"worst-case" conditions arises often when the air quality impacts of
indirect sources are to be estimated. The second problem—that of
evaluating the relative impact of sources at different distances—can
arise during the site selection process for monitoring stations. The
question of undue influence from certain sources is often an important
issue.
The specific computer program described here has been designed to
answer the two questions presented above, but the basic simple model
itself should have application to a much broader variety of problems.
It should be particularly suitable in any case where it is desirable to
calculate a long sequence of concentrations very rapidly. The simple
model also provides a very useful basis for qualitative understanding of
the physical factors affecting concentrations at a site.
B. The Model
1. General
The model described here descends from the model proposed by
Ott, Clarke, and Ozolins (1967) by way of the APRAC-1A model developed
by Ludwig et al. (1970). The following discussion is derived largely
from the latter source. Both models divide the upwind region into
sectors, each of which is treated as a uniform, ground-level area
source. The APRAC model uses geometrically spaced upwind intervals to
76
-------
provide better resolution for nearby sources. Figure A-l shows the
spatial partitioning that is used for the emissions. Nine segments are
used, numbered from 1, for those sources within 125 m of the receptor,
to 9, for those between 16 and 32 km from the receptor.
Concentrations are calculated from two basically different
formulations. For sources near the receptor, a Gaussian model is used.
This model postulates that unrestricted vertical diffusion of pollutants
from ground-level line sources of infinite crosswind extent results in
concentrations that have a Gaussian distribution of concentration in the
vertical. The standard deviation, az, of this distribution depends on
such things as the atmospheric stability, and the distance downwind of
the source. Several authors have proposed empirical relationships among
az, atmospheric stability and travel distance. Figure A-2 presents the
Pasquill (1961), Gifford (1961) values, which are appropriate for rural
conditions. Urban conditions differ from those in the country (Ludwig
and Dabberdt, 1973), and for urban areas the curves in Figure A-3 are
more appropriate. The urban relationships were derived by Johnson
et al. (1971) from the work of McElroy (1969) and the Stanford Aerosol
Laboratory (1952). In both instances, the relationships can be
reasonably well approximated within each of the upwind segments by an
expression of the form:
a = a..r
z ij
(1)
where r is the distance between the source and the receptor and a-^j and
b-H are constants applicable to upwind segment i and stability class j.
This Gaussian model uses the following equation to calculate the
concentration from one of the upwind segments in Figure A-l.
Cij ua.. (1 " bij' (ri+l ri ) ij
(2)
0.8 Q,.
ua \ r
i
where
= the ground-level concentration arising from the ith
upwind segment for the jth stability class (gm/m3)
a..,b.. = constants in Eq. (1) used to specify o~z , the vertical
standard deviation of dispersion.
Q. = emission rate in ith segment (gm/m s)
r. = distance to the downwind boundary of the ith segment (m)
u = wind speed (m/s)
77
-------
CO
c
r
o
o
a:
D
O
CO
CO
LU
LL
O
CC
<
Q.
_J
<
<
0.
CO
cc
O
LL
Q
LU
CO
CO
z
LU
LU
CO
LL
O
CC
LU
CC
D
78
-------
1000
100
10
1.0
1 l i iiiiM / I I i i Lri 11
I I I
A = Very Unstable
B = Unstable
C = Slightly unstable
D = Neutral
E = Slightly stable
F = Stable
100
SOURCE. Turner, 197O.
1000 10,000
DISTANCE DOWNWIND — meters
SA-3515-16
FIGURE A-2 VERTICAL DIFFUSION AS A FUNCTION OF TRAVEL
DISTANCE AND STABILITY CATEGORY FOR RURAL AREAS
79
-------
10"
10°
10
A - Extremely Unstable
B - Moderately Unstable
C - Slightly Unstable
D - Neutral
E - Slightly Stable
10 102 103 104
DOWNWIND DISTANCE — meters
SOURCE: Johnson et al., 1971.
II
10a
TA-8563-98R
FIGURE A-3 VERTICAL DIFFUSION AS A FUNCTION OF TRAVEL
DISTANCE AND STABILITY CATEGORY FOR URBAN AREAS
80
-------
When the layer into which the pollutants are being dispersed
is restricted, they will tend to become uniformly distributed in the
vertical after sufficient travel has taken place. Under these
conditions, the "box" model can be used. According to the box model,
the concentration arising from a uniform area source in the ith annular
segment is defined by (see Miller and Holzworth, 1967) :
i uh ^i
(3)
where h is the depth in meters of the layer into which the pollutants
are mixed. When the box model is used, the concentration is independent
of stability. The point of transition from the Gaussian to the box
formulations occurs at that point where the two—in their line source
formulations—would give equal concentrations. Using this criterion,
the transition distance, rT, is:
(4)
The composite model, defining the concentration that results from
emissions in all the upwind segments, is given by:
1-b.. 1-b..
(5)
0.3 '^ - ' r-rl , r - r
The transition from Gaussian to box models occurs at rT in the Nth
segment.
2. Simplifications
The rather simple model described by Eq. (5) can be further
simplified if we take advantage of the finite number of atmospheric
stability categories and introduce categories of mixing depth p. and wind
direction. Finally, we can assume that the source strengths in the
different segments can be expressed as the product of a time-dependent
factor and a time-independent factor.
81
-------
Qi,d,t = ptQi,d
where
Qi d't = the source strength within the ith segment, in
the dth direction, for the tth hour. (g/m2 s)
Q-L d = the average daily source strength within the ith
segment, in the dth direction (g/m2 s)
Pt = a factor that gives the source strength for the tth
hour. Pt is assumed to be independent of location in
the city, except that provision has been made in this
model for using a different daily emission cycle for
sources near, versus far from, the receptor.
If categories of mixing depth are introduced (see Table A-l),
the model can be simplified to the following:
where
C = CO concentration at the receptor
u = wind speed
(X/Q)-j_ -i m = rati° °f tne ^0 concentration received from the ith
segment to the emissions in that segment (for unit
wind speed) . The values of these ratios depend on
stability class, j, and mixing depth class, m.
Values of (X/Q)i,j,m calculated using a mixing depth equal to
the geometric mean of the mixing depth class intervals in Table A-l and
rural 0~z functions from Slade (1968) are given in Table A-2. The values
for urban 'C 's (Johnson et al. , 1971) were used to derive the values of
i j m given in Table A-3. Table A-2 or Table A-3 will serve
essentially as a model of pollutant concentration. Both tables indicate
that contributions from sources within about 1/2 km are very nearly
independent of mixing depth, while those from sources at greater
distances are independent of atmospheric stability when the vertical
mixing is sufficiently restricted. It is beyond the scope of this
discussion to analyze the entries in the tables further.
82
-------
Table A-l
MIXING DEPTH CLASSES
Mixing
Depth Class
1
2
3
4
5
6
7
Mixing Depth Interval
(meters)
<100
100-200
200-400
400-800
800-1600
1600-3200
>3200
Geometric Mean of
Mixing Depth, Class
70.7*
141
283
566
1131
2262
4525*
* For classes 1 and 7, the geometric mean was calculated
as though the classes were bounded.
However, note that the tables are quite revealing of the behavior of the
model, and—to the extent that the model represents—the nature of the
real atmosphere.
3. Special Model for Street Canyon Sources
The importance of nearby sources will be enhanced if the
sources are confined in street canyons or other special situations. The
model that is described here has a provision for treating street canyon
effects, using the empirical submodel from the APRAC-1A model (Johnson
et al., 1971; Ludwig and Dabberdt, 1972). For winds blowing toward the
side of street on which the receptor is located:
KQ „ ( Wind blowing
—— I toward receptor's
"close ~ W(u+0.5) H | side of the street
where
C
close = contribution of close (street canyon) sources to
the concentration.
H = depth of street canyon (m)
W = width of street canyon (m)
Qs = emission rate on street (g/m s)
u = wind speed at roof level* (m/s)
z = height of receptor (m)
K = nondimensional empirical constant (approximately = 7)
* When airport winds are used, the term (ufO.5) is halved to approxi-
mate the effects of urban roughness in reducing wind speed.
83
-------
Table A-2
CONTRIBUTIONS TO THE CONCENTRATION (ngm m~3) AT A POINT FROM EMISSIONS
OF UNIT STRENGTH (1 u.gm m-2s~1) AT VARIOUS UPWIND
DISTANCE INTERVALS (FOR UNIT WIND SPEED -1 m s"1)
RURAL
Stability
Class*
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
Mixing
Depth
Interval
<100 m
100 to
200 m
200 to
400 m
400 to
800 m
800 to
1600 m
1600 to
3200 m
>3200 m
Upwind Distance Interval
1-
125 m
4.60
6.89
10.38
16.11
23.0
4.60
6.89
10.38
16.11
23.0
4.60
6.89
10. 38
16. 11
23.0
4.60
6.89
10.38
16.11
23.0
4.60
6.89
10.38
16.11
23.0
4.60
6.89
10.38
16.11
23.0
4.60
6.89
10.38
16.11
23.0
125-
250 m
3.04
4.92
7.22
11.91
16.67
3.04
4.92
7.22
11.91
16.67
3.04
4.92
7.22
11.91
16.67
3.04
4. 92
7.22
11.91
16.67
3. 04
4.92
7. 22
11.91
16.67
3.04
4.92
7.22
11. 91
16.67
3.04
4.92
7.22
11.91
16.67
250-
500 m
3.60
4.99
7.49
13.48
18.96
2.62
4.99
7.49
13.48
18.96
2.61
4.99
7.49
13.48
18.96
2.61
4.99
7.49
13.48
18.96
2.61
4.99
7.49
13.48
18.96
2.61
4.99
7.49
13.48
18.96
2.61
4. 99
7.49
13.48
18.96
500-
1000 m
7.07
7.07
8.24
15.74
22. 50
3.54
4.60
8.08
15.74
22.5
1.96
4.55
8.08
15.74
22.5
1. 58
4. 55
8.O8
15. 74
22. 5
1. 58
4. 55
8.08
15. 74
22.5
1. 58
4. 55
8.O8
15.74
22. 5
1.58
4.55
8.08
15.74
22.5
1-2 km
14.14
14.14
14.14
19.17
27.6
7.07
7.07
8.79
19.17
27.6
3. 54
4.03
8.75
19. 17
27.6
1.77
3.60
8.75
19. 17
27.6
0.955
3.60
8.75
19.17
27.6
0.67
3.60
8.75
19. 17
27.6
0. 604
3.60
8.75
19. 17
27.6
2-4 km
28.3
28.3
28.3
28.5
34.9
14.14
14.14
14.14
24.2
34.9
7.07
7.07
9.65
24.2
34.9
3. 54
3. 54
9.65
24. 2
34.9
1. 77
2.38
9.65
24.2
34.9
0.884
2.88
9.65
24.2
34.9
0.442
2.28
9.65
24.2
34.9
4-8 km
56.6
56.6
56.6
56.6
56.6
28.3
28.3
28.3
31. 5
47.1
14. 14
14. 14
14. 16
31.1
47.1
7.07
7.07
10.89
31.1
47.1
3. 54
3.54
10.9
31.1
47.1
1. 77
1.83
10.89
31.1
47.1
0.884
1. 28
10.89
31.1
47.1
8-16 km
113.2
113.2
113.2
113.2
113.2
56.6
56.6
56.6
56.6
67.4
28.3
28. 3
28.3
42.0
67.4
14.14
14. 14
14.50
42.0
67.4
7.07
7.07
12.97
42.0
67.4
3. 54
3.54
12.97
42.0
67.4
1.77
1. 77
12.97
42.0
67.4
16-32 km
226.
226.
226.
226.
226.
113.2
113.2
113.2
113.2
113.2
56.6
56.6
56.6
60. 5
100.3
28.3
28. 3
28.3
59.3
100. 3
14.14
14.14
16.37
59.3
100.3
7.07
7.07
16.3
59.3
100.3
3.54
3.54
16.3
59. 3
100.3
1 = extremely unstable
2 = moderately unstable
3 = slightly unstable
4 = neutral
5 = slightly stable.
84
-------
Table A-3
CONTRIBUTIONS TO THE CONCENTRATION (^gm m 3) AT A POINT FHCM EMISSIONS
OF UNIT STRENGTH (1 M-gm m^s'1) AT VARIOUS UPWIND
DISTANCE INTERVALS (FOR UNIT WIND SPEED -1ms
URBAN
-1
Stability
Class
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
Mixing
Depth
Interval
<100 ra
100 to
200 m
200 to
400 ni
400 to
800 ra
800 to
1600 m
1600 to
3200 m
>3200 m
Upwind Distance Interval
1-
125 m
2.9
3.4
4.0
4.8
6.2
2.9
3.9
4.0
4.8
6.2
2.9
3.4
4.0
4.8
6.2
2.9
3.4
4.0
4.8
6.2
2.9
3.4
4.0
4.8
6.2
2.9
3.4
4.0
4.8
6.2
2.9
3 .4
4.0
4.8
6.2
125-
250 m
2.0
2.3
2.8
3.6
5.2
1.9
2.2
2.8
3.6
5.2
1.9
2.3
2.8
3.6
5.2
1.9
2.3
2.8
3.6
5.2
1.9
2.3
2.8
3.6
5.2
1.9
2.3
2.8
3.6
5.2
1.9
2.3
2.8
3.6
5.2
250-
500 m
3.5
3.5
3.6
4.3
7.3
1.9
2.1
2.9
4.3
7.3
1.5
2.0
2.9
4.3
7.3
1.5
2.0
2.9
4.3
7.3
1 .5
2.0
2.9
4.3
7.3
1.5
2.0
2.9
4.3
7.3
1.5
2.0
2.9
4.3
7.3
500-
1000 m
7.1
7.1
7.1
7.1
10.3
3.5
3.5
3.6
5.0
10.3
1.8
2.0
2.9
5.0
10.3
1.3
1.8
2.9
5.0
10.3
1.3
1.8
2.9
5.0
10.3
1.3
1.8
2.9
5.0
10.3
1.3
1.8
2.9
5.0
10.3
1-2 km
14~. 1
14.1
14.1
14.1
14.9
7.1
7.1
7.1
7.1
14.4
3.5
3.5
3.6
5.9
14.4
1 .8
1.9
3.0
5.9
14.4
1.1
1.7
3.0
5.9
14.4
1 .0
1.7
3.0
5.9
14.4
1.0
1.7
3.0
5.9
14.4
2-4 km
28.3
28.3
28.3
28.3
28.3
14.2
14.2
14.2
14.2
20.2
7.1
7.1
7.1
7.4
20.2
3.5
3.5
3.6
6.9
20.2
1.8
1.8
3.1
6.9
20 2
1.0
1.5
3.1
6.9
20.2
0.9
1.5
3.1
6.9
20.2
4-8 km
56.6
56.6
56.6
56.6
56.6
28.4
28.4
28.4
28.4
29.6
14.1
14.1
14.1
14.1
28.4
7.1
7.1
7.1
8.2
28.4
3.5
3.5
3.6
8.1
28.4
1.8
1.8
3.1
8.1
28.4
0.9
1.4
3.1
8.1
28.4
8-16 km
113.2
113.2
113.2
113.2
113.2
56.7
56.7
56.7
56.7
56.7
28.3
28.3
28.3
28.3
39.9
14.1
14.1
14.1
14.1
39.9
7.1
7.1
7.1
9.5
39.9
3.5
3.5
3.7
9.5
39.9
1.8
1.8
3.2
9.5
39.9
16-32 km
226.3
226.3
226.3
226.3
226.3
113.5
113.5
113.5
113.5
113.5
56.5
56.5
56.5
56.5
58.8
28.3
28.3
28.3
28.3
56.0
14.1
14.1
14.1
14.2
56.0
7.1
7.1
7.1
11.1
56.0
3.5
3.5
3.7
11.1
56.0
1 - extremely unstable
2 - moderately unstable
3 = slightly unstable
4 = neutral
5 = slightly stable.
85
-------
for the opposite side of the street:
^s ( Wind blowing away
r = —. ] from receptor's
close c\/o, /—5 , x ' side of the street
(u+0.5)(2+/x2+z2 ) (8b)
where
x = horizontal distance from the receptor to the nearest
lane of traffic (ro)
Finally, when the wind blows within 30 degrees of the street
alignment, the average of expressions (8a) and (8b) is used to calculate
concentration.
C. Applications
Once a model is available to calculate hour-by-hour concentrations
from near and far sources, then it is possible to apply methods
generally reserved for use with observed monitoring data. In
particular, it is possible to consider multi-hour averages directly and
without resort to statistical models. In this subsection, some specific
applications of this type will be considered briefly. One such
application is the identification of "worst-cases" for different
averaging periods. The sequence of calculated near and far
contributions, can be scanned and running means determined. The highest
values are easily identified along with the time interval during which
they occurred. Once a limited number of candidate worst cases have been
identified, then the specific cases can be studied in more detail and
with more comprehensive models.
Air quality impact assessments often require that "most probable"
and "worst" cases be identified and the impact of some unbuilt complex
be determined for these two categories. We will first consider the
"worst" case. "Worst" will depend on some unique distribution of
sources (often, as yet unbuilt), daily emission cycles, and sequences of
meteorological conditions. Because of differences in the surrounding
sources, the "worst" periods may not occur everywhere at the same time.
Therefore, it may not be possible to identify worst conditions on the
basis of observations taken at established monitoring sites, and the
simple model provides a useful tool.
For single-hour averages, the most probable and worst cases are
easily defined as the most common combination of stability, mixing
height, wind speed and wind direction, and the observed combination that
produces the highest concentrations, respectively. Identification of
the worst cases was discussed above; identification of the most probable
cases would probably require some subjective analysis by a professional
meteorologist, but the model could be used to reduce the complete set of
conditions down to some smaller subset that contained all those
86
-------
instances when calculated concentrations fell within some specified
interval around the mode. The general meteorological conditions that
accompany these instances could be studied and classified and perhaps
one, or a few, "most probable" prototypes could be selected for detailed
analysis.
Site selection for monitoring stations sometimes requires that the
measurements be representative of a large area and, in other instances
it is desirable to monitor the effects of nearby sources. Thus, it may
be desirable to evaluate the relative contributions of sources at
various distances from a proposed monitoring site. If multi-hour
averages are of importance, such as might be the case when a site is
serving to determine compliance with federal or state air quality
standards, this model can provide a method for solving the problem.
Running means can be calculated for the contributions from near sources
and from more remote sources. The contributions from the two areas can
be compared by calculating the average of their ratios, the frequency
distributions of the ratios, or the average of their differences; all
are possible approaches to the comparisions, once the running means are
available.
The following subsection describes a computer program that was
written to use the model presented above to identify worst cases and to
calculate a frequency distribution for the ratio of the contributions of
near and far sources to receptor concentration. The program listing, in
Control Data FORTRAN Version 2.1, is given at the end of this appendix.
D. Computer Program
A simplified flow chart of the program SMOCER (simple model of
concentration-emissions relationships) is shown in Figure A-4. Many
calculations are made in the initial stages and the resulting values are
stored for subsequent use. Once these preliminary calculations are
complete, hourly meteorological data is read and used to calculate
hourly concentrations, with the contributions from near and far sources
treated separately. The listing included here has provisions for
reading the meteorological inputs from tape with the program
modifications necessary to use card inputs given as comments.
Modification of the program might be required for use on small
computers. The version given here stores hourly values of the several
pertinent parameters for a period up to one year. After the values are
all stored, running sums and other calculations are made. If storage
were limited, such calculations could be made sequentially as the hourly
values were obtained. Such a modification should not be particularly
difficult.
The program is capable of obtaining two different kinds of output
from the sequence of hourly results. First, it can identify the ten
worst cases for the period. Initially, no restrictions were placed on
the selection process except that the highest multi-hour averages were
to be chosen. In practice, several of the worst cases will overlap.
87
-------
That is, the multi-hour averages for several cases may include some of
the same hours. Since it was intended that the identified cases be
relatively independent, overlapping cases are no longer considered. If a
sequence of high values of running means occurs, and they are separated
by less than the averaging interval being used, only the highest value
will be chosen even though other values in the sequence might exceed
some of the retained high values from other parts of the total list.
After all the data have been treated, the model will produce a list of
the ten highest running mean concentrations (with the restrictions noted
above) and the day and last hour of the corresponding averaging period.
The program will also calculate the ratio of the contributions from
near sources to those from sources farther away, again based on multi-
hour running means.
Table A-4 lists the FORTRAN variables used in the program and their
meaning, and Table A-5 gives the input requirements for using the
program. This is followed by the program listing. During the course of
this study the program was applied with Los Angeles inputs using one
year of hourly data on tape. The costs, using a CDC 6400 computer, were
less than four dollars, including compilation of the program. Central
processor time, excluding compilation, was less than 13 seconds.
88
-------
C rO CD
01 4-1
C O
O
QJ
en
0
jr
4-1
00 "4-1
0
JJ
Qj
to en
£ QJ
0 J=
U J-1
>-,
CTi
'D i-<
oi c
0
UI
"O
cO O
I-i ^
CD
3
CD
> GT3
QJ
>
.r-l
CO
bO
OJ
C
01
^
01
x:
4_l
o
cO
o
u
QJ
J-i
!-l
3
u
01
4_l
r-l
0-
T)
01
CO O
4-1 in
OJ "D
4-1 CO
CD Oi
QJ 1-4
00 cO
w en
13
t>0 >
c
C en
c o
d -c
CJS
o
«
c
O rH
4J CO
cO
C 'J-i C
QJ O
u tn -H •
C OJ 4-J X!
O U U 4-1
U i-l 01 Q,
D U 01
T3 O -H -D
QJ in ID
N 60
T-l QJ C T-f
frt LJ .-4 y
Precalculate norrni
for more remote a:
combinations of w:
, stability , and mi:
-^
"D i OJJ2
C •<-> C J-1
03 J-i >,.,-| OJ
m C ^ i-i 4J 01
4-1 O 03 CU JT
d u 01 4-1 en 4-i
Q- C OJ C
C "O "O O 01
•r4 01 TD T-l )-l
N E= 0 4J 0
r-l -H CO 4-1 3 4J
CO i-l J3 tn
u cQ Oi en -H
• H g 4-1 0) i-l
&C 1-4 O U 4J •
O O 6 H C w
•-I C QJ D O C
0 Jj 0 U 0
Read hourly meteor
use precalculated (
butions from more
(street or area) s
the corresponding
hourly concentratl
hourly values .
1-1
C -u
u E
QJ .
X i-i tn
7^§
3 Vi en
O O
•£• E 00
C
'W 13 -H
oca
CO C
in 3
0) >, i-l
u J^
Use stored sequen
butions from near
sources to obtain
t .
^
QJ
i-l
Qj r-l 4J
C 4J C
•H ro QJ
4-1 rH U
a 3 c
o 50
I-i 3 U
J3 U
3 U 4-1
CO < O
__^
QJ
X!
O
OI
cO
o
4J
VJ
0
3
dj
<
f
I 2
X«
C~'
-D
0>
4-4
r-l
C
-o
K
z
_^
{ i
V
\ 5
;
; &
^x ^
0
C_J
1 1
O
I—
OL
d;
1C
CJ
~^L
O
1
u_
Q
UJ
I— 1
U_
i — ^
— 1
a.
i — i
(ji
•3-
i
=c
LaJ
C£.
<~S
1 — 4
u_
89
-------
^
-^
o
=
IOUTINE
pa
D
en
L
TJ W
cu E
4J 3
01
to ac
QJ c
0. C
Ol C
Li 3
T3
O u
S-. Oi
Ol Ll
Q, Li
3
QJ CJ
^
4-1 U
jz;
01
O X
O J^
,
A
I 'J
Ua
^^1
01
o
£
4-1
c>-
4-t ID
O (U
L.
>i O
ca to
03 TJ
— I 03
Li 0)
> *H
O 03
,
k
)
S^\
V /
p
cu c oi
Li 03 J=
Ll £ 4J
3 J-J
u e
Li O
OJ Ol l-i c-.-
s: 4-1 14-1 TJ
4-1 CO O
Oi 01 "H
O tJJi-t 1>
co a
ca E oo
•w 3-0 C
O en Oi -^
4-1 l-i CL
M o a.
01 C 4J 03
4J C 5-i
C Oi 01
« d ^ >
M Ll 4J O
A
t
i"
"O C1J
C ' 3
T3 03 Li i— i
QJ O 03
J-i Ul 4-1 >
o e c
"a -> y^
O V4 4J
a -H
-i- 5
o c
0
"O -H
e i
£: 0 r-t
3 ^ 03
w C >
r-l
tO 4-J
C "O P
•i-j ai ai
c: LI LI
c: o LJ
3 4J 3
pi LI w u
r-l "O
CD
Q>
£
CO
Oi
^-1
00 03
E
tfi
o
4_)
3
W)
c
•iH
c
Li •— 1
ai O T-I
^
J
O
4J
01
4J q
o
4-1-,-,
O 4-)
3
bution
ontrib
-r( U
V4
to 0
•-( 14-1
"O ~^
Li
>! 03
U 01
C C
Ol
CT.C
TJ
01
C
1-1
E
Ol
01
T3
,£3
o-
3
b-
— »• s
01
C
•H
4J
3
O
,0
3
en
03 p
in 3
->o
W (J •
03 to
HI 4-1 01
C C en
O Ol
Ll
0 u
•^ c
fl "*
c
4* 03
4-)
ca w
3 U
ca o
O to
»,
^
C
a>
Ol
42
w
ai
d
>
>,
LJ
P
-C
m
0
flj
c
ai
3
D"
UI -O
ai u
JZ QJ
a.
« E
£d u
1
CJ3
§
a.
a;
LU
O
O
<:
<_>
00
O
O
03
90
-------
Table A-4
FORTRAN VARIABLES USED IN SMOCER
ADT = Average daily traffic (vehicles/day).
ANGLE = Angle normal to STDIR.
CFAR = Far concentration array.
GLEE = Concentration with unit wind speed for receptor facing.
away from the wind direction.
CLOSTYP = Type of source near receptor (street or area).
CMILGM = Average concentration mg/m3.
CNEAR = Near concentration array.
COS60 = Cosine of 60 degrees.
CWIND = Concentration with unit wind speed for receptor facing
toward the wind direction.
DP = Dewpoint (not used, but on meteorological tape).
DT = Temperature (not used, but on meteorological tape).
EMRATE = Average vehicular emission rate (g/veh.-mi.).
FACTOR = Used to convert emissions in g/mile-day to g/m s.
I = Mixing depth index.
IDIR = Upwind direction (1 = north, ...16 = north-northwest,
0 = calm).
IHR = Hour index.
ISEG = Upwind segment index.
ISTAB = Stability index.
MEANHR = Averaging interval for running means (hrs).
MHR = Hour (of NDAY) of end of averaging interval for a high
concentration case.
NCLOSE = Number of segments to be considered as near when
CLOSTYP = areas; when CLOSTYP = Street, then NCLOSE = 1.
NDAY = Number of the day in which last hour of averaging
interval occurs for a high concentration case.
NHR = Last hour of averaging period (running sequentially from
first hour) of the highest concentrations.
NN = Number of hours since start.
OC = Opaque cloud cover (not used but on meteorological
tape).
OUTPUT = Type of calculations to be performed.
PPM = Average concentration (parts per million).
PTFAR = Fraction of far source emissions occurring during hour.
PTNEAR = Fraction of near source emissions occurring during hour.
QST = Emission rate on street (g/m s).
RATIO = Frequencies of ratios of NEAR/FAR source contributions.
91
-------
Table A-4 (concluded)
FORTRAN VARIABLES USED IN SMOCER
SI = Stability index (1 = extremely unstable ... 5 = slightly
stable).
SIDE = Side (left or right) of receptor when facing STDIR.
STDIR = Street direction (degrees from north).
TOPRAT = Ratios of NEAR/FAR contributions—during same periods as
the highest concentrations.
TOP10 = Ten highest multi-hour concentrations.
VALUE = Class boundaries for cumulative frequency distribution.
(number greater than . . .).
WD = Wind direction (16 directions, 1 = NNE...16 = N,
0 = Calm).
W,H = Width and height of street canyon (m).
WS = Wind speed (m/s).
Z,X = Height and horizontal distance of receptor from near.
traffic lane (m).
92
-------
Table A-5
INPUTS FOR A SIMPLE MODEL OF CONCENTRATION/EMISSION RELATIONSHIPS
Card Number
Variables
Format
Remarks
2A
This card
used only
if CL0STYP
= STREET
3-7
13-28
29-31
32-34
35
Problem identification, 16A5
up to 80 characters
CL0STYP (either "STREET" A10
or "AREAS")
OUTPUT ("RATIO," "WORST," A10
or "BOTH")
MEANHR
NCL0SE
ADT
EMRATE
STDIR
15
15
F8.1
F8.1
F8.1
SIDE ("LEFT" or "RIGHT") AS
W, H 2F8.1
Z, X 2F8.1
X03 values for 9 segments 9F8.1
on one card, cards for
stability classes 1-5,
1st mixing category
X0Q-stability classes
1-5, 2nd mixing
category
9F8.1
9E8.1
PTNEAR fraction of dally 8F10.3
traffic/hour
PTFAR fraction of daily 8F10.3
traffic/hour
MIXTYP mixing type to be 2412
used with XOQ
Whatever this card contains is printed with
outputs for identification
Must start at Column 1, identifies whether
close source uses street canyon model
Calculates only ratios, only worst cases, or
both, depending on this input. Must start
at Column 11
Number of hours for averaging
If Clostyp not "STREET" tells how many segments
are considered as "near"
•
Average daily traffic-vehicles/day
Emission rate g/vehicle-mile
Street direction, degrees from north
Starts at Column 25-side when facing STDIR
Street canyon width, height (m)
Vertical, horizontal distances (m) to
nearest traffic
Cards 3 through 12 contain concentration/
emission ratios from Tables A-2 or A-3
Same as Cards 3-7.
16 cards containing emission rates
(g/m -s), 9 segments on a card, one card
for each direction 1, 16
From near sources (or street), hours run from
1 to 24
From far sources, hours run from 1 to 24
24 values (cither 1 or 2) for each hour from
1 to 24
Meteorological inputs follow, either from tape or cards, see program listing.
93
-------
PROGRAM LISTING
SMOCER (INPUT, OUTPUT»TAPF1)
C
C»**»»»*THIS SIMPLE MODEU OF CONCENTRATION RELATIONSHIPS IDENTIFIES
C»**»»»»«*«THE WORST CASES FOR DIFFERENT AVERAGING INTERVALS ANO THF
C»»******»«RFLATIVF CONTRIBUTIONS FROM NEAR ANO FAR SOURCES.
C»*»»»*»*#*F.L.LUDWIG ANO H, SHIGTESHl.
C
DIMENSION XOO (9,5,2) ,3(9,16) »CNEAR(97B4) »CFAR(8784) »CMNDR* ( 5 , 1 6, ?)
1,CFNORM{5,16.2) ."TNEMia*) ,PTFAR(24) ,MIXTYP FORMAT (3F8.1,A0«4F8.1)
7 FORMAT (1H ,#AOT s«F8.2»3X*EMRATE =»re.2t3X»sTOlR =*Fi),2i3X»STDE =
1»» AB«1X#W *»F8.??»3X
15 FORMAT (8F10.3)
1? FORMAT <9F8.i)
18 FORMAT (9F.8.1)
C
C**«»»»»**»FACTOR is USED TD CONVERT EMISSIONS IN GM/HILE-OAY TO
c
FACTOR = 1.0/<1609.35»3600.0»24.0)
C
C»»»*»»*»#»RFAD lOFNTTFYlNG HEADING
C
PRINT i $ RFAD 2,ALPHA $ PRINT 3
c
TYPES OF CALCULATION TO BE1 PERFORMED.
CCOSTYP=TYPE' OF SOURCE NEAR RECEPTOR (STREET DR AREAS),
OUTPUTsTYPE OF CALCULATIONS TO RE PERFORMED...
RAT10--FIND FREQUENCIES OF RATIOS OF NEAR/FAR SOJRCE
CONTRIBUTIONS.
WORST—FIMO 10 HIGHEST CONCENTRATIONS.
MEANHR—AVERAGING INTERVAL FOR RUNNING MF.nNS (HOURS),
MCLOSE--NJMBER OF SEGMENTS TO BE CONSIDERED AS NEAR
WHEN CLOSTYP=AREAS, WHEN CLOSTYPsSTREET, THEN NCir>SE*l
READ 4,CLOSTYP,0'JTPUT,MEANHR,NCLOSE
PRINT5.CLOSTYP,OUTPUT,MEANHR.NCLOSE
MCLOSE=NCLOSF*1
TF (CLOSTYP.NE.6HSTREET) GO TO 100
94
-------
c
C»**#*«**«*FnR STREET CASE READ
c***»»««««ft AOT»AVERAGE DAILY TRAFFIC (VEHICLES/DAY)
c»*«*»«*»** EMRATE = AVERAGE VEHICULAR EMISSIOM RATE
C»»«»»»*«*» STOIR « STREET DIRECTION (DEGREES FROM NORTH')
C»»*»»»»«»ft SIDE « SIDE (LEFT OR RIGHT) 0? RECEPTOR WHEM' rftCI^S
C»***»»»««* w,H = WIDTH JND HEIGHT OF STREET CANYON
C
READ 6.ADTtEMRATE»STDlR»SIDE,W,H,Z»X
PRINT7,ADT»EMRATE,STDIR»SIDE»W»H,Z»X
C
C»«*»»»*««*Q«;T»EMISSION RATE. ON STREET (2*»GM/M-S ) . NCLOSE«1»
C»*«»»*»»«»MrANS THAT THE STREET REPLACES THE 1ST SEGMENT -\S THE
C
QST=ADT»EMRATE«FACTOR $ NCLOSE*I $
100 PRINT a, (L»L«1»9)
c
C«#*»»»»»«»R5'AD RATIOS OF CONCENTRATION TO' EMISSIONS ( XOO--S/M) AT JwIT
C#*«»**»»*#WTND SPFEDIM/S) FOR VARIOUS UPWIND SEGMENTS dSESOi STAB-
C»»*«»*»«»«ILITIES (ISTAB) AND MIXING DEPTH CLASSES (I)i
C
READ 17 1 ( ( (XOQ(ISEG»ISTA8iI) »ISES«1,9) ,ISTAB = lt5) »I«1,E)
PRINT Q, « I STAB, (XOQ(ISEG.ISTAB,I) ,ISEG=1,9) ,ISTAB=l,5i) «I = 1»2)
PRINT 10.(L,L=1»9)
C
C«««»»»»«»»REAO EMISSIONS (a-,GM/M»M#s» IM DIFFERENT UPWIND DIRECTIONS
C»**»»«»«*»tTOlR) AND SEGMENTS (ISEG).
C
READ 1«, ( (QflSEG.IDIR) ,ISEG»1,9) iIOIR*l,16>
PRINT 9, (IDIR, (Q(ISEG,IDIR) ,ISEG»1»9) *I9IR»l*l6)
C
C»*»»««««»#RFAD FRACTION OF DAILY TRAFFIC OCCURRING DURING EISCH HOUP
C»»»»»»»»»»(IHR) FOR NEAR(PTNEAP) AND FAR(»TFAR) SOURCES.
C
READ 1«5, (PTN|F_AR(IHR) ,IHR»1,2*)
READ !<>» (PTFAR ( IHR) . IHRsl ,3^)
C
C»«*«*«»«««READ MIXING DEPTH TYPE FOR EACH HOJR,
c
READ 11, (MIXTYP(IHR) ,IHR=1,?*)
PRINT 1~2
PRINT 13, (lHR«PTMEAR(IHR) ,PTFAR(IHR) ,MIXTYP(lHR) ,IHR=1»2*)
IF (CLOSTYP.FQ.6HSTREET ) GO TO 125
C
C##»»»»»»#»CALCULATING NORMALIZED CONCENTRATIONS FROM NEAR, NON-STRF-ET
c
00 120 IDlRal.16
f>0 115 ISTAB.1,5
00 110 J»l,2
SUMCN«0,0
00 105 ISE3si,NCLOSE
SUMCN««?UMCN*XOQf I SEG, ISTAB, J)»Q(ISE9, IDIR)
10«5 CONTINUE
CNNORM(ISTA8,IDIR,J)»SUMCN
HO CONTINUE
115 CONTINUE
120 CONTINUE
SO TO 14.5
C
C»»»#*»»»#»CaLCULATIigB NORMALIZED CONCENTRATIONS FROM STREET SOURCE FOR
C»«»»**»#»»UNIT WIND SPEED.
C
00 UP IDIR*1,16
95
-------
CALL STREET
C
C»»««««««*»STREET CONCENTRATIONS ARE INDEPENDENT OF STABILITY AMD MIX-
C»«*»»»»»«»lM6 DEPTH
C
DO 135 iSTABsl.1?
DO 130 Jsl»2
CWORMifISTAB.IOIR»J)=XNORM
130 CONTINUE
135 CONTINUE
Un CONTINUE
145 00 165 IOlRrl,l6
00 160 ISTABsl.5
00 155 Jsl»?
SUMCF=0.0
C
C»«»»*#»»»«C&LCULATING NORMALIZED CONCENTRATIONS FROM FAR SOURCES.
C
DO 150 ISE6sMCLOSEt9
SUMCF=SUMCF+XOQ(ISEG. ISTAB. J) *Q (ISE9<, IDI»)
150 CONTINUE
CFNORMfISTAB.IDIR*J)sSuMCF
155 CONTINUE
160 CONTINUE
165 CONTINUE
NIN=0
C
C»«««»*»*»*FTLL CONCENTRATION ARRAYS (878* HOJRS»I LEAP YEARit NEAR
C»*»»»*»»#«(rNEAR)ANO FAR(CFAR) WITH NESATIVE VALUES
C
CALL MFMSETx (-99.0,CNEAR,8784)
CALL MfTMSETx (-99.0»CFAR t87B4)
C
C»»»»»»»»»»RFAD 2* HOURS OF METEOROLOGICAL! DATA...
DIRECTION(16 DlR.t 1«N..,16«NMW.,.0«CALM)
SPEED (M/S)
C»«#»*«»*«# OTtDP.3C=TE*1PERATUREtDEWPOlNT» CLOUD ..NOT jSfD.
c»«»»*##*#» SlaSTABlLITY INDEX (1»EXTR£MEL .DT,OP,OCfSl(L) »L«1»Z*-)
TF (EOF(1» ?00. 175
C
C»»»«»»»»»#FOLLOWING CA«DS CAN BE USED (WITHOUT COMMENTS) TO READ
C»#»*»«»»»»MFTEOROLOGICAL DATA FROM CARDS.
C 170 00 171 L=l»24
C READ l .L'T. o.o ) 30 TO 200
C 171 CONTINUE
C
17>5 00 195 1HR»1,24
C
C»##»»*»**#NN=NUMBFR OF HOURS SINCE START,
c
NN = NN+I
c
C»»»»»»»»»#SKIP MISSINB DATA...NEGATIVE VALUES.
c
TF (WD(IHR).LT.O.O.OR.WS(IHR).LT.O.O.OR.SI(IHR).LT.O.O) 30 TO 195
jsMIXTYP(IHR)
96
-------
C»###**#CALM WINDS SET= 1 M/S AND N^OSf HECLNTLY OBSERVED DIRECTION.
C
IF UD(IHH) .EQ.0.0) 60 TO 180
IF (WS(IHR) .LT.1.0) WS
195 CONTINUE
GO TO 170
200 NTOTsNN
C
c»»#«#««xnt*S£T ARRAYS NEGATIVE...
c*tt**»«ff«»« TOP10= 10 HIGHEST MULTI-HOUR CONCENTRATIONS
C»*««*»»»«* ERAGE).
c«##«##»»## TOPRA,T= RATIOS OF NEAR/FAN CONTRIBUTIONS ...SAME PERIODS
c#«tt««#
CALL MEMSETX (-9.0.TOPKAT. 10)
CALL MEMSETX (-9»NHR,10)
C
C#«#**#»»#«SET FREQUENCIES TO *ERO.
C
CALL MEMSETX
FARC*FARC*CFAR(NN)
IF (MM.LT.MEANHR) GO TO 2o5
CTOT*CLOSC*FARC
CALL CONFREQ
IF (OUTPUT. E0.5HWORST. OR. OUTPUT. EQ.4HBOTH) GO TO 215
IF (OUTPUT. EQ.5HRATIO> GO TO 220
PRINT 14
STOP210
97
-------
c
C»*»»»»»*»»CHECK TO SEE IF SUNNING SUM GREATER THAM THOSE ALREADY STORED
C
IF (CTOT.GT.TOPlOdOn CALL 8IG10
IF (OUTPUT ,EQ. 5HWORST) GO TO 222
C»*ftft»«»««« IF FAPC«0 SET TO VERY SMALL
C
220 IF (FAPC.LE.OtO) FARC»1.0E-20
CALL
222 CLOSC « CLOSC-CNEAR(NMHPl)
FARC « FARC-CFA«(NMHP1>
IF (NN.LT.NTOT) 30 TO 205
CALL CgNPNT
IF (OUTPUT. EQ.BHrtORST. OR. OUTPUT. EQ.M80TH) 230.235
230 CALL BTGPMT
IF (OUTPUT. ME. 4H80TH) STOP230
23s CALL RATPMT
STOP23S
98
-------
SUBROUTINE BTGPNT
C
C*»»»»«**»»THIS SUBPROGRAM PRINTS THE 10 LARGEST MEANHR-AVE*A3es
C»«**»*»»»#OTHER ITEMS.
C
COMMON /TOP/ NN,CTOT,CLOSC,FARC,TOPlOUO),NHR<10)iTOP3AT(lo).
ICOFREQ(IO)
COMMON /DAT/ ALPHA(16)tMEANHR,CLOSTr°.NCLOSE
1 FORMAT <1H1,5X16A5)
2 FORMAT (/1H ,«Hl3HEST»I3, *»HOUR AVERAGES*)
3 FORMAT
-------
SUBROUTINE
C
c»***»«**»«Tms SUBPROGRAM CHECKS MN TO SFF. IF IT is WITHIN MEANHR OF
C»«»»»*»»**ANY OF THE DETAINED VALUES OF MHR. IFNOT CTOT IS ADDED TO THE
C»»»»»»»..»LTST OF RETAINED VALUES OF TOPlO. TOPlO(lO) IS DROPPED AND TH
C»».*»«»»*»LTST Is REORDERED,LARGEST (1) TD SMALLEST (10). MN AND
C»»«»»*»«»«CLOSC/FARC ARE ADDED TO THE NHR AND TOPRAT LISTS ix THE SAMF
C*»»*»«»»»*PnSITlDN THAT CTDT WAS ADDED TO THE TOPlO LIST. IF NN IS WITH
C»***»»»«»«IN| MFANHR OF SOME NHR THEN THE L&ST IS NOT REVISED
c**»«»«»»«*ctoT EXCEEDS THE CORRESPONDING TOPIO.
COMMON /TOP/ NN.CTOT,CLOSC,FARC.TOPlO(lO) »NHP(10) .TOP^H
ICOFREQhO)
COMMON /DAT/ ALPHA(16) ,MEANH9,CLOSTyo,NjCLOSE
RATSO = CLOSC/FACJC
00 130 1=1,10
C
C»*i»»#»*»»CHEC< TO SEE IF MN IS WITHIN ME4NH3 OF ANY NHR.
C
IF {(NN-NHR(t)).ST. MEANHR) SO TO 130
C
C»#**»«n»»#*CMECK TO SEE IF CTOT IS GREATER THAN CORRESPONDIMG TOPlO.
C
TF (CToT.GT.TOPlOd) ) GO TO 1?0
RETURN
C
C»»»#»#*«»*IF CTOT IS LARGER, MAKE SUBSTITIUTIONS.
C
12n TOP10 s NHR
-------
SUBROUTINE CONFREQ
C
C»«#»»»»#*THIS SUBROUTINE ACCUMULATES FRQJENCIES OF TOTAL CALCULATED
C»»»«»«»**« CLASSBaCLASS 30UNDARIES OF FREQUENCY DISTRIBUTION (H9/M3).
C»##«»#»**# XXHRsAVERAGINS INTERVAL (MEAMHR).
C
COMMON /CLS/ CLASSB(IO)
COMMON /DAT/ ALPHA(16) ,MEANHR,CLOSTY»,NCLOSE
COMMON /TOP/ NN*CTOT»CLOSC»FARCtTOP10(10> fNHR(lO) tTOPRAT(lO) »
DATA (CLASS8aO.O»0.5*1.0»2tO»4.0i8.0tl2.0»16.0»32.0t6*»0)
XXHR = MEANHR
COFREQh)*COFREO(l 1*1.0
C
C»*»»»»«»»#CnNVERTING AVERA3E IN MS/M3 TO JNITS COMPARABLE TO SUMS.
C»»*»»»»*»»ACCUMULATIN3 FREaUENCIES.
C
DO 110 !=2tlO
CC = Cl'ASS8»XXHR«l,OE-3
IF (CToT .GT. co GO TO 100
RETURN
100 COFREQ(I)sCOFREQ(I)*1.0
11 (i CONTINUE
RETURN
END
101
-------
SUBROUTINE CONPNT
DISTRIBUTION OF"
INFORMATION.
ALCULATFD
C»#»*»«»*»THIS SUBROUTINE POINTS FREQUENCY
C»» .
i
?
3
4
5'
6
T
9
COMMOM
FORMAT
FORMAT
FORMAT
FORMAT
FORMAT
FORMAT
FORMAT
FORMAT
/DAT/
(1H1,
(/iH
(/.IH
!#PERCENT **iOFio.2)
C»#»**»»»*»PRINTINR HEADINGS.
c
PRTMTl. ALPHA
PRINT ?»MEANMR
IF (CLOSTYP.EQ.CJHSTREET) GO TO 100
PRINT 3 * GO TO 110
100 PRINT 4
110 PRINT =1
C
C»»»»»»»*»»PRINTIMR NUMERICAL FREQUENCIES.
c
PRINT MCLASSB
PRINT 7.COFREQ
C
C»*«*»#»*#»NORMALI7ING TO PERCENTAGES.
C
PERCENT=0.01*COFREQ(1>
00 120 1=1.10
COFREQ'f I ) sCOFREO ( I ) /PERCENT
120 CONTINUE
C
C»#»»»«»«»#PRINTIMG PERCENTAGE FRESUENCIES.
c
PRINT R.COFREQ
RETURN
END
(MS/M3)
*)
102
-------
SUBROUTINE RATFREQ
C»«»»««*»««THIS SUBPROGRAM ACCUMULATES THE* FREQUENCY DISTRIBUTION or THE
C»«»»**»**»RATIOS OF THE NEAR/FAR SOURCES (RATIO).
c
COMMON /RAT/ RATlOtFREQ(lO) .VAlUEUQ)
C»«**#»»#«#VaLUE=CLASS BOUNDARIES FOR CUMJLATIVE FREQUENCY 015TRI9JTlON
C»«##**»»«»(NUMBER GREATER THAN...)
C
DATA (yALUE«0.0»O.OH»0.(H»0.08t0.2iO.*t0.8»2.0»4.0«8.0)
FREQ
-------
SUBROUTINE RATPNT
C»»»*»»»»*»THIS SURPROSRAM PRINTS CUMULATIVE FREQUENCY DISTRIBUTIONS FOR
C»**#««*»«*RATIOS OF NEAR/FAR CONTRIBUTION' TO CONCENTRATION.
c
COMMON /OAT/ ALPHA(16)»MEANHR,CLOSTY».NCLOSE
COMMON /RAT/ RATIO.FREQ(10).VALUE(101
i FORMAT dHi,i5xi&A5)
9 FORMAT (/1H ,#RATIOS 3ASED ON»I3»»-HOUR AVERAGE*)
3 FORMAT »/IH ,»CLOSE * CONTRIBUTION FROM FIRST »n*» SEGMENTS*)
i FORMAT (/1H ,*CLOSE » STREET CONTRl3UTIONS*)
5' FORMAT (/1H ,10X»RATIO' OF CLOSE/FAR CONTRIBUTIONS GREATER THAN**.*
1)
*, FORMAT (/IN ,9X10F10.3)
7 FORMAT
-------
SUBROUTINE STREET
C
C»»»»»»»»«»THIS SURPROSRAM CALCULATES STREET CONTRIBUTION TO CONCENT^A.
C»»#»»»»*«»TTON FOR UNIT WIND SPEED....
c»#»*»»»### QsTsEMissiO'i RATE ON STREET (SM/M*S).
c»«»»»»**#* IOIR»WTND DIRECTION <16 POINT).
c*»«««*««*» STOIR»STREET DIRECTION JDESREES cw FROM NORTH),
C»***»««*»* SIDE (LEFT OR RIGHT). SIDE: OF RECEPTOR WHEN FACINS
C»»»*»*«#*« W,H=WIDTH AND HEIGHT OF STREET CANYON (M) .
C»»»*»»«*#» Z,X=HEIGHT AMD HORIZONTAL DISTANCE OF RECEPTOR' FRO>M
C#««»*#«*#» FIC.
C»»»»»*»*»» COS60*COSINEK60 DESREES).
C
COMMON /VAR/ XNORM.QST»SIDE»IOIR»STDIR»^.H»Z»X
DATA CnS60 /0.5/
C
C»#»»»#*##*ANSLE NORMAL TO STDIR,
c
TF (SIDE.NE.4HLEFT) GO' TO 100
GO TO T?0
inn IF (5lnE.EQ.5HRl3HT) 30 TO 110
PRINT 1
-------
-------
Appendix B
SOURCES OF TRAFFIC INFORMATION
107
-------
Appendix B
SOURCES OF TRAFFIC INFORMATION
Traffic information is generally available in urban areas where it
is collected for use in traffic control. In small towns and outlying
areas, data may not always be available, but a check with the local
highway, transportation, or public works offices would reveal any
special counts that may have been taken. State and federal highway
counts are maintained and updated regularly. The U.S. Department of
Transportation, Federal Highway Administration, publishes a directory,
"Urbanized Area Transportation Planning Programs," that lists traffic
agencies throughout the country, as well as key officials of these
agencies. The directory is divided into four sections:
• Section I: Lists the metropolitan area transportation
planning programs.
• Section II: Lists the Federal Highway Administration
regional planning engineers.
• Section III: Lists the Federal Highway Administration
Division planning and research engineers.
. Section IV: Lists the State Highway Department planning
engineers.
Copies may be obtained from the Federal Highway Administration, 400
Seventh Street, S. W., Washington, D.C., 20590.
Traffic information may appear in many forms. Examples are shown
in Table B-l and Figures B-l through B-5. Table B-l gives traffic
volumes at different locations along a state (California) highway. The
traffic during peak hours, the average daily traffic (ADT) during the
peak month, and for the whole year are all given. The highway segments
are identified by name and mile-post values.
Figure B-l shows a street map of Washington, D.C., drawn so that
the street widths are proportional to traffic volumes. Figure B-2 is an
example of the same graphic technique applied over a larger area
(northwestern Virginia). Figure B-3 illustrates another, simpler method
of designating traffic volumes along a road network.
Figure B-4 shows traffic volumes that are presented in still
another way. In this instance, the average daily traffic on all of the
streets in each of the squares have been used to determine the average
total number of vehicle miles traveled on surface streets each day, in
108
-------
Table B~l
EXAMPLE OF TABULATED TRAFFIC VOLUMES
Rte
Mile-
post
101, SM Co
Description
Peak
Hour
1971 TRAFFIC
ADT
Pk. Mo. Annual
VOLUMES
Mile-
post Description
Rte 101,
Peak
Hour Pk.
Mrn
ADT
Co
Mo. Annual
539 Redwood City, Jet Rte 114,
Chestnut-Spruce Street
Interchange
662 Redwood City, Whipple Ave-
nue Interchange
840 Holly Street Interchange .
9,55 Belmont, Ralston Avenue
Interchange
11 15 San Mateo, Hillsdale Boule-
vard Interchange
1190 San Mateo, Jet Rte 92,
19th Avenue Interchange
12 69 San Mateo, Kehoe Avenue
Connection
13 46 San Mateo, Third Avenue
Interchange
14 33 San Mateo, Poplar-Dore Ave-
nue Connections
14 69 San Mateo, Peninsular Ave-
nue Interchange
16 58 Burhngame, Broadway Inter-
change
17 95 Millbrae, Millbrae Avenue
Interchange
1912 San Francisco Airport
Interchange
9,700 110,000 102,000
10,500 119,000 111,000
11,200 127,000 118,000
11,600 132,000 122,000
12,100 137,000 127,000
12,800 146,000 135,000
12,800 146,000 135,000
13,100 149,000 138,000
12,700 144,000 134,000
12,700 144,000 134,000
12,600 143,000 133,000
13,000 147,000 137,000
13,400 153,000 141,000
20 40 San Bruno Avenue Inter-
change
2145 South San Francisco, South 11,300 129,000 119,000
Airport Boulevard Inter-
change
21 92 South San Francisco, Grand
Avenue Interchange .
22 55 South San Francisco, Linden
Avenue Connection
11,100 126,000 117,000
11,700 132,000 123,000
2271 South San Francisco, Oyster 11,000 125,000 116,000
Point Boulevard Inter-
change
2339 South San Francisco, Old
Bayshore Connection
(Sierra Point)
11,400 129,000 120,000
23 39 South San Francisco, Old
Bayshort? Connection
(Sierra Point)
25 70 Brisbane, Candlestick Park
Connections
26 1J San Mateo County (Brisbane
North City Limits)
= 000 San Francisco County (San
Francisco South City
Limits)
0 77 San Francisco, Third Street
Interchange
111 San Francisco, Paul Avenue
Connection
1 77 San Francisco, Silver Ave-
nue Interchange
1 9S San Francisco, Jet Hte
280, Alemany Boulevard
Interchange
2 92 San Francisco, Army Street
Interchange
4 10 San Francisco, Vermont
Street Connection
R4 24 San Francisco, Jet Rte 80
Eastbound Lanes
R5 14 San Francisco, Mission
Street-South Van Ness
Avenue Connections
R5 47 San Francisco, Market
= 454 Street
4 76 San Francisco, Oak-Fell
Street Connections
5 09 End Freeway
5 17 San Francisco, Turk Street-
Golden Gate Avenue
Connections End Freeway
5 35 San Francisco, Turk Street
5 94 San Francisco, California
Street
671 San Francisco, Jet Rte
480, Lombard Street
(Break in Route)
Marin County
0.00 North End Golden Gate
Bridge Begin Freeway
11,200 127,000 118,000
11,200 127,000 118,000
11,200 128,000 118,000
11,300 128,000 119,000
12,100 137,000 127,000
15,000 199,000 185,000
15,000 206,000 191,000
15,000 201,000 186,000
10,000 120,000 111,000
8,100 97,000 90,000
8,500 96,000 89,000
4,550 52,000 48,000
4,700 53,000 49,500
5,000 62,000 53,000
4,450 55,000 47,000
4,100 50,000 43,000
4,100 50,000 43,000
147570 441
Source: California Division of Highways
109
-------
•••• PROJECT BOUNDARY
SA-3515-18
FIGURE B-1 SAMPLE TRAFFIC MAP FOR A DOWNTOWN AREA
110
-------
COMMONWEALTH OF VIRGINIA
1968
TRAFFIC FLOW MAP
MTEMSWE, ARTERIAL AND PRMARY ROUTES
MMML «WA6Ea4-«X« TRAFFIC-YEAR ENDMSOtt. 31. »6«
VMWU DEPARTMENT OF HGHVMVS
Oinai V irtrnc MO
USOEnWTMCNT OF
rOOML WJNWtT ADMINISTRATION
4UMC4U OF IJflLIC MOADS
TRAFFIC SCALE
o 10,000 20,000
> " STHJ^SSB^&JJ&L'
l°\2rjf jS$V~"
fa* \J^
-------
TRAFFIC VOLUME MAP
COLORADO STATE HIGHWAY SYSTEM
MICH TIH PAVCO ROAD
LOW TYPE WTUMINOUJ IURFACI
gRAVCL IURFACC
CftAOtO AND ORAINtO . .
DELTA-MONTROSE AREA
SCALE OF MILEl
I 1 I 5
5TATE OF COLOHAK)
PL'«M»« II.D MI[*«H DWISIO»
SA-3515-20
FIGURE B-3 SAMPLE TRAFFIC MAP FOR AN INTERCITY AREA, COLORADO
112
-------
IT)
n
<
i/)
<
Q
O
Q
LU
O
Q
<
C/5
CO
LLJ
EC
113
-------
that 2-mile square. This type of traffic data is not as common as the
others, but such inventories are known to have been made for at least
three areas in California: the Los Angeles Basin (Roberts et al, 1971),
the San Francisco Bay Area (Ludwig and Kealoha, 1974), and the example
shown in Figure B-4, Ventura County (Ludwig et al, 1975).
In the late-1960s, a series of documents was prepared for many U.S.
urban areas by the U.S. Department of Health, Education, and Welfare.
The generic title of these documents is:
"Report for Consultation on the Metropolitan. . . Air Quality
Control Region"
The missing words in the title give the name of the appropriate
intrastate or interstate metropolitan area. Some of these documents
contained gridded estimates of daily average CO emission rates. Since
most CO is emitted from vehicular sources, these inventories will be
closely related to traffic, and could be used as data for such
estimations, if no better information were available.
114
-------
Appendix C
SOURCES OF CLIMATOLOGICAL AND METEOROLOGICAL INFORMATION
115
-------
Appendix C
SOURCES OF CLIMATOLOGICAL AND METEOROLOGICAL INFORMATION
One of the most helpful publications specifically designed to
assist potential users of climatological data is called "Selective Guide
to Climatic Data Sources," Key to Meteorological Records Documentation
Number 4.11, prepared by the staff at the National Climatic Center,
Ashville, N.C., for sale by the Government Printing Office, Washington,
D.C. Its format indicates the publication(s) in which various clima-
tological categories (temperature, precipitation, wind, humidity, and so
on) may be found. Although this publication refers primarily to
published climatological data, a wealth of unpublished climatological
summaries are also available on special order from the files of the
National Climatic Center An index to the summaries that can be ordered
is given in the "Guide to Standard Weather Summaries," NAVAIR 50-IC-534,
U.S. Navy, March 1968.
The National Climatic Center makes every effort to obtain a copy of
all meteorological records collected in the United States. These data
are available and can be ordered on microfilm, magnetic tape, hard
copies, or as copies of raw data. The address and phone number are:
Director, National Climatic Center
Federal Building
Ashville, North Carolina 28801
Telephone: (704) 258-2850
The Center answers inquiries and analyzes, evaluates, and
interprets data. Routine letters or telephone inquires are usually
answered without charge; other services are provided at cost.
The bulk of the data at the Climatic Center are meteorological
observations made at airfields by the National Weather Service, the
Federal Aviation Administration, and the Defense Department. Figure C-l
shows an example of the kind of information to be found on a three-
hourly tabulation for one month at one station. Climatic information is
seldom available to the extent that it will be desired, but ingenuity
can often be used to ferret out sources not generally available from the
usual public data repositories.
At the State and regional level, fire stations, highway and
transportation departments, environmental studies groups, air pollution
districts, and utility districts may have continuing meteorological
records or special weather studies available. A call directly to these
agencies may result in a data source not available anywhere else.
116
-------
OBSERVATIONS AT 3-HOUR INTERVALS
DAY 01
10 40 7 51 24 85l 21
81*7! & 26| 72, 20(
10 1 1 2J SM 7l 2b\ 88 24|
OAT 04
0 UNL ICj 8> 17 77 12,
0 UNL 10 Ol 19* 84 16'
Z UNL 10J 2 20) 74 15
10 60 7l d 28, 82 25l
10 30 TI 5* I1 30( 69) 28
10 20 f S 331 32 85! 29
10 26 r1 S 33> 32' 891 30l
DAT 07
a U\L la 20 it 74 u
0 UNL 8) IS 1* 84 11
1 UNL r 12: U 84, 08
1 UNL Z 8, K 22 20' 74 15
4 UNL 7| ,31 26| 56[ I7l
0 UNL 7] 251 23' 691 16
DAY 10
10 31 11 LF IT 37HOO 37
IQ 121 2! f 39) 38 89' 36
IQ 21 dl2 F 37^ 37100 }7
10 2S 51 OF 37l 36l 89 34
10 T 1, 8 5 2«* 24 92 22
10 71 *| S 24, 231 85 20
10) UNL 10 20< 18' 74, 13
2 UHL, 12. 21' 1 91 63' 1 1
1 UNLi 12 , 22J 19 57' 09i
4 UNLj 12 21 18 62 101
U 40i IQ IT 15 68 08
DAY 16
1Q *ft 71 Rw 391 36 76 32
a t. IP BI «F 4^ *i 93 40
DAY 19
0 UNL 71 32 31 85 28
101 Clff S ' * 54i 4»| 72 45
Iq CIR 6| K 32l 47! 72 43
DAY 22
10 4. 11 8 F 35 34, 92, 33
id 15. OJ 8 RSF 3*1 33' 92 32
101 * 1 8 SF 33| 33100 33
10; f S SF 32l 32I1001 }2
Id 4 ; 8 Z8F 32> 32 9A 31
iQJ 2' a|l2| ZLF sz| 32 96 Ji
DAY 29
0 UNLi 10| 291 241 01 20
oi UNL[ r 21 21 92 19
01 UNL 7 21 21, 92 19
L' UNL 10 35| 30 94 20
1 UNL 15 44{ 33 38 20
(j UNL T) 1 39| 35 65' 20
DAY 2B
2i UNLi 01 53) 47| 661 42
« CI«| 7 52 47 72 *3
8 C1H( T1 38 49 91 *0
0j UNLi 7 63 921 46 42
3 UNd 7 59 90| 94 42
IQ CIR 7 94 49 69 44
iq CIR) TJ 92 «J 37] 37
DAY 31
0 UNL 7] ** J6J 47 29
0 UNLl B 43 36 49 2»
4 CIR 19 31 40 30 26
01 UNL 6 71 5B 44 *0
1C 30 7 66 59 «9 3*
1C S 3 RMOF 59 37 90 56
DAT 02
25 15 0 UNL 7
DAY
241 12 i UNL 7'
2*1 <• 10 *5 7' S
24i 0 10 16 5 &
24 9 10 14, 10
24,1 1 10] 50! B
26, 3 10, 40, 10
27 4 10 100 8'
DAY
32| 6 0 UNL 7,'
15 8 UNL 7
19 7 UNL 8
2J' 8 UNL,1 T
23, 9 10' 100, 7
23 13 10; 21 8,
23 12 10 16 10 '
24) 0 29 15'
23 2 UNL 15
32 I UNL 10
DAY
03 8 1 40 8'
36 5 UNL, 8
32' 7 , iJNL 7 ,
32 1 U«L 7'
3l' 9 UNL' 10'
29 9 1 UNL 10'
26 6 1 ' 80 8
181 9 10' 3 Ij 4| F
26, 7 10 0' Oi 1| F
*'"n Ih
21 201 8*' 17! 32 19 10
' 14 11 53 00 31 13 2
09
32 31 09 29 26 1 10
31 30, e 28 27 10
1 30 « 28 24 5
2 31 8 28 03, 1
3 31 ij 28 02 10 2
1 Z9 8 ! 26 05 11 0
, 71 26 8 24 07 15 0
08
26 2* 69 17 23 I 10
31 29 B2 2
33 30 70 2
*' 03, 1 10
1*
20 18 74 13 26' 1 10
18 16 71 10 26 10
15 14 7T 09 25 1 0
22, 20 71 14 22 3 0
27 23 i5 13 24 6 0
30 2h' St. 16 24 0
27 23 55 13 24. 0
28 25' 61 lb 15' 0
33 33100 3
, 3«| 37 93; 3
1 34, 341100 3
C*i 20
13 4 10 100 8 51' *5 64 3
07 7 10' 0 0 4, F 4* 43 96 4
16 12 10' 22 * &F , 44 43 93 4
DAT 23
27 7 10, 3 1 8 LF 3* 33, 9Z, 3
23 3 10 20 2 RF 35 34' 2! 3
01 4 10 76
-------
Schools, colleges, industrial complexes (such as oil refineries),
agricultural research stations, radio-TV stations, and electrical power
plants may cooperate with a data collection program if asked.
The following publications provide important information concerning
useful data sources.
1. Air Pollution Control Association (1973-1974): "Directory,
Governmental Air Pollution Agencies", published in cooperation with the
Office of Air Programs, EPA. This directory lists federal, state,
regional, and county agencies conducting air pollution programs. Names
of officials, titles, addresses, and telephone numbers are given. Write
to W. T. Beery, Editor, Directory Governmental Air Pollution Agencies,
Air Pollution Control Association, 4400 Fifth Avenue, Pittsburg, PA
15213.
2. World Weather Records, Smithsonian Misc. Collections, Vol. 79,
Publication 2913, Assembled and arranged for publication by H. H.
Clayton, published by the Smithsonian Institution, August 1927. This
reference book contains monthly and annual means of pressure,
temperature, and totals of rainfall.
A more extensive collection consisting of climatological data for
selected airfields and for the climatic areas in which they are located
has been compiled by the USAF Environmental Technical Applications
Center (ETAC), Building 159, Navy Yard Annex, Washington, D.C. 20333.
This series is also being published by the U.S. Naval Weather Service,
Navy Yard, Washington, D.C. 20390, under the title "U.S. Naval Weather
Service World-Wide Airfield Summaries." Table C-l lists the available
volumes in this series. Volume VI11 contains summaries for the United
States. Information requests should be addressed to:
The National Technical Information Service (NTIS)
Springfield, Virginia 22151.
3. "The Climatic Atlas of the United States," (1968) is a
comprehensive series of monthly and annual analyses showing the national
distribution of mean, normal and/or extreme values of temperature,
precipitation, wind, pressure, relative humidity, dewpoint, sunshine,
sky cover, heating degree days, solar adiation and evaporation. It was
prepared by the Environmental Data Service, NOAA, U.S. Department of
Commerce, for sale by the Superintendent of Documents, Washington, D.C.
4. "Mixing Heights, Wind Speeds, and Potential for Urban Air
Pollution Throughout the Contiguous United States" by George C.
Holzworth illustrates seasonal and annual, morning and afternoon mean
mixing heights, wind speeds, and normalized pollutant concentrations
that were exceeded 10, 25, and 50 percent of the time for specified city
sizes. Copies of this report (Office of Air Programs Publication No. AP-
101) may be ordered from the Office of Technical Information and
Publications, Office of Air Programs, Environmental Protection Agency,
Research Triangle Park, North Carolina 27711.
118
-------
t-
O)
w
CD
.Q
I I I I I
NO CM CM C\J CM
O NO NO NO NO
[•—NO NO NO NO
I I I I I
O Q P Q p
O\
CM
OO
I
J-
NO
I
p
«:
00 U\
CM _^T
00 CO
I I
-3 <~i
NO C^-
NO NO
p «
CM CM H H r— o\^O r-
[>-O\ONI>-rHrHOO
~
_
I
-
I I
I
I I
i--
coooo\ONO\OOO
vOvO^OvO^O c^ c~- : —
I I I I I I I I
H
ON ON
I I
CM CM
co oo
MD xO
I I
Q Q
•a! <
o.
p.
W
fn
o
0)
&
CS
i
fn
O
W
w
•a
0)
CO
•rl
5
PS1
cd
•H
CO
«*! +*
CO
CO (3 -p
cO co
d) CD cO
rj i — ^ fjQ
-P 73
pi 73 £H
O -rl cd
CO S PH
CM-
H
5
h
CO
PH
M
M M
H M M
Q)Q)Q)
H H H
O O O
eenland -Iceland
-Antarctica
g So. Pacific Is. )
rica
merica
ates of America
t, Western Mtns . and Great Basin
s. and Northwest Basin
O -rl vH 6 CO CO -P
1 H T) -a! M O g
co co pi co T! o
73 f~i rH ^J £-* CD t*j
cd -P o -P -p -P -P x
P^ CO C 0 Cj -H CO O
CO £3 «rl O CD d CD O
0 <— CO 0 PD S «
CM
rH
H CM
CO
-P -P -P
(H ^ In
cd cd cd
r> j p t A j
- — M
M M
> M M M
M > > > >
CD Q) Q) CD 0)
H H H H H
O O O O O
> > > > t>
la ins
es
pi Valley
ern Region
t and Appalachian Region
d Hawaii
Half
P-. X P, -P CO C
CO -H CO CO CO C
H 1-4 CO CO O S
CO CO CD O CO CO CD
fi -P iH ,JC .id O ,r!
P (0 CO-P-P CO-rlT'
CCDCOplCOcO^H^H
CD ^H -rl O cti r — 1 CM O
oo^cow^^s
<^\-d"lAv*O ?~~ co H
-p .p 43 43 .p 43 4J
Jn £n ^i (H ^-t ^ £-1
(Tj CTJ CO Oj Cti Oj CO
pi j p t p|_. A . n . p . p .
,.
R
u
r-!
O
t>
Half
ia & Northern Europe
ries & British Isles
uthwest Europe
> P O
C cd C co
L, a 3
CD 0) -H O c3
jc a-d o
T3 o C co
pj h co S ft
O J3 o O H
co W CO i-3 «<
CM H CM ro
-P -P -P -P
^ fn !-, C-,
cd cO co cO
PL, PL, PL, PH
X
0)
H
O
c>
C
CD
C
cd
CD
.p
•H
73
0)
-3
-p
cd
a
H
g
O
CO
119
-------
The National Climatic Center will prepare special data summaries.
They also,.have standard computer programs available for special
summaries. One of the most useful summaries for air pollution studies
is that prepared by the STAR program. It is a joint frequency
distribution of atmospheric stability and wind speed aad direction. The
atmospheric stability is calculated objectively from the cloud cover
and wind data. This stability algorithm is based on the work of
Pasquill (1961). The summaries can be based on any extended period of
record with separate outputs for the months or seasons, as well as an
annual summary. There are some pollution models that use the output of
STAR program as part of their input requirements.
120
-------
Appendix D
SOURCES OF LAND USE INFORMATION
121
-------
Appendix D
SOURCES OF LAND USE INFORMATION
The extent and availability of land use data is dependent on the
specific area under study and what one chooses to call "land use infor-
mation." The more formal information can be obtained at different
levels of government. Some states have developed sizable data bases as
an aid in generalized state planning (i.e., Connecticut, Florida,
Hawaii, Maine, Vermont). The majority of states, however, are just
beginning the information gathering process. Land use information for
nonurban areas is best obtained from State Planning and regional
agencies.
Regional planning agencies (e.g., the Denver Council of Governments
in Colorado, Southeastern Virginia, Planning District Commission, and
the Comprehensive Planning Organization in San Diego) can be excellent
sources of information. These regional agencies gather socioeconomic,
existing land use, and transportation data. Comprehensive regional
plans can then be prepared to provide projections of long range
demographic growth and land use.
Cities and counties will usually have current, readily available
data on population, employment, existing and projected land use, general
development plans and zoning regulations. Also, they will be able to
provide basic transportation information and maps of major arterials and
proposed thoroughfares.
In cities with schools of urban and regional planning, planning
professors can help the researcher meet specific needs. Also, their
libraries can be researched for applicable graduate and doctoral theses
which are frequently case studies of the immediate vicinity.
There are other sources of land use information that are not
specifically directed to the topic. Good maps or aerial photographs can
provide a lot of useful information that may not be available from
conventional land use sources. Useful sources of information for the
United States are discussed next.
A. U.S. Bureau of the Census
Demographic and socioeconomic information of use to planners is
available from the Department of the Census. Data developed by census
tracts can be used to answer questions regarding a neighborhood's
population and characteristics. Census tract information can be
122
-------
outdated, so it should be supplemented by material developed by the
city, county, or regional planning bodies.
B. United States Geological Survey (USGS)
1. Topographic Maps
Topographic maps portray man-made and natural features, and
the shape and elevation of the terrain. The usefulness of topographic
maps is revealed in their accuracy, availability, economy, and wealth of
detail. All maps are classified according to scale. The map scale
expresses a ratio between the features shown on the map and the same
features on the earth's surface. A scale of 1:24,000 states that one
unit on the map represents 24,000 units on the ground. Figure D-l is an
example of three map scales of the same area showing the type of
information that is available in large, medium, and small scale maos.
Table D-l is a summary of the principal maps and their essential
characteristics. A booklet describing topographic maps and symbols is
available free upon request from the Geological Survey. To order maps
of a specific area, first obtain a state index map by asking for the
"Index to Topographic Maps of (state)." An order form is included with
each index as well as a list of local merchants that may stock
topographic maps. Map reference facilities are also maintained in many
public libraries. All maps for areas west of the Mississippi may be
purchased by mail or over the counter from: Distribution Section, U.S.
Geological Survey, Federal Center, Denver, Colorado, 80225; and for
areas east of the Mississippi: distribution Section USGS, 1200 S. Eads
Street, Arlington, Virginia 33303.
2. Photoimage Maps
Photoimage maps are available in the 1:24,000 scale. These
are a new standard product called the orthophotoquad maps. An
orthophotoquad portrays by aerial photoimagery a wealth of detail not
found in conventional line maps. Yet there is the same positional
accuracy as in standard topographic maps. Orthophotoquads are
reproduced in black and white as photographic, diazo, or lithograpic
copies. Diazo or lithographic products are comparable in price with 7.5
minute topographic maps. To obtain an index of orthophotoquad
availability, contact the:
National Cartographic Information Center
U. S. Geological Survey
National Center, Stop 507
Reston, Virginia 22092
(703) 860-6045
Figure D-2 shows a portion of the orthophotoquad index,
legend, and a coded portion of the state of Florida (USGS, 1974).
123
-------
C/2
^
§
C^J
HH
OH
ry2
MM
^ o
« £
" £
*
H (-5
^^
^—<
o
^M
— i co c>j
— _ M TT
CTJ
cq *-i on- -.
o 83 83 o 2 S
^ O O "^ ^ r,f
0 *J *-> 0 CO g
2 ^ a. o ^ co
•^ C^ • — 1 CM ^ C—
q
'i
C C .50
"i 1 1 p
^-1 ^
r r. 10 o => o
I> I> ^H (N CM «O
>< X x X X X
— ' — v_O lO ° °
^ t, ^ " r1 ^
Q; K
0) cfi 1J
^ £ 'e 'i E
, . '^-.
S 3 ~ " T! Tl
X O a P a! cS
°. _c aj ^ m a)
CM aS C >— i C C
s
§0000°.
O O JO O O
^r o' c>J co* vo ^-i
2
c
1
<^ o o
J_ to o °
o co" o o'
QJ 5J ^_Q ^- c^
2 3 "^ Cvl ,-H*
.E C c as ^ ^
6 tJ 'S "^ . .
i, a; 7 03 &o co
~ ^ »O ~~^ i — ^ i — '
y>
a>
o
c
r\l
X
_j-
— '
^H
nj
J3
x
D
-«_>
3
C
E
12
^
u
js
o
c
t-
CM
X
M
cS
•f.
O
^-
-2
c
1
r-
^_,
CO
3
^
• — '
^^
O
A
-*j
3
CO
s
_C
0
c
OJ
X
01
a
-r.
1
"/;
ai
!/J >—^
r^ I"fi
c5 ^
T^ *
ri
'& * £
-f ^ 1
C 1J C
^-.•^O-I
°^ £
X| X
',-*„
^ ^ ^
-g 'S tJ
•5 ^ "5
i-vi ^3 Cvl
CO C "*
T. C^ ^j
3 ^ ^
\£ _^ -^
-2
-------
J
1
* 2
"0 m
i
8
to
I
-
r- ,- < ,-
7
in
in
n
LU
CO
CO
Q
Q
LU
5
LU"
cc
£
N «
S .S £ «
^ r- < «£)
UJ
_l
<
Q
UJ
LL
O
111
O
CJ
tr
O
LL
LU
h- Q.
.. <
£
£
»>
^ - < ,-
en 0>
w 10
UJ *
a: w
< o
-I CO
D
UJ
u
DC
D
O
tn
s <
< a:
X O
LU o
Q
LU
OC
125
-------
-ORTHOPHOTOQUAD INDEX
December 1974
OF ORTHOPHOTOQUADS
Monocolor orthographic photograph or
photomosaic in quadrangle format with
minimal cartographic enhancement.
Planned
Aerial Photography
Advance copy
Note- No product or copy available.
High Altitude, mostly quad centered.
Final copv for approximately 90 percent
of orthophotoquads. {for ordenn
f
material index, available from appropriate
Mapping Center as indicated below or from
the National Cartographic Information
Center, L'. S. Geological Survev (507),
Reston, Virginia 22092.)
(For ordering instructions refer to state
index to topographic maps, available from
the Branch of Distribution, U. S. Geological
Survev, 1200 S Eads Street, Arlington,
Virginia 22202 or Federal Center, Denver,
Colorado 80225.)
SA-3515-22
FIGURE D-2 A PORTION OF THE ORTHOPHOTOQUAD INDEX SHOWING THE LEGEND
AND A PORTION OF THE STATE OF FLORIDA
126
-------
3. Earth Resources Technology Satellite
The Earth Resources Technology Satellite (ERTS) has the multi-
spectral sensors on board that "photograph" the earth's surface in the
visible through near-infrared range. The potential of such a capability
for land use mapping, updating, and projection is currently a subject of
extensive study. The images received from 'the satellite are available
for sale as individual frames each covering an area approximately 1000 x
100 nautical miles with a 10 percent overlap along the spacecraft track.
Table D-2
PICTURE PRODUCTS AVAILABLE FROM ERTS
Image Size Scale Material
70mm (contact size) 1:3,369,000 Resin coated paper,
film positive or negative
7 1/2" x 7 1/2"
15" x 15"
30" x 30"
1:1,000,000
1:500,000
1:250,000
Resin coated paper
Resin coated paper
Resin coated paper
Source: EDCDM Form 6
For more information, contact the ERTS Data Center, Sioux
Falls, South Dakota, telephone: (605) 339-2270. The ERTS Data Center
has substantial holdings of images acquired by aircraft throughout the
United States. They invite inquiry regarding availability of suitable
coverage of your area of interest (USGSEDC).
4. Sanborn Maps
Sanborn maps are land use maps prepared by the Sanborn Map Company.
They are plotted to a scale of 1 in. to either 50 or 100 ft showing
details such as streets, railroad tracks, lot lines, building dimen-
sions, nature of the building material, number of stories, height of the
building, and use of the building. Sanborn maps are used primarily for
fire insurance purposes. Local sources may be fire insurance offices,
realtors offices, city planning, and the county courthouse (Murphy,
1966). For information, contact:
The Sanborn Map company, Inc.
629 5th Avenue
Pelham, N. Y. 10803
Mr. G. Greeley Wells
Telephone: (914) 738-1649
Figure D-3 is an example of a Sanborn map for a section of Portland,
Oregon.
127
-------
53
'/« SfC 29ZS t 3028
44
IOD
N W GLISAN
2
a
(Oil
(*>d>
102
N W EVERETT
I
N W HOYT
99
!—•* 1 N
- tj;- I-, ^
^ *ift
Pi*.-
i
: G
-ff.ia..-% ..j.
' 918
54
" ""© i
;?.'?.'. N w FLANDERS-*"'"--*—r
„ fc.'i::
JK/»t/iMIK CO j i
4 £
n
^^-^
FIGURE D-3 A SANBORN MAP FOR A SECTION OF PORTLAND, OREGON
128
-------
Appendix E
BIBLIOGRAPHY
129
-------
Appendix E
BIBLIOGRAPHY
A literature review has provided a collection of papers and reports
on topics related to the measurement and distribution of CO concentra-
tions. A number of foreign translations were provided by the Contract
Monitor and are included in the bibliography. The bibliography is
arranged alphabetically by author and numbered consecutively. Each item
in the bibliography has been classified into the categories shown in
Table 2 (see Section II of the text). The resulting matrix is given in
Table E-l. The bibliography item number is entered in all the
applicable categories. Some of the more comprehensive reports may be
tallied in several categories. Table E-l can be easily used to
determine which of the reports in the bibliography are related to the
various sampling purposes and scales of measurement.
130
-------
Table E-l
BIBLIOGRAPHY INDEX, ARRANGED ACCORDING TO MEASUREMENT PURPOSES AND SCALES
(Numbers in Columns Refer to Numbers by Which Entries Are
Cataloged in the Bibliography)
Purpose for Which Data
Are to Be Used
1. Determine compliance with
ambient air quality
standards
2. Alert authorities to exist-
ing or impending critical
si tuations
3. Evaluate results of control
measures
4. Determine long-term trends,
urban and rural
5. Provide information for de-
veloping, evaluating, and
refining air pollution
models
6. Provide information for com-
parisons among locations of
the same general class,
street canyons, highways,
neighborhoods, rural areas
7. Serve as data base for city
and regional planners and
decision makers
8. Serve as a data base for en-
vironmental impact state-
ments, highway projects,
transportation plans, large
developments, indirect
sources
9. Provide measures of the
magnitude of sources and
sinks; anthropogenic,
biological
10. Provide measures of indoor/
outdoor concentration re-
lationships
Applicable Scale of Measurement
Middle
Downtown, Street
Canyon
36 69
41 72
57 111
4 103
45 107
73 113
86
36
98
113
126
131
18 64
26 75
49 77
50 99
55 116
7 41 86 107
21 44 96 119
32 54 97 112
34 59 98 128
36 66 100 133
37 72 101 134
39 83 103 135
40 85 106
96
32
38
44
130
Indirect
Source
43 111
57 136
69 138
72
43 103
45 113
73
98
113
126
49
50
64
67
99
20 98
21 100
44 101
59 103
66 106
85 133
96 134
67
96
105
21
22
104
115
130
Corridor
28 70
57 71
58 138
69
4
58
113
58
62
113
126
48
18 50 76
24 62 116
25 64 129
27 67
31 68
5 59
20 66
21 85
28 96
44 100
48 101
58 106
28 62
33 67
58 96
76
Neighborhood
1 69 82
28 70 96
43 71 136
57 81 138
4
45
113
102
117
126
1 102
17 117
48
66
82
20 68
50 99
56 111
61 116
67 129
5 59 101
20 66 102
21 81 106
28 84 125
43 85
44 96
48 100
2 63 114
16 66 117
17 67 120
19 79 123
28 80 132
33 96 137
43 105
87 92
88 93
89 94
90 95
91
51
52
53
Regional
28 69' 124
29 70 136
30 71 138
36 81
58 96
58
118
36
58
117
17 66
29 102
30 117
47 125
48
20 68
50 78
60 112
61 125
67
3 44 106
5 48 125
20 74
28 78
35 81
36 85
43 101
2 33 114
16 58 117
17 66 120
19 67 123
28 79 125
29 80 132
30 106 137
87 92
88 93
89 94
90 95
91
51 108
52 109
53
131
-------
BIBLIOGRAPHY
1 Akland, G. G., "Design of Sampling Schedules," JAPCA, Vol. 22, No. 4,
April 1972, pp. 264-266
2 Alan M. Voorhees and Asso., Inc., and Ryckman, Edgerley, Tomlinson and
Asso., "A Guide to Reducing Air Pollution Through Urban Planning,"
for the Env. Protection Agency, APTD-093, December 1971, 114 p.
3 Aschbacher, P. W., "Air Pollution Research Needs: Livestock Production
Systems," JAPCA, Vol. 23, No. 4, April 1973, pp. 267-272.
4 Babcock, L. R. and N. L. Nagda, "Cost Effectiveness of Emission Control,"
JAPCA, Vol. 23, No. 3, March 1973, pp. 173-179.
5 Bank, Marvin and Douglas M. McEachern, "A Carbon Monoxide Profile of a
Main Traffic Artery in Mexico City," presented at 63rd Annual
Meeting of the APCA, St. Louis, Missouri, June 1970.
6 Bastress, E. Karl, "Nature and Control of Aircraft Engine Exhaust
Emissions," presented at 62nd Annual Meeting of the APCA , June 1969,
(APCA No. 69-190).
7 Bayley, E and A. Dockerty, "Traffic Pollution of Urban Environments,'1
J. Royal Soc. Health, 92 (1) pp. 6-11, Jan-Feb 1972.
8 Beard, R. R. and G. A. Wertheim, "Behaviorial Impairment, Associated
with Small Doses of CO," Am. J. Pub. H. 57, November 67, pp. 2012-2002.
9 Beaton, J. L., J. B. Skog, and E. C. Shirley, "Traffic Information Require-
ments for Estimates of Highway Impact on Air Quality," Materials &
Research Department Air Quality Manual, Division of Highways, State
of California, April 1972, 29 pp.
10 Beaton, J. L., J. B. Skog, and E. C. Shirley, "A Method for Analyzing and
Reporting Highway Impact on Air Quality," Materials and Research
Department Air Quality Manual, Division of Highways, State of
California, July 1972, 30 pp.
11 Beaton, J. L., J. B. Skog, and A. J. Ranzieri, "Motor Vehicle Emission
Factors for Estimates of Highway Impact on Air Quality," Materials
and Research Department Air Quality Manual, Division of Highways,
State of California, September 1972, 58 pp.
132
-------
12 Beaton, J. L., J. B. Skog, E. C. Shirley, and A. J. Ranzieri, "Meteorology
and its Influence on the Dispersion of Pollutants from Highway Line
Sources," Materials and Research Department Air Quality Manual,
Division of Highways, State of California, April 1972, 159 pp.
13 Beaton, J. L., J. B. Skog, E. C. Shirley, and A. J. Ranzieri, "Mathematical
Approach to Estimating Highway Impact on Air Quality," Materials
and Research Department Air Quality Manual, Division of Highways,
State of California, July 1972, 65 pp.
14 Beaton, J. L., J. B. S-kog, E. C. Shirley, and A. J. Ranzieri, "Mathematical
Approach to Estimating Highway Impact on Air Quality, Appendix,"
Materials and Research Department Air Quality Manual, Division of
Highways, State of California, July 1972, 107 pp.
15 Beaton, J. L., J, B. Skog, E. C. Shirley, and A. J. Ranzieri, "Analysis
of Ambient Air Quality for Highway Projects," Materials and Research
Department of Air Quality Manual, Division of Hlghwats, State of
California, July 1972, 105 pp.
16 Berry, B. J. L., et al, "Land Use, Urban Form and Environmental Quality,"
Dept. of Geography, U. of Chicago, for the Office of Res. and Develop-
ment, Environmental Protection Agency, 1974, 442 pp.
17 Bisselle, C. A., S. H. Lubore, and R. P. Pikul, "National Environmental
Indices: Air Quality and Outdoor Recreation," The Mitre Corp., McLean,
Va., for the Council on Environmental Quality, Wash., D.C., 1972, 262 pp.
18 Chang, T. Y. and B. Weinstock, "Urban CO Concentrations and Vehicle
Emissions," JAPCA, Vol. 23, No, 8, August 1973, pp. 691-696.
19 Cooper, A. G., "Carbon Monoxide, A Bibliography with Abstracts," PHS
Pub. No. 1503, U.S. Dept. of HEW, 1966, 440 pp.
20 Copley, Charles M., Jr., Division of Air Pollution Control, City of
St. Louis, Missouri and D. A. Pecsok, Air Pollution Control,
St. Louis, Missouri, "St. Louis Regional Air Monitoring Network,"
63rd Annual Meeting - APCA, St. Louis, Mo., June 14-18.
21 Corning Laboratories, Inc., "Procedure for Constructing a Sample Station
Network," Corning Laboratories Inc. (formerly Doerfer Labs), Cedar
Falls, Iowa.
22 Dabberdt, W. F., and R. C. Sandys, and P. A. Buder, "A Population Exposure
Index for Assessment of Air Quality Impact," SRI Project 3364,
Final Report, July 1974, 43 pp.
133
-------
23 Dabberdt, W. F., R. C. Sandys, and P. A. Buder, "Assessment of the Air
Quality Impact of Indirect Sources," SRI Project 2947, Final Report,
March 1974, 97 pp.
24 Danard, M. B., "Numerical Modeling of Carbon Monoxide Concentrations
Near Highways," J. Appl. Meteor., Vol. 11, No. 6, September 1972,
pp. 947-957.
25 Danard, M. B., R. S. Koneru, and P. R. Slawson, "A Numerical Model for
Carbon Monoxide Concentration Near a Highway," Proceedings of
Symposium on Air Pollution, Turbulence and Diffusion, December 7-10,
1971, New Mexico State University, pp. 152-157.
26 Drivas, P. J. and F. H. Shair, "Probing the Air Flow within the Wake
Downwind of a Building by Means of a Tracer Technique," Atm. Env.,
Vol. 8, pp. 1165-1175.
27 Drivas, P. J. and F. H. Shair, "Dispersion of an Instantaneous Cross-Wind
Line Source of Tracer Released from an Urban Highway," Atm. Env.,
Vol. 8, pp. 475-485.
28 Elkus, B. E., and K. R. Wilson, "Air Basin Pollution Response Function:
The Weekend Effect," Dept. of Chem., U.C. San Diego, La Jolla, CA.
Submitted to Science.
29 Environmental Protection Agency, 1973: The National Air Quality Program:
Air Quality and Emissions Trends, Annual Report, Vol, II, 350 pp.
30 Environmental Protection Agency, "Monitoring and Air Quality Trends
Report, 1972," EPA-450/1-73-004, Monitoring and Data Analysis Division,
Durham, N.C., December 1973, 218 pp.
31 Eschenroeder, A. Q., "An Approach for Modeling the Effects of Carbon
Monoxide on the Urban Freeway User," Internal Memorandum, IMR-1259,
General Research Corp., Santa Barbara, Calif., January 1970.
32 Evans, D. W. , "Carbon Monoxide Levels (Letter to the Editor)," JAPCA,
Vol. 23, No. 12 , December 1973 .
33 Everett, M. D. , "Roadside Air Pollution Hazards in Recreational Land Use
Planning," J. Am. Inst. of Planners, Vol. 40, No. 2, March 1974,
pp. 83-89.
34 Fara, G. M., A Pagano, and G. Ziglio, "Investigation of Pollution from
Motorized Traffic in the City of Milan," Minerva Medica, Vol. 64,
No. 5, EPA Translation TR-265-74, January 1973, pp. 254-271.
35 Flynn, Ne. E., and W. R. Crouse, "Carbon Monoxide Emissions in the Bay
Area - 1963," Bay Area Air Pollution Control District, San Francisco,
Calif.
134
-------
36 Fresno Air Pollution Study, State of California, Dept. of Public Health,
(July 1960) .
37 Fukuoka, S. et al., "Results of a Round-the-Clock Investigation of Carbon
Monoxide at Some Intersections and their Vicinities," Tokyoto Kogai
Kenkyusho Nenpo, Vol, 3, pp. 47-56 (March 1972) EPA Translation
TR-259-74.
38 General Electric Co. , "Indoor-Outdoor CO Pollution Study, for the Env.
Protection Agency, EPA-R4-73-020, December 1972 .
39 Georgii, H. W., "Determination of CO Emission Concentrations Allowing
for Meteorological Factors," February 1972, pp. 32-36.
40 Georgii, H. W., E. Busch, and E. Weber, 1967; Investigation of the
Temporal and Spatial Distribution of the Immission Concentration of
Carbon Monoxide in Frankfurt/Main, Report No. 11 of the Institute
for Meteorology and Geophysics of the University of Frankfurt/
Main (Translation No. 0477, NAPCA).
41 Geronimi, J., and J. Duffaud, "Carbon Monoxide Operation," Laboratoire
Central Prefecture dePolice, Office of the Secreatry General,
Ref. 65.794 D/PA, April 1971 , EPA Translation TR-264-74.
42 Giever, P. M., "Significance of CO as an Air Pollutant," J. Occup. Med.,
Vol. 9, 1967, p. 265.
43 Gilmore, T. M., and T. R. Hanna, "Regional Monitoring of Ambient Air
Carbon Monoxide in Fairbands, Alaska," JAPCA, Vol. 24, No. 11,
November 1974, pp. 1077-1079.
44 Godin, G., G. Wright, and R. J. Shephard, "Urban Exposure to Carbon
Monoxide," Arch. Environmental Health, Vol. 25, November 1972,
pp. 305-313.
45 Goldsmith, J. R. and S. A. Landaw, "Carbon Monoxide and Human Health,"
Sciences, Vol. 162, 20 December 1968, pp. 1352-1359.
46 Haskell, E. H., "Land Use and the Environment: Public Policy Issues,"
Environment Reporter, Monograph 20, Vol. 5, No. 28, November 8, 1974,
32 pp.
47 Heck, W. W., O. C. Taylor, and H. E. Heggestad, "Air Pollution Research
Needs: Herbaceous and Ornamental Plants and Agriculturally
Generated Pollutants," JAPCA, Vol. 23, No. 4, April 1973, pp. 357-266.
48 Hindawi, I. J., "Air Pollution Injury to Vegetation," NAPCA Publications
AP-71, 1970 .
135
-------
^ .oydysh, W. G. and H. H. Chiu, "An Experimental and Theoretical Investi-
gation of the Dispersion of Carbon Monoxide in the Urban Complex,
Urban Technology Conf., AIAA Paper No. 71-523, New York Coliseum,
New York, N.Y., May 24-26, 1971.
50 Hunt, W. F., Jr., "The Precision Associated with the Sampling Frequency
of Log-Normally Distributed Air Pollutant Measurements," JAPCA,
Vol. 22, No. 9, September 1972, pp. 687-691.
51 Inman, Robert E., Royal B. Ingersoll, and Elaine A. Levy, "Soil: A
Natural Sink for Carbon Monoxide," Reprinted from Science, 18
June 1971, Volume 172, pp. 1229^1231.
52 Inman, R. and R. Ingersoll, "Tiny Fungi Eat Gas," SRI Intercom, No. 212,
September 28, 1973, pp. 1.
53 Inman, R. E. and R. B. Ingersoll, Note on the "Uptake of Carbon Monoxide
by Soil Fungi," APCA J., Vol. 21, No. 10, October 1971, pp. 646-647.
54 Iwasaki, K., S. Fukuoka, and T. Ohira, "On the Results of Continuous
Measurements of Automobile Exhaust Gas in the Vicinity of the
Ushigome Yanagicho Intersection," Tokyoto Kogai Kenkysha Kempo,
Vol. 2, 1971, pp. 62-67, EPA Translation TR-278-74.
55 Johnson, W. B., W. F. Dabberdt, F. L. Ludwig and R. J. Allen, 1971:
Field Study for Initial Evaluation of an Urban Diffusion Model
for Carbon Monoxide, Comprehensive Report of Coordinating Research
Council and Environmental Protection Agency, Contract CAPA-3-68
(1-69), Stanford Research Institute, Menlo Park.
56 Kahn, H. D., "Distribution of Air Pollutants (Note on)," JAPCA, Vol. 23,
No. 11, November 1973, pp. 973.
57 Kanagawa Prefectural Environmental Pollution Center and Kanagawa Prefectural
Pharamacists' Association, "Survey on Carbon Monoxide Environmental
Pollution Caused by Automotive Exhaust Gas," Kanagawaken Taikiosen
Chosaken Kyu Hokoku, No. 14, February 1972, pp. 33-45, EPA
Translation TR-282-74.
58 Kauper, Erwin K. and Charlotte J. Hopper, "The Utilization of Optimum
Meteorological Conditions for the Reduction of Los Angeles Automotive
Pollution," JAPCA Vol. 15, No. 5, May 1965, pp. 210-213.
59 Kinosian, J. R., and D. Simeroth, "The Distribution of CO and Ox Concentra-
tions in Urban Areas," Calif. Air Resources Board, Div. of Technical
Services, Oct 1973 .
60 Knox, J. B., "Numerical Modeling of the Transport Diffusion and Deposition
of Pollutants for Regions and Extended Scales," JAPCA, Vol. 24, No. 7,
July 1974, pp. 660-664.
136
-------
61 Knox, J. B. and R. Lange, "Surface Air Pollutant Concentration Frequency
Distributions: Implications for Urban Modeling," JAPCA, Vol. 24,
No. 1, January 1974, pp. 48-53.
62 Kurtzweg, C. L, and J. A. Kurtzweg, "Forecasting Carbon Monoxide Emissions
on an Urban Freeway: The Effect of Traffic Management Techniques,"
presented at: The 65th Annual Meeting of the Air Pollution Control
Association, Miami Beach, Florida 18-22 June 1972 .
63 Kurtzweg, C. L. and J. A. Kurtzweg, "Urban Planning and Air Pollution
Control: A Review of Selected Recent Research," Am. Inst. of Planners J.,
Vol. 39, No. 2, March 1973, pp. 82-92.
64 Kurtzweg, J. A. and D. W. Weig, "Determining Air Pollutant Emissions
from Transportation Systems," presented at: The Applications of
Computers to the Problems of an Urban Society, October 24, 1969,
New York, N.Y., U.S. Dept. of Health, Education and Welfare, National
Air Pollution Control Administration, Durham, North Carolina.
65 Kwok, H.C.W., W. E. Langlois, and R. A. Ellefsen, "Digital Simulation
of the Global Transport of Carbon Monoxide," IBM J. Res. Develop.,
January 1971, pp. 3-9.
66 Lahmann, E., "CO Concentrations in the Urban Areas of Berlin (Streets,
Tunnels, Residential Areas and Water Bodies)," Staub-Reinhalt.
Luft, Vol. 32, No. 2, February 1972, pp. 36-40.
67 Lamb, R. G., "An Air Pollution Model at Los Angeles," A Master's Thesis
No. 2749, Univ. of California, Los Angeles, Calif., 1968, 104 pp.
68 Lamb, R. G. and M. Neiburger, "An Interim Version of a Generalized Urban
Air Pollution Model," Atm. Env., Vol. 5, 1971, pp. 239-264.
69 Larsen, R. I., "Relating Air Pollutant Effects to Concentration and
Control," JAPCA, Vol. 20, No. 4, April 1970, pp. 214-225.
70 Larsen, R. I., "An Air Quality Data Analysis System for Interrelating
Effects, Standards, and Needed Source Reductions," JAPCA, Vol. 23,
No. 11, November 1973, pp. 933-940.
71 Larsen, R. I., "An Air Quality Data Analysis System for Interrelating
Effects, Standards, and Needed Source Reductions, Part 2," JAPCA,
Vol. 24, No. 6, June 1974, pp. 551-558.
72 Larsen, R. I. and H. W. Burke, "Ambient Carbon Monoxide Exposures," APCA
Paper 69-167, Presented at Annual Meeting of the APCA, New York,
N.Y., 22-26 June 1969 .
137
-------
73 Lauther, P. J., B. T. Comrains, and M. Henderson, "Carbon Monoxide in
Town Air, An Interim Report," Ann. Occup. Hyg., Vol. 5,
Pergamon Press Ltd. , 1962, pp. 241-248.
74 Lillie, R., "Air Pollutants Affecting the Performance of Domestic Animals,
A Literature Review," USDA Ag. Res. Service, Agriculture Handbook
No. 380, January 1972 .
75 Ludwig, F. L. and W. F. Dabberdt, 1972: Evaluation of the APRAC-1A
Urban Diffusion Model for Carbon Monoxide. Final Report, Coordinating
Research Council, EPAC Contract CAPA-3-68 (1-69), Stanford Research
Institute, Menlo Park, California, 147 pp.
76 Ludwig, F. L., W. B. Johnson, and R. E. Inman, 1975: Air Quality Impact
Study for a Proposed Highway Widening Near Ojai, Part 2:
Projected Impact, Final Report, California Department of Trans-
portation Contract J-7292, Stanford Research Institute, Menlo Park,
California, 111 pp.
77 Ludwig, F. L., W. B. Johnson, A. E. Moon and R. L. Mancuso, 1970: A
Practical Multipurpose Urban Diffusion Model for Carbon Monoxide,
Final Report, Coordinating Research Council Contract CAPA 3-68(1-68,
National Air Poll. Contr. Admin. Contract CPA 22-69-64, 184 pp.
78 Ludwig, F. L. and J. H. S. Kealoha, 1974: Present and Prospective
San Francisco Bay Area Air Quality, Final Report for Wallace,
McHarg, Roberts and Todd and the Metropolitan Transportation
Commission, Stanford Research Institute, Menlo Park, California, 110 pp.
79 Mahoney, J. R., "Fundamentals of Air Pollution Analysis for the Planner,"
source unknown; typed manuscript. July 16, 1972, pp. 1-8.
80 Mason, D. V., G. Ozolins, and C. B. Morita, "Sources and Air Pollutant
Emission Patterns in Major Metropolitan Areas," APCA #69-101,
Annual Meeting APCA, New York, N.Y., 23-26 June 1969.
81 McGuire, T. and K. E. Noll, "Relationships Between Concentrations of
Atmospheric Pollutants and Averaging Time," presented at the llth
Conference on Methods in Air Pollution and Industrial Hygiene
Studies, April 1, 1970.
82 McKee, H. C., J. H. Margeson, and T. W. Stanley, "Collaborative Testing
of Methods to Measure Air Pollutants," JAPCA, Vol. 23, No. 10,
October 1973, pp. 870-875,
138
-------
83 Mortstedt, S. E., K. E. Egeback, N. Walde, and K. Bjorkqvist, "Measuring
of Carbon Monoxide Contents in the Air in City Traffic in Stockholm
and Gothenburg, May-June 1966," Aktiebolaget Atomenergi, October 6,
1962, 22 pp., EPA Translation TR-266-74.
84 Munn, R. E. and I. M. Stewart, "The Use of Meteorological Towers in
Urban Air Pollution Programs," JAPCA, Vol. 17, No. 2, February 1967,
pp., 98-101.
85 Nader, J. S., "Developments in Sampling and Analysis Instrumentation for
Stationary Sources," JAPCA, Vol. 23, No. 7, July 1973, pp. 587-591.
86 Nakano, K. and T. Odaira, "Considerations on CO Cone, in Crossroads
Environs," Annual Report of the Tokyo Metropolitan Research
Institute for Environmental Protection, 1971, pp. 10-20.
87 National Air Pollution Control Administration, 1968 : Report for
Consultation on the Washington, D.C., National Capital
Interstate Air Quality Control Region, U.S. Department of HEW.
88 National Air Pollution Control Administration, 1968 : Report for
Consultation on the Metropolitan Los Angeles Air Quality Control
Region, U.S. Department of HEW.
89 National Air Pollution Control Administration, 1968 : Report for
Consultation on the Metropolitan Boston Intrastate Air Quality
Control Region, U.S. Department of HEW.
90 National Air Pollution Control Administration, 1968 : Report for
Consultation on the San Francisco Bay Area Quality Control
Region, U.S. Department of HEW.
91 National Air Pollution Control Administration, 1969 : Report for
Consultation on the Metropolitan Pittsburgh City Intrastate
Air Quality Control Region, U.S. Department of HEW.
92 National Air Pollution Control Administration, 1969 : Report for
Consultation on the Greater Metropolitan Cleveland Intrastate
Air Quality Control Region, U.S. Department of HEW.
93 National Air Pollution Control Administration, 1969 : Report for
Consultation on the Metropolitan Kansas City Intrastate Air
Quality Control Region, U.S. Department of HEW.
94 National Air Pollution Control Administration, 1969 : Report for
Consultation on the Metropolitan Baltimore Intrastate Air
Quality Control Region, U.S. Department of HEW.
95 National Air Pollution Control Administration, 1969 : Report for
Consultation on the Hartford-Springfield Interstate Air Quality
Control Region (Connecticut-Massachusetts), U.S. Department of HEW.
139
-------
96 National Air Pollution Control Administration, "Air Quality Criteria
for CO," NAPCA Pub. No. AP-62, U.S. Dept. of HEW, March 1970, 184 pp.
97 Nogami, J., M. Izumikawa, and S. Nakagishi, "CO Pollution from Automobile
Exhaust," J. Japan Soc. of Air Pollution, Vol. 5, No. 1, 1970,
pp. 237, EPA Translation TR-260-74.
98 Ott, W. R., "Need for Adequate Monitoring Siting Criteria," Manuscript,
2 April 1974, 6 pp.
99 Ott, W. R., "Development of Criteria for Siting Air Monitoring Stations,"
(draft) to be presented at the 68th Annual Meeting of the APCA,
Boston, Mass., June 1975.
100 Ott, W. R., and R. Eliassen, "A Survey Technique for Determining the
Representativeness of Urban Air Monitoring Stations with Respect
to Carbon Monoxide," JAPCA Vol. 23, No. 3, August 1973, pp. 685-690.
101 Ott, W. R. and D. Mage, "The Representativeness of Urban Air Monitoring
Stations with Respect to CO," Proceedings of the Second Annual
Environmental Engineering and Sciences Conf., Louisville, KY.,
April 1972, pp. 379-394.
102 Ott, W. R. and D. T. Mage, "Trend Assessment of Air Quality Over Large
Physical Areas by Random Sampling," presented at the 4th Annual
Environmental Engineering and Science Conference, Kentucky,
March 1974, p. 19.
103 Ott, W. R. and D. T. Mage, "A Method for Simulating the True Human
Exposure of Critical Population Groups to Air Pollutants,"
presented at Int. Symp.: Recent Advances in the Assessment of
the Health Effects of Environmental Pollution, Paris, June 1974,
11 pp.
104 Patterson, R. M. and F. A. Record, 1974: Monitoring and Analysis of
Carbon Monoxide and Traffic Characteristics at Oakbrook, EPA
Pub. No. EPA-450/3-74-058, 458 pp.
105 Perkins, N. M., 1973: Do Air Monitoring Station Data Represent the
Surrounding Community Exposure? Int. J. Biometeorology, 17,
pp. 23-28.
106 Pooler, F., Jr., "Network Requirements for the St. Louis Regional Air
Pollution Study," JAPCA, Vol. 24, No 3, March 1974, pp. 228-231.
107 Public Damage Countermeasures Branch, Automotive Public Damage Sub-
committee, "A Survey of Environmental Pollution by Automotive
Exhaust Gases, Report No. 1, Pollution by CO," PDCB, Construction
Division, Nerima District, Tokyo, 1971, 51 pp.
108 Robbins, R. C., K. M. Borg, and E. Robinson, "Carbon Monoxide in the
Atmosphere," JAPCA, Vol. 18, No. 2, February 1968, pp. 106-110.
140
------- |