EPA-450/3-77-018
June 1977
Revised December, 1977
SELECTING SITES
FOR MONITORING TOTAL
SUSPENDED PARTICULATES
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
CONTENTS
ABSTRACT i
PREFACE ii
LIST OF ILLUSTRATIONS v
LIST OF TABLES vii
ACKNOWLEDGMENTS viii
SUMMARY ' ix
I INTRODUCTION . 1
A. Need for Careful Selection of Monitoring Sites 1
B. The Approach to the Problem 1
II SPECIAL CHARACTERISTICS OF PARTICULATES 3
A. Method of Measuring TSP „ 3
B. Particle Size 5
C. Sources of Particulates 10
III DECIDING THE TYPE OF TSP MEASUREMENTS THAT ARE TO BE MADE .... 13
A. Uses of Total Suspended Particulate Measurements 13
B. Classifications of Monitoring Purposes 14
1. Appropriate Spatial Scale of Representativeness .... 14
2. Most Important Particle Sizes 17
3. Source Oriented Monitoring 18
C. Relative Importance of the Different Classifications of
Monitoring Objectives 20
D. Relating TSP Monitoring Purpose to the Appropriate Scale of
Measurement and the Appropriate Particle Size Range 22
1. Determine Compliance with Air Quality Standards .... 23
2. Provide Information for the Preparation of Environmental
Impact Statements 23
3. Determine Impact of Specific Sources 23
4. Determine Effects on the Environment ..... 26
5. Provide Information for Better Understanding of the
Processes Affecting TSP Concentration 26
ill
-------
CONTENTS (Continued)
6. Evaluate Results of Control Measures 26
7. Determine Long-term Trends in TSP Concentration 26
8. Evaluate Effects on Humans 27
9. Evaluate Effects on Animals and Plants 27
10. Assess Representativeness of Existing Sites 27
IV SELECTING STATION LOCATION
29
A. Regional Stations 29
B. Neighborhood Stations 38
C. Source Oriented Stations 50
D. Middle-Scale Stations 55
1. Ceneral 55
2. Special Roadway Sites 55
3. Other Special Problems 61
V RATIONALE FOR SITE SELECTION CRITERIA 63
A. Background . . 63
B. Sampler Locations 63
1. Height 63
2. Horizontal Separation from Sources 66
3. Separation from Unpaved Roads 67
4. Separation from Urban Regions 68
5. Effects of Obstructions 70
C. The Importance of Sources at Various Distances from
the Sampler . 70
APPENDICES
APPENDIX A: MODELING THE POLLUTION CLIMATOLOGY ASSOCIATED
WITH STACK EMISSIONS . . 77
APPENDIX B: BIBLIOGRAPHY 97
REFERENCES ....... 115
iv
-------
EPA-450/3-77-018
SELECTING SITES
FOR MONITORING TOTAL
SUSPENDED PARTICULATES
by
F.L. Ludwig, J.H.S. Kealoha, and E. Shelar
Stanford Research Institute
Menlo Park, California 94025
Contract No. 68-02-2053
EPA Project Officer: E.L. Martinez
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
June 1977
Revised Leceaiber, IS77
-------
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 - in limited quantities - from the
Library Services Office (MD-35) , 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. 68-02-2053. The contents of this report are reproduced herein as received
from Stanford Research Institute. The opinions, findings, and conclusions express-
ed 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-450/3-77-018
-------
Criteria are sii<',;>e.'-:ted for locating hi j'.h-volure total suspended
parti ctilate (TSi1) neasuroronl. sires based upon sanplim? needs. These
needs are ret en.'ined and classified according to the purposes for which
measurements are made. The first step in the site selection process is
thus to identify the purpose' of the r '.on i tor in;.', one' relate it to the size
of the area for which the r.easuri'irents are !.o be representative. Atten-
tion must also be p.iven to particle size and the special requirements of
monitoring the impacts of lar;.:e, individual sources. A matrix is in-
cluded to help the reader relate different purposes to appropriate spa-
tial scales and to the other factors that are important to the site
selection process.
Procedures are given for selecting, locations that are representa-
tive of urban neighborhoods and interurban regions; selecting, sites
along traffic corridors is also discussed, "ethods are al.so given for
finding, locations where the impact of nnjor individual sources are most
pronounced. The importance of smaller particles is emphasized because
of their greater health and environmental effects relative to their
mass. Specific recommendations for son.pl in p. heights, distances from
sources, and placement relative to urban areas are <-,ivt-n alonp witli the
rationale behind these recommendations.
-------
PREFACE
The intent of this report is to provide a comprehensive and up to
date technical resource document to assist EPA, state and local air pol-
lution control agencies, and other users in developing better and more
effective TSP monitoring networks. The information irsay be used by EPA
in the future for developing nore definitive guidelines and criteria for
TSP monitoring. However, this report in itself does not constitute the
official monitoring guideline of the Agency.
ii
-------
ILLUSTRATIONS
1 HI-VOLUME SAMPLER AND SHELTER 4
2 CHARACTERISTICS OF PARTICLES AMD PARTICLE DISPERSIONS ...... 6
3 FRACTION OF PARTICLES DEPOSITED I'! THE THREE RESPIRATORY
TRACT COMPARTMENTS AS A FUNCTION OF PARTICLE DIAMETER 8
4 SOUE OBSERVED PARTICLE-VOLUME SIZE DISTRIBUTIONS 9
5 IDEALIZED I1ASS/SIZE DISTRIBUTION FOR. URBAN AEROSOLS 1Q
6 CALCULATED IMPACTS OF A HYPOTHETICAL 160 ?! STACK
LOCATED IN THE SOUTHWESTERN UNITED STATES 19
7 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING REGIONAL
MONITORING STATIONS 50
8 A SATELLITE (LANDSAT) PHOTOGRAPH AND A MOSIAC OF SUCH
PHOTOGRAPHS 31
9 AERIAL PHOTOGRAPH OF A RURAL AREA . 32
10 EXAMPLE OF WIND ROSES 34
11 EXAMPLE OF A NEDS OUTPUT FOR A POINT SOURCE 35
12 SCHEMATIC DIAGRAM OF APPROPRIATE AREAS FOR
A REGIONAL MONITORING SITE 36
13 SCHEMATIC DIAGRAMS OF APPROPRIATE SITING AREAS FOR
REGIONAL MONITORS WHEN TWO SITES ARE PLANNED 37
14 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING
NEIGHBORHOOD MONITORING STATIONS 39
15 AN EXAMPLE OF OUTPUT FROM THE NATIONAL CLIMATIC
CENTER'S STAR PROGRAM . . . 40
16 A SAWTORN J1AP FOR A SECTION OF PORTLAND, OREGON 42
17 SAMPLE TRAFFIC MAP FOR A DOUNTOUN AREA 43
18 CENSUS TRACTS IN THE SAGIKAU, MICI!. SMSA 44
-------
(ILLUSTRATIONS Continued)
19 AERIAL PHOTOGRAPHS OF URBAN RESIDENTIAL NEIGHBORHOODS 46
20 A TYPICAL URBAN NEIGHBORHOOD DEPICTED ON A TOPOGRAPHICAL MAP. . . 47
21 HIGH VOLUME SAMPLER LOCATION IN RESIDENTIAL NEIGHBORHOOD PARK . . 49
22 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING A MONITORING
SITE TO ASSESS THE EFFECTS OF A LARGE, ELEVATED POINT SOURCE . . 51
23- NORMALIZED GROUND-LEVEL CONCENTRATION UNDER EXTREMELY
UNSTABLE CONDITIONS 52
24 NORMALIZED GROUND-LEVEL CONCENTRATION UNDER NEUTRAL CONDITIONS . 54
25 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING SPECIAL SITES
IN STREET CANYONS 56
26 SCHEMATIC DIAGRAM OF AIR CIRCULATION WITHIN A STREET CANYON ... 57
27 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING SPECIAL SITES
NEAR TRAFFIC CORRIDORS. 60
28 NORMALIZED GROUND LEVEL CONCENTRATION FROM AN ELEVATED
ROADWAY 63
29 TSP CONTRIBUTION FROM ROADWAYS ACCORDING TO RECORD'S EMPIRICAL
RELATIONSHIP 65
30 NORMALIZED CONCENTRATIONS DOWNWIND OF A CITY COMPUTED WITH A
GAUSSIAN DISPERSION MODEL FOR TWO STABILITY CLASSES 69
31 SCHEMATIC REPRESENTATION OF THE AIRFLOW AROUND AN OBSTACLE ... 71
32 HIGH-VOLUME SAMPLER LOCATED ON THE ROOF OF AN INSTRUMENT
SHELTER 72
A-l VARIATION OF Oz AND ay WITH DISTANCE AND ATMOSPHERIC STABILITY . 79
A-2 VARIATION OF WIND SPEED WITH HEIGHT AND ATMOSPHERIC STABILITY . . 81
VI.
-------
TABLES
1 SIGNIFICANT SOURCES OF PARTICULATE EMISSIONS IN Tilt: UNITED
STATES .11
2 CATEGORIES OF TSP MONITORING OBJECTIVES RACKED ACCORDING
TO IMPORTANCE 21
3 IMPORTANT APPLICATIONS OF DIFFERENT TYPES OF MONITORING SITE. . . 24
4 EXAMPLE OF A TABULATED WIND SUMMARY 33
5 MINIMUM SEPARATIONS BETWEEN A NEIGHBORHOOD SAMPLER AND
DIFFERENT TYPES OF EOADV.'AY 67
A-l FORTRAN VARIABLES USED IN THE PROGRAMS POLFREQ AND POLAVE .... 84
A-2 INPUT PARAMETERS FOR THE PROGRAMS POLFREQ AND POLAVE 86
B-l BIBLIOGRAPHY INDEX, ARRANGED ACCORDING TO MEASUREMENT
PURPOSES AND SCALES 100
vli
-------
ACKKOULF.HCMr.NTS
The authors have been greatly assisted by the comments and techni-
cal advice of many people, particularly Messrs. E. L. Martinez, Neil
Bern and Alan Hoffman of the Environmental Protection Agency and Hisao
Shigcishi of Stanford research Institute. !>.Te are. also indebted to Ms.
I'.athey Mabrey, Ms. Linda Jones, Ms. Sue Skrabo, Ms. Marilyn Fulsaas, and
Mr. Gary Parsons for their contributions to the preparation of this re-
port.
viii
-------
A review of air pollution literature reveals that atmospheric total
suspended particulates (TSP) are collected and monitored Cor a wide
variety -of. purposes. • F.urthcrinore, it is only . seldom that any attempts
are trade to ensure that the location v/here the samples are collected
matches the purpose of the sampling. Little has been done to classify
the purposes of TSP i:,ensureroent in such a way that a set of standardized
station types can he proposed to meet the various purposes. The need
for a set of uniform site definitions is certainly great. The costs of
equipping the stations, the purchase or rental cost for housing the
equipment, the servicing costs, and the data collection analysis costs
can total many tens of thousands of dollars. If the site of the station
is poorly chosen, most of this money could he wasted, especially if the
data cannot be used for the intended purpose.
Most of the uses of TSP monitoring data fall into one of the fol-
lowing general categories:
. Evaluating compliance with air quality standards
Determination of air quality trends
Development and evaluation of control measures
. Public health studies
Scientific research
Miscellaneous
The preceding list of categories of monitoring objectives is based on
the problems to which the data are relevant, but other systems of clas-
sification are also possible. In particular, we could classify the data
according to what we want the data to represent. For instance, data
that will be used to determine the effects of TSP on public health
should represent the conditions in areas of public exposure. Data that
are to be used to assess the impact of a single, large source should
represent areas where the impact is most pronounced. If the classifica-
tion of monitoring purpose is recast in terms of what the data should
represent, then it v;ill be much more closely related to the physical re-
quirements of the monitoring sites.
ix
-------
The concept of spatial representativeness provides a useful basis
for classifying stations and the uses for the data. Furthermore, the
concept has a physical basis that can serve to define station charac-
teristics. In general, the measurement scales that are of greatest im-
portance are:
Regional scale, to provide a measure of 1ST concentrations
typical of a large, usually rural area of reasonably
homogeneous geography and that extends for tens to hundreds
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.
. Neighborhood scale, defining concentrations within some
extended area of the city that has relatively uniform
land use; dimensions are of the order of kilometers.
. 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; typical areas can be elongated,
measuring tens of meters by hundreds of meters or even
kilometers. Evaluating the effects of many kinds of fugitive
dust sources can also fall in this category.
Kicroscale, to define concentrations in volumes with
dimensions of the order of meters to tens of meters.
Of course, the concept of scale can be extended to include the glo-
bal scale, but for this discussion, those scales enumerated above are
sufficient. In fact, the discussion emphasizes the neighborhood, and
regional scales. Urban scale conditions, as noted, cannot be specified
with measurements at a. single location, but can be synthesized from
measurements representing neighborhood scales. When the neighborhood
scale is well represented, it will not be necessary to find the ideal
urban scale site.
The microscale is not considered in detail in this report for
several reasons. The most important of these arise because the err.phasis
is on the standard high-volume sampler method of measurement, which re-
quires that the samples be collected over extended time periods— usual-
ly 24 hours. An average over such a time period has a temporal resolu-
tion that is inconsistent with a microscale spatial resolution. The tem-
poral variability that matches the spatial variability that one seeks to
x
-------
define with microscale measurements will he totally obliterated with
24-hour sampling. Other reasons exist for not proposing criteria for
ipicroscale measurements. For instance, such measurements are often made
for very specialized research purposes that rake a generalized set of
siting requirements nearly impossible to devise.
The middle scale is not emphasized for reasons similar to those
cited above for excluding the ndcroscale. The inconsistency of the spa-
tial and temporal scales applies. The primary air quality standards are
based on health effects and apply to 24-hour or annual periods. Exposure
on these time scales will generally involve areas much larger than • the
middle scale. Secondary standards are most applicable to visibility
reduction, which is a phenomenon that usually manifests itself on scales
larger than the middle scale.
Some TSP monitoring does not fit readily into any of the different
classes of spatial representativeness. Cases exist where the samples
are collected to determine the impacts of some large individual source
of particulates—elevated sources are especially frequent objects of
such monitoring. Although the scale to be represented will usually be
the neighborhood scale the methods used to locate these "source- orient-
ed" sites are so suffiently different they require a separate
category, outside the categories devised solely on the basis of spatial
representativity.
The most important basis for air quality standards is the public
health impacts of a pollutant, which leads to one further consideration
in the classifying of TSP monitoring sites and in the derivation of phy-
sical criteria for those sites. For monitoring purposes related to air
quality standards or the evaluation of health effects, the sites should
be located so that they portray public exposure reasonably realistically
and so that the samples are also representative of those particulates
that are the most damaging to human health. Studies have shown the
smaller particles—with diameters less than a few microns—to be more
respirable and a greater health hazard. The high-volume sampling method
prescribed for measuring TSF to assess compliance with air quality stan-
dards emphasizes the larger particles. Mass per unit volume is meas-
ured. Since the mass of a single 10 ym particle is equal to that of a
million 0.1 u. m particles, the larger particles are weighted by the meas-
urement method in a fashion inconsistent with their health effects. The
smaller particles also have more pronounced effects (per unit mass) on
visibility, attenuation of solar radiation, and cloud condensation
processes than do the larger particles. Inasmuch as the prescribed sam-
pling methodology emphasizes larger particles, which could be misleading
for sone monitoring purposes, the siting process should try to minimize
such effects whenever appropriate.
-------
Factors other than scale of representattveness, source evaluation
requirements, and particle size are important to the definition of the
required characteristics for the different types of monitoring site. One
of these other factors arises because virtually all routine TSP monitor-
ing is related directly.or indirectly to health or public exposure and,
therefore, samples taken at breathing, level are most appropriate. In-
lets at breathing level will constitute an obstacle in many locations
and will also be. subject to vandalism. It has been suggested that a
height of 2- n is a reasonable compromise between these two conflicting
requirements. Obviously, the sappier cannot always be placed exactly 2
meters high. If it is between about 2 meters and 15 meters high, and if
it is well removed from any nearby source, the breathing zone concen-
trations will be reasonably portrayed.
The adoption of uniform criteria for locating air quality monitors
would provide a consistency not now found in available data. Consistency
will allow data from different locations to be compared with the as-
surance that differences in the data sets will reflect something other
than anomalous site characteristics. In suggesting siting criteria for
TSP monitors, an attempt has been made to maintain consistency wherever
possible with other site classification schemes, especially those
developed at the 1976 EPA Siting Workshop and those of Ott (1975) and
Ludwig and Kealoha (1975). The latter authors have focused on the prob-
lems of carbon monoxide monitoring so that some unique differences rela-
tive to T5jP siting will be inevitable, but they are few and not serious.
In proceeding from concepts such as spatial representativity or em-
phasis on different particle size ranges to a set of concrete siting re-
quirements, some 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 con-
centrations at a monitor and the distance to that source.. From this re-
lationship, maximum allowable contributions (in yg m ) could be con-
verted to minimum separations (in in) between the monitor and certain
kinds of sources. For Instance, regional sites should be far enough re-
moved from cities so that the concentrations arising from emissions
within the city will be less than five percent of the secondary NAAQS,
or less than 3 yg m • Similarly, neighborhood sites should not be so
close to large sources that the emissions from those sources increase
measured concentrations more than about ten percent of the primary annu-
al -standard, or 7.5 yg m • Among the sources of TSP in urban areas is
the street dust raised by traffic. Examination of collected samples
(e.g., Draftz, 1975) has shown the presence of large amounts of such ma-
terial, so samplers must be separated from nearby roadways if they are
to be representative of something more general than street side condi-
tions.
xii
-------
The relationships between the minimum separation of sources and
monitors, and the maximum concentration impacts of those sources has
been derived largely fron empirical relationships found in past data,
but standard Gaussian diffusion equations have also been used to estab-
lish some of the siting specifications. The relationships, empirical or
theoretical, have also been used to derive the minimum separations
necessary to keep the influence of individual sources below specific
levels at neighborhood or regional monitoring sites.
In those instances where a source-oriented monitor is being sited,
the same relationships can be applied for virtually the opposite
purpose—that is, to identify areas where the impact of a single source
will be greatest. In the discussions of source-oriented monitoring
presented in this report, two kinds of impact are identified. The first
is the short-term effect corresponding to the 24-hour standard, and the
second is the long-term, or annual, effect. The reason for distinguish-
ing between the two is that the maximum long-te.rm impacts of elevated
sources may be quite different from maximum short-term effects, not only
in magnitude, but also in their location. Maximum short-term effects
often occur much closer to smokestacks and other elevated sources than
do the longer-term effects.
The text of this report provides the details of how the areas of
maximum impact can be identified for source oriented monitoring. Step-
by-step procedures for the identification of suitable neighborhood, re-
gional, and traffic corridor sites are also given and the reader is en-
couraged to read the complete report if actually faced with the problem
of selecting suitable TSP monitoring sites. The major physical charac-
teristics of the two most important site types are tabulated below.
Whenever applicable, we have also included the corresponding site type
and nomenclature proposed by Ott (1975).
Neighborhood Monitor
Ott type: "C" - Residential population exposure
station.
Scale of repre-
sentativeness: Neighborhood
Sampler location: 2 to 15 m above ground level removed from large
obstacles by at least twice the height of the
obstacle, 20 n from nearest traffic*,
400 m from nearest major roadway with traffic
of 50,000 vehicles per day or more.
* Ott (1975) specifies a minimum distance of 100 m to nearest street
with more than 500 vehicles per day, but this seems unduly stringent.
xiii
-------
Other requirercents;
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 neighborhoods in the town. The
surface in the immediate area should have low
ground cover to reduce interference from
fugitive dust. Traffic counts on nearby
streets and wind measurements would be
valuable.
Regional Monitor
Ott type:
Scale of repre-
sentativeness :
Sampler location:
- tionurban background stations.
Regional
2 to 15 n above ground level, removed from
large obstacles by at least twice the height of
the obstacle. At least 10 km from nearest city,
in direction that is least frequently downwind;
if more than one such station is planned, see
discussion in text for preferred directions;
0.5 km from nearest intercity roadway of 50,000
or more vehicles per day, 4 km from any
unpaved public road, and at least 40 or 50
meters from even very lightly traveled paved
public roads. Other requirements: An under-
lying surface as dust free as possible, low
ground cover desirable. Low lying areas,
subject to cold air drainage and stagnation are
to be avoided, (iind measurements at the site
would be useful.
xiv
-------
Source-Oriented Monitor
Ott Type:
Scale of repre-
sentativeness :
Sampler location:
Other requirements:
"F" - Specialized source survey station. Ott's
Type F Station appears directed more toward
ground-level sources than is envisioned for a
source-oriented TSP monitor.
Generally supposed to represent the neighbor-
hood scale effects of a single source.
2 to 15 m above ground level, removed from
large obstacles by at least twice the obstacle
height. Should be placed in area of maximum
impact of the source—this will differ for
short-term and long-term impacts. Methods of
using source characteristics and climatological
information to locate areas of maximum impact
are discussed in the text. If possible, the
site should be away from major roadways or
other sources that will obscure the effects of
the source of interest.
Much the same as for Regional and
Neighborhood Monitors. Wind Measurements
at the source would be useful.
The above summaries of physical characteristics mention factors
such as nearby tra.ffic, homogeneous land use in the surrounding neigh-
borhood, local ground cover, and areas of maximum source impact. As the
text makes clear, these are some of the many considerations that are re-
quired to properly locate a monitoring site. You don't just go out and
put a high volume sampler 2 m in the air and turn it on. Some places
are much better for this than others, and this report describes step-
by-step procedures for identifying the better locations for the dif-
ferent types of site. In all cases, the first step of the selection
process is to gather demographic, traffic, and land-use data, topograph-
ic maps, and climatological information. This infomtion is used to
identify generally desirable areas, and to reduce their numbers and
their size until a final selection is carefully made. Vlhen 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 or relationships to individual sources. In fact, the
application of the concept of representativeness to the siting problem
has also involved relationships between the monitor and sources so that
neighborhood or regional sites are located with the intention that they
xv
-------
should represent the composite effects of emissions over areas extending
kilometers or tens of kilometers away from the sampler. Essentially,
the purpose behind the site classification system presented here is to
provide an identifiable constituency of emitters for each monitor. For
neighborhood stations we have estimated that sources within about 2 ki-
lometers will provide more than half the measured TSP concentrations
about 30 or 40 percent of the time unless there are elevated individual
sources that impact on the monitor.
In summary, the guidelines presented here will serve as a good
technical basis for selecting sites that can be classified into a limit-
ed number of comparable types. The. standardization of physical charac-
teristics will ensure that comparison among sites of the same type will
not be clouded by anomalies in the data arising from peculiarities in
the siting. Furthermore, the guidelines are sufficiently consistent
with those already proposed elsewhere that there should be little diffi-
culty in combining the schemes to jointly locate monitors for different
types of pollutants.
Use of the classification scheme does nore than ensure compati-
bility of data and allow reasonable comparisons among stations of the
same type. 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 represent. 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.
xvi
-------
IfJTROD'JCTION
A. Meed for Careful Selection of Vonltoring Sites
Air pollution lias become a matter of considerable public concern
because of its adverse effects on public health arc1 property. Laws have
been passed to provide a legal basis for instituting, actions to reduce
pollution to acceptable levels. In the case of TSP, the primary Nation-
al Ambient Air Uuallty Standards (MAAQS) have been set at 260 yg m~3
for a 24 hour average and 75 yg m~3 for the annual average. The
corresponding secondary !'AAQS are 150 and 60 yg m~^« Monitoring can
tell whether acceptable levels are being exceeded and provide the meas-
urements necessary to develop or evaluate those control plans that will
be effective.
Obviously, the monitoring data have great importance and their col-
lection deserves considerable attention. The question naturally arises:
Are some measurements more suitable for their intended purpose than oth-
ers? Of course, the answer is "yes"; otherwise, this report would not
have been written, its objective being to define the characteristics of
the most suitable locations. Some extreme examples of unsuitability are
easy to imagine, e.g., a particulate monitor in an open, dusty field
providing data from which regional control plans are derived, or the
collection of samples for assessing respiratory health effects in an
area where the total dust loading is dominated by very large, nonrespir-
able particles.
The uses of monitoring data are important, and their importance can
be belittled by improper siting. The importance of the data requires
that equivalent importance be attached to the selection of sites. For
this reason, we would expect that those selecting sites should be wil-
ling to invest an appropriate amount of time and effort in the task. The
procedures recommended here are relatively simple, but they do require
acquisition and interpretation of considerable information.
P. The Approach to the Problem
Finding the appropriate siting criteria for monitoring stations
starts by finding why the pollutants of interest are monitored. Siting
criteria are rules describing the proper physical location of a monitor,
and they should be physically related to the reasons for monitoring. The
second step toward finding the appropriate siting criteria for different
monitoring purposes is the categorization of those purposes according to
a physically based classification syster.. Then it is necessary to iden-
tify physical siting criteria corresponding to each set of objectives.
-------
As a prelude to the development of siting criteria, the uses of to-
tal suspended particulate (TSP) data were surveyed. Generally, the .in-
terpretation given to TSP data assumes that the measurements represent
areas and volumes that extend well beyond the limited area and the snail
volume that is actually sampled. The area presumed to be represented
may range from the intermediate in size, like a neighborhood, to 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 for estab-
lishing 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. This purpose provides a basis for selecting s station
type, based on the area that-the measurements should represent. Final-
ly, t procedures arc followed that lead to sites that represent areas of
the appropriate size.
Total suspended particulate data have another characteristic that
affects the connection between monitoring purpose and site characteris-
tics: particle size. Size is important for three reasons:
(1) Mass concentration varies with size
(2) Particle removal processes depend on size
(3) Health effects, particularly respirability, are influenced
by size.
Thus, spatial scale as a basis for classifying monitoring sites is
not enough, and the system has to be extended to include consideration
of the particle sizes of greatest interest. The incorporation of parti-
cle size into the list of factors to consider in the monitoring sites is
consistent with the philosophy that the monitoring goals derived from
social, scientific, or legal requirements should be translated into a
corresponding set of .desirable, physically defined goals, which in turn
can be related to actual physical characteristics of an appropriate mon-
itoring site.
-------
II SPECIAL CHARACTERISTICS UF PARTICULAr.LT,S
Before proceeding to the derivation, of siting criteria, it will be
worthwhile to review briefly sone of the important characteristics of
natural and pollutant aerosols, their sources and methods of measure-
ment • "
A. Method of Measuring TSP
Basically, the measurement of TSP concentration entails drawing air
through a filter to remove the particles. The sampled air volume must
be measured, as must the amount of collected particulate. According to
the regulations published in the Federal Register (1971), the approved
apparatus must be capable of drawing air through an 8 by 10 inch (20.3
by 25.9 cm) clean glass fiber filter at a rate of at least 1.7 m^ min~l.
Average air flow through the high-volume (Hi-Vol) sampler can be es-
timated from the average of the flow rates at the beginning and end of
the sampling period, and the duration of the sampling. The mass col-
lected is determined from the weight of the filter (to the nearest mil-
ligram) before and after sampling. Both weighings are to be made after
the filter has had time to equilibrate with an environment of controlled
temperature and humidity.
The Federal Register (1971) states that Cor 24-hour samples, the
method is adequate to measure average concentrations as low as 1 yg m~ •
Comparisons of simultaneous Hi-Vol sampling (Clenents et al., 1972; Lee,
Caldwell and T'organ, 1972) indicate that for more than 80 percent of the
cases, the duplicate samples differed from their mean by less than ten
percent. Correlations greater than 0.99 were found between concurrent
samples by Lee. et al. (1972). All these results suggest reasonable
reproducibility with the method. tJhen filter materials other than glass
fiber are used, results may differ. Lee et al. (1972) found a tendency
toward higher measured concentrations when membrane filters were used,
especially in those areas where particle sizes were smaller, suggesting
somewhat higher collection efficiencies for small particles with the
membrane filters than with the glass fiber.
The approved TSF sampling methodology (as given in the. Federal, i'e-
gister) specifies the acceptable flow rates and sar.pling materials to
provide a degree of standardization. Similarly, the shelter for the
sampler is also prescribed. Figure 1 is a schematic representation of
the shelter and a properly mounted Hi-Vol sampler. Tho dimensions shown
in the Figure are approximate; there i.
-------
7"
cm
Source: Federal Register, 1971
FIGURE 1 H'I-VOLUME SAMPLER AND SHELTER
-------
Whatever its shortcomings, the !'i-V'ol sampler in a standard housing
is the most u'idely user' instrument for sampling TSP. It is the current-
ly accepted "standard" liethod (Hoffman of nl, 1975), although rany other
methods exist, based on rany principles other than Hi-Vol filtration—
e.g., inertial iripac tion, Light scattering, light transmission, centri-
fugation, ability to serve ns droplet nuclei and so forth. This report
is directed solely to the siting of the "standard" instrument described
in the Federal Kegistrr. Many of the. sair.e principles can be extended to
other methods, should the reader care lo do so. • •
B. Particle Size
As noted, the TSP monitoring method considered here is supposed to
be effective over all size ranges up to about 100 ym. The measurements
are based on total particle mass. For monitoring purposes of this kind,
a single 100-um-diameter particle is equivalent to 10° 1-um-diarneter
particles, or ±Q9 0.1-um-diameter particles. This emphasis on mass
results in some conflicts of purpose that have to be resolved in the
development of siting criteria. These will be discussed later.
The size of a particle affects its behavior in a multitude of ways
from light scattering to settling velocity. Figure 2 (Lapple, 1961)
provides a convenient summary of size-related characteristics of aerosol
particles. Among the more important characteristics is the terminal
velocity for gravitational settling in air. Figure 2 shows that submi-
cron particles fall very slowly, less than about 25 en hr~l. The termi-
nal velocity increases rapidly with size, reaching about 2000 cm hr~l
for a 10-um-diameter particle. Obviously, submicron particles are like-
ly to stay airborne longer and be transported farther than the 10 um and
larger particles, a fact of some importance to the siting process.
The settling velocity of a particle is a reflection of the way in
which the particle moves through the air when a force, or acceleration is
applied. Small particles move very little relative to the air; large
particles move a lot. When moving air encounters an obstacle, its flow
is changed and accelerated around the object. Small particles will tend
to follow the air and bypass the obstacle, but larger particles are much
more likely to continue along their original path and may impact and be
renoved from the air stream. The theory of impaction of particles is
beyond the scope of this work (see for example, P-anz and Wong,, 1952); it
should suffice to note that larger particles are more subject to removal
by impaction on leaves, twigs, buildings, or any other obstacle to air
flow. Again, the importance of this phenomenon to aerosol sampling
should be apparent.
If particles are subject to removal by accelerations, around build-
ings or vegetation, it seems likely that they might also be subject to
impaction and removal during, their movement through the many passages to
the respiratory system, and they are. Figure 3 shows the fraction of
-------
CHARACTERISTICS OF PARTICLES AND PARTICLE DISPERSOIDS
Equivalent
Sizes
Electromagnetic
Waves
Technical
Definitions
Common Atmospheric
Dispersoids
Typical Particles
and
Gas Dispersoids
Methods for
Particle Size
Analysis
Types of
Gas Cleaning
Equipment
Particle Diametef. microns (ji
0.0001 0.001 0.01 0.1
-, ;-;;•,':,
(1mm.)
10 100 1.000
! 1 ! 5,000 1.2
1 10 100 1.000 10,000 2,500
i l l
Angstrom Units. A j I Theoretical Mesh
j i I (Used very infrequen
! 1 i :
X-Rays *
-- - Ultraviol
c^j,, Atlerbefg or International Std. Classification System
00 • adopted by Internal Soc. Soil So. Since 1934
-
Visible j
et ,^t . »4-«j_Near infrared *
^ Solar Radiation
— — Smog— J- -
-• Rosin Smoke — »-
0. CO C.H.
H. 1 F. i Cl. 1
i xA;/1" c.
I "--??Vr'VC. Molecules' "
! N i CH, | SO.
'CO H I) HC CfH
••Mole, uljr diamfteri calculdled
Irom viwusity data al 0 C
•* — - Tobacco Smoke-- *•
- - Metallurgical Dusts and
K- Ammonium Chloric
«•
Si li-
— -M- — Clouds
, Coal
Fumes
e Fume-H" — Ce
Sulturic
Concentrator Mi
act
„., Sulfuric Mist
k Paint Pigments- - — »•
-* — Zinc Oxide Fume--H t*- Insecticide Du
^ K-| -Groun
5llica ff — .. Spray Dried Milk —
Aitken -- — -™M,,run«
Nuclei i
- Sea S
Combustion
"Nuclei
i » Viruses -
h
- x-
alt Nuclei H t* Nebulize
.Lung Damaging
Dust
Red Blood Cell Diame
-*-- - - -- Bacteri
"*• - Impinge
Ultramicrosocope '- --»t*- -
-Electron Microscope
•*• — Centrifuge -
u- . - - - . Turbidi
tey Diffraction *-*** - - —
Adsorption' •
- - ; — -Light Scatt
Nuclei Counter t >-
Mi
Tietry1-
-Permeabili
sts-*
Talc
Willed
r Dro
I
er(A
3
Elec
:roscc
...
Sedir
-
y-
50 i «*> " \v i
625 II
V- rt(l I/O
ly) iViTf
J.".. .30 'K-
1 1 ' i
•* Far In
*
>' 65 35 2ol 1 10 6
Tyler Screen Mesh
100 48 28 1 114 1 8
1
. 60 40 20! 1 i: 6
U.S. Screen Mesh
!00 50 30 1 116 1 8
rared ••
d-—- - Coarse Si
and Fog — ^*Mist-wOrizzeH —
i
- — Fertilizer. Ground L mestone— *•
Dust -+•
ment Dust-: »
5^ N
•Pulverized Coal-
* — Flotation Ore
__„ ^
Plant
Spores' '
Flour *H
' Beach Sand
is -*-, K— ; Hydraulic Nozzle
Pneumatic
"*" Nozzle Drop?*j
Jults):7.5^±0.3/J
1 >f Human Hairn
troformed
Sieves '
pc »
,--
— Elulriation
nentation
,,__J
;ring'-'— -- -~»4-
~* Electrical Conductvity »
Ultrasomcs i
•
ivrry limited md
M'idl applicalionl
- - - -- - —--Liquid Sen
- - Cloth Collectors
-- Packed Beds —
u _ .. .. j —
High Efficiency Air Tillers - f - • — *•
Electrical Precipitators- ~ -
i Rev-nolds N.imwi BO " 10 ". 10 "' 10 " 10 " 10 10"
'"„*". ' 1
•Terminal
Gravitational
Settling*
[for spheres, 1
,p.gr.2.0J '
Particle Diffusion
cm '/sec.
lain) Senlinp Vflootv JQ
''"'"•"' ' '
j
Reynolds Number QQ '"lO '"10 '
InWatet r J •
Sfflln^Vetocitv. )0 '"^ 10
at'^'c '. -1 . 10 : 10 '" 10
1 aim (• ' - * • • ' - - " ' •
In Water
at 10 ' 10
M-C .'•••' ••:' ' •'
' . 10-
10 'MO 10 ;
10 \ .. 10 , . .
. 10 ;: ; . 10
10
•Stofces Cunningham ' /'«•'<•*• .' .' •» "•* # . • -i -P ?•
factor included m 0.0001 0001 0.01 0
values Riven tiif air ' ilm«l
but not included for water Particle Dia
.- - - io '. . .
10 ;' 10 ; 10 ,
10 '. . . 10 '. , s
" 10
10 '§
.1
meter, microns (p.
10 '; 10 ;'
10 ''. , 10
10 ; 10 ,
10 ". , ^ 10
10
-*
entril
bbers
Comrr
10
0 '
'•< s
., .
,
ion Air Filters — ••
Impingeme
- -Mechanical Sei
10 '': 10 \ 10"
10" , ., 10', ,
- -H
-H
1
Machine Tools W
Settling Char
arators — *^
•10' ( 10'
., 10' , , s
10 .' 10 10 ! jio1'. 10' ( 10
10 ; , „ 10 ! , .
10 "t . ,
10 ', 10 '".
(1C
10.C
3
4 V
3
t ',
n.
-00
1 .
•f 1-
'/,- r
'/:
\
-•Microwaves (Radar, etc.)
1
'
— Rain *
Drops 1
* Furnishes average
diameter but no s
distribution
* *Sue distribution rr
obtained by spec
cat bration.
crometers, Calipe
ibers - - -
10 '
10'
particle
K
ay be
1
rs. etc.) •
.4 !>
.- 10' ,„
•' , 10
10 ' , , 10: ? .
, i 10" ,
, -
10" 10- 10
::;
-------
the total breathed particles retained in different parts of the respira-
tory system as a function of the mass median particle size (Task Croup
on Lung Dynamics, 1966). The extremes represent different standard
geometric deviations of the inhaled aerosol partirles. Deposition in the
lungs is a much greater hazard with particles of suhnicron diameter than
with the larger particles that tend to be trapped in the nasal and upper
respiratory passages.
The role of particles in visibility reduction is another of the
reasons often given for TSP monitoring and here, too, particle size is
an important factor. Reduction in atmospheric visibility is caused by
light scattering in the aerosol. According to Foxvog (1975), single
"absorbing particles scatter light most efficiently for diameters near
0.2 urn, whereas for nonabsorbing particles, the scattering per unit mass
peaks in the 0.4 to 1,0-um-diameter range." The scattering per unit
mass falls rapidly toward 10-ym-diameter. The most important
visibility- reducing particles are those below 10 um in size.
The health and visibility consequences of small particles are more
severe than those of the larger particles, which would be of no impor-
tance if there were no large particles in ambient air; then TSP would be
equivalent to small suspended particulate, and TSP measurements would be
reasonable indicators of the visibility or the health degradation poten-
tial of the atmospheric aerosol. Since removal processes like settling
and impaction are more effective for large particles than for small, we
might expect the measured particulates to be dominated by the smaller
particles. This is the case on a number basis. However TSP monitoring
measures mass and as noted before, a single 10-um particle has the mass
of a million 0.1-ym particles. Therefore, the distribution of mass with
particle size is of importance.
Figure 4 shows several volume size distributions observed in vari-
ous U.S. locations. These should be taken as approximations because
several have been converted from the number distributions given in the
original sources (e.g., Junge, Robinson, and Ludwig, 1969; Ludwig and
Robinson, 1971; Peterson, Paulus and Foley, 1969). The results are
quite interesting in that several examples show particle volume distri-
bution increasing with size toward a large particle mode at several mi-
crons diameter or greater. The Los Angeles measurements of Midy ct al.
(1975) extend to larger sizes and confirm that a binodal distribution is
common. Figure 5 shows the idealized urban size distribution hy-
pothesized by llidy et al. (1975). According to that reference, the mode
at smaller sizes represents the anthropogenic contribution to the urban
aerosol while the larger material with a volume mode at about 10 ym c'i-
ameter is of natural or "quasi-natural" origin. The chemical elements
that are usually found in the larger particulates, the "blowing dust,"
although referred to as natural, is as likely to be from road dust, con-
struction activity, and so forth.
.7
-------
Flocchini et al. (1976) have found similar size separation of ele-
ments in their California samples. Silicon compounds, prime dust com-
ponents, are almost all found in particles larger than 3.6 urn; about 80
percent of such materials were between 3.6 and 20-urn diameter. On the
other Viand, about 80 percent of the lead compounds were found in the
size range 0.11 urn to C.65 um, consistent with the findings of others
(e.j,:., Lee, Patterson and Vaginan, 1968; Robinson and Ludwig, 1967) in
other parts of the country. Lead has often been assumed to be associat-
ed with auto exhaust.
S£$^/U"^/A^
MASS MEDIAN DIAMETER, JU
Source: Task Group on Lung Dynamics, 1966
FIGURE 3 FRACTION OF PARTICLES DEPOSITED IN THE THREE RESPIRATORY
TRACT COMPARTMENTS AS A FUNCTION OF PARTICLE DIAMETER
-------
1
1
90
80
70
00
90
40
30
20
10
0
0.09
I • • • • I I • I • • • • I
a. Data from Peterson et al, 1969 —
V
MINNEAPOLIS/ST. PAUL
90
80
TO
60
90
40
30
20
10
0
0.09
0.1
0.2
0.9 I
RADIUS —
c. Data from Ludwig and
Robinson, 1971
SAN FRANCISCO BAY AREA
0.1
0.2
0.9 I
RADIUS —
10
10
20
20
90
80
TO
Hf
1 "
I 60
e «°
_ b. Data from Junge et al, 1969
1
ft
I
g
i
SO
20
10
O
0.09
CAPE BLANCO, OREGON
\
0.1
RAOIU9 —
10
20
90
80
TO
60
90
40
30
20
10
_ d. Data from Hldy et al, 1975
FT.
ARGUELLO
\
GOLDSTONE
-r.".l"***t»-'':
HARBOR
FREEWAY.
0.09
0,1
0.9 I
RADIUS —
10
20
FIGURE 4 SOME OBSERVED PARTICLE-VOLUME SIZE DISTRIBUTIONS
-------
MASS
CONCENTRATION
PER UNIT
DIAMETER RANGE
PRIMARY
ANTHROPOGENIC
AND SECONDARY
I
4
NO"
PRIMARY NATURAL
OR QUASI-NATbRAL
I
0.5
1
DIAMETER
10
/urn
SA-4349-1
SOURCE: Hidy, et al., 1975.
FIGURE 5 IDEALIZED MASS/SIZE DISTRIBUTION FOR URBAN AEROSOLS
In summary, particle size is of importance in aerosol sampling be-
cause it is related to several of the purposes of such sampling. Health
effects are most pronounced for the smaller size:;, as are visibility ef-
fects. Control neasures will often he nost applicable to the sources of
the smaller particles. Observations have suggested that on a volume
(or mass) basis, the larger particles make up a very important fraction
of the total suspended particulate. It is important that these facts be
acknowledged when siting criteria are derived.
C. Sources of Particulates
Table 1 summarizes the magnitude of the various anthropogenic
sources of particulates in the United States, t.'hile natural dusts con-
stitute nearly half the particulate emissions, their generally large
sizes limit their lifetime in the atmosphere, as well as their health
and visibility effects. In general, high atmospheric concentrations of
large particles are limited to areas near their sources. The gaseous
pollutants will also exhibit strong concentration gradients in the vi-
cinity of sources, but not as great as those associated with large par-
ticles. The diffusion processes that cause concentration gradients in
the vicinity of gaseous sources also operate on particulates, but arc
augmented by settling and other removal processes that will emphasize
the gradients. This suggests that the arua '..'hero a particulate source.
substantially affects observed concentrations ray be smaller than for a
comparable source of gaseous pollution.
10
-------
Table 1
NATIONWIDE EMISSIONS
ESTIMATES FOR 1974
SOURCE CATEGORY
EMISSIONS
106 t/yr
PERCENT OF
TOTAL
Transportation
Highway
Non-highway
Stationary Fuel Combustion
Electric Utilities
Other
0.9
0.4
3.4
3.6
4
2
17
18
Industrial Processes
Chemical
Petroleum Refining
Metals
Mineral products
Other
0.2
0.1
1.5
6.3
2.5
7
31
12
Solid Waste
0.6
Miscellaneous
Forest wildfires 0.5
Forest managed burning 0.1
Coal refuse burning 0.1
Structural fires < 0.1
Other 0.1
Total 20.3
100
Source: EPA-450/1-76-002, Nov 1976, National Air Quality and Emissions
Trends Report, 1975.
11
-------
One final comment is in order regarding the interpretation of the
figure;-; ^ivpn in Table 1. The table says nothing about recntrainment of
larj/e particles after they have been removed. Soil erosion may repeat-
edly involve the same particles. In the extreme, this is "saltation,"
where large particles are eroded by the wind hut not truly suspended in
air so that their progress is made through a series of hops and jumps
over the surface. It is not always clear how the evaluation of emis-
sions should treat reentrainrnent. Cowherd et al. (1974) have adopted a
rininnun significant travel distance of about (> m in their evaluations of
emissions from fugitive dust sources, \ccording to their calculations
this criterion eliminates most of those particles larger than 100 urn at
wind speeds less than about 5m s~l—consistent with the sampling size
limitations of the Ui-Vol sampler shelter. Effectively, this consti-
tutes a definition of particulate emissions from ground-level sources—
they are those which become sufficiently well suspended in the air to be
transported at least 6 n. If they travel beyond that distance, settle
to the ground, and later become resuspended in the wind, then presumably
that reintroduction constitutes a new emission to he added once again to
the year's total emissions. Obviously, considerable potential exists
for ambiguity in the interpretation of the annual emissions p.iven in
Table 1. Fortunately, this ambiguity is generally limited to the large
particulates, especially those in the category of natural and fugitive
dusts. As has been discussed earlier, these are usually the materials
least important to health or to visibility reduction.
12
-------
Ill DECIDING THE TV!1!! OF TSP MEASUREMENTS TMAT ART. TO F.F.
A.
Uses of Total Suspended Participate "casurenents
The objective of tbis section is to present a unifier? themo for
classifying monitoring stations. Cur approach to tbis objective has
been to start with a simple list of the uses of ambient TSP concentra-
tion data, which was compiled by searching the literature for reports
and papers that made use of such data (see Appendix B). 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 gen-
eral categories:
1. Evaluate Air Quality and Air Quality Trends
Determine adherence to air quality standards:
- Federal primary
- Federal secondary
- State or local.
2. 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 TSP concentrations:
- In an urban area
- Provide information for city and regional planners
and decision-makers.
3. Public Health
Evaluate the effects of human' exposure to ambient
aerosols.
. Determine the relationship of ambient outdoor TSP levels
to those inside buildings.
Evaluate TSF effects on animals and- plants.
13
-------
4. Research
Determine effects on the environment, e.y. ,
- Visibility
- Condensation processes
- Attenuation of solar radiation.
Provide information for better understanding of the
processes affecting TSP concentrations:
- F.emoval mechanisms
- Transport processes
- Objective modeling: development, evaluation, and
refinement.
Provide information for evaluating fugitive dust sources
5. 'Miscellaneous
Test monitoring equipment and methods.
. Assess representativeness of existing or proposed
monitoring sites.
The list of purposes given above is quite diverse, and it may not
be immediately obvious that a coherent, physically based classification
system is possible. Ludwig and Kealoha (1975) used some of the ideas of
Ott (1975) to classify a similar list of carbon monoxide monitoring ob-
jectives according to the spatial scales of representativeness that were
most appropriate to each of the monitoring objectives. Such a scheme
satisfies the need for a physical basis to the classification system;
the physical basis is important because the classification system has to
be related to a set of appropriate physical site characteristics.
Any useful measurement of TSP concentration is supposed to
represent some volume and some period of time. The measurements always
entail tine averaging, because the sample is collected over a finite
period; at the saive time, sor.ie spatial averaging occurs because of the
finite volume of air that is collected. In general, this volume does
not correspond to the. objectives of the monitoring. The objectives may
require representation of volumes considerably different from those frorr
which the samples are drawn. Site selection, when viewed fron this per-
spective, consists of selecting a location or locations where concentra-
tions in the volume sampled can be related to concentrations representa-
tive of the volume that is required to v.eet objectives.
14
-------
B. Classification of f'onitoring Purposes
1. Appropriate Spatial Scale of Representativeness
The categories of spatial representativeness used for classi-
fying 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.
a. 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 TSP distribu-
tions within parking lots, within street canyons, over unpaved highways,
or around fugitive dust sources will require microscale measurements.
The development and testing of models to describe the processes that
produce the observed details of such concentration distributions would
also require data of this scale.
b. Middle Scale
lluch of the short term (up to a few hours) public expo-
sure to particulates is on this scale. People moving through downtown
areas, or living near major roadways, encounter particulates that would
be adequately characterized by observations of this spatial scale.
Thus, measurements of this type would he very appropriate to the evalua-
tion of some of the possible short term public health effects of parti-
culate pollution, but the air quality standards are based on 24- hour
and annual averages.
This class also includes the characteristic concentra-
tions 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 do not produce pollutants themselves but which
attract significant numbers of pollutant producers, such as autos.
Shopping centers, stadiums, and office buildings are examples of in-
direct sources. In the case of TSP, unpaved or seldom-swept parking
lots associated with such -attractions could he an important source in
addition to the vehicular emissions themselves.
This category also may be applicable to streets that are
several kilometers long, if sources and land use are reasonably homo-
geneous along the street, but inhomogeneous in other directions, such as
is the case with strip development, freeway corridors, or downtown
street canyons. We have included in this category special measurements
to characterize the TSP concentrations from emissions and road dust
along such traffic corridors. A site for monitoring the conditions in n
single street canyon, or a block of strip development, will be represen-
tative on a scale of tens of meters by hundreds of meters. If we try to
characterize street-side conditions throughout the downtov;n area, or
along an extended stretch o,f roadway, then the characteristic dimensions
may be tens of meters by kilometers.
15
-------
c. f'eighhorhood
'•'easnrenients 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. ! or.ogeneity refers to T'SP concentration, hut it probably also ap-
plies to land use ami land surface characteristics. 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. Stations of these kinds provide the
best information about health effects and compliance with regulations
because they represent conditions in areas where people commonly live
and work for periods comparable to those specified by the NAAQS. In the
sense used here, this category includes industrial and commercial neigh-
borhoods, as well as residential.
Neighborhood-scale data could provide valuable informa-
tion 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. The
neighborhood scale measurement could also be used for interneighborhood
comparisons within, or between, cities. This is the most likely scale
of measurement to meet the needs of planners.
d. Urban
This class of measurement would be made to typify the
particulate concentration over an entire metropolitan area. Such meas-
urements would be useful for assessing trends in citywide air quality
and, hence, the effectiveness of larger-scale air pollution control
strategies. Measurements that represent a citywide area would serves as
a valid basis for comparisons among different cities.
e. 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 homo-
geneity in the area. For this reason, the class of regional measure-
ments would be r.iost applicable to sparsely populated areas v;ith reason-
ably uniform ground cover, at least as it relates to surface dust en-
trainment. Data characteristic of this scale would provide information
about larger-scale, processes of particulate emissions, losses, and tran-
sport.
f. rational
Measurements that defined concentrations on this scale
would characterize the TSP .level of the nation as a whole and would pro-
vide trend data to allow assessment of national policies. Such data
night also be useful for studying international and global transport
processes.
16
-------
g. Global
Such neasurervnts vould provide information useful, to the
identification of worldwide trends in aerosol concentrations.
The abovo list categorizes measurements according to the spa-
tial scale to be represented. 'Virtually nil uses of TSP measurements
entail the characterization of concentration on one or rore of these
scales. Some scales would he 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 compo-
sites of measurements characterizing smaller areas.
2. Most Important Particle Sizes
It should be clear from discussions in Section II that dif-
ferent monitoring purposes may be satisfied by measurements of concen-
trations of particulates in different size ranges. Unfortunately, the
methodology most conunonly used for TSP measurements .does not discrim-
inate among different sizes—at least for those particles smaller than
about 100-ym diameter. Nevertheless, the requirement for some size
discrimination persists and must be satisfied to whatever extent is pos-
sible through proper siting.
Recognizing that it would be unrealistic to expect to achieve
fine resolution of particle size through siting criteria alone, we have
chosen to consider only two categories, "small", or "respirable" versus
"total." Nominally, "total" includes all those particles which the
standard Hi-Vol systera normally collects. The most obvious application
of this category is to meet the legal requirements associated with air
quality standards. Another possible application inigrit be to determine
potential effects of dust loadings on equipment or materials.
Above 5-ymdiameter, most of the inhaled particles are depo-
sited in the nose and throat (see Figure 3). Such particles also have
appreciable settling velocities and are subject to relatively efficient
removal by inertial processes, as noted before in this report. Finally,
as Figure 5 shows, the bulk of the material from natural and "quasi -
natural" dust-raising processes falls in this size range (Hidy et al.,
1975). Therefore, it seems appropriate to try to consider monitoring
purposes according to whether they are best served by measurements nomi-
nally limited to the "snail particles"—smaller than 5 ym—or by meas-
urements that cover the total size range—at least up to 100 ym.
Many of the purposes for Monitoring aerosol concentrations
would 'or better server' by measurements of the small, respirable frac-
tion ,.rather than of the total mass per unit volume. The most obvious
of these are the measurements made to assess health effects, but others
also fall in this category. For instance, the environmental effects of
17
-------
the small particles are apt to be wore profound than those of the large
particles; aerosol effects on visibility, attenuation of solar radiation
and cloud condensation processes are all more pronounced, iiciss for mass,
for particles in the smaller size range than Cor the larger particles.
Finally, if the hypothesis of Hidy et a.l. (1975) is true, and most of
the primary and secondary aerosols fror> anthropogenic sources are in the
smaller size range, then measurements made to evaluate the effectiveness
of controls on anthropogenic emissions would be best limited to the
smaller sizes.
Unless considerable care is taken during, the selection of
sites for TSP monitoring, the data may be overly influenced by the large
particles that are not the major threats to public, health and environ-
mental quality. Furthermore, data that are overly influenced by the.
large particles may not be as'useful for .the formulation of control
strategies as are data that are more representative of the concentra-
tions of small particles. While it nay not always be possible to over-
come the large particle effects by siting, recognition of the problem
should improve data interpretation and reduce the impact of the problem.
3. Source-Oriented Monitoring
Sometimes, 1ST is monitored to assess the effects of a large
source on its surroundings. If this source is at ground Level, then it
niay be possible to define the objective in terms of scale of representa-
tiveness. For instance, effects from large traffic sources will be ade-
quately assessed by measurements that describe the middle-scale traffic
corridor conditions.
Large, elevated particulate sources will usually have their
greatest impact at distances well removed from the source. Furthermore,
there will he differences between the areas of maximum long-term average
impact and those where shorter-term effects (in this case., 24-hour aver-
age) are felt. Figure (> shows the impacts of a hypothetical 160 in stack
calculated with the model described in Appendix A. The climatological
inputs, used were those from Tucson, Arizona. The units are relative,
since the numerical results.will depend on the source strength. For
purposes of monitoring, the important features to note in Figure 6 are
the differences between the location and relatively large extent of the
area of maximum long-term average concentration and the location and
comparatively small size of.the areas where very high, 1-hour concentra-
tions are most frequently observed. When locating a TSP monitoring sta-
tion to evaluate the impacts of a large, elevated point source, it in
important to determine which type of impact is more important, long-term
averages or most frequent high concentrations of sliert duration. The
usual averaging period for TSP sampling in 24 hours. The difference
between the location of the maximum 24-!iour average, nnd that of the
greatest annual average are not apt to be so pronounced as the differ-
ences in Figure 6, between the locations of the maximum annual average
and most frequent high 1-hr average, but niust still be considered in the
siting of monitors.
18
-------
40 30 20
10km
(a) LONG TERM AVERAGE IN ARBITRARY UNITS
STACK
10 km
(b) PERCENTAGE FREQUENCY OF CONCENTRATIONS EXCEEDING 1300 ARBITRARY UNITS
FIGURE 6 CALCULATED IMPACTS OF A HYPOTHETICAL I60 m STACK LOCATED
IN THE SOUTHWESTERN UNITED STATES
19
-------
C. F.elative Importance of the Different
Classifications of ?'on.i.toring Objectives
In the preceding sections, several different factors were discussed
relating to the classification of TSP-nonitorinr objectives. Uhen taken
in combination, a very large nmi.ber of categories could result—too many
to ruikt: feasible a special type of monitoring site for each. Therefore,
it is essential to e.staMisl1, a hierarchy of objectives and focus on only
the most important types. Such a ranking and selection process will
necessarily be subjective, but can nevertheless be rational if it is
based on some of the considerations already discussed.
Table 2 summarizes the different, categories of monitoring objec-
tives that are most important and assigns ranis to them. Only the mid-
dle, neighborhood, and regional scales of representativeness have been
considered. This arises in part because of the dominant importance we
attach to monitoring purposes that are related to the NAAOf..
The primary ambient NAAOS are defined as "those which, ... based on
the air quality criteria and allowing, for an adequate margin of safety
are requisite to protect the public health." (Federal Register, 1971).
The emphasis on public health gives the air quality standards great im-
portance. The emphasis on public exposure is further shown by the de-
finition of ambient air that is used in connection 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 am-
bient air quality standards as the major motivation for monitoring, then
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.
Most people are exposed to pollutants, at least over the 24-hour
periods, on neighborhood-scale and, to a lesser extent, middle-scale
This accounts for the evident importance of these scales in Table 2.
A mixture of neighborhood and middle scale exposures can be con-
sidered as respresentative of the urban environment as a whole, and
hence some attempt must be made to typify urbanwide concentration, but
it would be virtually impossible to represent the hodgepodge of neigh-
borhood and middle-scale concentrations with measurements at a single
site; no "urban-scale" measuring station can be described, in spite of
the acknowledged importance of this scale of measurement. Urban areas
must be characterized by networks of stations covering a range of condi-
tions within the area. For this reason, the urban scale has not been
included in Table 2.
20
-------
A lar^-c sej'.r.ient of tlio population dwells in areas that could be
characterized by regionally representative sites, and so this scale must
be included in the table. However, it is prol-nhly not as important as
the smaller scales because "regional" areas will he rore rural and ;',en-
erally have lower concentrations of people and pollutants, thereby being
less critical for consideration.
Microscale measurements for 'health-related purposes are not very
appropriate, because very few people remain within a rather small area
for the 24-hour period of the standard.. The numbers of people affected
by ISP exposures on the nicroscale are very small compared to other
categories. There is another reason for omitting objectives wet by ini-
croscale observations from consideration in Table 2. Our survey .of the
literature suggests that those raking microscale measurements, and even
cany middle scale measurements, have specialized requirenents that would
be difficult to generalize. The users are most often 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 of this
report—to provide guidelines for the siting of more routine monitoring
stations.
Table 2
CATEGORIES OF TSP MONITORING OBJECTIVES
RANKED ACCORDING TO IMPORTANCE
Class of Important Scales Ranking
Objective To be Represented
Determination of Middle or 3
the impact of specific Small area
sources or categories Neighborhood 2
of sources
Determination of Neighborhood 1
prevailing ambient
conditions Regional 6
Development and Middle 5
evaluation of
control strategies Neighborhood 4
21
-------
National- and global-sea] e measurements have not been incorporated
into Table 2 because they are representative of such large areas that
they do not relate to the exposure of the general public. Furthermore,
they are largely redundant. To some extent this is also true of the ur-
ban scale. If the middle- and neighborhood-scale measurements have ade-
quately characterized the city's streets, shopping centers and neighbor-
hoods, then all the necessary information is available for characteriz-
ing the city as a whole. Similarly, if the cities and regions are ade-
quately described, their descriptions can be synthesized into a national
description, and on to the global scale.
Basically, the ranking shown in Table 2 arose because we attached
the .greatest importance to the acquisition of data for defining health
hazards and for determining adherence to air quality standards. Thus,
only the middle, neighborhood, and regional scales of representativeness
are considered. Of these, three, the objectives related to the determi-
nation of prevailing ambient neighbornood conditions are judged rrore im-
portant than those related to the assessment of the impact of more
specific sources, because the problems associated with prevailing, am-
bient conditions are, inherently, more widespread than those associated
with the local impact of an individual source. The middle-scale moni-
toring of particulates is most likely to be associated with specific
sources, such as roads, dusty storage areas, and parking lots. Finally,
the measurement of neighborhood conditions was given precedence over re-
gional (generally nonurban) conditions, because as noted before, the
numbers of people exposed to neighborhood conditions are likely to be
higher, as are the concentrations of particulates.
P. Relating TSP Monitoring Purpose to the Appropriate Scale of
Measurement
Table 3 summarizes some of the more inportant applications of the
various kinds of monitoring that have been discussed to this point. The
purposes are not necessarily listed in order of importance. The X's In-
dicate important general applications; those in parentheses, less gen-
eral applications. The table is intended to serve as a guide to the
selection of site types that are appropriate to the listed monitoring
purposes. The principles enunciated in preceding sections should help
the reader pick site types appropriate to purposes that are not listed
specifically in the table. Procedures are given in the next section for
locating the various types of site. Certain features in Table 3 reflect
the rankings given in Table 2.
The entries for the neighborhood and regional scales indicate where
the smaller size particles are of greatest importance and where, particu-
lar care should be taken to avoid biasing the measurements with larger
particles. In general, riddle-scale sites cannot be oriented specifi-
cally toward smaller particles. It will not be possible, within the di-
mensions of the middle ;3cale, to be sufficiently removed from sources so
that larger particles can be removed by natural processes. There are
22
-------
only a few instances where total particulate monitoring is useful for
typifying neighborhood or regional TSP concentrations, because the irre-
gularities in the distribution of the larger particles and their lesser
impacts on health and the environment nal-'e emphasis on tlie smaller par-
ticles nore reasonable. The reasons behind the entries in Tahle 3 are
discussed individually below.
1. Determine Compliance with Air Duality Standards
The air quality standards stress public exposure over 24-hour
or annual periods and health, effects, BO the most important locations
for measurements related to oir quality standards will be in areas that
combine high concentration with high public exposure. The most likely
locations for such a combination are the neighborhoods where people
spend great periods of time. Thus, high density residential neighbor-
hoods are likely candidates for stations devoted to this purpose. In-
dustrial neighborhoods are. also prirre candidates for monitoring because
of the many dust sources often present in such areas. Similarly, the
possibility that large elevated sources may be causing violations of
standards is likely to provide impetus for source-oriented monitoring.
Many people spend at least some of their time in the downtown street
canyons or near roadways, so conditions in such areas must also be con-
sidered. Wherever possible, the emphasis will be on the smaller parti-
cles, but the parenthesized X's in Table 3 suggest the possibility that
the legal requirement for total particulate monitoring may sometimes
override the rational underpinnings of the established air quality stan-
dards.
2. Provide Information for the Preparation of 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 be-
fore the project is begun. Sometimes i.t will be necessary to rake spe-
cial middle-scale measurements for the purpose. For example, if the
project is a highway widening, measurements may be required in the area
where the widening is planned. Of course, the implementation of plans
that involve stack emissions should probably be prefaced by neighborhood
scale measurements in the areas where the greatest impacts are expected,
and perhaps special source-oriented measurements in the vicinity of any
similar sources that might already exist in the region. All these meas-
urements would usually be made to provide estimates of current condi-
tions against which calculations, and subsequent observations, of post-
project conditions could be compared.
3. Determine Impact oC Specific Sources
Except for elevated emissions, the impacts of a source --'inin-
ish at greater distances froi" the source. Thus, for roadway emissions,
or fugitive dust emissions, it is ivost appropriate to select n monitor-
ing site that will characterize, middle-scale conditions in the imrrei' in te
vicinity of the source. Hence, the different kinds of middle-scale mon-
itor are emphasized for this purpose. Of course, the effects of on
elevated source are best defined by a source-oriented nonitor.
23
-------
Table 3
IMPORTANT APPLICATIONS OF DIFFERENT TYPES OF TSP-MONITORING SITE
Monitoring Purpose
1. Determine compliance
with air quality
standards
2. Provide information
for preparation of
environmental impact
statements
3. Determine impact of
specific sources:
Roadways
Ground level area
sources
Elevated sources
4. Determine effects on
the environment :
Visibility
Attenuation of solar
radiation
Condensation processes
Middle Scale
Special Near
Roadway Sites
(X)+
(X)
X
Small
Area
(X)
(X)
X
Neighborhood
Small
Particle
X*
X
X
X
X
Total
Particle
(X)
Regional
Small
Particle
X
X
X
X
Total
Particle
(X)
Source
Oriented
X
X
X
-------
Table 3 (concluded)
Monitoring Purpose
5. Provide information for
better understanding of
processes affecting TSP
concentrations
6. Evaluate results of con-
trol measures
"Hot spots"
Urban conditions
Larger scale
7. Determine long-term trends
in TSP concentrations:
Urban
Rural
8. Evaluate effects on humans
9. Evaluate effects on
animals and plants
10. Assess representativeness
of existing sites
Middle Scale
Special Near
Roadway Sites
X
X
(X)
Small
Area
X
X
Neighborhood
Small
Particle
X
X
X
X
X
X
Total
Particle
X
Regional
Small
Particle
X
X
X
X
X
Total
Particle
(X)
X
Source
Oriented
X
X
X
X* = general applications
(X)+ = more specialized applications
-------
4. Determine Effects on the Fr.vi romnent
Most environmental effects, especially those, noted in Table 3,
occur on a fairly large scale and, therefore, neighborhood and regions]
monitoring are most important. For instance, visibility reduction is a
cumulative effect, taking place over distances of several ki.lor-vterf.;
the major atmospheric effects that arise fron solar attentuotion or
cloud formation processes also extend over distances of kilometers and
more. The emphasis on smaller particles has been discussed already—all
the effects noted are much more pronounced with numerous small particles
than with the same nass of large particles.
5. Provide Information for Better Understanding of the
Processes Affecting TSP Concentration
All categories of monitoring are likely to be useful for de-
fining processes affecting TSP concentration. Total participate moni-
toring, at all scales, can be useful to the understanding of the effec-
tiveness of processes that tend to remove the larger particles preferen-
tially. Such monitoring is also essential to the understanding of
processes that introduce large particles into the atmosphere and govern
their spatial distribution near streets and small area sources.
In the case of source oriented monitoring, interest in both
the total size spectrum and the smaller material will arise when one.
tries to understand the differences in behavior, and in resultant ground
level concentration patterns, between the large, rapidly settling ma-
terial and the smaller sizes with negligible settling.
When the interest is in longer-range transport processes, the
emphasis will be. on the smaller particle, sizes and the larger scales of
representativeness because only the smaller particles are subject to
such transport, and turbulent mixing over long distances will obscure
the smaller-scale features.
6. Evaluate results of Control Measures
Control measures range from local, small-scale actions, such
as street sweeping, to citywide measures, such as restrictions on types
of fuel to be used. For specific, smaller-scale control measures,
middle-scale monitor ing is most appropriate. The type of middle-scale
monitoring to be done will be governed by the control measure being
evaluated. Controls on a specific, elevated source v;ill require source
oriented monitoring, for evaluation, but the monitoring need not neces-
sarily be conducted over long periods; the effectiveness of the controls
can be established with a limited number of measurements. For evaluat-
ing citywide control measures, larger -scales of representativeness and
longer monitor ing, periods will be required.
7. Determine Long-tern Trends in TSP Concentration
Middle-scale measurements will, riot be as useful for this pur-
pose as neighborhood-scale 'measurements', because the middle scale sites
26
-------
are highly influenced by large, nm! often erratic, fluctuations in emis-
sions over a relatively small ar^a. \.'hen the emissions influencing a
location are averaged over a larger area, then the fluctuations in con-
centration wi.1.1 rnore 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 measure-
ments of smaller particles are likely to be most suitable for evaluating
long-term trends in TS" concentration. There nay be instances in rural
areas, when it is appropriate to consider the total participate loading
because of its relation .shin to the nature of the ground surface and ero-
sion processes.
8. Evaluate Effects on i'unans
If one were specially designing a monitoring program to evalu-
ate effects on humans, one probably would not use conventional. TSP moni-
toring methods, but would use techniques that would provide separate in-
formation about the respirable and nonrespirable fractions and about
specific chemical constituents. However, if forced to use existing
data, the best choices would be those data collected at sites where peo-
ple were most frequently found and where the larger particles were not
likely to dominate the samples.
9. Evaluate Effects on Animals and Plants
The same arguments apply to the study of particulato effects
on plants and animals as to the effects on humans, but in the case of
plants respirability is not a factor. The larger particles are more
subject to deposition on the foliage than are the smaller materials.
10. Assess Representativeness of Existing Sites
Table 3 indicates that only neighborhood- and regional-scale
sites that emphasize the smaller particles are suitable for assessing
the representativeness of similar, existing sites. This reflects the
difficulty in defining and measuring representativness for a parameter
with great, small-scale spatial variability. In such cases, techniques
other than conventional TSP monitor in;;—such as mobile nepheloriet ry
(see, e.g., Charlson et al. , !%'.>), or widespread use of .snail membrane
filters, or lidar (laser radar) probing—right provide the necessary in-
formation .
Table 3, relating general monitoring purposes to the various
scales and particle sixes to be characterized, supports the rankings
given earlier in Table 2. Most purposes are best served by neighborhood
measurements, especially those emphasizing the small particles. The
following sections, therefore, emphasize the selection of sites that
will represent neighborhood conditions, especially the concent rot ions of
the smaller particulntes. Regional scale and source oriented monitoring,
also warrant considerable attention. The monitoring of middle-scale
conditions trust also he discussed, although this scale is not as con-
sistent with the averaging periods specified for assessing compliance
with the ilAAOS.
• 21
-------
TMI; STATION LOCATION
The purpose of this section is to provide step-by-step procedures
Cor locating monitoring stations that will, serve the most important
measurement purposes. For the more general purposes, this means identi-
fying a site'that is 'characteristic of the appropriate spatia1 scale and
which, for neighborhood- or regional-scale measurements, does not
overemphasize the larger particles. Tor source-specific monitoring, it
requires the itientif icat ion of a location where the source is likely to
have its greatest impact. The major steps of the site-selection pro-
cess are presented in flowcharts, with accompanying discussions. The
reasoning behind the recommendations made in this section is deferred to
Section V so that the procedures themselves can be presented more con-
cisely.
A. Regional Stations
Figure 7 shows schematically the site-selection process for a
regional-type TSP monitor. The form of the flow chart reflects the
differences between selecting a single site and a set of sites. Differ-
ences between procedures for selecting sites to define background TSP
concentrations for a specific city, and those for selecting regional
stations for more general purposes are also evident in the chart.
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.
Three basic kinds of information are required: geographical, emissions
and cliinatological. The geographical material is used to determine the
distribution of natural features—forests, rivers, lakes—and the works
of man. Useful sources of such information may include: road and topo-
graphical maps, aerial photographs, and even satellite photographs, par-
ticularly those from the LANDSAT (U.S. Geological Survey, 1974). The
site-selection process is one that winnows out unsuitable regions and
proceeds to fewer and smaller areas, so the photographs and maps will
generally be used in an order that procee'ds from those that depict very
large area?, such as shown in LANDSAT photographs like Figure 8, to
those that show smaller areas in greater detail, like the aerial photo-
graph in Figure 9.
The climatologlca1 summaries of greatest use are the frequency dis-
tributions of wind speed and direction. This information will usually
come in one of two forms. One of these is a tabulated joint freoucncy
distribution like that shown in Table 4, an example of material that is
available fror the National Climatic Center. The wind rose is an easily
interpreted graphical presentation of the directional frequencies. Ex-
amples of wind roses are shown in Figure 10, fror, the National Climatic
Atlas (National. Oceanic anri Atmospheric Administration 19(i<°). Other
29
-------
Acquire and make preliminary analyses
of necessary background material:
Regional maps and aerial photo-
graphs showing:
Regional maps and aerial
photographs showing:
• Topography
• Settlements
• Major industries
• Highways
Emissions inventory data
Climatological data.
Is station to define the background
TSP concentrat ion for a specific c ity
or more general regional TSP
concentrations?
Background for speci fie c it
Regional concentrations
Will more than one monitoring
station be available for this
Select tentative siting areas
that are downwind of the
nearest urban area for the
least frequently occurring
wind directions.
( One station . J
(Two or more "N
Stations J
Should the influence of large
particles be deemphasized?
Select tentative siting areas
that are downwind fur "the
least frequently occurring
wind direction.
Select tentative siting areas
that are upwind for the most
frequent wind directions.
K 1 iminate unsiti tab U* sit ing areas .
Suitable areas vi11:
• lie more than ten km from the
nearest urban nren
• Be more than one km from the
nea rest maj or roadway
• Be relat ive1y uninf1uenced by
major point sources (see text)
• Have reasonably Uniterm surface
characteristics
• He we 1 1 away from dust sources.
Eliminate unsuitable siting areas.
Suitable areas will:
• Be more than ten km from the
nearest urban area
• Be more than one km from the
nearest major roadway
• Be relatively uninfluenced by
major point sources
• Be in an area whose surface
characteristics and usage is
typical of the region
• Be well away from atypical dust
sources.
Eliminate specific sites in heavily
vegetat^ areas, within about 150 m,
of ;i surfaced road or four m of an
unpaved road with traffTc groater than
a few hundred vehicles per day. Avoid
sitos near inn jor topographical
obstacles.
Locate monitor at a height of about
2 - 15 m, well away from buildings
or other obstacles.
FIGURE 7 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING REGIONAL
MONITORING STATIONS
30
-------
SOURCE: USGS INF-74-22 (R.1).
SA-3515-5
FIGURE 8 A SATELLITE (LANDSAT) PHOTOGRAPH AND A MOSAIC OF SUCH PHOTOGRAPHS
-------
N— 200 m —»-l
SOURCE: Dabberdt and Davis, 1974
FIGURE 9 AERIAL PHOTOGRAPH OF A RURAL AREA
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.
Emissions inventories will be most useful for identifying large
point sources of particulate that night otherwise be overlooked. The
Environmental Protection Agency has a computerized National Emissions
Data System (NEDS; see e.g., Bosch, 1975). All large point sources of a
given pollutant, in this case TSP, greater than some specified source
strength can be retrieved for a given area. Obviously, a list of large
sources will be useful in the siting process so that proximity to large
sources can be avoided. Figure 11 illustrates the kind of information
provided by NEDS for each identified source. Countywide summaries of
emissions from all stationary and mobile sources are also available from
the NEDS inventory (Bosch, 1975).
After the background material has been assembled, a decision must
be made regarding whether or not to use the monitor to supply informa-
tion on TSP concentrations that enter a specific city. For selecting a
purely regional site, maps or aerial photographs should be used to iden-
tify general areas that are suitable for locating such stations. Uni-
form topography and surface characteristics are desirable.
32
-------
Table 4
EXAMPLE OF A TABULATED WIND SUMMARY
PERCENTAGE FREQUENCIES
OF WIND DIRECTION AND SPEED:
WBKTION
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 HOUf.
0 - 3
4-
+
4-
+
4-
4-
•f
4-
1
•f
4-
•f
4-
•f
4-
4-
4-
4
4 . 7
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
13
8 12
2
2
3
2
1
1
2
2
2
1
1
3
3
-2
2
1
30
13 16
1
1
2
1
4-
•f
1
1
3
3
3
4
5
3
2
1
30
19 . J4
4-
4-
•f
•f
4-
4-
+
2
2
2
4
3
1
1
4-
17
35 31
*
4-
4-
4-
+
1
1
1
1
+
•f
+
•f
5
3} 38
+
•f
•«•
4-
4-
4-
1
3V • 46
+
4-
4-
4-
47
OVH
4-
4-
TOTAl
4
4
6
4
3
2
4
4
10
8
8
14
12
7
6
4
4-
100
AV
tfKD
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.
The cliraatolopical information should be incorporated into the
selection process to determine those areas that will be least frequently
downwind of the closest major urban areas. If possible, the prospective
siting areas should be at least 10 ki" fron the outer lirits of that ur-
ban area as illustrated in Figure 12. It also illustrates that the sit-
ing area should not be within about 1 km of the nearest ir.ajcr intercity
freeway and about 1/2 kn fron other major roans.
The requirements for purely regional monitoring and for rnoni tor ing
background concentrations entering, urban areas are nuch the. sane. Thcv
are identical when only one background site is planned. If nore than
one upwind urban background site is planned, then the potential siting
areas should be chosen to be upwind for the two r.ost frequent
directions. Figure 1 3a illustrates this schcrat i ca l.ly. If the two
frequent wind directions are within ')(.) der.rees .of each. other (T'ipure
13b), then the most frequent direction '..'ill be used, ;.ilonf with the most
cornxion direction fron: ainonj-, .ill those that are nore titan 90 decrees dif-
ferent from it. Fij'.ure 13b illustrates this. ."o noted before, any pro-
33
-------
u>
-p-
SURFACE WIND ROSES, ANNUAL\
WIND ROSES SHOW HEHLENTMjE
OF TIME WIND BLEW PROM THE
16 COMPASS POINTS OR WAS C
* INDICATES LESS THAN 0.5% CALM
i 25 HOURLY PCMCCNTAOCS 25
• —-
SOURCE: National Oceanographic and Atmospheric Administration, 1968
SA-6600-9
FIGURE 10 EXAMPLE OF WIND ROSES
-------
FILE CPEATEC CN KtCNESCAY CdCtMBtf
S, 1S73
NATIONAL t
Pa INT
* I S S I C
SOURCE
ST«Ti(l
<•••••••*•*•*•
^CT SPECIFIC
COMPLliNCi: ST«TUS
LPCAIi: / /
CM.CRG..U _Y (. >MK :L
ACTION FLAN
STATUS UNNNCnN
UTM
CCCRLINATtS
*************
UTM 26Nc: 15
HJKHGNTAL: 7ie.7 HP
V^KT KAL: 4. 3C5.4 KH
ST«CK
250 FT
15.5 FT
32^ F
STACK HCIGhT:
ST*CK OIAME1EK:
GAS .tHPtRATUFc:
G-.S rLCln RATc:
PLO^t HT (NC STACK) : 0 FT
SAHC STACK VENTS FCINTS 01-031
(.CNTPCL OtV ICfc/MfcTHOD IDtNT I F ICAT ION
S':CCNC. PART: NO CCNT5CL
PR IWAHY 'S'JX: NG CGNTCuL ;CUIPMENf
StCCNO. SCx: NC CCNTfiCL EQUIPMENT
PHlf'ASY NCX: NC CUNTKOL tULIPMI.NT
SiCCN3. NCX: NC CCNTPCL
P6!HA(-Y KC: NC CONTROL
SCCONCJ. HC: NO CCNTPCL !-CU|P^t^T
CC: NC CCNTRCL c<
S-T.JND. CC: NC CCNTC'L -C
FUEL CHARACJEFISTICS
r-D. L SULFL* f.GNTcNT: C.OC <
FUSL AJH CONTENT : CO.O %
HAND CALCULATED tMISSION ESllNATcS
PARTICULATE: 11,200 TONS/YP
sox: e.CAC IONS/YP
NOX: 2,650 TCNS/YR
HC : 156 TONS/YR
CC: Ti T'JHS/YS
EMISSION ESTIMATION MJ-THOC
PART: EMISSION FACTOPIAP-'
SOX: : MISSION FACT3a i ,950 TCNS/YP
NOXI 2,500 TONS/Yf-
HC: 0 rQNS/Yf-
CCil TCNS/YF
COPUTf-P CALCULATED EMISSIONS
PART: <.3 TONS/YK
SOX: 2 TONS/Yt-
NOX: 2,050 TCNS/YR
HCJ 3 TCNS/Yt.
ANNLAL CPl-l-ATINO R A T •; :
JLPLY *tXP CEiIGN "t~i-.
EC1LEF CtSIGN CAPAC ITY:
6,820 MLLICN CUBIC FE£T
.579 MILLICN CUBIC FE<"T
bOO MILLION BTU/HF
CCNTENf: 1.C35 KILLUN ETU/MILLIjI. CL8!C
DLF NS-0
FIGURE 15 AN EXAMPLE OF OUTPUT FROM THE NATIONAL CLIMATIC CENTER'S STAR PROGRAM
-------
u>
URBAN
AREA
PROSPECTIVE
SITING AREA
LEAST FREQUENT
WIND DIRECTION
1 km
FROM HIGHWAY
(4 km FROM UNPAVED PUBLIC ROADS
SA-3515-7
FIGURE 12 SCHEMATIC DIAGRAM OF APPROPRIATE AREAS FOR A REGIONAL MONITORING SITE
-------
NEXT MOST FREQUENT
WIND DIRECTION
MOST FREQUENT
WIND DIRECTION
(a)
MOST FREQUENT
WIND DIRECTION
SECOND
MOST FREQUENT
WIND DIRECTION
SECOND WIND DIRECTION
FOR LOCATING SITING AREA
PROSPECTIVE
SITING AREA
(b)
SA-3515-8
FIGURE 13 SCHEMATIC DIAGRAMS OF APPROPRIATE SITING AREAS FOR REGIONAL
MONITORS WHEN TWO SITES ARE PLANNED
37
-------
spective siLinj', area should be more than 1 km fror; the nearest major in-
tercity arterial roa.:1. The emissions survey should be used to ensure
that no n.ajor point sources are near enough that they will unduly influ-
ence the observer' concentrations. The identification of areas of
greatest point source ir.pncf is discussed later, in connection with the
siting of source oriented noni.tcrs.
The final part of the site, selection process is the elimination of
unsuitable locations within the suitable regions. "o surfaced roac! with
traffic of more than a few hundred cars per day should be within AD or
50 n of the. site. Tf at all possible, no unsurfaced public road should
be within 4 k;n of the site. It is quite important that the sampler be
located away fron; oSstacl.es such as buildings, large trees, and so
forth. The distance between obstacles and the sampler should be at least
twice the height of the obstacle.
The height of the bigh-vobmc sar-pler above ground should he. 2 to
15 r;:. The underlying, surface should be as free of dust as possible. Low
ground cover is preferable to pavement, or other impermeable surfaces.
Sites with ground cover 'i;ay be unsuitable during seasons of pollination
or other botanical dust genera tin;.- processes. On the other hand, if the
[•round cover is typical of the area, then the measurements will still be
representative, even if they may be misleading to anyone thinl:ing in
terms of anthropogenic causes.
A TSP-monitoring site can also be used for measuring other environ-
mental factors that are related to the air pollution problem. In par-
ticular, winds are quite important, and they should be nieasured at a
height of about 1C m above the. general level of the surrounding surface.
!• . Keighborhood Stations
Figure 14 is schematic diagram of the procedures used to select a
neighborhood monitoring station. As with a regional station, the first
step of the selection process for neighborhood stations is the acquisi-
tion of necessary background materials. Climatological information will
be required, especially a joint frequency' distribution of winds and at-
mospheric stability. This is provided by the output of the National
Climatic Center's "STA?" program.
The National Climatic Center is located in Asheville, 'iorth Caroli-
na. Figure 15 provides an example of the type of infornation generated
by the STAR program. Relative frequencies of occurrence such as those
shown in Figure 15 for Pasquill's (lrK>l) neutral stability, are avail-
able for each of the other .stability classes. The results given in Fig-
ure 15 are derived fro;- si.-ri ng ,t ire O'nrch, April, ''ay) data, but similar
information can be produced for inonthly or annual frequency distribu-
tions. The National Climatic Center alsc provides the output on punched
cards suitable for use wi th the Clinato.loeical dispersion !';odel (CDM;
"i'sse and ','. inreriran, 1973).
38
-------
Background Informati
mi use maps
ria1 photographs
te I 1 ite photographs
issions Inventories
•at"fie maps
imaLuljgical suimiiarLc
eighb
nflue
i'hood that is heavi 1 v
ced hy a single soure
n
No, the site is to typify
neighborhood-scale conditions
uninfluenced by any major
individual sources if possible
Yes, the situ is to evaIn
the neighborhood seale
effects of an individual
Use procedures for locating
source oriented site
Is the Lype of neighborhood to be
represented commere ia 1 or industria 1 ,
and if so, would a street canyon,
traffic corridor or sma 11 area source
type site be better?
Yes, one of the middle scale
sites would he more
appropriate
\
No, this is a true
neighborhood type site
Use the procedure for
selecting the correct
middle-scale site
Identify neighborhoods of the desired
tvpe on a map
Use simple c 1 imatological -type
diffusion model to estimate relative
distribution of TSP concentrations .
Plot results on map with identified
neighborhoods
Conduct optional small filter samp ling
program, or nephclometry program to
improve the model estimates
T
Is the site to be typical of the
particular type of neighborhood o
should it represent the worst cas
Should be typical
neighborhood type
of this
Should be the worst of this
neighborhood type
Identify neighborhoods with
modeled, or measured TSP
concentrations nearest the
average for neighborhoods of
this type
Identify neighborhoods of
this type with the highest
modeled, or measured TSP
concentration
\
Find areas within identified neighbor-
hoods where horizontal gradients are
small and the effects of individual
sources is minimal
Locate Hi-volume sampler at least 20 m
from any street and at least 400 m from
roadways with average daily traffic
greater than 50,000. The sampler
should be at a height of 2 - 15 ra, at a
distance from any obstacle that is at
least two times the height of that
obstacle. The ground surface should
have ground cover or pavement. If
pavement, it should be kept free of
loose dust^
FIGURE 14 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING NEIGHBORHOOD
MONITORING STATIONS
39
-------
SEA:
MAM
RELATIVE FREQUENCY DISTRIBUTION
STATIDN -13882 CHATTANOOGA, TN 24 0»S 1960-6*
SPEFDIKT*)
QJRECTION
N
NNE
NE
ENE
E
ESE
SE
SSE
-> S
O
ssw
sw
wsw
w
WNW
NW
NNW
TOTAL
RELATIVE
RELATIVE
0-3
0.000912
0.000800
0.000731
0.000385
0.000069
0.0002*5
0.000919
0.0014*9
0.001386
0.000655
0.000696
0.000280
0.000160
0.000135
0.000*2?.
0.0005*7
0.009789
FREQUENCY OF
FREQUENCY OF
4 - 6
0.002719
0.00*169
0.002900
0.001813
0.001269
O.OOO997
0.001088
0.003807
0.006163
0.003263
0.002266
0.001631
0.001178
0.000725
0.000725
0.001269
0.03*983
OCCURRENCE OF 0
CALMS DISTRIBUTED
(NEUTRAL
7-10
0.00*079
0.0068B8
0.002175
0.00063*
0.000*53
0.0005**
0.001178
0.0053*8
0.017765
0.006526
0.003625
0.002175
0.001994
0.003807
0.001269
0.003988
0.062**9
STABILITY
ABOVE WITH
/ DAY)
11-16
0.00761*
0.0i06f)5
0.0035?5
0.0005**
0.000091
0.000091
C. 000*53
0.006707
0.02619*
0.01**H
0.00*713
0.0066i7
0.005962
0.010333
0.003529
0.009*26
0.1128*3
= 0.2*7530
D STABILITY
17 - 21
0.000725
0.00063*
0.000181
0.000000
0.000000
0.000000
0.000091
0.000997
0.00*260
0.001903
0.000725
0.001722
0.001088
0.003625
0.002538
0.002357
0.0208*7
= 0,002357
GREATER THAN 21
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000000
0.000272
0.000363
0.000181
0.000091
0.000997
0.000906
0.001269
0.001178
0.000363
0.005620
TOTAL
0.016048
0.023096
r,. 009523
0.003376
0.001882
0.001877
0.003729
0.018579
0.056131
0.0269*0
0.012117
0,013*22
0.011308
0.019894
0.0) 1661
0.017949
FIGURE 11 EXAMPLE OF A NEDS OUTPUT FOR A POINT SOURCE
-------
Aerial photographs and topographic raps will ho quite useful to
determine the. distribution of Land use through the city. Special naps,
showing the details of land use, are sonetii les available for study in
planning agency offices. Figure If; shows a section from a nap of this
type produced by the Sanborn Map Company. *
Another valuable type of nap is the traffic r.iap. Figure 17 is an
example of this kind of data. Traffic data are available in several
forms, but the. information itself is usually limited to the average
numbers of cars per day on a given section of roadway. Ludwig and
Kealoha (1975) have presented a brief summary of the usual types and
forms of traffic data that are available. They also discuss sources of
other kinds of data, e.g., land use, clinatologlca], and so forth.
Emissions inventories are another important type of background in-
formation that must be assembled during the initial stage of the site.
selection process. The NEDS data have already been cited (see Figure 11)
and they provide valuable information concerning the larger point
sources, but the spatial resolution for area sources is not usually any
finer than county size. The area source emissions rcust be distributed
into snaller areas where possible. Sometimes this must be done rather
crudely—for example, on the basis of population or housing units. Popu-
lation and housing data are available for the census tracts within 241
Standard Metropolitan Statistical Areas (SMSA)o Figure IP, from one of
these Eureau of the Census (L972) documents, shows the size of some
tracts. In general, the tracts are smaller i.n areas of dense population
than in less densely populated regions. The Office of Air Quality Plan-
ning and Standards of EPA ** has prepared computer programs that will
apportion area source emissions fron the NTHS into gridded areas accord-
ing to population.
Before proceeding with the site selection process, it is necessary
to answer two questions. First, is the site being established to deter-
mine the impact of a specific major source or the neighborhood? If so,
then the procedures used for locating, source-oriented sites would be
more suitable. The procedures for selecting source-oriented sites are
discussed in a later section of this report. The second question relates
to scale. If the purpose does not require characterization of conditions
throughout a neighborhood, then the procedures for selecting middle scale
sites would be more appropriate.
* Sanborn Map Company, 629 5th Avenue, Pelham, New York 10803
** Mr. J. Mersch, EPA, Research Triangle Park, N.C., personal
communication.
41
-------
COUCH'S A DON
\
. ..«-"r£
^
51
i *
. ^
i
i
! ^
I l!
I ";
(4
f'f"
i
1
3
V
f' '
1 I
%-- N. W. HOYT
." •' ' "' ~"j
'— '"- ': ,f I54g3
^~™wc*f^
«-•"•-"• — *•;
"—
"1 '
T'i
r'. 1
366)
e
T3J-1
- .••»•» a.'i
!&N -:
M. y
,-.-''
ft/C **oro»
*f*0
** • ' ' "'•K-*'
"j'.r,'-s*T>?..j !' J"
S ,'
p
//>P GBfiSLfc CO
in?3'T-
'f387>
4,' -
, .'."'.'"v
riit/. £-ov/«s7v«^'
»,. ,,n »«J- •••
---,* - -*.**.'.1...N W FLANDERS V
@
" >J«
i
'I
1
S
i?
e.
?
•V.H
® *>
---'-"•
i
^^Vra^,
^•^-;
>— j--.1 -• ^
S. ' --, o ,
•SL,-^
^frr'ltsl
u — >f •-** -+
j^^-^-U.
d
l\ ' : '•
Sj --i»i_ :
,
*Aff* IT6tf>^ t
1 ("'"'f*,t' •*
e •' ' ..,.,,-.;.: •.
,7.« ^«:5^ r-iS1^-;
— i.: *. -'^ J- (S
O
! 44 j
« ' ' ::
'"V . :
: • •') i:
:: * ,
: IOD
(39|J
(>-
1 ! ; *uj
j -J_l
'"' .
2
D
S
? f
< . 'ul
joi
r (390
» J,
<£> ;
a^
. . 1
• |
"'3.' '
1 ' [ ,
•1
! i
/ :
^ j
102
. / t3^
' /
.1
';• *' — : l
y!) N.W. EVERE
"-1 ! " &
i
||
[
/ i ! ',
i i i
si
• •'•
!
; i
', : !
i
1 i
< i
"'i
j
''
1 ', '
>
i \ *
l_ . ... ._. ...L^ ,
FT ®
r. - =p,
!
|
1
y :-
I
.?
fi;
V.r
c
3
:,
n
C
F
'<7"*v *
•
A
#'•"
|
• ©
•. ;? N. w.
T~;-" v"
'eiiaf-tot i**tstu Co
».j?l.ZC.. — ,
(-
I \(>r&
\ [W
't: ; • -'•-*"
•-•• ': i •' ! ' •--•
., ' :
t * •
!i ^
< :
; i
t*
*!
ti«
*i®
. . j
• .»
5
P
1 3
5 S
i z;
3
.!®
i!
X
i
f
\ i
i S
»• •<
S iff'
>
54
fSAV. .-/-.-UO. ..»..., •
^@.
FIGURE 16 A SANBORN MAP FOR A SECTION OF PORTLAND, OREGON
42
-------
•••• PROJECT BOUNDARY
SA-3515-18
FIGURE 17 SAMPLE TRAFFIC MAP FOR A DOWNTOWN AREA
After acquiring the necessary background materials and deciding
that a neighborhood scale site is appropriate, the next step is to
use the maps and arterial photographs to identify neighborhoods of
the type that are of concern to the particular monitoring objective.
Figure 19 shows aerial photographs of two different kinds of residential
neighborhood. Figure 19a shows an area that was built more recently
than that shown in Figure 19b, with its more mature trees and shrubbery.
Figure 20 shows a section of a topographic map corresponding to Figure
19b.
43 .
-------
Soginow Twp.
103
Xf/, = •.•.»! luCn "0
"^ 1
^S'-, • *
1
1
L
Corrollton Twp
107
*•
a*
104
105
»\
H
21
14
106
109
110
1
1
Source: Bureau of the Census, 1972
FIGURE 18 CENSUS TRACTS IN THE SAGINAW, MICHIGAN SMSA
-------
After potentially suitable TSP-monitoring neighborhoods have been
identified on a map, a simple numerical simulation model should be ap-
plied to provide an estimate of annual or seasonal average TSP concen-
tration throughout the city. Several suitable models are available.
Probably the climatological dispersion model (COM; 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 point source
and pridded area source emissions. Although the use of the computer
model is most desirable, the gridded emissions inventory could be used
to identify neighborhoods where the distributed sources have their
greatest impact because concentrations are generally closely related to
nearby ground level emissions. Areas strongly influenced by large
elevated sources will be exceptions to this rule. The model given in
Appendix A of this report could be used to identify the areas of
greatest influence for the largest sources. The graphs and methods
given by Turner (1969) could be applied for the more frequently occur-
ring combinations of wind and stability if the facilities for using the
model in Appendix A are not available.
If time and resources are available, a limited sampling program
would be highly desirable to provide better data on the relative concen-
trations in the various candidate neighborhoods. Studies in connection
with the National Air Sampling Network (U.S. Public Health Service,
1958) concluded that biweekly (alternate weeks) sampling, randor'ized
with respect to day of the week, would define median TSP concentrations
sufficiently we.11.
Thus, a limited sampling program would be justified. Furthermore,
it might be sufficient to use nonstandard sampling methods for this pur-
pose. Tor example, smaller pumps with smaller filters would be more
portable and easier to use for preliminary studies. Another approach to
the problem of a preliminary particulate survey might employ a nephelor.—
eter, moved from one location to another during the sampling period.
The nephelometer measures light scattering which is related to particu-
late concentration (Charlson et al., 1969).
Once the spatial distribution of TSP concentration has been es-
timated from modeling, a preliminary survey, or from a combination of
the two methods, then it should be superimposed on the map showing the
locations of the candidate neighborhoods. If the objective of the moni-
toring is to obtain data for the particular neighborhood that is likely
to have the highest average TSP concentrations among all the neighbor-
hoods of the chosen type, the measured or modeled spatial distribution
will serve to identify the appropriate area. Similarly, the distribu-
tion of average TSP concentration can be used to judge which of the can-
didate neighborhoods is nearest the average for all of them.
45
-------
(a) RECENT RESIDENTIAL
(b) OLDER NEIGHBORHOOD WITH MATURE TREES
SOURCE: Dabberdt and Davis, 1974
FIGURE 19 AERIAL PHOTOGRAPHS OF URBAN
RESIDENTIAL NEIGHBORHOODS
46
-------
APPROXIMATE AREA SHOWN"
IN FIGURE 12(b)
FIGURE 20 A TYPICAL URBAN NEIGHBORHOOD DEPICTED ON A TOPOGRAPHICAL MAP
47
-------
The above discussion represents rather minimal preparatory efforts
in the site selection process. Obviously, modeling is subject to con-
siderable uncertainty, which can be reduced by a limited sampling pro-
gram. Even the limited sampling program provides only estimates of
average concentrations, and averages do not allow as complete a compari-
son among sites as would be desirable. A more extensive sampling pro-
gram could determine more reliable frequency distributions, which would
he better bases for comparison.
Once the neighborhoods that most nearly meet the monitoring objec-
tive have been identified, then pore specific locations must be select-
ed. These locations must be at least 400 m from the nearest highway
carrying 50,000 or more vehicles per day. This spacing should li-mit the
contribution of the roadway to less than about 7 or 3 .ug m~3 . Lower
traffic, volumes will allow for closer spacing. An example of the calcu-
lation of a roadway's contribution is presented in Section IV. The in-
let must be at least 100 m from any local street having daily traffic of
about 20,000 vehicles. The sampler should be at least 20 m from neigh-
borhood streets with traffic of 2000 vehicles per day or less. A limit-
ed sampling program could be undertaken to define gradients (especially
horizontal) in the area. A nephelometer would be particularly appropri-
ate for this purpose. If the gradients are very strong, then a location
more distant froir. the nearest street might be required. Conversely,
weak gradients would allow closer placement. The Hi-Vol sampler should
be at a height of 2 to 15 n.
The sampler should be located well away from obstructions; in gen-
eral, the distance between the. sampler and the nearest large obstruction
should be about twice the height of the obstruction. Unless the sampler
is supposed to collect large particles that might be typical of the
neighborhood, the sampler should be located in an area with a ground
cover that will minimize the incidence of blowing dust. If samples are
to be collected that represent the full size range, then the site
selected should be surrounded by a surface that is typical of the neigh-
borhood as a whole. If the surrounding area is paved, and the emphasis
is on small particles, then it should be cleaned often enough to prevent
the settled large particles from biasing the TSP samples. In any event,
the area directly beneath the l!i-Vol sampler should be. such that the ex-
hausted air from the sampler itself does not raise spurious dust. Fig-
ure 21 shows a TSF sampler located in a small park within a residential
neighborhood.
Some auxiliary measurements, such as wind speed and direction, are
desirable for a neighborhood station. 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
nearby streets and the wind measurements could both provide valuable in-
formation for the interpretation of the measured TSP concentrations. For
example, the causes of anomalously high values might be identified with
unusual traffic conditions or with entrained dust during periods of high
wind.
48
-------
FIGURE 21 HIGH VOLUME SAMPLER LOCATION IN RESIDENTIAL NEIGHBORHOOD PARK
49
-------
C. Source Oriented Stations
Figure 22 is a schematic diagram of a procedure for locating a mon-
itoring site that will provide information about the effects of a
specific elevated point source of particulates. We refer to this type
of source when we speak of source-oriented monitoring in this report.
Monitoring the effects of ground-level sources—highways and fugitive
dust emissions—is addressed in connection with middle-scale stations in
the next section.
Uith reference to Figure 22, it shows that the site selection pro-
cess for a source-oriented monitor begins with the familiar admonition
to assemble the necessary background information; in this case, it is
climatological summaries like the STAR program output, maps of the area,
emissions inventories and the characteristics of the source, of interest.
The characteristics of the. source that are of particular interest in-
clude stack height, gas temperature and flow rates. Most such informa-
tion is available from the National Emissions Data System (KEDS), as
shown in Figure 11, but it might be useful to verify these facts in-
dependently and to find out the extent to which the emissions might,
from time to tinie, deviate from the average values.
It was noted earlier that the impact of an elevated point source
can be defined in different ways and that the areas where the impact is
greatest will differ according to the definition that is used. In gen-
eral, the highest short-term average concentrations are riost likely to
occur nearer the stack than the areas of greatest long-term average con-
centration. A decision has to be made between these two kinds of im-
pact. A simple computer model like that described in Appendix A can be
used to identify either kind of area if the climatological factors and
the stack characteristics are known.
Seasonable estimates can also be made without the computer. The
maximum long-tern impact is most likely to be associated with the most
frequent combination of wind direction, speed, and stability, as deter-
mined from the STAR program output. Obviously, if this is the case the
downwind direction for the most frequent combination is the most likely
direction for maximum long-term impact. The most likely distance can be
estimated by calculating the plume rise for the stability and wind speed
of concern. The wind speed at the top of the stack is likely to differ
from that measured at a lower height, typically 10 m. The speed should
be corrected to stack height; Eeals (1971) gives information for such
corrections, and they are summarized in Figure A-2 of Appendix A.
Rriggs (1973) gives equations for the determination of plume rise. Once
the height of the plume is known for these conditions, then the distance
to the maximum concentrations for that case can be estimated from graphs
given by Turner (1969). Figures 23 and 24 show two of these graphs.
The ordinate in these graphs is a concentration normalized for emission
rate, Q, and wind speed, u.
50
-------
ASSEMBLE BACKGROUND INFORMATION
• Source characteristics
• Climatological summaries
• Maps
• Emissions inventories
Is the impact of the source to be
determined where it has the greatest
effect on long term average TSP
concentrations or where the occurrence
of very high, short-term concentrations
is most likely?
At location of greatest long
term effects
Identify areas where long
term average concentrations
are greatest:
• Computerized models
• E s t ima t e s f r om
climatological
summaries
At location of most frequent,
very high, shcrt-term
concentrations
Identify areas where highest
concentrations are most
likely to occur:
• Computerized models
• .Estimates from
Climatological
summaries
Are there apt to be appreciable
contributions to the TSP concentrations
from sources other than the one of
interest?
Yes, appreciable contributions
from other sources
No appreciable background TSP
concentrations
Select neighborhood or
regional type site in the
direction that is least
frequently downwind of the
source .... to be used to
define "background"
concentration to which the
point source concentrations
are added
Locate source oriented sampler near
center of area of greatest impact at a
height of 2-15 m, at least 20 m from
the nearest street and 400 m from
roadways with ADT of 50,000 or more.
The sampler should be well removed
from any obstacles (about two obstacle
heights or more). The surroundings
should have ground cover or pavement
to prevent surface dust from affecting
the measurements
FIGURE 22 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING A MONITORING
SITE TO ASSESS THE EFFECTS OF A LARGE, ELEVATED POINT SOURCE
51
-------
2
Q
UJ
LU
> 10"
DC
I-
z
UJ
o
o
CJ
O.I
100
DISTANCE - km
Source: Turner, 1969
FIGURE 23 NORMALIZED GROUND-LEVEL CONCENTRATION UNDER EXTREMELY
UNSTABLE CONDITIONS
-------
The actual ground level concentration is of less importance to the
siting problem than the location of the r.aximur.; concentration. The fig-
ures show that the concentration rises rapidly to a maximum and then
falls more gradually as distance from the source increases. If the. par-
ticles are small so that their fall is negligible, then it would proh-
ably be wise to locate the sampler slightly beyond the maximum to avoid
the possibility of being too close and hence find in.}: concentrations that
are much too low. Particles of appreciable size will tend to settle and
thereby move the location of the maximum concentrations closer to the
stack. In such instances, sampling at the location of the maxip'um
predicted for the negligible settling case would still provide a margin
for error. It would also tend to bias sampling against the larger par-
ticles, which will tend to reach their maximum concentrations closer to
the source.
Although the highest short-tern concentrations are most likely at
ground level under extremely unstable conditions, these unstable cond-
tions can only persist for a few hours a day, during periods of maximum
insolation. The short-term NAAQS specify 24-hour averaging, so for this
type of monitoring those conditions that can persist for day long
periods and still lead to the reasonably high ground level concentra-
tions provide the best guidelines for the siting procedure. The only
atmospheric stability class that can occur at any time of the day or
night is the neutral category (Pasnuill, 1961). The wind direction:, and
wind speeds occurring most frequently with neutral stability will be
available from the output of the STAR program. The effective plume rise
should be calculated for the most frequent wind directions and speeds
and then Figure 24 can be used as a basis for locating the distance from
the source to the region of maxinure TSP concentration under neutral
conditions—remembering that the ordinate values are normalized for wind
speed. The direction to the best monitoring site will be the most fre-
quent direction occurring during neutral conditions. If there are
several common directions, then more than one site nay be necesssary.
Figure 22 shows that after an area of maximum impact has been
identified—whether it is short-tern or long-term impact that is of
interest—then it is necessary to check to see if other major TSP
sources impact on the same area. If so, then it will be necessary to
provide an additional station to measure "bad-ground" concentrations. In
this case, a background concentration is supposed to represent the con-
centration that would prevail if the source of interest were not
present. Such a site, would usually be a neighborhood site in a city, or
a regional site in a rural area. It should !-e as similar as possible, to
the source-oriented site, except that it should be located, where the
source will influence it least frequently.
After the general area for the source-oriented monitor lias been
identified, it is then necessary to select a specific location within
that area. The charticteristi.es of the specific location will he the
same as those for a neighborhood site, v;hen the, source of interest i.s in
an urban area, or as a regional site, when the source is in a rural
area. A iligh-voluro sampler' between 2 to 15 p i.n height, well recover'
from roadways, fugitive dust sources, nnd obstacles to air flow is
desirable.
53
-------
DISTANCE - km
Source: Turner, 1969
FIGURE 24 NORMALIZED GROUND-LEVEL CONCENTRATION UNDER NEUTRAL CONDITIONS
54
-------
'.Seasurere.nts of i.'iml speed and direction at the sanplin;; site or
near the source, of interest would be extremely valuable. Such r.easure-
nents v/ould he invaluable for the interpretation of the collected TSP
data. The wind speed and direction can be used to determine whether the
site was actually exposed to the raxirnutr. inpact of the source. Fluctua-
tions in wind direction can also be used to estimate atmospheric stabil-
ity. If aval labl e, records of the operal in;1. condi tionr, at the source
would also be very valuable as a supplement, to the TSP data.
D. Middle-Scale Stations
1. General
As noted earlier, there are different types of monitoring
sites that mi^ht be considered for measuring middle-scale TSP concentra-
tions. These are those special sites that characterize conditions in a
street canyon or along other types of roadway and traffic corridors.
There are also needs for special sites to characterize the effects of
small area sources. Although this section provides procedures for
selecting special sites in street canyons and alonp, traffic corridors,
and provides guidelines for locating monitors that v:ill determine con-
centrations around the small area sources, it should be recognized that
such sites have rather uncertain application in routine, monitoring. In
general, the concentrations measured at such sites will not be represen-
tative of typical population exposures over the 24-hour averaging period
specified as part of the NAAOS for TSP. f'everthcless, it will sonetimes
be desirable to monitor concentrations in such areas.
2. Special Poadway Sites
Figure 25 provides a schematic diagram of the procedure for
locating a TSP monitor in a street canyon. As with the selection of
other types of sites, the first step of the procedure is to acquire the
necessary background information. This includes the average daily
traffic on all the streets in the area, wind roses, and maps showing
street widths, and buildinj-. heights. These can he obtained fron land
use naps or by having personnel survey the area and estimate a typical
height for each side of each block. Emissions inventories will also be
required, especially those identifying, the location and magnitude of. im-
portant nearby point sources. \n emissions inventory of distributed
source emissions would be desirable, but in lieu of that, some surrogate
such as eir.ployrent, office and retail space, or population way be used
to estimate emissions from space heatinj',.
As indicated in Figure 2?, this type of station may have one
of two basic, special monitoring objectives. One n'iyht be to typify the
v.'orst possible conditions to which the city's street canyons rni
-------
ASSEMBLE BACKGROUND DATA
• Maps
• Land use
• Aerial photographs
• Point source emissions
• Traffic data
• Climatological data
Is station to represent the "worst"
conditions in the area or more typical
conditions?
Represent typical conditions
for the area
Represent "worst"
canyon conditions
Identify and avoid areas where TSP
concentrations are apt to be dominated
by emissions from individual point
sources
Determine Average Daily Traffic (ADT)
on streets within the area of concern
Determine average daily traffic on
streets within the area of concern
Identify areas where the street traffic
emissions are likely to produce the
greatest concentrations
Estimate the distribution of area
source emissions within the area of
concern
Identify streets with ADT near the
average for all streets in the area
LFrom
are i
omiss
From those streets select ones that
are in areas of typical ;irea source
omiss i on rates
Dn the frequencies of wind direction ]
and speed have nearly svmetrical j
distributions (see text)? j
Svmetrical wind rose
Select street that parallels
the axis of wind symetry.,.
if possible
Unsymetrical wind rose |
possible)
Select a street parallel to
the most frequent wind
direction
Locate high volume samp ler
on the more convenient side
the street
Locate
the me
the s
instn
build
membrt
oppos
r
» i
h volume sampler on
the more convenient side of
. Install wind
instrument on top of nearby
building or use supplemental
membrane filter sampling on
opposite sides of the street
Use wind data to identify the side of
the street where higher concentrations
are most likely to be found
Conduct an opt ica1 nephelometer or
1 small filter sampling program to verif>
I or mod ifv conelus ions
1
Locate sampler about 3 -t 0.5 m high,
miiih lock , over sidewa Ik and at 1 east
2 m from building. Should be away from
areas of unusual traffic (e.g. bus
s tops , 1 Oil ding zones , etc.) or other
particulate sources
FIGURE 25 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING SPECIAL
SITES IN STREET CANYONS
56
-------
:BUILDING
TRAFFIC
LANE
-W-
MEAN
WIND
(U)
BACKGROUND
CO CONCENTRATION
-------
If t!\e rointc-rin;; site if-: .supposes' to typify the streets of an
CT.tire downtown area, the averar/- daily traffic should he calculated
hascd en all the important, struct segments in the area. Then those
blocks '..'here the daily traffic is nearest the average for the whole area
ohou.l'! be identified. If possiMe, avoid street segments that have, rad-
ically different widths or building heights from those typical of the
area. Areas whore, the emission densities fron: distributed sources are.
near the downtown average should also be identified and those streets
with average-: traffic that fall within typical distributed source areas
will he the ir.ost suitable for this type of TSP sampling. The next step
for selecting a typical street canyon location is to study the frequen-
cies of the various wind directions. If. they are distributed with rea-
sonable symmetry about some axis, select a street nearly parallel to
that axis. For example, as Figure 10 shows, Salt Lake City, Utah, has a
reasonably symmetric wind rose with an axis of symmetry running approxi-
mately north-northwest to south-southeast. 'North-south streets would be
sufficiently parallel to this axis to satisfy the requirement. Most,
hut not all, of the wind roses in Tipure 10 are nearly symmetric; Albu-
querque, "lev? Mexico, and Kansas City, ''issouri, are examples of asym-
metric wind roses. A word of caution is in order regarding the usually
available wind summaries. The leather Service station for which wind
frequencies are available pay not always be typical of conditions in the.
part of the city where the site is to be. locate;!, especially in regions
with complex topography. Therefore, it will often he wise to seek ad-
vice from local meteorologists and make adjustments where necessary.
Vhen the site is to provide a measure of typical. TSP concen-
trations, then it should he located so that the effects of cross-street
concentration gradients discussed earlier will be minimized. This is
different from the case where the site is supposed to measure worst-case
concentrations and an attempt is trade to locate the sampler on the. side
with highest values. To minimize the effects of the 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
aliened with the a::is of symmetry of the wind rose. In such a case, the
side of the street on which the inlet is located is not of much, impor-
tance. If the wind rose is not symmetric or no typical street parallel
to the axis of symmetry car. he found, then the next best choice is a
street parallel to the most frequently occurring wind direction. In the
latter case1, it i.-'.y -he desirable to augment the res-ular hiph-voluwe san-
r>.lin;:, with' other measurements takun on opposite .sides of the street and
with roof level wiru! measurements. These supplemental data can he used
to estimate the extent to which, the samples collected at the regular
samnlir^. location are atypical of the average within the street canyon.
A limited sampling program usin,". smaller filters or nephelom.eters can
provide information about whether the tentatively selected locations are
indeed representative of conditions in the downtown area. Such a pro—
S'.raiT' could also be used to find other representative locations that
mivht be more convenient for one reason or another; for example, space
."i^ht be more readily available, and less expensive, security mi<;,ht be
better, and so forth.
58
-------
Tf the special monitoring purpose is to characterize public
exposure, the location of the !!i-Vo.l sampler should ho near the breath-
ing zone, say at a height between 2 and 15 ir over the sidewalk, but not
closer than about 2 r. to t!u- buildings. It should not be closer than 10
n.eters to a cross street. Several days should be spent observing, ac-
tivity around any tentatively selected location to make sure that vehi-
clcs do net corirronly atop nnd .spend extended periods of time idling or
that other unexpected dust sources art- not present. '"us stops, loading
zones, ant' areas where lines of cars form regularly should be avoided.
If the purpose were to measure vertical gradients then a variety of
heights would be used.
A vehicle counter would be a valuable instrument to .use in
conjunction with the TSP ironitor. Its records could be used to identify
unusually heavy traffic conditions. T,,'i nd measurements at roof level
might be used for measuring, general air flow to evaluate the street
canyon air circulations and their possible effects or, the observations.
Identification of periods with strong winds that mi;;ht increase blowing
street dust would be very useful to data interpretation.
Figure 27 schematically illustrates a procedure for selecting
high-volume sampler locations to characterize conditions near roadways.
Many similarities exist between this procedure and that suggested for
locating street canyon sites (Figure 25). Poth procedures begin, as
should all site selection, with the acquisition of background informa-
tion, e.g., traffic data, street naps, emissions inventories, and c.liina-
tological information. A decision must also be made whether the purpose
of the monitor is to determine the highest particulate concentrations in
the vicinity of roadways or to monitor inore typical conditions. In both
cases, atypical locations dominated by large point source emissions
should be avoided, especially when the desire is to characterize roadway
sources.
If the monitor is to measure high IS" concentrations arising
froin traffic, then it will be necessary to identify the areas of
greatest daily traffic volumes. As with the ctreet canyons, road seg-
ments witli the greatest traffic should provide suitable locations for
sampling TST concentrations that are characteristic of tiie highest con-
centratiors generated on the particular type of roadway being con-
sidered. The most appropriate side of the roadway for such monitoring
is that side which is most frequently downwind.
If the monitor is supposed to characterize concentrations typ-
ical of those found near many roadways in the area, then average traffic
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 dis-
cussed in connection with the selection of street canyon sites.) If it
•does, then the sampler can he placed on either side of the roadway. The
basis of selecting one side over the other might then be convenience, or
greater population exposure. If the wind rose is not symmetric then a
roadway parallel to the most frequently occurring wind direction should
be chosen. Again, the choice of the side of the road is arbitrary.
59
-------
ASSEMBLE BACKGROUND INFORMATION
• Traffic
• Maps
• Aerial photographs
• Emissions inventories
• Climatological data
Is the station to be representative
of "typical" or "worst case" TSP
concentrations?
Typical of more
generally representative
TSP concentrations
B } f
I \.
Identify and avoid areas dominated by
point source emissions
Typical of "worst case
concentrations
")
)
Identify and avoid areas dominated by
point source emissions
Identify at-grade sections of roadway
with traffic volumes that are near the
average for the road type of interest
(e.g. expressways, arterials, limited
access freeways, etc.)
Identify areas, with distributed source
emissions that are near average for the
region of interest
Are wind direction frequencies
distributed more-or-less symetrically?
(
Symetrical
wind rose
Unsymetrical
wind rose
Select a section of roadway
parallel to the axis of wind
rose symmetry
Identify at-grade sections of the
roadway with greatest traffic volumes
Choose side of the'roadway that is most
frequently downwind
Select a section of roadway
parallel to the most
frequently occurring wind
direction
Select a location on the most
.convenient or fhe most populated side
of the roadwav
' Conduct a limited sampling program to verify that th
I selected site will represent the conditions of interest I
IT
Make f ina 1 se 1 c-c- tion of r-amp I ing
location; at a height of 2 to 15 meters
outside the highway right-of -Vvuy and at
least two obstruc Lion heights from any
major obstructions. Avoid anomolous
sources such as toll gates, on-ramps,
etc.
FIGURE 27 SCHEMATIC DIAGRAM OF A PROCEDURE FOR LOCATING SPECIAL
SITES NEAR TRAFFIC CORRIDORS
60
-------
If it is not possible to locate appropriate.ly aligned, road-
ways, then a brief supploriental rioasuremert program, collecting sarnies
at several locations, or using nephelor et ry to identify aroas of large
spatial gradient, would ho very useful to establish the represcntativity
of s selected location, or at least, the1 relationships atv.ong measure-
ment a frori the selected location and other sites. Such a supplemental
measurement program would be useful in any case, but particularly so in
those instances where something less than ideal sites have to be used.
Once a general location for a monitor has been selected, the
inlet should be placed just above breathing height. If there is a need.
to characterize the typical population exposure, the site should be at a
distance frorr the roadway that is about equal to the average building
setback from, the roadway. If a worst exanple is sought, then the moni-
tor should be as near the edge of the right-of-way as possible. If gra-
dients are being; measured, then different heights and setbacks v/ill be
used.
Finally, monitors should not be placed in the vicinity of pos-
sibly anomalous source areas unless, of course, the purpose is to evalu-
ate their effects. Examples of such anomalous areas include toll gates
on turnpikes, metered freeway ranps, and drawbridge approaches. As with
other types of sampling, the roadway site should be well away frcr. ob-
structions.
Corollary data collection should be considered in connection
with TSP sampling a Ion;? roadways. Wind measurements are always useful
to the interpretation of the data ami to provide inputs for modeling or
other research-oriented studies. Traffic monitoring vi.ll often be use-
ful when the TSP observations are being interpreted because those data
can help to identify occurrences of unusual traffic conditions.
3. Other Special Problems
There are occasions when fairly detailed descriptions of pol-
lutant concentrations are required for the area surrounding some spe-
cial kind of source. Such sources include j.aved and unpaved roadways,
aggregate storage piles, and activities such as fanning and heavy con-
struction. Tor longer—term, routine monitoring near such sources, the
same principles will apply as in the sampling site selection methods
that have already been given. The sites will be. located dowm/ind for
the most frequently occurring wind direction. If uiaxi.:nnv.! concentrations
ore required, the sampler will be located at the closest publicly acces-
sible sites when the source is at. ground level. For elevated sources,
such as roadways, the maximum concentrations will occur farther away.
r. eat on ct al. (1972) have prepared a number of graphs relating ground
level concentrations to roadway configurations rind meteorological condi-
tions; Figure 28 shows an exar.ple of these graphs. c'>uch graphs would be
useful for locating high-volume samplers in the most likely places for
finding i.iaxi"ium concentrations of snail particles. The n.'.:xirinn;i concen-
trations of larger •[.•.articles wit.b nppreciaMf? settling velocities -..-ill
be found closer to an elevated source than indicated !-y c a Icuation.s that
have assumed negligible settling.
61
-------
30r
ELEVATED SECTION
STABILITY CLASS B
2 Oh
1 I I I I I I I \ 1
0=67.5° ANGLE OF INTERSECTION BETWEEN —
WIND DIRECTION AND HIGHWAY _
ALIGNMENT IN DEGREES
.—t:" I
H = Height of pavement above
surrounding terrain in feet
_OH = 0
:a = 50
0 = I 00
=150
= 200
O = 250
0001
0 600 1200 1800 2400 3000
NORMAL DISTANCE FROM DOWNWIND EDGE OF SHOULDER-FEET
3600
sions
SOURCE: Beaton et al. 1972
FIGURE 28 NORMALIZED GROUND LEVEL CONCENTRATION FROM
AN ELEVATED ROADWAY
If the objective of the monitoring is to evaluate TSP emis-
'roiri fugitive dust sources, then special monitoring will be re-
quired. It is beyond the scope cf this report to descri.be such sampling
methods. Laurence Herl.eley La'.voratory (1^1., 1975) has prepared a
comprehensive rev lev/ of part i culatc sai:!p]inj\ rethods and equipment that
contains iruch valuable information. Cowherd and his associates at
'lidwest research Institute have developed ;iothc.!ologies for evaluating
fugitive source e.r.isp i.ons (see c.«., Cov;hcrc! et al. , 1974) that night
have application to other TSP-sanplin;- objectives.
62
-------
r'ii:
S::U:CTION
A. Packgruurul
The site selection procedures in the preceding section of this r<--
pcrt contain sore- very specific rocorineti-lat ions concerning such things
as the heights of inlets and distance hot1"1 eon ron.itors and TSP sources.
The recommendations have been derived through a variety of methods. In
nost cases, an a priori judgment Ls required at some point, hut nn at-
teiiipt has been r.arle to iaal>e those a priori jiulgrents as recognizable and
consistent as possible. This section presents the reasoning, and judg-
iT:ents that ve.re used to arrive at the re.corrondat ions.
In sjone cases, such as the recomnended heights of .the. sar-pler, the
choices are straigbtforv;ard. The. importance of population exposure to
TSP concentrations derands sanpling near average breathing heights.
Practical, factors, like prevention of vandalism and possfble obstruction
to pedestrians, require that the sampling lie elevated; the recommended 2
meters is an admittedly a priori compromise between these two require-
ments. Practical considerations will almost always prevent the. sampler
Iron being located at exactly a 3-r; height, so an acceptable ran.c.c of
heights had to be specified. If we can translate this range of heights
into some corresponding measure of the. expected variability of concen-
trations over that height interval, then it v;ill hr much easier to de-
cide if the results are acceptably close to those that would be observed
at the "standard" 3 n hei.fht.
Similarly, the recornendcd spacing between neighborhood or regional
sites and specific sources can be more clearly understood if it is re-
stated in terms of the expected maximum contributions of the source to
the measured TSP concentrations at the site. Acceptable levels of in-
terference by a specific source can be defined, and then the irinimum
spacing between the source and the monitoring site can be determined so
that level is not likely to be exceeded.
P. Sampler Locations
1. Background
It lias been recoumeiu'ed that most sampling should be. done at a
height between 2 and 15 n. The choice of 2 r, for the ninimur. height has
already been explained. It is a compromise between the representation
of breathing height and the prevention of vandalisr--, and it is con-
sistent with recommendations :nade for the. ronitori p.;; of other pollu-
tants.
63
-------
The recommended 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 spec-
ified, and 13m should provide adequate leeway to meet most requirements.
Record (1976") has examined annual mean TSP concentrations from
several stations and found that the amount of the average TSP concentra-
tion attributable to a nearby roadway is approximately proportional to
the reciprocal of the distance of the sampler from the roadway, i.e.,
C = k T/r
_3
where C .= average contribution of paved road to measured TSP—ug m
T = average daily traffic—vehicles day
r = slant distance between monitor and roadway—m
-2 -1
k = proportionality constant =0.046 yg day m veh
Converting units and making trigonometric substitutions gives:
C/T = 0.046/Vz2 + x2
where height, z, and horizontal separation, x, are in meters. The equa-
tion will estimate the impact of nearby streets, or more importantly,
the distances at which those impacts will be at acceptable levels. It
can also be used to estimate the impact of changing the height of the
sampler. The expression given above gives equal weight to the effects
of vertical and horizontal separation between the sampler and the
source, when in fact the vertical effects are probably more pronounced.
Thus, application of the expression to the effects of changing sampler
height is likely to underestimate the effects somewhat.
Figure 29 shows how the average contribution of a roadway varies
with traffic, and with the vertical and horizontal separation between the
road and the sampler, according to the equation given above. It has been
noted by Record (1976) that there is a tendency to overestimate at the
higher concentrations. This implies that the introduction of roadway
dust into the air is less than proportional to the traffic volume, but
data for very high traffic volumes have not been included among the data
from which the equation was derived.
An overestimate is not very important to this application which
focuses on keeping contributions from an individual roadway low. How
low the contribution should be is one of those arbitrary decisions that
is necessary. It seems that a reasonable limit might be ten percent of
the federal standard, or about 7.5 ug m , for urban stations. For re-
gional stations, the maximum allowable contribution from an individual
source might be five percent of the secondary standard—3 yg m
64
-------
o
U-
oc
I-
D
III
DC
UJ
>
<
u
z
O
o
u.
o
o
"V
.c
V
,>
-------
2. Separation from Paved Roads
a. Horizontal
With the criteria described above it is possible to
use the equation, or Figure 29, to obtain estimates of the minimum accept-
able separation between a high volume sampler and the nearest roadway.
Assuming a height of about 4 m for the high-volume sampler, the
values given in Table 5 are appropriate for various road types. If the
criteria that have been used to define the maximum acceptable contribu-
tions, i.e., 5 or 10 percent of the federal annual standards, are not
satisfactory, then other values could be chosen. The distances in Table
5 could be changed accordingly by using Figure 29 with the value considered
more appropriate.
b. Vertical Placement
As has been noted before, the most desirable height
for a monitor is near breathing height. Because of the general imprac-
ticality of placing a high volume sampler at breathing height in many
locations, the height of 2 m has been adopted as a most desirable com-
promise. How far from this most desirable height can the sampler be
placed and still provide measurements similar to those at the standard
heights? Figure 29 shows that the height of the sampler is not a very
important factor at locations that are sufficiently far from a roadway
source; the various sampling heights show quite similar contributions.
However, as the distance between source and monitor declines, the impor-
tance of the vertical placement increases. Figure 29 makes it clear that
if the sampler is at least 20 m from a street, tradeoffs are possible
and there will be considerable freedom in choosing the height at which
the sampler can be placed, while still remaining certain that the measure-
ments will not differ by much from those at the "standard" 2-m height.
Anywhere in the 2 to 15 m range should provide similar results, as long
as the sampler is set back by more than 20 m.
Pace et al. (1977) have obtained results similar to
those shown in Figure 29. They did not consider the effects of horizon-
tal separation between site and sources, but did find strong vertical
gradients up to 10 m or so. The implications are that the sites must be
well separated from sources, if the vertical effects are to be avoided.
It should be noted that Record's expression applies
to long-term average results. Some of the individual 24-hour averages
are likely to have exhibited more than the average influence from the
local sources. Therefore, the distances cited in Table 5 should be
taken as minima and greater separations are preferable.
66
-------
Table 5
MINIMUM SEPARATIONS BETWEEN A NEIGHBORHOOD AND REGIONAL
SAMPLERS 4-METERS HIGH FOR DIFFERENT TYPES OF ROADWAY
Type of Roadway
Typical Daily
Traffic Volume
(vehicles day )
Minimum Distance
to Sampler (meters)
Neighborhood Regional
Site Site
Major arterial,
expressway
Other arterial
Neighborhood street
20,000-50,000
2,000-20,000
< 2000
125-300
15-125
15
300-750
30-300
30
3. Separation from Unpaved Koads
It has been assumed that no unpaved roads will be near a
neighborhood site, but this ray not be so for regional sites that are
located in rural areas. The presence of dirt roads in an area will
raise TSP concentrations .-ippreciably. Calculations baser! on line source
equations and estinates of emissions from unpaved roads (Cowherd et al.,
1974; v.ann and Cowherd, 1075) indicate that sanplors would have to be
reroved tens of kilometers fror: such roads if they have nore than a hun-
dred cars per day traffic to ensure that their contributions to TSP con-
centrations would be only a few yg m~3. These results suggest that the
effects of dirt roads can extend over very lar^p areas and influence
concentrations on the regional scale. Therefore it is not appropriate
to try to place a regional TSP monitor so far from the unpaved roads
that it does not reflect their contributions at all, but it should be
placed where the gradients are snail.
If the gradient arising fron a specific source is expressed
relative to the concentration arising from that source, it is indepen-
dent of source strength. If a site is sufficiently far from a parttcu-
Lato source, then the gradient of concentration will not be affected ap-
preciably by- Ruttlirj: particles—nil the Lirro particles will have set-
tled closer to the source. *.ccording to Mann and Cowherd (l'J75), parti-
cles with diameters: larger than about ?n ym—and settling speeds greater
thnn about 5 cir. s"1—will fall to the surface within about 100 r. of the
source. Thus, for greater distances the effects can be approxinated by
formulae that assure negli>;ib]e t-.tttlin,-,. With this in n-ind, it is pos-
sible to use Ludwir- and Kealoha'r. (1975) formulations. L'sin;: a value of
67
-------
20 percent km~ for the n-axirun; acceptable, "radient in the concentration
arising from a specific source, they estimated that the rinir.Him separa-
tion between sampler am- lino, source should he about 4 kn for neutral
atmospheric stability. As noted earlier, neutral stability is the post
commonly occurring condition and, the only stability class that conic'
occur at any hour of a 24-hour TSP sairpl iry interval.
4. Separation fron Urban Regions
Our approach to defining the distance that a regional.
station should he outside of an urlrin area h.as been based on reasoning
similar to that used above in connection with the determination of
minimum separations between the sites and roadways. That is, nn attempt
has been rade to ! indt the influence of the city on the. nearby regional
monitors. We have taken advantage of similar calculations made for ot!1-
er purposes. Tijure 30 (fron Stanford research Institute, 1°72) shows
calculated relative concentrations downwind of a circular area , 70 l..p
in diameter. U'ithin the area, source onvission rates are highest at the
center, decreasing in Caussian fashion, to 24 percent of the central
value at the Cf'j;e. The figure shows that concentrations outside the
city under the slightly unstable conditions that are frequent during the
day are nuch lower than occur with the sliphtly stable conditions connon
at night. For purposes of calculation, we have assumed that the 24-hour
averaging that is used for TSP sampling will result in concentrations
somewhere between those shown in the two parts of the. figure. A series
of reports* indicates that although the emission rates uo occasionally
exceed 10 tons mi~2 d.ay~l» tne area;:i wlicre. this occurs are limited.
When the emissions are averaj-ed over a larger port of the metropolitan
area the emission rates art- generally less than 1 ton mi~2 day~~l, or
about 4 ug m~ . s"^-. If we choose to linit the urban inpacts on regional
stations to about f> pg m~^, and if we assume that i::inin!ui:> v/ind speeds
averaged over the day :ire al.out 2 n s~^-, then Fiy.ure 30 can be used to
estimate the minimum distances between the regional samplers and nearby
cities. Sur.'nari zinp,
(; = fpission rate = 4 Mg m~2 s-l
u = i.'i rid speed.= 2.p s~^
M = "aximur acceptable concentration
fron city emissions = 3 yg m~3
X«/Q = ?'axi "H.nVi acceptable nornalized concentration = 1.5
* published in I'.-'Oc' and I'd') under the general title of "Report for Con-
sultation on .... ", where the ri.ssinj- vordr, nar'e a i'.S. city or rej'i.on.
Those regions consi;!erc<' were "'altirore, I'oston, Buffalo, Cine inn.it 1,
f'leveland, l'artfor-1, Kansas ('ity, Los An;-eles, Minneapolis, Vew '.'or!--,
I'hoenix, Pit tsburv,
68
-------
VO
RELATIVE SURFACE CONCENTRATIONS (XQU/Q)
GAUSSIAN AREA SOURCE AT 10 m HEIGHT
35
160km
SA-1365-37
FIGURE 30 NORMALIZED CONCENTRATIONS DOWNWIND OF A CITY COMPUTED WITH A GAUSSIAN DISPERSION MODEL
FOR TWO STABILITY CLASSES
-------
Figure 30 shows that an average normalized concentration of less than
1.5 occurs at distances of only a few km fror the city's edge. r'e have
arbitrarily chosen 10 !:;:>, as a distance that should provide a reasonably
conservative separation between sites ant' urban area sources. If other
pollutants are to be measurer! along with TSP, th.en much larger separa-
tions will be required. Tor example, Ludwig and Kealoh? (1975) have
specified 35 V.in as the minimum separation for rer.ional-scale v.ionitoring
of carbon monoxide.
5. Effects of Obstructions
Particulates .ire subject to inertia! forces that do not affect
gaseous pollutants; every tine an obstruction is introduced into an air
stream, it is possible that some of the larger particles will not nego-
tiate the sharp apples am! will be impacted. The smallest particles
(less than 0.1-ym diameter) are generally unaffected, but the larger,
nore massive particles with diameters of a feu microns can be lost more
easily. Although the effects are not well understood, it does seem rea-
sonable to expect measured ISP concentrations to be altered if the
samplers are placed too close to an obstruction. Shelar (1974) observed
that TSP concentrations measured downwind of a rooftop shed were sig-
nificantly (at the 0.05 level) greater than those measured upwind of the
same obstacle at the same time.
Figure 31 is a schematic representation of air flow around a
sharp-edged building. It is based on the work of Halitsky (1961), Eriggs
(1973), and Gifford (1973). It is apparent fror. the figure that the.
greatest accelerations and the most complex airflow is within the region
labeled "cavity zone". It would be prudent to avoid this area if at all •
possible. According t.o Briggs, the cavity zone extends to roughly 1-1/2
building heights downwind of the building. I'sing this as a guide, we
have suggested a spacing of two "heights" fror, an obstruction.
Often it is desirable for practical reasons to place the Hi-.
Vol sampler on top of a snull structure used to house o.ther instruments. ••
Figure 32 shows an example of such a- location. Although it is desirable •
to have the sar.pler separated from the building, a location on top of
the building apparently would not introduce serious bias if. the building'
is not large. Judging from Figure 31, a sampler that is about 1-1/2 n .
above a 2 m building wil1 be reasonably well removed from the worst of
the turbulent effects.
C. The Importance of Sources at Various Distances from.the Sampler
The concept of representativity,has .been translated into siting
criteria by trying to:minimize the effects of individual sources on the
sampler. In. the case of the neighborhood and. regional •. sites., this
.minimization has been done by specifying how far the sampler must be re-
moved from the sources. In essence, this approach emphasizes the lack
of nearby sources. In the case of street canyon or traffic corridor
sites, the nature of the adjacent source dominates the site selection
process; the site is supposed to be. located where, the adjacent street
70
-------
and other surroundings are typical of the conditions that the data are
supposed to represent. These approaches tacitly assume that the meas-
urements will be disproportionately affected by nearby emissions. For a
relatively inert gaseous pollutant whose sources are almost exclusively
near ground level, the importance of nearby sources is well documented.
Carbon monoxide is such a pollutant and several investigators (Ott,
1971; Kinosian and Si meroth, 1973; Perkins, 1973) have concluded that
sources near the monitor have an influence on readings that is dispro-
portionate to their strength.
CAVITY ZONE
SA-3400-1
FIGURE 31 SCHEMATIC REPRESENTATION OF THE AIRFLOW AROUND AN OBSTACLE
Ludwig and J'ealoha (1975) used a very simple model to estimate rela-
tive contributions of carbon monoxide sources at different distances.
They concluded that about one-fourth of the time, downtown street canyon
monitors measured concentrations where more than half of the. carbon
monoxide cane from the local street. The. frequency of local traffic
dominance increases to more than half of the time for street canyons
away from the center of the city. Lur'wig and ICealoha's (1975) calcula-
tions suggest that the dominance of local sources is, as expected, less
pronounced for neighborhood type sites. Sources within 2 km of the mon-
itor contribute more than half the observed concentration on about a
quarter of the occasions. Thus, the nearer sources contribute a major
portion to nei £,hborhood-scale observations, but not as much as for the
middle-scale observations.
The preceding discussion applies to gaseous pollutants whose main
sources are at ground level and closely related to traffic. Drafts
(1975) lias concluded that around 90 percent of the mass collected by !!i-
Vol samplers near Chicago streets were minerals and auto exhaust parti-
cles; the minerals have been attributed to street dust or local soil.
Thus, it appears that urban TSP is analogous to carbon monoxide to the
71
-------
FIGURE 32 HIGH-VOLUME SAMPLER LOCATED ON THE ROOF
OF AN INSTRUMENT SHELTER
72
-------
extent that its ground level • sources are traffic dominated. Therefore,
observed TSP concentrations should also be dorinated by local sources
except where najor elevated point sources overv:l:eln the otlicr sources.
If we include the effects of the settling of particulates, the do mi nnr.cc
of local sources will be reinforced.
Although Oraftz (1975) did not measure size distribution:; of the
collected particles, he did indicate that particles with diameters of
several tens of microns were fount' in samples collected near city
streets. As we have noted before, only a few such particles are re-
quired to equal the mass contained in literally hundreds of thousands of
subrcicron particles. The height to which these particles are raised
with their initial injection into the atmosphere will affect the dis-
tance that they travel before, they settle to the ground again. Cowherd
et al. (1974) estimate an average injection height of five feet (150 cm)
for dust from unpaved roads. This can be used as a guide to estimate
that these larger particles, with settling velocities of one or two
cm s~l, are injected at an initial height of 150 to 200 era. This means
that on the average they will remain airborne for a few hundred seconds.
At typical wind speeds of a few meters per second these particles will
not affect concentrations at distances of much irore than a kiVneter
from their source.
If we return to Ludwig and Kealoha's (1975) estimates of the impor-
tance of ground level gaseous sources at various distances from a noni-
tor, we can make some assumptions that will allow us to alter those es-
timates so that they are more applicable to TSP. From rigures 4 and 5
it appears that about ten to twenty percent of the aerosol mass may be
associated with particles that could fall out within about 2 km of their
source—particles larger than 5 to 10 ym diameter. If Ludwig and
Kealoha's (1975) calculations are adjusted to account for the loss of
about ten to twenty percent of the material emitted before it travels
more than two kilometers from the source, then it appears that emissions
within 2 km of the sampler would account for rore than half the observed
TSP concentration about .30 to 40 percent of the tine.
Of course, all the preceding discussion applies only to areas where
the TSP concentration.'; are not greatly influenced by emissions fron
elevated point sources. The influence? of an elevated point source will
depend on emission rates, plume rise, stack height, nnr! distance fron
the stack With this many variables, it is not possible to give any
generalized estimate of elevated point source effects on observed TSi.'
concentrations. However, the Climatologies] Dispersion t'odel (CHM) of
Busse and Zimmerman (1973) lias provisions for separating point source
effects from area source effects, and it could be used, to evaluate rela-
tive contributions for specific, cases. Even in those areas where not
many contributions come from specific, elevated sources, the model right
be applied to ensure that the effect of some, point source, or n particu-
larly strong area source, had not been overlooked.
73
-------
''odeling could also serve to evaluate the appropriateness of emis-
sions control strategies based on data fron an exist in;1, TSP monitoring
site. It could tell if the c-nissions to be controlled were actually the
most Important contributions to the observed concentrations. The esti-
mates of the importance of sources at different distances fron the
sampler support our notion that siting criteria based largely on the
distribution of nearby sources wi.1.1 !>e adequate to define, at least
qualitatively, the representativeness of a station tbat is not intended
to be a source-oriented monitor. Once a well located site exists, tbe
argument can be reversed, and ve can assume that the data from tbat site
strongly reflect the influence, of emissions within its area of
representativeness—e.£., a few MJornetcrs for a neighborhood site, a
few tens of kilometers for a regional site. Then any control measures
that are based on data fron that site can be focused on tbe appropriate
area. Obviously, data from source-oriented monitors will be most-useful
for defining control measures required of the source toward which their
monitoring is directed.
74
-------
Appendix A
MODELING THE POLLUTION CLIMATOLOGY
ASSOCIATED WITH STACK EMISSIONS
75
-------
Appendix A
MODELING THE POLLUTION CLIMATOLOGY
ASSOCIATED WITH. STACK El'I SS IONS
A. Background
The average ground level concentrations of a pollutant and the fre-
quency of occurrence of specified concentrations depend on the charac-
teristics of the sources and on the occurrences of the different atmos-
pheric conditions that transport: and dilute the emissions. The heights
of the sources and their emission rates are also of great importance.
The most influential atmospheric parameters are wind and stability.
Wind direction determines where the pollutants will be carried, and wind
speed determines the volume of air available for the initial dilution of
the pollutants. Atmospheric stability governs the rate of subsequent
dilution through turbulent mixing. If a particular pollutant is chemi-
cally reactive so that its total mass changes with time, this fact is
also important in the determination of expected concentrations. This
Appendix provides two simple computer programs, one for estimating aver-
age ground-level concentrations, and the other for calculating concen-
tration frequency distributions around an isolated, elevated sorrce.
From this information, the most important monitoring locations can be
identified.
Diffusion theory provides the basis for a systematic approach to
estimating the effects of stack emissions on the environs. That is, the
theory can be used to calculate the concentration patterns to be expect-
ed during each of some finite number of meteorological conditions. It is
thus possible to determine average concentrations and frequency distri-
butions, if the frequency of occurrence of each of the input meteorolog-
ical conditions is known. This is the approach that has been used here.
B. Description of the Pollution Model
The behavior of pollutants being emitted from a continuous source
is commonly described by assuming a Gaussian distribution of concentra-
tion within the plume (e.g., Peals, 1971; Smith, 1968; Turner, 1969).
For a continuous plume from a smokestack, the distribution of the con-
centration (C) in space is given by:
Q
C = exp
rrua -j
yz
Where
Q is the emission rate of the source (,gm s~l)
u is the wind speed at stack height (m s~l)
ay, az are the standard deviations of the distribution of
concentration in the horizontal and vertical (m)
77
-------
y,h are the distances from the center of the plume in the
horizontal and vertical (rn)
The values of ay and Oz, depend on the atmospheric stability and the dis-
tance downwind of the source. This dependence has been determined from
field experiments that have, released tracer materials in the atmopshere.
Figure A—1 illustrates relationships among ay, and oz, distance from
source, and atmospheric stability. These relationships are widely used
(e.g., Beals, 1971; Turner, 1969). They were originally pub-
lished by Cifford (1961), anci they fon;i a basis for some of the inputs
to the model described here.
Because the concentrations of interest are those at ground level,
the term h in Eq. (1) can be considered the effective height of the
source. Cases usually leave a stack with some finite velocity and buoy-
ancy, because they are often warmer than the ambient atmosphere. There-
fore, the effective height of the stack is generally higher than the ac-
tual height by some increment. This increment, Ah, is determined by
the equations below, suggested by Smith (1968).
For stable atmospheric conditions:
0.33
— W
For neutral or unstable atmospheric conditions:
3 (3)
Ah = 150F/U
(^)
•f = g V r I I is the buoyancy flux
S S * rr, I .
g is the gravitational acceleration, 9.8 m s~^
T is the stack gas temperature at exit (°K)
S "*
T is the ambient air temperature (taken to be 300° K)
V0 is the velocity of the stack gases (m s )•
S
The term C relates to the rate of change of temperature with
height. For stable atmospheric conditions a value of about 6 x 10 is
appropriate (see example given by Smith, 1968) and is used with this
78
-------
2
UJ
CJ
O
u
g
CO
cr
CO
Q
z
UJ
(J
Z
g
CO
cc
UJ
a.
CO
I
CC
O
I
4 x 10'
A - Extremely Unstable | | ;
B - Moderately Unstable Zp-t. *"
C - Slightly Unstable
D - Neutral
E - Slightly Stable
F - Moderately Stable
A - Extremely Unstable
B - Moderately Unstable
C - Slightly Unstable
D - Neutral
E - Slightly Stable
t F - Moderately Stable
i i——r-p* •
DISTANCE FROM SOURCE (ml
SA-1567-8
FIGURE A-1 VARIATION OF oz AND oy WITH DISTANCE AND ATMOSPHERIC STABILITY
79
-------
model. For low wind speeds and neutral or unstable atmospheric condi-
tions, the values of Ah become quite larj;e and unreliable. To riini^-
ize the effects of this artifact, a raximino. value of effective stack
height should be chosen, such ns the typical, afternoon nix inf. depths for
the area and season of concern. \verage values for the. contiguous Unit-
ed States have been given by Lolzworth (1971). The afternoon values
should he used because at nij'ht low i/ind speeds accompany stable atmos-
pheric conditions, and the equation used to calculate effective stack
height during stable conditions is much Jess sensitive to wind speed.
Some pollutants (e.g., sulfur dioxide) undergo transformations that
can be approximated by an exponential decay term. Provision for such
transformations have been included. Uith exponential decay, liquation
(1) becomes:
Tiua z
y
Q
C = ; exp „ , ,, . _ 2
z
(5)
where
x is the distance downwind of the source (m)
T is the time required for 63.2 percent (1/e) of the
material to be transformed.
Equations (2), (3), and (5) are sufficient to calculate ground lev-
el concentrations, if the proper stack emission and meteorological data
are available. The required meteorological data are atmospheric stabil-
ity and winds at stack height. Although atmospheric stability is not
reported from weather stations routinely, algorithms have been developed
to determine this parameter from conventional surface weather data.
Turner (1964) presented a method for determining stability; the required
inputs are wind speed, the sun's elevation, cloud cover, and ceiling
height. Turner's technique is used by the National Climatic Center to
prepare summaries of the frequency of occurrence of various combinations
of stability, wind speed, and wind direction. These summaries, prepared
from hourly weather data with the Climatic Center's STAR program, pro-
vide a nearly ideal input for studies of pollution climatology. Wind
directions are divided into 16 classes, wind speeds into six classes ( <
3, 4-6, 7-10, 11-16, 17-21, and > 21 knots), and the atmospheric stabil-
ities into five (or for some locations, six) classes. Since the wind
speeds are measured at low levels, it is necessary to make some correc-
tion to approximate the speeds at stack height. The corrections used in
the model were based on a graph given hy Reals (1971). Figure A-2 shows
his suggested relationships among height, vind speeds, and stability;
the very unstable, slightly unstable, and slightly staM.e curvuc hnve
been added to ttie three curves values "i von by !'eals. The r-idranj-e
value for each wind speed class c;in be converted to a stncl heij-.ht value
for the appropriate stability class ami input to tin- model using n fac-
tor obtained from Figure A-2. These stncl.-lici y,h t wind s pee dr. ,nre used
to calculate the concentrations associated with '\->ch of the coi..M nal t ons
of wind speed, wi'.id direction, .mci ;ili su.spher ir st.-ibi.lif.' represented in
the STAR program output.
80
-------
Certain simplifications have been ;;iade to permit the use of polar
coordinates and to reduce the conputational complexity. The rodel h?s
provided for tre.itin;; tv;o stacks; both are assured to be at the origin
of the. polar coordinate system. This sii'iplif ic.it Lon is justifiable if
the distance between the stacks is small compared with the downwind dis-
600
o
<
i.
o
UJ
I
1.0
1.5 2.0 2.5 3.0
WIND INCREASE FACTOR
- 150
- 4
5.0
SA-1 567-9
FIGURE A-2 VARIATION OF WIND SPEED WITH HEIGHT AND
ATMOSPHERIC STABILITY
81
-------
tances to areas where higher ground level concentrations are expected to
occur. One stack is nssuined to have relatively constant emission
characteristics. The second can have three different levels of emis-
sions.
The fraction of the time spent by the variable stack in each
operating mode can be specified. It is assumed that the different
operations have no meteorological bias, i.e., that the likelihood of the
stack being in a particular operating mode is the same for all meteoro-
logical conditions.
For each of the 16 different directions, the model calculates con-
centrations at numerous distances (up to 40) from the stacks. These
concentrations are determined for each of the 1440 meteorological-
operational conditions (i.e., 16 wind directions x 6 wind speed classes
x 5 stability classes x 3 stack operating modes). In one version of the
nodel (POLAVE) the calculated values for a particular meteorological-
operational class are multiplied by the frequency of occurrence of that
class; these products are summed over all possible operational combina-
tions to determine average concentrations in the area surrounding the
stacks. In the other version (POLFREQ), the concentrations calculated
by the model are categorized for the determination of the frequency that
various critical concentration levels (such as standards) might be ex-
ceeded. This determination is based on the calculated concentration
values and the frequency of occurrence of the meteorological-operational
conditions leading to such concentrations.
The concentrations along the plume center line extending in the
downwind direction are easily calculated from the equations by setting y
equal to zero in Eq. (5). For the concentrations on adjoining radii,
22.5° from the downwind direction, the following approximation is used:
y = x sin (22.5°)
(6)
In essence, this results in about an
-------
C. Computer Prograr.is
The computer program for computing average pollution concentrations
around an elevated source is called POLAVE, while that for computing
frequency distributions of concentration around an elevated source is
called POLFREq. A subroutine called STCKHT is used with both of these
programs t^o calculate plume rise. The FORTRAN variables and the input
requirements are identical for both programs. Table A-l defines the
variables used in the programs; Table A-2 describes the inputs required
to operate either of the programs. The data that are read into the com-
puter consist of: (1) mixing depth and categorized values of variables
derived from Figures A-l and A-2; (2) stack height and diameter, emis-
sion rate, gas temperature, and pollutant concentration in the stack
gas—for both a constant and a variable source; and (3) frequency dis-
tributions at the site for all combinations of stability, wind direc-
tioni and wind speed. These data provide the necessary information for
computing either the average or the frequency distribution of pollutant
concentrations about an isolated elevated source, using either POLAVE or
POLFREQ. A listing for these two programs and subroutine STCKHT follows
Table A-2.
83
-------
Table A-l
PORTION VARIABLES USE!) IN THE PROGRAMS POLFREO AMD POLM'E
CL = Concentration on plume axis and at two distances off
axis (g m~3)
DEC = Provision for decay—currently set to be negligible
(sec)
Dl,02 = Stack diameter (m) for constant and variable source
TRACT = Tine spent in each of the three different operating
conditions of the variable source (percent)
FFEO = Frequency of occurrence of various categorical
combinations of stability, wind direction, and speed
(percent)
F1,F2 = Emission rate (scfm) for constant and variable sources
SEASON = Season or annual identifiers (alphanumeric)
HIIUM = Mixing depth (m)
ID = Wind direction category index
ISP = Kind speed index
IX = Distance index
R = F.adia.l distance (rc)
PCT1,PCT2 = Stack gas pollutant concentration (g ft~3) for
constant and variable source.
STAND = Lower limits on concentration categories for frequency
calculations (g m~3)
SUM F = Frequency of occurrence (POLFRF.P) or averages (FOLAVF)
of various concentration categories for the different
distance and direction categories (percent)
84
-------
Table A-l (continued)
SY,SZ = Horizontal and vertical standard deviations of plunc
concentration (m)
T1,T2 = Stack gas temperature for constant and variable
source (°F)
WS = Wind speed (m s~l)
85
-------
Table A-2
INPUT PAPWETERS FOR THE PROGRAM POLFREQ AND POLAVE
CARD
J.'UUBF.P VARIABLES
FORMAT COLUMNS
REMARKS
1-5 1ST
IX, 12 2-3 Stability category
index (total number of
categories 5 or 6; 5 used
in listing given here)
US values for . 6F6.2
6 speed classes on
one card, cards for
5 (or (•>) stability
classes
4-39 Wind speeds at stack height
for different surface wind
speed and stability cate-
gories (m/s). Obtain from
Figure A-2 and STAR program
categories.
6-45 IX
12
R; one distance
value per card.
SZ values for 5
(G) stability
classes on one
card, cards for
40 distance values.
F6.0
F6.0
SY values for 6
stability classes
on one card, cards
for 40 distance
values.
6F6.0
1-2 Distance index (total
number of distance valuns
to be specified by user is
40).
3-8 Radial distance (m)
3-8 Standard deviation of plume
concentration in vertical
(m). Obtain from Figure
A-l. If 5 stability cate-
gories are used (as here),
then last field is not
used.
45-80 Standard deviation of plume
concentration in hori-
zontal (m); see comment for
SZ.
86
-------
Table A-2 (continued)
CARD
NUMBER VARIABLES FORMAT
46 Fl E8.1
Tl F.8.1
PCT1 F8.1
Dl F8.1
Hll F8.1
D2 F8.1
1122 F8.1
47-49 F2 E8.1
COLUMNS
1-8
9-16
17-26
25-32
33-40
41-48
49-56
1-8
PEMARKS
Emission rate (SCFM) for
constant source.
Gas temperature (°F) for
constant source.
Pollutant concentration
(g/standard ft3).
Stack diameter (m) for
constant source.
Stack height (m) for
constant source.
Stack diameter (m) for
variable source.
Stack height (m) for
variable source.
Emission rates (SCFM) for
T2
PCT2
TRACT
F8.1 9-16
F8.1 17-24
F8.1 25-32
variable source, one card
for each of three operating
conditions.
Cas temperature (°F) for
variable source.
Pollutant concentrations
(g/ft3) for variable
source.
Time (percent) spent in
each of the three
operating conditions of
the variable source.
87
-------
Table A-2 (concluded)
CARD
NUMBER VARIAHLES
FOP.?!AT COLUMNS
50
51-
130
SEASON
HNUM
1ST
A 10
1-10
F10.1 11-20
IX,14 2-5
FRECi, values for 6F7.5
6 stability clas-
ses on one card,
cards for all
possible com-
binations of wind
direction and speed
classes.
11-:
Season or annual
identifier.
Afternoon rr.ixing depth (n).
1ST is the stability class.
There are f?0 cards, 16 wind
directions and 5 stability
classes. If six stability
classes are used, there
will be °6 cards.
Frequency of occurrence
of various categorical
conibinations of stability,
wind direction, and
speed (percent).
88
-------
LIST OF PROGRAMS
PiiOG-JAM i'ULFUFO ( I NPLT , OUTPUT )
c **********
C***»**#^«* TMI c. PU'OC.K A" CALC JLAItlS F i< H.' J iT N C Y O I ST iJ I .> ,H I , ri2 ( 3 ) . L < I bT , 15 P .Ml 1 . H2i!, Dl 1 , 02?! HNUM
D 1 M EN S I ON N C i .3 > . V S V. ( 3 > » J k> ( 3 ) , .'-' ^ AC r ( 3 > , F K EO ( 6 . 1 6 , 6 ) , w S ( ri . f> ) . S Z ( Ci , 4 0
1 ) . SY( 6 .4 0) .K< 40 ) .JHZJM 3 ) . S JMF (40 , 1 6 . 1 0 I • F,'( .'I ) ,T2 ( J )« PCT 2( .T ) . IJ ( ? >
2. XXX < J ) . STAN!'J( o )
C» *** ******
p****««««4* 3TANO = CONCr-.NTIVAT 1 Gi'v C A TEOOR IPS ( OM/ CU . M )
C 4 *********
DATA ( STANO = 0 .0 ,2 .SH-'+ . .3 . hbL-* . S . ot:-4. 1 . 3F-3. I .0 )
DATA ( XX.X=0.0,C..'ia,? 7, 0. 7U71 )
DATA (!J = -lO.-^.-di -7 ,-f,,-5.-4.- Jt-2 )
C * »**#J»****TH^GUGH STATfiMtNT 22 VAR1GI.E DATA ARE READ A MJ UMTS ARt
O**********CONVfcPTtU.
C* »*******$
C********»*IN1 Tl ALLY - RtZAD WIND SI'MthO AT STACK HT. FG^i DIFFERENT SPEED
I SP ) AND STABILITY (1ST) CATtGOHItS. THIS IS A TABLt BASED
f- 1C 2 OF APPtNOIX. ST ACK-HE: I GHT WINDS ARC ENTERED FOR EACH
I TY CLASS ANC EACH W INU SHt'ED CLASS OSED BY STAi-i
***«******PRUGRAM (M/S). 6 CARDS
C**********
DO 19 1=1.6
READ 1 0. I ST. ( lnS{ I ST. I SP ) .ISP = 1 .6)
19 CONTINUE
£*********«
C*******<<**ttF. AD DISTANCES (R) AND CC R KESHONC I NG STANDARD DEVIATIGN
C**********VALUtS (SiC.SY) TO BL JSt:D IN SUGSeOUENf CALCULATIONS. SIGMAS
C ****#*****TAKFN FRO.M FIG 1, FOK kADlAL OISTANCtS CHOSEN dY OSEH. ALL
r* «*****J(I**(.JN I TS IN METERS. 40 CARDS
DO iiO 1=1.40
REAL) 1 4. 1 X. * ( I X> . I SZ ( I S . I X) . 1 S= 1 .(> ) . ( SY ( I S . i X) . I S = l ,6)
20 CONTINUE
r ******** »»
C**********HK AD FLOW (SCFM), GAS Tt.vi PEH ATUR f. (DUG. F), POLLUTANT
C**********CONCENTt^ATI(!N (GM/CU.FT), ID (FT). HT (FT) FCR CONSTANT
C**********Sl.)UKCE . AND IO (FT), HT. (FT) F;JI< VARIABLE SCURCF.
C**********
RtAD 7 ,F 1 ,T1 .PC Tl .1)1 ,M1 1 ,D2 ,HX?
C*»********
C**********CONVERT ING TJ Mfc'TMIC JNITS AND GETTING VAL>JFS FOR PLUMr RISE
C **********
0 1=0.01 7*F 1*PCT1 t 01=0. .105*01 S HI 1 =0 .30S*Hl 1 f D? = 0.305*J2
T l=0.bS5o*{ T 1- J2. 0 ) + 2 7''. . 0 S H2;? = 0 . ^0 b*f-' 2 i'
VS 1 - 1 . 1*4. 7«JC'.-4*t-' 1 »T I/ I <'9 j . 0*0 . /85*Dl «-t) 1 )
C **********HF. AO t-'LCW (^CFM), GAS T L -1 P F R A T LV; f (CEG. F), POLLUTANT
f. ****ITIUN
i;** + *****«*( PCT J FGt) EACH CF J O.''t-fc AT IM-. CONDI T I f;\S ON VARIARLF. ".JTACK.
C **********.{ CARDS
C* *********
****>
DO a^e i =1 , J
PiT AC ^ .K .PC Ti! ( I ) .1- RACT { I )
C ****** t* » *C PiN V(:i-' T I Nl« T !"! vt; T« 1 C l/M T b flNU GET! [fiij VAllJES FOR tJt.:JMt. K 1 S f.;
C. * ** *»«:*5-. * «,; , ALC ',IL >' ! I (. '.i'jS .
l" ***#****« T
U 2( I ) = 0 . I.) I 7« I- ? ( ! ) *:• PC I ?' I, 1 )
I 2( I ) = •) . -ii? i-*( T.:( ; )-j,-.. o ) t- jr.'. ;;
VS.! ( I ) =1 . I *4 . ?:;.-. --.•;. ,' i *r'r}AC T ( i )
89
-------
f Si'A..; Oh l = \*'if- t'Ci< 6 STABILITY CLASSES.
r. ********* *~;-AI: rCLi. t.t- f.xcj^'JE'MO ..IF- c L h U C L A c S ( IJ I .
r. **********
fi BAD 0 t 1ST. ID . ( FWr.:) ( 1ST . I '.". . J ) . J = 1 . o )
^.> CONTINUE
{. ********* *;^KC =PROVl o I ON t-'CK Jt":CAY - St:T TO Fff NCCLIGIRLK.
(; 4 » ********CALCUI_AT ING ••> AiiMr: T E.RS Fl1.'? 'JLU^r P I Sii- ASSUMES AMt3IFNT TbMT OF
r.#«******** 100 DtG K.
C *»»*******
DtC = (>. DOK6 tFS 1 =0 1 *?- .4b*V , '^Cl«{ r 1-.?00 .0 )/300. 0
C. **********
Ct ***** **»*ShT1 ING INITIAL VALUi-. ' TiF l-Kc Ol-uNC I FS -0.
(^* « t « *t 4 At f
CALL MhMSET ( 0. 0 . SLiVK .6400 )
C #»***»****CALC JL AT ING PAkvltTt.WS POi^ PL.JMF KISF-ASSJMtS AMLJIENT TEMP UF
(7»*********JOO OtG K.
C_ <+****»*»- I I ) - JOO.OJ/300.0
30 CONTINUE
C**********rHE LOOMS THROUCH STATFMENT 900 CALCULATE CCINCEN T« A T I ONS FOR
f.»*#******* VARIOUS STA3I I.I TIES (1ST). U I ND SPEEDS (ISP), DISTANCES FROM
C*********#STACK (IX). AND TH;i OPF.RATING C.CNDITICNS (NCP). OCCURRENCE
C.*****»****FR£OUENCV IM f)IFF£RENT CONCENTRATION INTERVALS AfA T*SYC>,M ) s GICOEI =ai*cnFF
Hi 1 = 0. *:*Hl*Hl/r, ^;.'
DO 9JO Nil;-'= 1 . i
HZ.-! ( NJP) -0.5* H^ (Nil P ) «h^ (N '.II ')/',",/?
r. **********
f; *4********;>i:TE1:i''- I M NG CHN C ENT H A T I !]\ (Ct-C) CN PLUMC AXIS AN'O AT ;.' OI3T-
f. ******** **ANCLf. (Y) DFr AXIS.
*»*<<
DH tiCO J=l .-)
^ ( j .Ni; . .") ) GO ru \. 10
90
-------
(,:, C'f.' MU'.T DtG. L.I- r Cr.NTuK I. INC. CO\'C F.M THAT ION IS ASSUM!-)
.'ALL 1 .'-I TiiP. LClWf-'ST CONCF.NT,-.
NCLC = ! -i GO TC rifiO
510 Y = X*XXX(J) 6 Y <2 = C . '"> * Y * Y / S Y 2
Cl=OlCOtr*FXP<-H/l-Y2)
02=U2( NUP ) *COF.f-~*E XP (-HZ2 (NUP )- Y2)
CLC = C1+C2
C»***-******-",:if:C IF Y ING C'.JNCENTRAT I UN CATEGORY FCR CuNTtWLINEI AND +/- ;'.?. 5
C *«#**»*«*«
540 00 553 JJJ=lt6
IF (CLC.CK.STAMJ ( JJJ ) I NCLC - J J J
5£5 CONTINUE
560 NC(J)=NCLC
6CC CCNTINUF
DO 9UO ID=1 , 16
C*******»»*FKAFRt = FRLO OF OCCURRENCE CF COMPINATIONS OF STABILITY,
C**********'A IND OIHECTIUNi ANO SMEbO. ANO PLANT OPHHATING CONDITIONS.
FRAFh*F.=FRACT (NOP) *FWEQ( I ST. ID. I SP)
I F (FREQ( I ST . ID, ISP ) .EC .0 .0 ) GO TO 900
DO 900 Jl =1 ,1 6
I F( Jl .CU. l.U«.J 1 .EC. 15) GO TO 610
IF(Jl.LQ.ie) GU TO 6C5
J=3 J GO TO 615
605 J=l 6 GC TO 615
£10 J=2
tl 5 J0= ID+J 1 t7
JD=MOD( JD , 1 6»
IF ( JD .EO. 0) JD = 16
Nl-NC( J)
C 4** + ******Ar CUMULATI KG F W IT CJUE.NC 1 1 S FCK OIFFbKtNT LOCATICNS.
DU 900 MCLM-l.Nl
SUMF( IX. JD. MCUM ) = SUMF ( I X , JU, "CUM ) + Ff?AFRE
SOO C ON T I:\IUE
f. **********
C **»******#THHrjUGM STATEMENT 950 FKEGCENCIES AMK PRINTcD.
DO Q50 10=1 . 16
UI = ID S UI = ^2
PRINT .3 ,DI . SPA 3ON,HM.N * PRINT 9
PRINT 1 1 . (STANUC JJ J » ,J JJ= 1 . 6)
DO 950 1 X=l .JO
PRINT =>,«( I X), ( SUM-"( I X, IO.MC) ,' NC - 1.6)
950 CONTINUE
1 FORMAT ( /\ 1 C ,F 1 0. 1 )
J FORMAT 11HO . *i)JSRCTION FRuM SCcLTRrt = * ,F6 . 1 , *DbOr?EE S* . 1 OX ,
1 A 1 0 , 1 OX*(V I X INv. rIT. =+F10.1/>
5 FORMAT { li- . T. 1 1 . 3, et-" 14 .6 )
6 FORMAT ( IX . 14 , I'.:>,6F7 . j, )
7 FORMAT (te. I . c.FH. 1 ) •
t3 FORMAT ( 4Ff . ! )
9 Fi.!(
-------
HUL-O ;AV POLAVt < INMUT , OLTPJT )
C**********
C»********»C ALC'JL AT L'5 AVfcl .Fhf #CT ( 3 ) .FIxK'.aa. 1 6 .5 ) .«S (6 •<:• I . UZC-). 40
1 ) .SYl6.40),K(40).h?2(3),SLr4F(40.1ail(>)»Fc( ? ) . T >» I ,'S > . :>CT;M .< I . I J( » )
2. XXX( 3) , XX.MC( Ji)
DATA 27.G.7C/1)
DATA CIJ = -! 0.-V.--K. -7. -o. -S .-4.- j ,-? )
r * * * «*****»
C ********** THROUGH STATfA/ENT ?J V A HI C) L S DATA AfiK RcAD ANO UNITS A:!1L
C »******#*6CON VLfiTLO.
£ 4 *** ******
C********** I NI T I Al.l.Y - K'ffAC » INU SHiirl) AT STACK HT . \-ty-t DIFFE^EINT S^C'O
C**********( ISP I AND STABILITY (1ST) C A T t(i C. (• I f e, .
I) U 19 1=1.6
KdAt; 10 . 1ST. ( W'J( 1 ST. I SrJ ) , I Sf- 1 , ft )
CUNT I,MU f?Str: CU C NT CALCULATIONS.
C****»***»*
OO ? 0 I = 1 . t 0
HcAD 14.IX,P(IX).(<;Z(IS,IX).lS=l.o).(SY(IS.IX),IS=1.6t
20 XCNT I MUt
c **********
c »****»**»*KF AD FLOW (F,CF:«), GA» Th ^PLWA TUi"; (Ccf,. F>. PCLLUTANT CONOFN'-
C**********TWAT ICN (GM/CJ. FT.», ID (FT), I- T (FT) FO>-! CONSTANT SOJHC^
C* »*»******ANO ID (f'T). ANC HT. (I f) FCR VAKIAtLF: SOU»Ct.
C* *********
HhAC 7 . Fl .T 1 .PCT1 ,O1 .111 1 .0, V.i 12^
C *»**+*****CCNVEfiT ING TO METWIC UNITS.
C**********
Ci 1 = 0 .0 1 7*F 1 *PCT 1 * O1 = C.2C5*1>1 S Hi 1 =C. 3C5*H1 1 $ D2=0.30'j*O2
T 1 =0.6^56* { T 1-J2 .0 ) »?73 .0 i Hd 2 = 0 . 30 5 *H 2 2
VSl=l. l*«. 72L- 4 *Fl*Tl/(2<> 3. 0*0. 765*01*01)
c**********KeAO ( FOH \/A^ . source cPCf<. CONO.. i>. FLOW (SCFMI. TF.MP.
C«»********(OeG. F). POLLUTANT C CNCti N 1 >i A T I CN (GM/CU.I-T). T I VE SPENT IN
C**********OP(£R AT ING CONDITION (PCT).
C**********
DU 22 1= 1 . J
HEAD 7 .f=2( I ) ,TiM I ).PCT2 ( I ) .FRACT( I )
C**********
c**********crjNvE^T i NC; TO MLTHIC UNITS.
c **********
Q2( I I=0.013*F 2( I ) *PCT2( I )
T2( 1 ) =0.55e>6*l T2( I I-J2.0) +27J.O
VS2( I )= 1 . 1 *4.7,?L-4*F2( I )*T2(l)/(293. 0* 0 . 7a5*D2 *O2 )
FHACT(I )=0.01*FKACT(1 )
22 CONTINUE
C**********
C»*********RF AD StASON ICENTIFItR ANIJ AVERAGE P. M. MIXING HT.
C **********
RF-:AO i . SEASON, HNUM
DO 23 1=1 .BO
C **********
C* »********SHOULL> HC 1=1. y£ fCU 6 i T A Ei I L I T Y CLASScS.
C**********l", (f-K^U( If.T. 1D,J).J=I.6I
23 CONTlNUf
Ol-:C=t>.OOL4 SFS1 =U t *2.4b*V:>l »C1 * ( T 1 - JTC) . 0 ) / 300 . O
c **********
C«»*» ******ShTT INC, INITIAL. VALUTA Li- F l.'t. JUL N C I f. S =0 .
£*********«
CALL MtiMSL'T ( 0. C . ?LVF ,6400 )
DO J.) t = 1 . 3
F s.'i I > =?.<*s* vsj ( i > *;>^»oi'» ( T;'(!)- 300 .n )/:uio .0
.iO CO Ml liMUE
92
-------
C *****»****THF LOOPS THROUGH STATEMENT 900 CALCULATE CONCENTRATIONS FOP
C**********VAR IOU5 STAHILITIFIS (1ST). WIND SPEEDS ( I SP I . DISTANCES FROM
C**********STACK (IX) ANO PLANT OPERATING CONDITIONS (NUP).
C**********
or: 900 isT=i.s
C**********SHOULD bE IST=1.6 FOR fi STABILITY CLASSES.
c **********
DO 900 I SP=1 ,6
U = WS(IST.ISP) 1 UPI=3.1416*U
C********** ........ . .
C* *********CUHRECT ING WIND SPEED TO STACK HEIGHT.
e**********
CALL STCKHT
DO 900 IX=2.36
MMM = MOO { IX .2 )
IF ( IX .LT. 28 .,ANt). MMM.GT. 0 ) GO TC 900
SZDAT = SZ< 1ST , IX I I SYO AT= SY ( I ST , IX )
SZ2=SZDAT*SZDAT t SY2=S YD A T* SYO *T
C»*********
C**********oeTeRMINING VALJfcS OF VARIOJS FACTORS NECESSARY TO CALCULATE
C***+******CONCENTRATi ONS. •
X=R(IX) * T
= EXP(-T/DEC: )
= DECAY/(UPI""SZr!AT*SYOAT > $ Q1COEF =
HZ 1 = 0.5*H1 *Hl/Si:2
DO 900 NOP=1,3
I/SZ2
C**********DfcTE«MINING CONCENTRATION ( CL C ) ON PLUME AXIS AND AT 2 OIST-
C **********ANCES (Y) OFF AXIS.
DO 600 J=I,3
Y=X*XXX( J )
Y?=0.5*Y*Y/SY2
C1 = Q ICOEF*fcXP(-HZl-Y2)
C2 = 02( NOP>*COEF*t:xP(-HZ2(NOP )-Y2)
CLC = CI+C2 S XXNC
-------
AH'I- PF(lNTP.i),
C********** THROUGH STATEMENT 3r>0
r**********
DO 950 ID-1 . 16
DI •= Ft) $ 01 = ? 2.5* 01
PRINT 3,01 . SEASON. HNLW * PKINT
SMtLTFR = * . F h . 1 . "OrfGWCH ;i * . I OX ,
5
6
7
8
9
10
11
14
16
FORMAT
FORMAT
FORMAT
FORMAT
FORMAT
FORMAT
FORMAT
FORMAT
FORMAT
STOP
END
( 1H , 2E 1 5 . 3 )
{ 1 X, 14, I5.6F7 ,5, )
(F.8.1. 6F>3.1)
(4F8. I i
C/50X. *AVEHAGE CCKCf NTR A T I CNS - GM./CU. M*)
( IX. 12, 6F ft . 2 )
(7X, *D ISTANCd M *9X * AVCH AGE * )
(I2.13F6.0)
( 3F9.2.2E9 .1 >
94
-------
SL'BRUUTINE STCKHT
C**********
C ********** THI S ROUTINE CALCULATFS EFFECTIVE STACK t-f.MGHT
C**********HNUM = MAX ALLUWAULt EFFtCTlVE STACK HT. («)i USING AVFRAC
C**********P. M. MIXING DEPTH.
C*«** ******
COMMON /A/ FS1 t PS? (3) . H 1 . h2 ( 3 ) , t, 1ST, 1 5P . H 1 I ,H22,V I 1 iD2?.HN'JM
IF (IST.LT.5) GO TO 4*0
G = 5.9E-4
Hl= Hll +2.0 *{ (FSl/('J*G) )**0 .33 )
(F(HI.GT.HNUM) H1=HNLM
DO 410 NCP=1,3
H2(NOP)=H22»2.0 *((FE2(NOP)/(U*G))**0.33>
IF { H2( NOP) .GT.HNUM) H2 ( NUP )'=HN UW
410 CONTINUE
GO TO 450
440 H1=H11 +( 150.0*FSl/dJ*U*U ))
IF (HI.GT.HNUM) H1=HMJM
443 DO 450 NOP=I.3
H2(NOP)=H22 + (150.0*FS2(NUP
IF (H2
-------
Appendix B
BIBLIOGRAPHY
97
-------
Appendix R
MBLIOGRAPliY
A literature review has provided a collection of papers and reports
on topics related to the measurement and distribution of TSP concentra-
tions. The bibliography is arranged alphabetically by author and num-
bered consecutivelyo Each item in the bibliography has been classified
into the categories shown in Table B (see Section III of the text). The
resulting matrix is given in Table B-l. The bibliography item number is
entered in all the applicable categories. Some of the more comprehen-
sive reports may be tallied in several categories. Table B-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.
99
-------
Table B-1
LIBRARY 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 standards
2. Provide data for en-
vironmental impact
statements
3. Evaluate impact of
specific sources
4. Determine environ-
mental effects
5. Provide basis for
describing processes
that affect particu-
late concentration
6. Provide basis for
measures
7. Determine trends
8. Evaluate effects
on human exposure
9. Evaluation effects
on vegetation
10. Tests of methods
and equipment
11. Assess repre-
of existing
monitoring sites
12. Miscellaneous
Not Related
to Scale
2 74
53 97
4
98
72
7 47
9 52
10 70
31 89
33 94
34 134
43 135
7
1 5
39
53
27
7 62
10 97
36 118
61 138
117
1 50 90
7 54 91
8 57 103
13 59 104
14 64 HI
21 67 113
34 73 114
40 81 142
46 83
49 84
16
30
28
f.6
121
128
147
12 87 115
28 88 123
30 91 125
33 97 139
38 101 145
42 102
Applicable Scale of Measurement
Regional
Respir-
able
41
62
129
Total
53
74
4
37 77
51 107
76
65
37 78 112
51 79 116
52 95 120
58 103 136
63 106 143
65 107 148
70 108
76 109
77 110
53 75
60 82
63 100
74
23 27 130
24 80 131
25 103
26 116
5
117
129
5
17
109
18 128
77 133
86 136
95 147
96
121
Neighborhood
Resplr-
able
4
41
35
62
35
3
35
129
138
146
127
141
Total
53
Middle
Street
Canyon
!
i
144
22 56
48 68
51 69
55 76
7 79
22 95
48 108
51 109
55 116
58 136
63 143
71 144
76 148
78
7 74
53 75
60 82
63 122
68
67
85
115
5
7
48
117
129
5
17
109
124
126
11 89
16 95
18 96
20 99
45 121
55 124
56 132
85 133
86 136
22
51
10
22
32
51
137
143
82
3
10
117
18
85
Traffic
Corridor
98
6
82
3
18
85
Source Oriented
Elevated
120
19
69
80
19
126
140
141
29
19
19
Surface •
120
i
19
44
69
80
19 77
32 126
33 140
34 141
44
52
71
29
19
34
77
99
19
33
100
-------
BIBLIOGRAPHY
1 Achinger, W. C., and R. T. Shigehara, 1968: A Guide for Selected
Sampling Methods for Different Source Conditions, JAPCA,
18(9), pp. 606-609.
2 Akland, G. G., 1972: Design of Sampling Schedules, JAPCA, 22(4),
pp. 264-266.
3 Amdur, M., 1969: Toxicologic Appraisal of Particulate Matter,
Oxides of Sulfur and Sulfuric Acid, JAPCA, 19(9), pp. 638-
644.
4 Babcock, L. R., Jr., 1970: A Combined Pollution Index for
Measurement of Total Air Pollution, JAPCA, 20(10), pp. 653-
659.
5 Barrett, L. B., and T. E. Waddell, 1973: Cost of Air Pollution
Dan age, a Status Report:, EPA Publ. #AP-85.
6 Bayley, E., and A. Dockerty, 1972: Traffic Pollution of Urban
Environments, J. Royal Soc. Health, 92(1), pp. 6-11.
7 Benson, R. B., J. J. Henderson, and D. E. Caldwell, 1972; Indoor-
O-tdocr Air Pollution Relationships: A Literature Review,
EPA Publ. #AP-112.
8 Benson. F. B., W. C. Nelson, B. A. Newill, J. E. Thompson, M. Terabe,
S. Oomichi, and M. Nagata, 1970: Relationships Between Air
O^iality Measurement Methods in Japan and the United States - II •
Suspended Particulate Matter, Atm. Env., vol. 4, pp. 409-415.
9 Bliffcrd, I. H., Jr., (editor), 1971: Particulate Models: Their
Validity and Application, NCAR-TN/PROC-68, 21 Aug. 1970.,
Boulder, CO.
10 Bullock, J., and W. M. Lewis, 1968: The Influence of Traffic on
Atmospheric Pollution,, Atmos. Environ., 2^, 517-534.
11 Byers, H. R., et al., 1973: An Assessment of Representativeness and
Significance of Air Quality Measurements in Harris County,
Prepared for: Consortium of Gulf Universities, College Station,
XX, July 1973, 15 pp.
101
-------
12 Cadle, R. D., 1975: The Measurement of Airborne Particles, A Wiley-
Interscience Publication, John Wiley and Sons, N.Y., 342 pp.
13 Cannon, T. W., 1970: High Speed Photography of Airborne Atmospheric
Particles, J. App. Meteorol.. 9(1), pp. 104-108.
14 Carpenter, T. E., and D. L. Brenchley, 1972: A Piezoelectric Cascade
Impactor for Aerosol Monitoring, Am. Industrial Hyg. Assoc. J.,
pp. 503-510.
15 Charleson, R. J., 1970: Note on the Design and Location of Air
Sampling Devices, JAPCA. 23(10), pp. 881-886.
16 Charleson, R. J., N. C. Ahlquist, H. Selvidge, and P. B. MacCready, Jr.,
1969: Monitoring of Atmospheric Aerosol Parameters with the
Integrating Nephelometer, JAPCA, 19(12), pp. 937-942.
17 Clements, H. A., T. B. McMullen, R. J. Thompson, and G. G. Akland,
1972: Reproducibility of the HI-VOL Sampling Method under
Field Conditions, JAPCA. 22(12), pp. 955-958.
18 Clifton, M., D. Kerridge, W. Moulds, J. Pernberton, and J. K. Donoghue,
1959: The Reliability of Air Pollution Measurements in Relation
co the Siting of Instruments, Int. J. Air Poll., vol. 2,
pp. 188-197.
19 Cowherd, Jr., C., et al., 1974: Development of Emission Factors for
Fugitive Dust Sources, EPA-450/3-74-037, Research Triangle Park,
:;.c., 172 pp.
20 Curran. T. C., and W. F. Hunt, Jr., 1975: Interpretation of Air
Quality Data With Reppect to the National Ambient Air Quality
Standard, JAPCA, 25, 711-714.
21 Davies, G. 25.s 1968: The Sampling of Aerosols, Staub-Reinhalt. Luft.
28(6), 1-9.
22 Draftz, R. G., 1975: Types and Sources of Suspended Particles in
Chicago, Rpt. No. IITRI-C9914-C01, IIT Res. Institute, Chicago,
111.
102
-------
23 Environmental Protection Agency, 1973: National Air Quality Levels
and Trends in Total Suspended Particulates and SC>2 Determined
by Data in the National Air Surveillance Network, EPA, Office
of Air Quality Planning and Standards, Monitoring and Data
Analysis Division, RTP, N. C. 27711
24 Environmental Protection Agency, 1973: The National Air Quality
Program: Air Quality a.nd Emissions Trends, Annual Report,
vol. II, 350 pp.
25 Environmental Protection Agency, 1973: Monitoring and Air Quality
Trends Report, 1972, EPA-450/1-73-004, Monitoring and Data
Analysis Division, Durham, N. C. 218 pp.
26 Environmental Protection Agency, 1974: Administrative and Technical
Aspects of Source Sampling for Particulates, EPA-450/3-74-047,
RTP, N. C. 27711.
27 Environmental Protection Agency, 1974: Guidelines for the Evaluation
of Air Quality Trends, EPA OAQPS No. 1.2-014.
28 Environmental Protection Agency, 1974: Air Quality Monitoring Site
Description Guideline, EPA OAQPS No. 1.2-019.
29 Environmental Protection Agency, 1975: Guidance for Air Quality
Monitoring Network Design and Instrument Siting (Revised)
EPA OAQPS No. 1.2-012.
30 EPA Announces Partial Results of Collaborative Test Measurements
Program, JAPCA, 1974, 24(8), pp. 783.
31 Esrcen, N. A. and M. Corn, 1971: Residence Time of Particles in
Urban Air, Atsos. Env., J5, pp. 571-578.
32 Feeney, P. J., et al», 1975: Effect of Roadbed Configuration on
Traffic Derived Aerosols. JAPCA, 2_5, 1145-1147.
33 Fennelly, P. F., 1975: Primary and Secondary Particulates as
Pollutants - A Literature Review. JAPCA, 25, 697-704.
34 Gatz, D. F., 1975: Relative Contributions of Different Sources
of Urban Aerosols: Application of a New Estimation Method
to Multiple Sites in Chicago, Atm. Env., vol. 9, pp. 1-18.
103
-------
35 Giever, P. M., and W. E. Ruch, 1971: Statistical Analysis of Air
Pollution Sampling Frequency in an Epidemiological Study,
Am. Ind. Hyg. Asso. J., pp. 260-266.
36 Giocomelli-Maltoni, et al., 1972: Deposition Efficiency of Mono-
disperse Particles in Human Respiratory Tract, Am. Ind. Hyg.
Assoc. J., pp. 603-610.
37 Hamburg, F. C., 1971: Some Basic Considerations in the Design
of an Air Pollution Monitoring System, JAPCA, 21(10),
pp. 609-613.
38 Hagen, L. J., and N. P. Woodruff, 1973: Air Pollution from
Dust Storms in the Great Plains, Atmos. Env., 7_, pp. 323-332.
39 Herpertz, E., 1969: A Simple Long-Term Measuring Procedure for
Determining the Dust Concentration in the Near Ground Air
Layer (Lib-method), Staub-Reinholt Luft. 29(1), pp. 12-18.
40 Herrick, R. A., 1971: TR-2 Air Pollution Measurements Committee
Looks at EPA's Proposed Test Methods, JAPCA, 21(10), pp. 652-
653.
41 . Eicy, G. M., et al., 1975: Summary of the California Aerosol
Characterization Experiment. JAPCA, 25, 1106-1114.
42 Ecffnan, A. J., et al., 1975: EPA's Role in Ambient Air Quality
Monitoring. Science, 190, 243-248.
43 Horbath, H., and R. J. Charlson, 1969: The Direct Measurement of
Atmospheric Air Pollution, Am. Ind. Hygiene Assn. J., 30,
pp. 500-509.
44 Hudson, J. L., J. J. Stukel and R. L. Solomon, 1975: Measurement
of the Ambient Level Concentration in the Vicinity of Urbana-
Charapaign, Illinois. Atmos. Environ., £, 1000-1006.
45 Hunt, W. H., Jr., 1972: The Precision Associated with the Sampling
Frequency of Log-Normally Distributed Air Pollutant Measurements,
JAPCA. 22(9), pp. 687-691.
46 Husar, R. B., 1974: Atmospheric Particulate Mass Monitoring with a
Radiation Detector, Atmos. Env., 8, pp. 183-188.
104
-------
47 Jutze, G. A., and K. E. Foster, 1967: Recommended Standard Method
for Atmospheric Sampling of Fine Particulate Matter by Filter
Media -- High-Volume Sampler, JAPCA, 17(1), pp. 17-20.
48 Kneip, T. J., M. Eisenbud, C. D. Strehlow, and P. C. Freudenthal,
1970: Airborne Participates in New York City, JAPCA. 29(3),
pp. 144-149.
49 Knight, G., 1972: A Simple Method for Determining Size Distribution
of Airborne Dust by Its Settling Velocity, Am. Ind. Hyg. Assoc.
Ji, pp. 526-532.
50 Kramer, D. N., and P. W. Mitchel, 1967: Evaluation of Filters for
High Volume Sampling of Atmospheric Par.ticulates, Am. Ind.
Hyg. Assoc. J.. 28, 22.4.
51 Larsen, R. I., 1970: Relating Air Pollutant Effects to Concentration
and Control, JAPCA, 20(4), pp. 214-225.
52 Larsen, R. I., 1973: An Air Quality Data Analysis System for Inter-
relating Effects, Standards and Needed Source Reductions, JAPCA,
.23, 933-940.
53 Larsen, R. I., 1974: An Air Quality Data Analysis System for Inter-
relating Effects, Standards and Needed Source Reductions — Part 2,
JAPCA. 24(6), pp. 551-558.
54 Lavrence Berkeley Laboratory, 1975: Instrumentation for Environmental
Monitoring, LSL-1, vol. 1, Part 2, Air Environmental Instrumenta-
tion Group, U. of California, Berkeley, CA.
55 Leavitt, J. M., F. Pooler, Jr., and R. C. Wanta, 1957: Design and
Interim Meteorological Evaluation of a Community Network for
Meteorological and Air Quality Measurements, JAPCA, 7(3),
pp. 211-215.
56 LeClare, P. C., et al., 1969: Factors Affecting the Accuracy of Average
Dust Concentration Measurements, Am. Ind. Hyg. Assoc. J., Jul.
Aug. 1969, pp. 386-393.
57 Ledbetter, J. 0., and B. R. Fish, 1972: The Jet Filter - A Single-
Stage Size-Selective Sampler for Airborne Particulates, Am.
Ind. Hyg. Assoc. J., pp. 90-93.
105
-------
58 Lee, R. E., Jr., and R. K. Patterson, 1969: Size Determination of
Atmospheric Phosphate, Nitrate, Chloride, and Ammonoim
Particulate in Several Urban Areas, Atmospheric Environment,
vol. 3, pp. 249-255.
59 Lee, R. S., J. S. Caldwell and G. B. Morgan, 1972: The Evaluation of
Methods for Measuring Suspended Particulates in Air, Atmos. Env.,
6, pp. 593-622.
60 Levaggi, D. A., et al., 1975: Detailed Chemical Analysis of Airborne
Particulates, paper presented at the International Conference
on Environmental Sensing and Assessment, Sept. 14-19, 1975,
Las Vegas, Nevada. •
61 Lippmann, M., and R. E. Albert, 1969: The Effect of Particle Size
on the Regional Deposition of Inhaled Aerosols in the Human
Respiratory Tract, Am. Ind. Hyg. Assoc. J., May-June 1969,
pp. 257-275.
62 Lippmann, M., 1970: "Respirable" Dust Sampling, Am. Ind. Hyg. Assoc.
J_._, Mar-April 1970, pp. 138-159.
63 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, CA., 110 pp.
6^ Luna, R. E., H. W. Church, and J. D. Shreve, 1971: Variability
cf Air Sampler Data, Atmos. Env., 5_, pp. 579-597.
65 Ludngren, D. A., 1970: Atmospheric Aerosol Composition and Concen-
tration as a Function of Particle Size and of Time, JAPCA,
20(9), pp. 603-608.
66 Lundgren, D. A., 1971: Determination of Particulate Composition and
Size Distribution Changes with Time. Atmos. Env., 6_, pp. 645-651.
67 Lynian, D. R., J. 0. Pierce and J. Cholak, 1969: Calibration of the
High Volume Sampler, Am. Ind. Hyg. Assoc. J., 30, 83.
68 Macphee, R. D., and A. H. Bokian, 1967: Suspended Particulates in
the Los Angeles Atmosphere, JAPCA. 17, 580.
106
-------
69 Marchesani, V. J., T. Towers and H. C. Wohlers, 1970: Minor Sources
of Air Pollutant Emissions, JAPCA, ,20, 19-22.
70 McCaldin, R. 0., L. W. Johnson, and N. T. Stephens, 1969: Atmospheric
Aerosols, Science, 166, pp. 381-382 (MGA 21.5-269).
71 McCormick, R. M., and J. H. Ludwig, 1967: Climate Modification by
Atmospheric Aerosols, Science, 156, pp. 1358-1359.
72 McCormick, R. M., 1971: Air Pollution in the Locality of Buildings,
Phil. Trans. Roy. Soc. Land. A., pp. 515-526.
73 McKee, H. C., et al., 1972: Collaborative Testing of Methods to
Measure Air Pollutants, JAPCA, 22(5),
74 McMullen, T. B., R. B. Faoro, and G. B. Morgan, 1970: Profile of .
Pollutant Fractions in Nonurban Suspended Particulate Matter,
JAPCA. 20(6), pp. 369-372.
75 Megonnell, W. H., and S. Smith Griswold, 1966: Federal Air Pollution
Prevention and Abatement Responsibilities and Operations,
JAPCA. 16(10), pp. 526-529.
76 Munn, R. E., and B. Bolin, 1971: Review Paper - Global Air Pollution -
Meteorological Aspects, Atm. Env., vol. 5, pp. 363 -402.
77 Muan, R. E., 1973: A Study of Suspended Particulate Air Pollution
at Two Locations in Toronto, Canada, Atmos. Env., 7_, pp. 311-318.
78 Munn, R. E., D. A. Thomas, and A. F. W. Cole, 1969: A Study of
Suspended Particulate and Iron Concentrations in Windsor,
Canada, Atmospheric Environment, Vol. 3, pp. 1-10.
79 Myrup, L. 0., et al., 1974: Numerical Models of the Urban Atmospheric
Volume III, Gaseous a.nd Particulate Pollution, Contributions in
Atmos. Science, U. of California - Davis.
80 Nader, J. S., 1973: Developments in Sampling Analysis Instrumentation
for Stationary Sources, JAPCA, 24(6), pp. 587-591.
81 Neustadter, H. E., et al.3 1975: The Use of Whatman - 41 Filters for
High-Volume Air Sampling, Atmos. Env., vol. , pp. 101-109.
107
-------
82 O'Donnell, H., T. L. Montgomery and M. Corn, 1970: Routine Assess-
ment of the Particle Size-Weight Dis tribution of Urban Aerosols
Atmos. Environ., 4_, pp. 1-7, (APCAA 12943).
83 Ogle, H. M. , 1968: Particle Counting Techniques Available for Aerosol
Research, JAPCA, 18(10). pp. 657-659.
84 Oliri, J. G., and G. J. Sera, 1971: Piezoelectric Microbalance for
Monitoring the Mass Concentration of Suspen ed Particles,
Atmos. Environ,,, !>, pp. 658-668.
85 Ott, W. R., 1975: Development of Criteria for Siting Air Monitoring
Stations, presented at the 68th Annual Meeting of the APCA,
Boston, Massachusetts.
86 Ott, W. R., and D. T. Mage, 1974: Trend Assessment of Air Quality
Over Large Physical Areas by Random Sampling, presented at
the 4th Annual Environmental Engineering and Science Conference,
Kentucky, pp. 19.
87 Ott, W. R., and D. T. Mage, 1974: A Method of Simulating the True
Human Exposure of Critical Population Groups to Air Pollutants,
presented at Int. Syrap.: Recent Advances in the Assessment of
the Health Effects of Environmental Pollution, Paris, 11 pp.
88 Pasceri, R. E., 1954: The Size Distribution of Atmospheric Aerosols,
dissertation, The John Hopkins University, Dept. of Engineering,
University Microfilms, Inc., Ann Arbor, Mich.
89 Pate, J. B., and £. C. Tabor, 1962: Analytical Aspects of the Use
of Glass Fiber Filters for the Collection and Analysis of
Atmospheric Particulate Matter, Am. Ind. Hyg. J., 23, 145.
90 Paulas, H. J., and C. M. Peterson, 1967: Urban Aerosol Count-Size
Related to Meteorological Data, JAPCA, 17, 575.
91 Pedace, E. A., arid E. B. Sanson, 1972: The Relationship Between
"Soiling Index" and Suspended Particulate Matter Concentrations,
JAPCA. 22(5), pp. 348-351.
92 Feeder, M. A., 1972: The Correlation of Aitken Nucleus Character-
istics v/ith Mass Concentration in Atmos. Aerosol., Atmos. Env.,
6(10), pp. 735-742.
108
-------
93 Perkins, N. M. , 1973: Do Air Monitoring Station Data Represent
the Surrounding Community Exposure? Int. J. Biometeorology,
vol. 17, pp. 23-28.
94 Peterson, E. M., 1968: Measuring and Relating Atmospheric Pollution
to Meteorological Parameters, JAPCA, 18(10), pp. 654-656.
95 Peterson, C. M., and H. J. Paulus, and G. H. Foley, 1969: The
Number-Size Distribution of Atmospheric Particles during
Temperature Inversions, JAPCA, 19(10), pp. 795-801.
96 Phinney, D. E., and J. E. Newman, 1972: The Precision Associated
with the Sampling Frequencies of Total Particulate at Indianapolis,
Indiana, JAPCA, 22(9), pp. 692-695.
97 Pierrard, J. M., and discussion by J. P. Lodge, Jr., and P. W. West,
1969: Environmental Appraisal — Particulate Matter, Oxides of
Sulfur, and Sulfuric Acid, JAPCA, 19(9), pp. 632-637.
98 Pierson, W. R., and W. W. Brachaczek, 1975: Note on In-Traffic
Measurement of Airborne Tire-Wear Particulate Debris, JAPCA,
25(4), pp. 404-405.
99 Pooler, F., Jr., 1974: Network Requirements for the St. Louis
Regional Air Pollution Study, JAPCA. 24(3), pp. 228-231
100 Porch, W. M., R. J. Char1son and L. F. Radke, 1970: Atmospheric
Aerosols: Does a Background Level Exist, Science, 170,
??. 315-317.
101 Public Health Service, 1966: Air Pollution - A National Sample,
USPHS Publ. No. 1562.
102 public Health Service, 1969: Air Quality Criteria for Particulate
Matter, AP-49, U.S. Dept. of HEW, Wash., D.C., 211.
103 Pueschel, R. F., and B. G. Mendonca, 1972: Sources of Atmospheric
Particulate Matter on Hawaii, Tellus, 24(2), pp. 139-148.
104 Preference Method for the Determination of Suspended Particles in the
Atmosphere (Hi-Vol Method) Appendix B "National Primary and
Secondary Air Quality Standards," Federal Register, 36(84)
Part II, April 30, 1971.
105 Robson, C. D. and K. E. Foster, 1962: Evaluation of Air Particulates
Sampling Equipment, Am. Ind. Ilyg. Assn. J., 23, 404.
109
-------
106 Rodhe, H., and Jan Grandell, 1972: On the Removal Time of Aerosol
Particles from the Atmosphere by Precipitation Scavenging,
Tellus, 24(5), pp. 442-454.
107 Rodhe, H., C. Person, and 0. Akesson, 1972: An Investigation into
Regional Transport of Soot and Sulfate Aerosols, Atmos. Env.,
6_, pp. 675-693.
108 Ronicke, G., and R. Graul, 1970: The Deposition of Air Pollutants
on Water Surfaces , Staub-Reinhalt Luft, 30(9), pp. 21-24.
109 Rosen, J. M., 1969: The Vertical Distribution of Particulate Matter
1,'ear the Surface of the Earth, JAPGA. 19(2), pp. 106-108.
110 Rubin, E. S., 1974: The Influence of Annual Meteorological Variations
on Regional Air Pollution Modeling: A Case Study of Allegheny
County, Pennsylvania, JAPCA, 24(4), pp. 349-356.
Ill Ruping, I. G., 1968: The Importance of Isokinetic Suction in Dust
Flow Measurement by Means of Sampling Probes, Staub-Reinhalt.
Luft. 28(4), pp. 1-11.
112 Rydell, C. P., and B. Schwarz, 1968: Air Pollution and the Shape
of Urban Areas, J. Am. Inst. of Planners, pp. 50-51.
113 Saltznan, B. E., 1968: Standardization of Methods for Measurement
of Air Pollutants, JAPCA. 18(5), pp. 326-329.
114 Saltraan, B. E., 1970: Significance of Sampling Time in Air
Monitoring, JAPCA, 20(10), pp. 660-665.
115 Saltznan, B. E., 1972: Simplified Methods for Statistical Interpreta-
tion of Monitoring Data, JAPCA. 22(2), pp. 90-95.
116 Schaefer, V. J. 1970: The Inadvertent Modification of the Atmosphere
by Air Pollution, Bull A.M.S., 50(4), pp. 199-206.
117 Schimel, H., and L. Greenburg, 1972: A Study of the Relation of
Pollution to Mortality, New York City, 1963-1968, jAPCA. 22(8),
pp. 607-616.
118 Schroeder, H. A., 1970: A Sensible Look at Air Pollution by Metals,
Arch. Env. Health, vol. 21, Dec., pp. 798-806.
110
-------
119 Schulze, R. H., 1973: The Economics of Environmental Quality
Measurement, JAPCA, 23(8), pp. 671-675.
120 Scriven, R. A., and B. E. A.. Fisher, 1975: The Long Range Transport
of Airborne Material and Its Removal by Deposition and Washout -
I. General Considerations, Atmos. Env., _9, pp. 49-58.
121 Seinfeld, J. H., 1972: Optimal Location of Pollutant Monitoring
Stations in an Air Shed, Atm. Env., vol. 6, pp. 847-858.
122 Severs, R. K., 1971: Use of Statistical Techniques to Assess Progress
Toward a Clean Air Environment: Houston, TX, Atmos. Envir.. .5,
pp. 853-861.
123- Shah, G. M., 1970: Study of Aerosols in the Atmosphere by Twilight
Scattering, Tellus XXII, 1, 82-93.
124 Shelar, E., 1974: Variability of High-Volume Sampling, Masters
Thesis, Texas A&M University, College Station, Texas.
125 Siegel, R. D., J. R. Ehrenfeld, and P. Morgenstern, 1975: A Strategy
for Reduction of Particulate Emission in the Boston Area, JAPCA,
25(3), pp. 256-259.
126 Saffian, G., and P. R. Westlin, 1975: Particulate Emissions from
Apartment House Boilers and Incinerators, JAPCA. 25(3), pp. 269-
273.
127 Sccka, A., R. Marek, and L. Gnan, 1975: A New Approach to Roof
Monitor Particulate Sampling, JAPCA, 25(4), pp. 397-398.
128 South Coast Air Basin Coordinating Council, Air Monitoring Site
Criteria, Air Monitoring Committee, Tec. Advisory Committee,
10 pp.
129 Speizer, F. E., and discussion by I. J. Selikoff, 1969: An Epidemiolo-
gical Appraisal of the Effects of Ambient Air on Health:
Particulates and Oxides of Sulfur, JAPCA, 19(9), pp. 647-656.
130 Spirtas, R., and H. J. Levin, 1970: Characteristics of Particulate
Patterns, 1957-1966, U.S. Public Health Service Publ. AP-61,
101 pp.
Ill
-------
131 Spirtas, R., and H. J. Levin, 1971:' Patterns and Trends in Levels
of Suspended Participate Matter, JAPCA, 21(6), pp. 329-333..
132 Stalker, W. W., and R. C. Dickerson, 1962: Sampling Station and
Time Requirements for Urban Air Pollution Surveys, Part II:
Suspended Particulate Matter and Soiling Index, JAPCA, 12,
pp. 111-128.
133 Stalker, W. W., R. C. Dickerson, and G. D. Kramer, 1962: Sampling
Station and Time Requirements for Urban Air Pollution Surveys,
JAPCA. 12(8), pp. 361-375.
134 Stanford Research Institute, 1972: Regional Air Pollution Study:
A Prospectus, Part III - Research Plan, Final Report EPA
Contract 68-02-0207. Stanford Research Institute, Menlo Park,
CA., 278 pp.
135 Stanford Research Institute, 1972: Regional Air Pollution Study:
A Prospectus, Part III - Research Facility, Final Report EPA
Contract 68-02-0207. Stanford Research Institute, Menlo Park,
CA., 167 pp.
136 Stewart, I. M., and Matheson, D. H., 1968: Methods of Relating High-
Volume Sampler Particulate Loadings to Wind Direction. Atmos.
Env., 2, pp. 181-185.
137 Stukel, J. J., R. L. Solomon and J. L. Judson, 1975: A Model for
the Dispersion of Particulate or Gaseous Pollutants from a
Network of Streets and Highways. Atmos. Environ., 9, pp. 990-999.
138 Task Group on Lung Dynamics, 1966: Deposition and Retention Models
for Internal Dosinetry of the Human Respiratory Tract. Health
Physics. 12, pp. 173-207.
139 Thompson, C. R., E. G. Hensel, and G. Kats, 1973: Outdoor-Indoor
Level of Six Air Pollutants, JAPCA, 23(10), pp. 881-886.
140 Vandergrift, A. E., L. J. Shannon, E. E. Sallee, P. G. Gorman, and
W. R. Park, 1971: Particulate Air Pollution in the United
States, JAPCA, 21(6), pp. 321-328.
141 Walther, E. G., 1972: A Rating of the Major Air Pollutants and
Their Sources by Effect, JAPCA, 25(5), pp. 352-355.
112
-------
142 Warner, P. 0., L. Saad, and J. 0. Jackson, 1972: Identification
and Quantitative Analysis of Particulate Air Contaminants
by X-Ray Diffraction Spectrometry, JAPCA. 22(11), pp. 887-890.
143 Webb, J. C., J. C. Kinchen, and J. E. Scarberry, 1970: Aerial
Sampling by Helicopter Using a High Volume Sampler, JAPCA,
. 20(7), pp. 453-455.
144 Weisraan, B., D. H. Matheson, and M. Hirt, 1969: Air Pollution
Survey for Hamilton, Ontario, Atmospheric Environment,
vol. 3, pp. 11-23.
145 Williams, J. D., J. R. Fanner, R. B. Stephenson, G. G. Evans, and
R. B. Dalton, 1968: Air Pollutant Emissions Related to Land
Area - A Basis for a Preventive Air Pollution Control Program,
NAPCA Publ. No. APTD-68-11.
146 Winkelstein, Warren, Jr., and Michael L. Gay, 1971: Suspended
Particulate Air Pollution, Relationship to Mortality from
Cirrhosis of the Liver, Arch. Env. Health, vol. 22, Jan.,
pp. 174-177.
147 Yamada, V. M., 1970: Current Practices in Siting and Physical
Design of Continuous Air Monitoring Stations, JAPCA, 20(4),
pp. 209-213.
148 Zinner, C.. E., E. C. Tabor and .A. C. Stem, 1959: Particulate
Pollutants in the Air of the U.S., JAPCA, 9, 136.
113
-------
REFERENCES
115
-------
PEFEP.KNCES
Beals, C. A. 1971; Guide to Local Diffusion of Air Pollutants; Air
Weather Service Tech. Report 214, Scott Air Force Base, Illinois,
Nat. Tech. Info. Service. (NTIS) No. An 726-984, 85 pp.
Beaton, J. L. ct al. , 1972: >'athenatical Approach to Estimating
Highway Impact on Air Quality, Appendix; California Division of
Highways, NTIS' Mo. TID25075, 81 pp.
Blifford, I. II., Jr. and 1,. D. Ringer, 1969: The Size and Number
Distribution of Aerosols in the Continental Troposphere, J.
Atmos. Sci., 26, 716-726.
Bosch, J. C., Jr. 1975: Aerometric Emissions Reporting System; EPA,
Office of Air and Waste Management, National Air Data Branch,
Durham, North Carolina, 96 pp.
Briggs, C. A., 1973: Diffusion Estimation for Small Emissions,
Draft No. 79, in 1973 Annual Report of the Atmospheric
Turbulence and Diffusion Laboratory (MOAA), Oak Ridge, Tennessee
Report No. ATDL-106, pp. 83-195.
Bureau of the Census, 1972: Census Tracts, Saginaw, Michigan, Standard
Metropolitan Statistical Area, Publication No. PHC(1)-179.
F.usse, A. D., and J. R. Zimmerman, 1973: User's Guide for the
Climatological Dispersion Model, EPA Report EPA-R4-75-024,
131 pp.
California Department of Transportation, 1975: 1974 Traffic Volumes
on the California State Highway System, 182 pp.
Charlson, R. J., N. C. Ahlquist, H. Sclvidge, and P. B. MacCready, Jr.,
1969: Monitoring of Atmospheric Aerosol Parameters with the
Integrating Nepheloneter, J. Air Poll. Cont. Assoc., 19, 937-942.
Clements, H. A., T. F,. McMullen, P. J. Thompson, and C. F. Akland,
1972: P.eproducibility of the UI-VOL Sampling Method Under Field
Conditions; J. Air. Poll. Cont. Assoc., 22, 955-958.
Cowherd, C. C., Jr., K. Axetell, Jr., C. M. Cuenther, and C. A. Jutze,
1974: Development of Emission Factors for Fugitive Dust Sources;
EPA Contract 68-02-0619. Midwest Pesearch Institute, Kansas City,
Missouri 64110, EPA Pub. No. 450/3-74-037, 172 pp.
117
-------
Dabberdt, U. F. and P. A. Davis, 1972: Do.termination of Energetic
Characteristics of Urban-Rural Surfaces in the Greater St. Louis
Area, Final Report EPA Contract 68-02-1015;, EPA Report Ho. EPA
650/4-74-0070, Stanford Research Institute, venlo Park,
California, 101 pp.
Draftz, R. G., 1975: Types and Sources of Suspended Particles in
Chicago; Prepared for Chicago Dept. of Environmental Control,
IIT Research Institute, Chicago, Illinois, 33 pp.
Environmental Protection Agency, 1972: Air Quality Data for 1968 from
the National Air Surveillance Network and Contributing State and
Local Networks; Office of Air Programs Pub. Mo. APTD-0978, 231 pp.
Federal Register, 1971: Volume 36, Part 50 — National Primary and
Secondary Ambient Air Quality Standards, pp 22384-22397.
Fennelly, P. F. 1975: Primary and Secondary Particulates as Pollutants,
A Literature Review, J. Air Poll. Cont. Assoc., 25, 697-704.
Flocchini, R. G., et al., 1976: Monitoring California's Aerosols by
Size arid Elemental Composition, Environ. Sci. and Tech., 10,
76-82.
Foxvog, F. R., 1975: Optical Scattering per Unit Mass of Single
Particles, Appl. Optics, 14, 269-270.
Georgii, 11. W., E. Rusch, and E. V.'eber, 1967: Investigation of the
Temporal and Spatial Distribution of the Immission Concentration
of Carbon Monoxide in Frankfurt/Main, Report TIo. 11 of the
Institute for Meteorology and Geophysics of the University of
Frankfurt/Main; Translation No. 0477, Nat. Air Poll. Cont.
Admin.
Cifford, F. A., 1973: Building and Stack Aerodynamic Effects,
Unpublished lecture notes; Atmospheric Turbulence and Diffusion
Laboratory (NCAA), Oak Ridge, Tennessee.
Ualitsky, J., 1961: Estimation of Stack Height Requirements to Limit
Contamination of Ruilding Air Intakes, J. Amer. Indust. Hygiene
Assoc., 26, 106-115.
llidy, C. M. et al., 1975: Summary of the California Aerosol Characteri-
zation Experiment, J. Air Poll. Cont. Assoc., 25, 1106-1114.
Hoffman, A. J., et al., 1975: EPA's Pole in Ambient Air Quality
Monitoring, Science, 190, 243-248.
Holzworth, C. C., 1971: Mixing Heights, Wind Speeds and Potential for
Urban Air Pollution Throughout the Contiguous United States;
EPA Div. of Meteorology, Raleigh, North Carolina, 146 pp.
118
-------
Johnson, T.'. B., \1. F. Dabberdt, r. L. Ludwi.g and P. J. Alien, 1971:
Field Study for Initial evaluation of an Urban Diffusion Model
for Carbon Monoxide, Comprehensive Report, HPA/CPC Contract
CAPA-3-68 (1-89); Stanford research Institute, "enlo Park,
California, NTIS No. PH 2C3469, 240 pp.
Junge, C. E., E. Robinson, and F. L. Ludwig, 1969: A Study of Aerosols
in Pacific Air Masses, J. Appl. Meteorol., i;, 340-347.
Kinosian, J. F.. , and D. Sineroth, 1973: The Distribution of Carbon
Monoxide and Oxidant Concentrations in Urban Areas; California
Air Resources Board, 19 pp.
Lapple, C. C., 1961: Characteristics of Particles and Particle
Dispersoids; Graph pub. by Stanford Research Institute, Menlo
Park, California.
Lawrence Berkeley Laboratory, 1975: Instrumentation for Environmental
Monitoring, AIR, Part 2; Environmental Instruments Group, LBL,
Univ. of California, Berkeley, California.
Lee, R. E., Jr., J. S. Caldwell, and G. E. Morgan, 1972: The
Evaluation of Methods for Measuring Suspended Participates in
Air, Atmos. Environ., 6, 593-622.
Lee, P.. E., R. K. Patterson, and J. Uagman, 1968: Particle Size
Distribution of Metal Components in Urban Air, Environ. Sci. and
Tech., 2, 13-17.
Ludwig, F. L., and J.H.S. Kealoha, 1975: Selecting Sites for Carbon
Monoxide Monitoring, Final Report, CPA Contract 68-02-1471;
Stanford Research Institute, Menlo Park, CA, 149 pp.
Ludwig, F. L. and E. Robinson, 1971: Observations of Aerosols and
Droplets in California Rtratus-II, Tell.us, 23, 164-175.
Mann, C. 0., and C. C. Cowherd, Jr., 1975: Fugitive Oust Sources; in
Supplement No. 5 for Compilation of Air Pollutant Emission Factors,
Second Edition. EPA Office of Air Quality Planning and Standards,
Research Triangle Park, North Carolina, pp 11.2-1 to 11.2-4.
Midwest Research Institute, 1971: Particulate Pollutant System Study,
Vol. I, Mass Emissions, Peport for EPA Contract Ho. CPA 22-/>9-104.
national Oceanographic and Atmospheric Administration, 196": Climatic
Atlas of the United States, 80 pp.
Ott, W., 1971: An Urban Survey Technique for Measuring the Spatial
Variation of Carbon Monoxide i.n Cirip.s, Phi) Thesis, ^ept. of
Civil Engineering, Stanford University, Stanford,
California, 153 pp.
119
-------
Ott, W., 1975: Development of Criteria for Siting Air Monitoring
Stations, paper presented at 68th Meeting of Air Poll. Cont.
Assoc., Boston, Massachusetts, June 1975.
Pace, T. G., W. P. Freas and E. M. Afify, 1977: Quantification of Rela-
tionship Between Monitor Height and Measured Particulate Levels in
7 U.S. Urban Areas, Pres. at 70th Meeting Air Poll. Cont. Assoc.,
Toronto. Paper 77-13.4
Pasquill, F., 1961: The Estimation of the Dispersion of Windborne
Material Meteorological Mag., (London), 90, 33-49.
Perkins, N. M., 1973: Do Air Monitoring Station Data Represent the
Surrounding Community Exposure? Int. J. Biotneteorology, 17,
23-28.
Peterson, C. M., H. J. Paulus, and C. H. Foley, 1969: The Number Size
Distribution of Atmospheric Particles During Temperature
Inversions, J. Air Poll. Cont. Assoc., 19, 795-801.
Ranz, W. E., and J. B. Wong, 1952: Impaction of Dust and Smoke
Particles, Ind. Eng. Chem., 44, 1371-1380.
Record, F. A., 1976: National Assessment of the Urban Particulate
Problem, Final Report, EPA Contract 68-02-1376, Task Order
18; EPA Report 450/3-76-024, GCA Corp., Bedford, Massachusetts,
Robinson, E., and F. L. Ludwig, 1967: Particle Size Distribution of
Urban Lead Aerosols, J. Air. Poll. Cont. Assoc., 17, 664-668.
Shelar, E., 1974: Variability of High-Volume Sampling, Thesis for MS in
Meteorology, Texas A & M University, 78 pp.
Smith, M. (ed.), 1968: Recommended Guide for the Prediction of the
Dispersion of Airborne Effluents, Amer. Soc. Mech. Eng., New
York, 85 pp.
Stanford Research Institute, 1972: Regional Air Pollution Study: A
Prospectus, Part III - Research Facility, Final Report EPA
Contract 68-02-0207; Stanford Research Institute, Menlo Park,
California, 167 pp.
Task Croup on Lung Dynamics, 1966: Deposition and Retention Models for
Internal Dosimetry of Human Respiratory Tract, Health Physics,
12, 173-207.
Turner, D. B., 1969: Workbook of Atmospheric Dispersion Estimates,
Nat. Air Poll. Cont. Admin., 84 pp.
U. S. Geological Survey, EROS Data Center, 1974: Pamphlet INF-74-22
(R.I).
U.S. Public Health Service, 1958: Air Pollution Measurements of the
National Air Sampling Networks, Pub. No. 637, 259 pp.
120
-------
TECHNICAL REPORT DATA
/'/iv.sv rcaJ luunn'ii<»is on the /rivn'i before i'i'"i/
1. REPORT NO.
EPA-450/3-77-018
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Selecting Sites for Monitoring Total Suspended
Particulates
5. REPORT DATE
June, 1977, Revised Dec, 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
•*F.L. Ludwig, J.H.S. Kealoha, and E. Shelar
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Stanford Research Institute
Menlo Park, CA 94025
10. PROGRAM ELEMENT NO.
2AF643
11. CONTRACT/GRANT NO.
68-02-2053
12. SPONSORING AGENCY NAME AND ADDRESS
OAQPS
Monitoring & Data Analysis Division
Research Triangle Park, N.C. 27711
13. TYPE OF REPORT AND PERIOD COVERED
r I r\*x.i
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Abstract enclosed within documemt.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS c. COSATI Field/Group
18. DISTRIBUTION STATEMEN1
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
141
Unlimited
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
121
------- |