EPA-450/3-77-013
April 1977
OPTIMUM SITE
EXPOSURE CRITERIA
FOR SO2 MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
EPA-450/3-77-013
OPTIMUM SITE
EXPOSURE CRITERIA
FOR SO2 MONITORING
by
Robert J. Ball and Gerald E. Anderson
The Center for the Environment and Man, Inc.
275 Windsor Street
Hartford, Conn. 06120
Contract No. EPA 68-02-2045
EPA Project Officer: Alan Hoffman
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
April 1977
-------
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 The
Center for the Environment and Man, Inc. , 275 Windsor Street, Hartford,
Conn. 06120, in fulfillment of Contract No. EPA 68-02-2045. The contents
of this report are reproduced herein as received from The Center for the
Environment and Man, Inc. The opinions, findings, and conclusions
expressed are those of the author and not necessarily those of the Environ-
mental Protection Agency. Mention of company or product names is not
to be considered as an endorsement by the Environmental Protection Agency.
Publication No. EPA-450/3-77-013
11
-------
ABSTRACT
This report presents procedures and exposure criteria for selecting SO2
monitoring sites. General monitoring program elements and data uses are first
reviewed and summarized; from this summary a list of specific siting objectives
is developed. Site selection procedures were then prepared for specific site
types each of which was associated with either a grouping of siting objectives
or with an individual objective.
Because of the variety of SO2 siting objectives, the averaging times asso-
ciated with the air quality standards, the physical environments in which sites
could be located, and spatial scales of representativeness, an SC>2 monitoring
"universe" was first constructed from which the final list of monitoring site
types and associated site selection procedures was developed.
Detailed procedures are provided for selecting sites to measure regional
mean concentrations, interregional SC>2 transport, representative concentrations
for areas of various sizes, peak concentrations in urban areas, and emergency
episode levels. Procedures for selecting sites to monitor impacts from iso-
lated point sources in a variety of physical settings including valleys and
coastal areas are also provided. A general guideline for locating sites in
mountainous terrain is included. Recommendations for inlet height and orien-
tation, and for minimizing undue influence from nearby sources are presented.
The rationale behind the various procedures and other support documentation is
given.
Sources of special information and data relevant to selecting specific
sites and guidelines for determining locations of sites for satisfying speci-
fic objectives are provided in a series of appendices. A bibliography, con-
veniently arranged according to specific subject areas, is included.
111
-------
ACKNOWLEDGEMENTS
The authors would like to thank Messrs, Alan Hoffman, Norman Possiel, and
Neil Berg of the Environmental Protection Agency for their comments and advice.
Also, the assistance provided by Messrs. Edward Sweeton and Joseph Sekorski of
GEM in modifying computer program is acknowledged. Other CEM staff providing
assistance include Ms. Margaret G. Atticks, Ms. Mary-Ellen Albert, and Ms.
Carmella Miller who carried out the formidable task of preparing the figures,
typing, and other contributions to the preparation of this report.
IV
-------
SUMMARY
Studies (e.g., Yamada, 1970; and Jutze and Tabor, 1953) show that little
consistency exists among monitoring site locations, their surrounding physical
environments, and the uses for which the data are intended. This report was
prepared to provide a means for achieving such consistency, for S02 monitoring,
by establishing a logical procedure by which various data uses can be related
to opta -urn site locations and instrument inlet exposures. The physical charac-
teristics of the area surrounding the site were also considered in developing
the relationships.
To give a national perspective to SO2 emission patterns and problems, in
the northern areas the residential space-heating and power plant emission cate-
gories are the largest. In the south, emissions from transportation, power
plants, and industrial processes dominate; while in the west the industrial
process emission category is the largest. Over the entire nation, SC>2 problems
can range in complexity from single source impacts over flat terrain to impacts
from multi-source urban-industria.1 complexes located in complex terrain set-
tings.
There are twelve major uses for which SC>2 data are required.
1) Judging attainment of SO2 NAAQS.
2) Evaluating progress in achieving/maintaining the NAAQS or
state/local standards.
3) Developing or revising state implementation plans (SIPs)
to attain/maintain NAAQS; evaluating control strategies.
4) Reviewing new sources.
5) Establishing baseline air quality levels for preventing
significant deterioration (PSD) and for air quality main-
tenance planning (AQMP).
6) Developing or revising national SC>2 control policies [e.g.,
new source performance standards (NSPS), tall stacks, sup-
plementary control systems (SCS)].
7) Providing data for model development and validation.
8) Providing data to implement the provisions of the Energy Supply
and Environmental Coordination" Act (ESECA) of 1974.
-------
9) Supporting enforcement actions.
10) Documenting episodes and initiating episode controls.
11) Documenting population exposure and health research.
12) Providing information to:
a) public - air pollution indices; and
b) city/regional planners, air quality policy/decision
makers - for activities related to programs such as
air quality maintenance planning (AQMP), prevention
of significant deterioration (PSD) and the prepara-
tion of environmental impact statements.
The above uses of SO2 data were determined from a literature survey. For the
most part, they are broadly program oriented and constitute all of the monitor-
ing requirements that are necessary to successfully implement thost federal and
state clean air policies that require the use of SO2 ambient monitoring data.
The broad program orientation of the above uses made it very difficult to
associate each with a specific monitoring site selection procedure. To obviate
this problem, a list of specific monitor siting objectives was developed (also,
mainly via literature survey)tto provide a link between an intended use of the
data and a specific site selection procedure. The siting objectives are couched
in terms more reflective of the means by which the appropriate data will be ob-
tained rather than in terms having a broad program connotation. The siting ob-
jectives and associated data uses are listed below.
Determination of peak concentrations in urban areas
- major data uses: 1, 2, and 3.
- other data uses.- 8, 9, and 12.
Determination of the impact of individual point sources in multi-source
urban settings
.- major data uses: 3, 4, 6, 8, and 9.
- other data uses: 12.
Determination of the impact of isolated point sources
- major data uses: 3, 4, 6, 8, and 9.
- other data uses: 5 and 12 ,
Assessment of Interregional SO? Transport
- major data uses: 2, 3, 5, and 12.
Determination of base concentrations in areas of projected growth
- major data uses: 5 and 12.
Initiation of Emergency Episode Abatement Actions
-.- major data uses: 10 and 12.
Determination of Population Exposure
- major data uses: 11 and 12.
VI
-------
Assessment of Background Concentrations in Rural Areas
- major data uses: 5 and 12.
- other data uses: 2 and 3.
Diffusion Model Calibration and Refinement*
- major data use: 7.
Each of the above siting objectives can be associated with a spatial scale.
For example, a regional mean concentration is meant to represent conditions over
a large area while the peak concentration zone in an urban area may be repre-
sented by a spatial scale no more than about a city block or so, in size. The'
spatial scales of interest in SC>2 monitoring (and, perhaps, other pollutants as
well) can be classified as follows:
Microscale. Ambient air volumes with dimensions ranging from
meters up to about 100 meters are associated with this scale.
Middle Scale. This scale represents dimensions of the order
from about 100 meters to 0.5 kilometers and characterizes areas
up to several city blocks in size.
Neighborhood Scale. Neighborhood-scale measurements would
characterize conditions over areas with dimensions in the 0.5
to 4.0 kilometer range.
Urban Scale. Urban-scale measurements would be made to repre-
sent conditions over areas with dimensions on the order of 4
to 50 kilometers.
Regional Scale. Conditions over areas with dimensions of as
much as hundreds of kilometers would be represented by regional-
scale measurements. These measurements would be applicable
mainly to large homogeneous areas, particularly those which are
sparsely populated.
National and Global Scales. These measurement scales represent
concentrations characterizing the nation and the globe as a
whole.
Urban scale conditions would, in general, require more than a single monitoring
site to characterize them. For this reason, monitoring sites established for
measuring concentrations representing volumes of this scale were not addressed
in this report. National and global scale measurements are not of sufficient
interest to state and local agencies to justify specific treatment. The re-
maining scales are relevant to SC>2 monitoring.
Since it is difficult to anticipate the monitoring requirements of diffusion
model calibration and refinement projects, and impossible to generalize re-
lated siting guidelines, an attempt to do so was considered beyond the scope
of the objectives of routine monitoring which this report addresses.
Vll
-------
In SC>2 monitoring, a distinction must be made between the spatial scale
desired to be represented by a single measurement and the scale actually re-
presented by that measurement. The former is a function only of the intended
use of the data and associated siting objective while the latter is a function
mainly of the horizontal concentration gradient prevailing around the site.
For long averaging times, urban, concentrations generally decrease outward from
maxima near the center with steep concentration gradients prevailing over the
central portions of a city, while in the outer portions (adjacent suburbs and
rural areas) the gradients would be moderate to flat. Therefore, over the
central portion of the city> middle and neighborhood spatial scales would be
those most likely to be represented by single measurements; similarly, single
measurements would represent neighborhood scales in suburban areas and region-
al scales in rural areas.
SC>2 monitoring sites may be classified as either proximate or general
level. Proximate sites are oriented toward measuring concentrations resulting
from a specific source or group of sources; general level sites are associated
with measurements of total concentrations where contributions from an individ-
ual source or a group of sources are either not required to be known or do not
predominate. Generally, siting objectives (and associated monitoring sites)
are either "pattern" oriented, such as sites established to measure urban peak
concentrations, or they are associated with specified geographical areas such
as areas of projected growth or locations of specific population groups.
S02 measurements should, depending on the specific siting objective and
related data use, represent 3-hour, 24-hour and annual averaging times,- which
are those of the National Ambient Air Quality Standards for SO2-
All of the above parameters regarding spatial scales, siting objectives,
averaging times, etc. when combined with basic land use types and topographi-
cal settings, represent an S02 monitoring universe. From this universe, a set
of five monitoring site types were established for which site selection pro-
cedures were developed.
1. General-Level, Regional-Scale Stations
a. General site location; located in areas of flat concentration
gradients and low emission densities; usually homogeneously
rural settings.
b. Siting objectives and related data uses:
Assessment of background concentrations in rural areas
- major data uses: 5 and 12.
- other data uses: 2 and 3.
Assessment of interregional SO2 transport
T major data uses: 2, 3, 5, and 12.
II. General-Level, Neighborhood-Scale Stations
a. General site location; located in areas of moderate SC>2 background
concentration gradients, generally in areas adjacent to the central
business districts (CBDs) of cities and suburbs.
CONTINUED
Vlll
-------
b. Siting objectives and related data uses:
Determination of population exposure
- major data uses: 11 and 12.
Determination of base concentrations in areas of projected growth
- major data uses: 5 and 12.
Initiation of emergency episode abatement actions
- major data uses: 10 and 12.
III. General-Level, Middle-Scale Stations
a. General site location: located mainly within CBDs of cities or
industrial districts> where SO2 concentrations are the highest and
concentration gradients are steepest.
b. Siting objectives and related data uses:
Determination of peak concentrations in urban areas
- major data uses: 1, 2, and 3.
- other data uses: 8, 9, 10, and 12.
IV. Proximate, Middle-Scale Stations
a. General site location; located in the area of the highest ground-
level concentration associated with an individual major SO2 source
or group of S02 sources. Sources may be located in either urban,
suburban, or rural areas.
b. Siting objectives and related data uses:
Determination of the impact of individual point sources in
multi-source urban settings
major data uses: 3, 4, 6, 8, and 9.
other data uses: 12.
Determination of the impact of isolated point sources
major data uses: 3, 4, 6, 8, and 9.
other data uses: 5 and 12.
V. Proximate, Microscale (mobile sampling) Measurements
General Comments. Microscale measurements are obtained via mobile
sampling in regions of complex terrain and in urban areas. They
are used mainly to define certain features of the SC>2 pattern such
as peak concentrations associated with major point sources, par-
ticularly under atmospheric stagnation conditions, and impacts due
to plume downwash. These measurements may be used in support of
the siting objectives and related data uses associated mainly with
Site Type IV above.
Each of these site types (except V, which is associated with mobile sampling)
is associated with a basic procedural siting approach with variations from
the basic approach being functions of the individual siting objective, averag-
ing time and physical setting. The siting approaches are basically step-by-
step procedures through which a siting area is first selected, then site
choices are gradually reduced until the best choice site is selected.
IX
-------
Diffusion modeling results and graphical solutions to the Gaussian diffu-
sion equation provide the basic initial guidance for locating the monitor sit-
ing area for nearly all objectives. An elimination process then follows, which
is essentially a procedure for choosing a site location such that undue influ-
ences from nearby sources are eliminated or minimized, in this regard, several
"interference distances" (ID) or minimum separation distances were calculated
via Guassian diffusion equations. These IDs were determined such that the
maximum concentration resulting at the monitoring site from a single source
would be less than 2.6 yg/m3 (which is the natural S02 background level) for
sites located in rural areas,or less than 10 yg/m3 (level or cleanest rural
air) for sites located in urban or suburban areas. Different sets of IDs were
calculated for urban, rural, and suburban diffusion conditions. Also, half-
life values for S02 of one hour and three hours were used for urban and rural
areas, respectively. The IDs are summarized in the tables below.
Source Types and Related Interference Distances*
For Regional-Scale Monitoring Stations.
Source Type
Interference Distance
Large point source (e.g., a 400 MW power
plant) .
Industrial Source (500 tons S02 per year).
Towns (various size population)
50,000
25,000
12,500
6,000
Individual home
30 km
10 km
22 km
15 km
10 km
7 km
0.6 km
* Based on an undue influence concentration of 2.6 yg/m
3"
Interference Distances for Three Development Intensities.
Interference Distances
Minor Sources (MSID)
Urban 200 m
Suburban 100 m
Rural1^ 60 m
Q
Point Sources (PSID)
1,000 m
2,200 m
3,200 m
a Based on undue inrluence concentration of 10.0 yg/m3.
b Minor sources include individual office building in the urban area
(105 gal/yr #2 oil) , small building in the suburban areas
gal/yr) and an individual home in the rural area (103 gal/yr) ,
Point sources include any source that uses 10 gal. of fuel oil per
year or more (see text) .
Rural settings adjacent to cities where steeper concentration gradi-
ents preclude regional scale stations.
x
-------
It was inferred from the literature survey that the generally turbulent
and well-mixed urban atmosphere produces SC>2 concentrations that are essential-
ly uniform with height up to about mean building height; therefore, the exact
vertical placement of the instrument inlet below this height was not important
for averaging times of several hours or more. For guidance purposes, we chose
a height of 0.8 of the mean building height in the urban area below which a
site could be established (or instrument inlet placed).
The most important elements of the siting methodologies for establishing
monitoring stations associated with each of the four major monitoring site
types are tersely summarized below. Also included are some of the desirable
physical characteristics of the site. These methodologies are described in
much more detail in the body of this report (Section 4.0).
I. General-Level, Regional-Scale Stations
a. Siting Objectives; To measure regional mean concentrations and
concentrations resulting from interregional transport processes.
b. Spatial Scale of Representativeness; Regional (up to 100s of km).
c. General Site Locating Methodology;
Regional mean concentration measurements - site should be no
closer than 30 km from nearest city in the direction that
is least frequently downwind.
Interstate-urban transport stations - locate site near state
line on out-of-state urban area side.
Interstate-general transport stations - locate si'te near wind-
ward state line.
Intercity transport stations - locate site upwind of city in
most frequent winter wind direction (to measure SC>2 entering
city.
Topography should be uniform. Sites no closer than 30 km to a
power plant, 22 km to a 50,000 population town, and 0.6 km to an
individual home,(see Section 4.2.1.1 for details).
d. Inlet Exposure Criteria; Avoid low-lying areas; choose open or
sparsely forested areas; since S02 (away from point .source) is
well-mixed in vertical, inlet height of 3-10 m above ground is
reasonable.
II.. General-Level, Neighborhood Stations
a. Siting objectives; To monitor emergency episodes, to measure
exposures of specific populations to SO2» and to measure base
concentrations in areas of projected growth.
b. Spatial Scale of Representativeness; Neighborhood (0.5-4.0 km).
c. General Site Locating Methodology;
Emergency episode stations - locate site near center of emission
maxima of city. (Since models are unreliable in near zero
wind, some verification via mobile sampling may be required.
Population exposure stations - general location from pop. maps.
Projected growth stations - general location from land use maps
CONTINUED
-------
Concentration gradient (from modeling) determines number of sta-
tions required and whether middle-scale siting procedures are to
be used (see text, Section 4.3.2). Sites should be no closer than
interference distance (ID) to specific sources in upwind direction
(see Table 4-4).
d. Inlet Expsosure Criteria;
Emergency episode stations - locate instrument inlet no higher
than about 80% of the mean building height in local area, away
from dusty areas (as a general rule) and between 1-2 m above
the roof (see Table 4-3 for additional criteria).
Population exposure/projected growth sites - additional criteria
include locating inlet on windward side of building, 1-2 m out
from building for non-rooftop locations.
All sites - no significant SC>2 emission points on roof (see
Table 4-5). In suburban areas, choose building of low height,
preferably one story.
III. General-Level, Middle-Scale Stations
a. Siting Objectives;
Major objective - to measure peak concentrations in urban areas.
Secondary objective - to measure exposure of specific popula-
tions to S02 and determining base concentration in projected
growth areas where steep concentration gradients prevail.
b. Spatial Scale of Representativeness; Middle (100-500 m).
c. General Site Locating Methodology; Diffusion model results in-
cluding annual and 3-hour/24-hour near "worst case" patterns pro-
vide most of the basic guidance for locating the urban peak sta-
tions. Diffusion meteorologists should be consulted regarding
exact form of model, inputs and assumptions; 100-m distance near
maximum resolving power of most models. Mobile sampling recom-
mended to confirm locations of where the 3-hour/24-hour peaks
occur and effects from downwash (see Section 4.4 for more details
of methodology). Methodology for population/growth stations is
similar to that for neighborhood stations except that IDs appro-
priate to middle scales are used (see main text).
d. Inlet Exposure Criteria; Exposure criteria are the same as those
for the non-episode neighborhood scale stations (see Table 4-5);
inlet height, 80% of mean building height in local area; no signi-
ficant SC>2 emission points on roof. Locate inlet on winter wind-
ward side of roof. If trailer site, avoid parking lots in general
and lots around which there are buildings using large amounts of
fuel oil(to prevent undue influences due to plume downwash effects).
IV. Proximate, Middle-Scale Stations
a. Siting Objectives; To measure impacts from major point sources
located in multi-source urban settings and impacts from isolated
point sources.
b. Spatial Scale of Representativeness; Middle (100-500 m).
CONTINUED
xll
-------
General Site Locating Methodology; Diffusion model results in-
cluding annual-and 3-hour/24-hour near "worst case" patterns pro-
vide most of the basic guidance for locating urban stations (see
Section 4.5). For isolated point source sites, use graphical
solutions to diffusion equations (see Section 4.6). Special pro-
cedures are utilized for locating sites in valleys and in near
coastline settings (see Sections 4.6.2, 4.6.3). In extremely
rough terrain, only general guidance is given; major points to
consider include the following -
In regions subject to at least occasional periods of low mixing
depths, locate'monitors in basins that have inlets for S02
source plumes.
Site monitors at ridgetop locations in the general downwind
directions from the source, or perhaps at ridgetop locations
surrounding the source, particularly those nearest the source
at near effective plume height elevations.
Site monitors in passes that may receive the plume advected
either by drainage or channeled winds.
A complete survey of the entire area influenced by the SC>2
source would almost certainly be required in all situations.
Visual observations, aerial photography, mobile sampling,
remote sensing, etc. would probably be the most important
means for conducting such surveys.
Inlet Exposure Criteria; For urban stations, the inlet exposure
criteria are shown in Table 4-5. Criteria for isolated point
source monitoring sites are similar to those of regional sta-
tions; exceptions are -
The height of the inlet should be no higher than about 3-5m
above ground in all cases.
No major fossil fuel-burning sources between source and site.
No need to avoid wake effects of small buildings or clumps of
trees.
In irregular/rough terrain, choose well exposed areas. Establish
monitoring site off to one side of very large obstacles.
A set of appendices are included to provide guidance for selecting back-
ground information and data appropriate to the siting objective. Information
on meteorology, land use, topography, demography, diffusion models, "worst
case" meteorology and associated pollutant patterns, and mobile sampling is
included. Such information is used as an aid to identifying the general sit-
ing area then in selecting the final site location within the area.
The rationale underlying the site selection procedures are embodied in
three basic elements:
1) Determining the general location of the monitoring site,
mainly via diffusion modeling,
2) Refining the location to minimize undue influences from
nearby sources, including meteorological effects.
xiii
-------
3) Placing the instrument inlet in such a location to avoid
local contamination.
Multi-source Gaussian models and graphical solutions to the Gaussian dif-
fusion equation provide the basic means of ascertaining the general location
of the monitoring site, except in areas of extremely rough terrain. In areas
characterized by terrain of moderate roughness, and in valleys, modification
of the meteorological input to the model may be necessary. A diffusion mete-
orologist should be consulted in such situations to help determine the nature
of such modifications and also in performing the modifications. Regarding
the monitoring of the 3-hour/24-hour short-term peak concentrations in urban
areas, the simulations of the 3-hour and 24-hour near-worst case patterns are
recommended to estimate appropriate site locations. Again a diffusion mete-
orologist is recommended to be consulted to aid in determining such patterns,
as well as in judging the potential for any plume downwash effects. Some
guidance regarding the former is provided in Appendix B. Emergency episode
stations should be located in the very heart of the maximum 502 emission
density zone of an urban area; during air stagnations wind speeds are low and
directions are variable so the maximum concentration should occur where the
emission density is a maximum. Most of the impact at ground level will be
from low-level SC>2 sources. Appropriate site locations can best be found by
using gridded emission inventory data with most of the weight being given to
the area source fraction of the inventory. The heat island mechanism may pro-
duce maximum concentrations near the wind inflow convergence point which also
may be located near the center of maximum S02 (and heat) emission zone of the
city. This justifies considering the emergency episode station as an alterna-
tive site for measuring the 3-hour peak concentration,
The natural S02 background level as reported by Robinson and Robbins
(1970), 2.6 yg/m3, was deemed to be the undue influence level for determining
IDs in rural areas. The IDs were calculated via solutions to the Gaussian
diffusion equation using that concentration and the assumptions and meteoro-
logical conditions as shown in the following table.
Configurations and Emissions for Typical Source Types Assumed in
Determining Interference Distances for Regional-Scale Stations
Source
Type
Power Plant
(400 HW)
Industrial
Space rteat
(500 T SOa/yr)
Small Town
(25,000 pop.,
6000 homes}
Individual
Home
Characteristic
Emission Period
365 days/yr
Winter Quarter
(DEC, JAN, FEB)
Winter Quarter
(DEC, JAN, FEB)
Winter Quarter
(DEC, JAN, FEB)
Fuel Rate
S Content
W
280 x IQS gai
16 oil 9 IX S
14 x 106 gal
16 oil (80.55! S
103 gal/home
#2 oil 8 0.2X S
103 gal
n on e o.2t s
Source
Configuration
Point, uniform
wind over 22.5"
sector
Point
Area Source
4 ml2
Point
Emission
Rate
(a/sec)
575
58
10
.0016
Meteorology
Wind
Soeed
5 m/sec
5 m/sec
1 m/sec
1 m/sec
Stability
Class
0
D
0
F
Effective
Ht. On)
300
200
0
0
XIV
-------
With respect to the physical characteristics of the area in the immedi-
ate vicinity of the site, their effect on critical meteorological variables
and how these relate to the final site location and the placement of the in-
strument inlet, the literature survey yielded much information from which the
following conclusions could be inferred.
Mixing produced by mechanical turbulence and wake effects of larger
obstacles over moderately rough, natural terrain averages the pollu-
tant over space which lessens the concern of exact site location and
inlet placement in rural areas.
Micro-scale urban features substantially increase mixing and promote
uniformity of pollutant levels from mid- and far-distant sources.
This mixing between source and monitoring site reduces the monitoring
site selection problem to the consideration of only near sources (less
than the interference distance). The bulk of the plume material was
assumed to be held within a 10° sector downwind of the source and a
20° sector downwind of the sources nearest to the prospective moni-
toring site.
A plume entering a building wake cavity will be rapidly mixed throughout
the cavity. The resulting volume source configuration can be approxi-
mated by considering the source as having been released from a virtual
point source upwind of the building.
The uniform mixing principle is not absolute and cavity flows often
build, get swept away, and reform, leading to large "puff" type re-
leases .
Except for near the windward edge of a city, a vertically uniform
S02 distribution up to at least the mean building height over the
area of interest in the city can be assumed. The choosing of 0.8
H (or lower) for inlet location above the ground is somewhat arbi-
trary, but was meant to insure that the instrument (or inlet) would
be placed at a point in the vertical where the measured levels would
approximate those existing near the breathing zone 5-6 feet above
the ground.
If pollutant release is known to be well within a cavity (e.g.,
emissions from a vehicle in a deep street canyon) averaging will
not be complete and concentration fluctuations and gradients are apt
to be found within the flow. Minimum velocities and maximum concen-
trations should be found near the ground on the leeward side of the
obstacle. This is the justification for avoiding trailer locations
just downwind of buildings with large stacks. This situation is
generally precluded, however, if the interference distance criteria
are satisfied.
Pollutants from sources located downwind of a building may be emitted
into the wake cavity behind the building. The reverse flow of the
wake may advect pollutants up to the roof of the building to at
least one-half of the width of the building from the downwind edge.
xv
-------
This observation is the rationale for recommending that inlet place-
ment locations be on the windward side of the building. It also
justifies the recommendation of not having SC>2 sources on the roof
of the building chosen for the monitoring site' or inlet location.
In summary, the guidelines presented here provide a basis for selecting
monitoring sites to satisfy specific siting objectives to ensure the compati-
bility of the data obtained with the ultimate use of the data. Also, the
guidelines require specific procedures to be followed; these procedures in-
volve diffusion model analyses, consideration of the effects of natural and
urban topography on plume-carrying winds, and nearby source effects. All of
these elements will provide a physical basis for interpreting monitoring data.
This will ensure the site selector and data user that the data reasonably re-
present actual conditions over the appropriate spatial scales. This ensurance
is critical since decisions regarding control plans and strategies, which may
have severe economic impacts, are usually based on the interpretation of such
data.
xvi
-------
TABLE OF CONTENTS
?aqe
ABSTRACT iii
ACKNOWLEDGEMENTS iv
SUMMARY V
LIST OF FIGURES xix
LIST OF TABLES xxii
1.0 INTRODUCTION 1
1.1 General 1
1.2 Geographical and Source Characteristics of SO2 Emissions. . . 4
1.3 .Site Location Standards 5
1.4 The Organization of This Report 5
2.0 S02 DATA USES AND RELATED SPECIFIC MONITOR SITING OBJECTIVES ... 7
2.1 General 7
2.2 Uses of S02 Monitoring Data 8
2.3 Monitor Siting Objectives 9
2.3.1 Siting Objective 1 9
2.3.2 Siting Objective 2 10
2.3.3 Siting Objective 3 11
2.3.4 Siting Objective 4 11
2.3.5 Siting Objective 5 12
2.3.6 Siting_ Objective 6 12
2.3.7 Siting Objective 7 13
2.3.8 Siting Objective 8 13
2.3.9 Siting Obj ective 9 13
3.0 SPECIAL CHARACTERISTICS ASSOCIATED WITH SO2 MONITORING 15
3.1 Monitoring Network Concepts 15
3.1.1 Target Networks 15
3.1.2 Area Networks 16
3.2 Spatial Scales of Representativeness 16
3.2.1 Measurement Scales Relevant to SO2 Monitoring 19
3.3 Monitoring Site Types and Associated Siting Ojbectives and
Data Uses 21
3.4 The SO2 Monitoring Universe 21
3.5 The Five Relevant Monitoring Site Types 24
4.0 SITE SELECTION PROCEDURES AND CRITERIA ' 27
4.1 Description of Site Selection Aids and Background Material. . 27
4.1.1 The Critical Role of Diffusion Modeling in the Site
Selection Process 29
4.1.1.1 Siting Objectives Associated with Fixed Geo-
graphical Areas 29
4.1,1.2 Siting Objectives Associated with Features of
the S02 Pattern 30
4.2 General-Level Regional-Scale Stations 30
xvii
-------
4.2.1 Regional Mean (Background) Concentration Stations. . .
4.2.1.1 Local Characteristics, Interferences and In-
let Placement
4.2.2 SO2 Transport Stations
4.3 General-Level, Neighborhood-Scale Stations. .........
4.3.1 Emergency Episode Stations .
4.3.2 Population Exposure and Projected Growth Monitoring
Stations 40
4.4 General-Level, Middle-Scale Stations 46
4.4.1 Peak Concentration Stations 47
4.4.1.1 Winter or Annual Peak Concentration Station . 47
4.4.1.2 24-Hour and 3-Hour Maximum Concentration Sta-
tions 50
4.4.2 Population Exposure and Projected Growth Stations. . . 51
4.5 Proximate, Middle-Scale Stations - Urban Sources 52
4.5.1 Annual Peak Concentration Stations 55
4.5.1.1 General-Level Urban Peak Station 55
. 4.5.1.2 Proximate Station 55
4.5.2 24-Hour and 3-Hour Maximum Concentration Stations. . . 55
4.'5. 3 Data Interpretation 57
4.6 Proximate, Middle-Scale Stations - Isolated Sources 59
4.6.1 Monitoring in Flat Terrain Settings 60
4.6.1.1 Peak Concentration Stations 61
4.6.1.2 Background Stations 61
4.6.1.3 Fumigation Effects 63
4.6.1.4 Role of Mobile Sampling and Final Site Selec-
tion 63
4.6.1.5 Site Characteristics and Inlet Placement. . . 63
4.6.1.6 Instrument Type and Supplementary Instrumen-
tation 64
4.6.2 Monitoring in Near Coastline Settings 64
4,6.2.1 Site Characteristics and Instrument Inlet
Placement 67
4.6.3- Monitoring in Ridge/Valley Settings 67
4.6.3.1 S02 Problems 68
4.6.3.2 Siting Procedures 70
4.6.3.2.1 Fumigation Concentration Stations. 71
4.6.3.2.2 Valley-Wall Impact Stations. ... 73
4.6.3.2.3 Worst-Case Conditions for Along
Valley Flow 74
4.6.3.2.4 Supplementary Monitoring Stations
and Concluding Comments 74
4.6.4 Monitoring in Rough, Irregular Terrain Settings. ... 76
4.6.4.1 Monitor Siting Procedures in Terrain of Up to
Moderate Roughness 76
4.6.4.2 Conditions in Extremely Rough Terrain .... 78
4.6.4.3 Implications for SC>2 Monitoring 80
5.0 RATIONALE AND SUPPORT DOCUMENTATION FOR SITING CRITERIA 83
5.1 Undue Influence Effects _ 83
5.2 Meteorological Processes Pertinent to Site Location Refine-
ment and Inlet Placement 86
xviii
-------
Page
5.2.1 Effects of Natural Topography 89
5.2.1.1 Effects of Above Considerations on SC>2 Dis-
tribution 91
5.2.2 Effects of Urbanization 91
5.2.2.1 Building-Induced Turbulence and Wakes .... 92
5.2.3 Relevance of Above Considerations on Siting Criteria . 95
5.3 Miscellaneous Considerations 99
5.3.1 Temperature 99
5.3.2 Chemical-Physical Interactions 99
5.3.2.1 Reactions of SC>2 with Atmospheric Liquid
Water 100
5.3.2.2 Catalytic and Photochemical Oxidation Reac-
tions 100
5.3.2.3 Reactions With Ground and Water Surfaces. . . 100
5.3.2.4 Residence Times and Half-Life 101
5.3.2.5 SC>2 Reactivity and Monitoring Device Inlet
Tubes 102
5.3.3 Relevance of Above Considerations to Siting Criteria . 102
6.0 REFERENCES 103
APPENDIX A. Sources of Climatological and Meteorological Information . . A-l
APPENDIX B. Suggested Approaches for Determining Worst Case SC>2 Pat-
terns and Associated Meteorology B-l
APPENDIX C. Mobile Sampling C-l
APPENDIX D. Sources of Land-Use and Topographical Information D-l
APPENDIX E. Available Air Quality Simulation Models Appropriate for
SO2 Monitor Siting Activities E-l
APPENDIX F. Bibliography F-l
LIST OF FIGURES
Figure Page
1-1 Concentration contours from an array of sources with steady
conditions 1
1-2 Instantaneous and potential regions of significant pollution
concentration from single source 2
1-3 Superimposed plumes from multiple sources 3
3-1 Illustration of various spatial scales of representiveness. . 18
3-2 Relative locations of sites for measuring concentrations re-
presenting several spatial scales of measurement in an urban
complex, with respect to annual averaging times 20
3-3 Portion of SC-2 monitoring universe. 23
4-1 Flow chart showing procedures for locating general-level re-
gional-scale stations 31
xix
-------
Figure Page
4-2 A typical wind rose with wind-speed information 32
4-3 Schematic illustration showing tentative siting area for re-
gional mean background concentration stations 34
4-4 Schematic illustration showing primary and secondary siting
areas for interstate urban transport stations 35
4-5 Schematic illustration showing primary and secondary siting
areas for intercity background stations 36
4-6 Flow chart showing procedures for loca-ting general-level
neighborhood-scale stations 38
4-7 General location of emergency episode siting area 39
4-8 Schematic illustrating typical concentration pattern over de-
lineated population or growth areas of interest 41
4-9 Schematic illustration of intermediate step by which neigh-
borhood-scale stations are located 42
4-10 Plan view blowup of siting area of Fig. 4-9 illustrating the
technique by which final candidate sites are selected .... 43
4-11 Blowup of Fig. 4-10 illustrating the technique by which the
final site is selected 43
4-12 Oblique view of siting area of Fig. 4-11 showing site loca-
tions and urban structure 45
4-13 Schematic illustration of an idealized SC>2 concentration pro-
file and its associated ground-level pattern and example site
locations 46
4-14 Flow chart showing procedures for locating general-level mid-
dle-scale stations 48
4-15 Schematic illustration of middle-scale siting procedure for
peak concentration stations 49
4-16 Schematic showing population exposure or projected growth
area divided into middle-scale parcels and recommended siting
areas 51
4-17 Schematic illustration of impact of point source in an urban
setting for two averaging times 53
4-18 Flow chart showing procedures for locating proximate middle-
scale stations 54
4-19 Schematic illustration showing annual impact pattern due to
urban point source alone, the siting area, and final candi-
date sites for assessing the annual impact from the point
source 56
4-20 Schematic illustration of the concept of the source contribu-
tion profile in an urban area for the annual pattern 58
xx
-------
Figure Page
4-21 Illustration of possible monitoring site configuration around
an isolated point source in flat terrain 62
4-22 Schematic illustration of a sea-breeze fumigation situation . 64
4-23 Mixing depth as function of stability, wind speed, and inland
travel distance 66
4-24 Vertical mixing depth adjusted for terrain 66
4-25 Example of computations of A6 67
4-26 Inversion aloft-above stack 68
4-27 Plume behavior near a steep bluff when the air is unstable
and when it is very stable 69
4-28 Plume dispersion in a deep valley with wind left from right . 69
4-29 Plume dispersion in a deep valley with wind parallel 70
4-30 Diurnal variation of valley winds during the summer in the
Columbia River Valley near Trail, B.C 71
4-31 Illustration of plume configurations under a variety of mete-
orological conditions and relative locations of sampling
sites 72
4-32 Illustration of plume configurations under a variety of mete-
orological conditions and relative locations of sampling
sites . . 75
4-33 Distribution at height of 40m from ground surface of concen-
tration of smoke emitted from a source with height 400 m
which is derived from the computer experiment 77
4-34 Schematic illustration of the dilution of an airborne plume
as it approaches and flows over nearby elevated terrain ... 80
4-35 Schematic illustration of turbulent wake effects caused by
obstacles protruding into the primary flow pattern 81
5-1 Normalized concentrations computed with a Gaussian dispersion
model 84
5-2 Schematic illustrating undue influence of nearby sources on
measurements at three sampling sites 87
5-3 Topography effects on wind 88
5-4 Asymmetry of flow approaching and leaving steep topography. . 89
5-5 Distortions of the wind flow by topographic obstacles .... 90
5-6 General arrangement of flow zones near a sharp-edged building 93
5-7 Cavity flows 93
5-8 Wake region in flow past bluff obstacle 94
5-9 Urban circulation and dispersion before sunrise 95
xxi
-------
Figure Page
5-10 The dispersal of a narrow plume passing over a single build-
ing 96
5-11 The dispersal of pollutant from a flush vent onthe top of a
complex building structure 96
5-12 Schematic illustration of vertical distribution of SC>2 within
a vertical column of air passing through an urban area. ... 97
5-13 Processes involved in the relationship of sulfur oxide emis-
sions to air quality 101
B-l Graphical presentation of persistence data, NW wind case from
Table B-l B-3
B-3 Distance of maximum concentration and maximum xu/Q as a
function of stability and effective height of emission. . . . B-3
C-l Ideal plume and measurement pattern; plume dispersion pattern
at Chalk Point power plant C-2
C-2 Ground-level pattern downwind of point source C-2
D-l Example of the informational content of large, medium, and
small scale topography map D-3
D-2 A portion of the orthophotoquad index showing the legend and
a portion of the state of Florida D-4
D-3 A Sanborn map for a section of Portland, Oregon D-6
LIST OF TABLES
Table Page
Source Types and Related interference Distances for Regional-
Scale Monitoring Stations x
Interference Distances for Three Development Intensities. . . x
Configurations and Emissions for Typical Source Types Assumed
in Determining Interference Distances for Regional-Scale Sta-
tions xiv
1-1 Sulfur Oxide Emission Inventories for the U.S. and for Selec-
ted Air Quality Control Regions 4
2-1 NAAQS for S02 8
3-1 Relationships Among Siting Objectives and Related Data Uses,
Site Types, and Scales of Representativeness. . . . 22
3-2 Relationships Among Table 3-1 Elements and Associated Rele-
vant Averaging Times 24
xxi i
-------
Table Page
3-3 Matrix of Topographical and Land Use Types 25
4-1 Example of a Tabulated Wind Summary 32
4-2 Source Types and Related Interference Distances for Regional-
Scale Monitoring Stations 34
4-3 Site Characteristics and Inlet Placement Criteria for Emer-
gency Episode Stations. .... 39
4-4 Interference Distances for Three Development Intensities. . . 44
4-5 Site Characteristics and Inlet Placement Criteria for Neigh-
borhood Stations 45
4-6 Percentage Frequencies of Sky Cover, Wind, and Relative Humi-
dity 47
5-1 Configurations and Emissions for Typical Source Types Assumed
in Determining Interference Distances for Regional-Scale Sta-
tions 85
5-2 Rationale for PSIDs and MSIDs of Table 4-4 86
A-l Example of Meteorological Records Available from the NCC. . . A-l
A-2 Published Volumes of World-Wide Airfield Summaries A-4
A-3 List of Stations for Which Stability-Wind-Rose Tables have
been Prepared A-5
B-l Tabulated Persistence Data for NW Wind Situation B-2
B-2 Illustration of Use of Stability-Wind-Rose for Determining
Site Locations for Monitoring Isolated Point Sources B-5
D-l National Topographic Maps D-5
D-2 Picture Products Available from ERTS D-5
XXlll
-------
-------
1.0 INTRODUCTION
x x
X
0
o «
An array of
near-ground
level S02
sources.
1.1 GENERAL
Sulfur dioxide (SO2) is a natural constituent of the air. Globally, about
one-half of all SC>2 in the atmosphere comes from natural sources (Robinson and
Robbins, 1968). These natural sources are, however, quite diffuse and lead to
background concentrations estimated to be
a small fraction of a part per billion
(parts of air). In contrast, the emis-
sions from anthropogenic processes are
relatively quite intense. As S02 is
dispersed in the atmosphere, its concen-
tration is reduced from noxious levels
near the sources to levels comparable to
that of SC>2 from natural sources. The
general nature of SC>2 (or any pollutant)
monitoring, then, is to measure the time
and space variability of concentrations
from the region of the source, or sources,
to where the pollutant has become suffi-
ciently dilute. Such measurements are
required to satisfy monitoring program
goals or data uses such as determining
population exposures and ascertaining
compliance with air quality standards.
Contours of
for sources
shown above.
max
Contours of
1/2 Xn
Contour of
1/100 X,
max
FIGURE 1-1. Concentration contours
from an array of sources with
steady conditions.
Since the dominant anthropogenic
sources of SO2 emissions are from sta-
tionary combustion devices, the most
striking characteristic of typical S02
concentration patterns if that the con-
centration peaks reproduce the source
patterns (see Figure 1-1). Monitoring
for the concentration maxima averaged
over any time scale may be accomplished
with great accuracy by putting an in-
strument in every chimney. Obviously,
it is of more interest to determine
time and space patterns of S02 concen-
tration away, but not too far away,
from one or more sources. Concentra-
tions very far away from all sources
must be low; i.e., they approach the
global average, and their patterns
-------
would contain features associated with only the longest time scales or the
broadest space scales.
In the region between "too near" and "too fax" from a source, a signifi-
cant concentration may be expected over only perhaps 5 percent of the area at
any one time (see Figure 1-2), because the wind comes from only one direction
at a time. This is an order of magnitude rule-of-thumb, no matter what the
minimum concentration of interest is. Any given monitor permanently placed
in hopes of defining such a region would have one chance in twenty of detect-
ing any S02 above the limit, even if all wind directions were equally likely.
Also, the varying inner and outer limits of the region, due to the varying dis-
persive power of the wind would reduce the chances further. In situations
where a given wind direction is significantly more likely than others, the
chances are significantly increased, but still not as large as one would like.
Therefore, instrument siting to monitor a single source even in an ideal en-
vironment free of micro- or meso-scale local effects such as topography, cavi-
ty wakes, or localized thermal effects is not a straightforward procedure.
/ R. - Maximum Radius X2 Leve1
of Interest at
Maximum Radius \
of Interest at v
. Plume
\ Impact Pattern
»at Ground Level
\
_ Minimum Radius
0 of Interest
a *max
\
\
\
\
x- = Minimum x of
Interest /
.05
.05
'
FIGURE 1-2. Instantaneous and potential regions of
significant pollution concentration
from a single source.
-------
If many sources are near enough to each other so that their "circles of
concern" (see Figure 1-2) overlap, then sources relatively far from a potential
monitoring site will contribute a relatively steady "background" concentration
upon which the relatively narrow plumes from nearer sources will be superim-
posed more randomly as they undulate past the site (see Figure 1-3). For a
widespread and reasonably dense array of homogeneous sources, as, for example,
in a large urban residential area, long-term mean concentrations can be deter-
mined quite accurately at any site. If the source array contains one or a few
sources that are much larger than the rest which are relatively homogeneous,
the problem of finding the background levels is as straightforward as for the
homogeneous sources alone, but the problem of locating the peaks is as diffi-
cult as for the single source.
= Limits of Plume from Nearest Source
= Plume Limits
X = Source
= Monitoring Site
FIGURE 1-3. Superimposed plumes from multiple sources.
-------
1.2
GEOGRAPHICAL AMD SOURCE CHARACTERISTICS OF SO? EMISSIONS
The relative importance of the various S02 source categories varies geo-
graphically as shown in Table 1-1. In the colder areas of the north, exempli-
fied by the Boston Air Quality ControJ Region (AQCR), commercial and residen-
tial heating is the largest category with 48.6 percent of the total for that
AQCR. Farther south, the transportation, power generation, and industrial
process emission categories are dominant, as indicated by the Atlanta AQCR
summary. In the western states, the industrial process category is the largest
with 40.1 percent and 37.6 percent of the total SO2 emitted in the Denver cind
Los Angeles AQCRs, respectively. Transportation sources are also very signifi-
cant in the west. In the Dallas/Fort Worth AQCR, SO2 emissions from power
generation are very small (3 percent of the total) because of the use of rela-
tively sulfur-free natural gas. The largest category here is transportation
(43;2 percent), followed by industrial processes (23.7 percent). In Arizona
and Texas (as a whole) more than 80 percent of the SO? is emitted from smel-
ters and refineries (Cavender, et al., 1973).
TABLE 1-1*
Sulfur Oxide Emission Inventories for the United 'states
and for Selected Air Quality Control Regions,
(NEDS Data for 1972)
Geographical Area
Total Sulfur Oxide Emissions
Source Cagetory
Stationary Source Fuel
Combustion
Electric Power Plants
Industrial
Commercial and Residential
Industrial Processes
Other Stationary Sources
Transportation Sources
United
States
Boston
AQCR
Atlanta
AQCR
St. Louis
AQCR
Dallas,
Ft. Worth
AOCR
Denver
AQCR
Los
Angeles
AOCR
SO? Emissions in 103 Tons/Year
32,000
332
94.7
1,234
17.3
33.5
238
Percentage of Sulfur Oxide Emissions by Source Category
54.3
15.3
7.1
21.1
0.2
2.0
41.6
8.2
48.6
0.5
0.1
1.0
70.8
5.6
5.7
12.3
0.5
5.1
76.2
6.0
1.9
15.3
0.1
0.5
3.0
5.0
19.8
23.7
5.3
43.2
34.2
10.4
5.3
40.7
0.2
9.2
16.8
14.6
18.8
37.6
1.6
10.6
* Taken from NAS (1975).
About two-thirds of S02 emissions occur in urban areas, with very large
fractions contributed by industrial, commercial, and residential heating. These
sources are also emitted near the ground which increases their ground-level im-
pact. In rural areas, much of the SO2 is emitted by a relatively small number
of large sources such as smelters. Also, about one-half of the nation's power
plants are located in rural areas. Although power plants comprise the largest
emission category, their S02 is emitted from tall stacks which reduce ground-
level impacts.
-------
It will be seen later (in 'Section 4.0) that the physical configuration of
the SC>2 source (i.e., point versus area source), whose SC>2 air quality impact
is to be monitored, is important in regard to specific siting procedures.
Point sources include large individual sources such as power plants and cer-
tain industrial processes. Commercial/residential heating and transportation
categories are considered collectively as "area" sources.
1.3 SITE LOCATION STANDARDS
Most of the literature reporting air quality data and data summaries (e.g.,
EPA, 1973) emphasizes that interpretation of the data must be tempered by an
understanding of the limitations imposed by inadequacies of surveillance meth-
odologies. These inadequacies include inconsistencies between the specific
objectives for which a monitoring station is established and the intended use
of the resulting data and sampling maldistributions in both a geographical and
temporal sense; these have"been brought about by non-uniform siting proce-
dures and/or a lack of an understanding of the atmospheric processes that
affect the temporal and spatial distributions of pollutants. To illustrate
these points, EPA (1973a) shows maximum 24-hour S02 concentration measure-
ments within individual cities varying typically by a factor of from 5 to 10,
and in extreme cases by 100 or more. Ott (1975) has shown similar variations
in carbon monoxide measurements in United States cities. Clearly, data from
most of these sites are not representative of the cities as a whole, but
merely reflect what is occurring in the immediate vicinity of the sites.
Yamada (1970) showed that little consistency existed among sampling site
locations and instrument inlet exposures. His study was based on a national
survey of monitoring site characteristics. Further, the early Continuous Air
Monitoring Program (CAMP) stations had inlet locations from 10 to 15 feet above
the ground (Jutze & Tabor, 1963) while most state network station inlets were
located on building roofs. A similar situation presently exists, although the
development and deployment of instrumented trailers, of generally uniform di-
mensions, has reduced the problem somewhat.
From the above discussion, it is clear that a need exists for objective,
uniform procedures for locating and categorizing S02 monitoring stations con-
sistent with the intended use of the resulting data.
1.4 THE ORGANIZATION OF THIS REPORT
In Section 2.0, the major uses of S02 monitoring data is reviewed and a
list of siting objectives, each consistent with a specific data use or group
of uses is developed. It will be seen that the siting objective (along with
the spatial scale of representativeness) is the major controlling factor in
determining the desired physical characteristics of a site and its surroundings.
Section 3.0 discusses two basic monitoring network concepts, the spatial
scales of representativeness relevant to S02 monitoring, and an S02 monitoring
-------
universe. A full review of these topics provides a basis to proceed with the
development of the siting procedures which are discussed in Section 4.0.
Section 4.0 is the working part of the report and provides detailed step-
by-step procedures for locating monitoring sites and the exposure of instrument
inlets to satisfy the requirements for the various siting objectives. The dis-
cussion proceeds from the largest spatial scale to the smallest considered.
In Section 5.0, the rationale behind the site location procedures and oth-
er support documentation are presented. Topics include some of the relevant
meteorological aspects of air pollution, topographical effects, urban modifi-
cations, washout/rainout, and chemical/physical interactions.
In determining monitoring site locations, the site selector will be re-
quired to use information and/or techniques with which he may be unfamiliar.
To obviate this problem, a set of appendices has been included which describe
the various kinds of data and techniques required as well as the sources from
which these may be obtained. Topics addressed include a general approach for
determining worst case meteorolo'gical conditions, the sources of meteorological
and land use data, a list of available air quality models which may be useful
in selecting a site, and some concepts of mobile sampling.
A bibliography (see Appendix F) is included showing a sample of the body
of information available on all relevant topics covered in this report.
-------
2.0 S02 DATA USES AND RELATED SPECIFIC MONITOR SITING OBJECTIVES
In this section, general SC>2 monitoring program elements and uses of SC>2
data are first reviewed; then, based on the review, a list of specific monitor
siting objectives is developed. The main thrust of this section is to give
some perspective to the various data uses and to relate them to the specific
siting objectives. This latter point is important since it was from the sit-
ing objectives that a relatively small group of monitoring site types was de-
veloped for which site selection procedures and instrument inlet exposure
criteria were prepared (Section 4.0).
2.1 GENERAL
Selecting and/or redistributing SC>2 monitoring sites on a priority basis
is becoming critical in view of issues that have arisen since the promulgation
of the Clean Air Act Amendments of 1970.* For example, air quality maintenance
planning (AQMP)(Federal Register, 1973a), prevention of significant deterior-
ation (PSD)(Federal Register, 1974), transportation control plans (Federal
Register, 1973b), supplemental control systems (SCS) or intermittent control
strategies (Federal Register, 1973c), and the Energy Supply and Environmental
Coordination Act (ESECA) of 1974 have resulted in, either directly or by im-
plication, requirements for expanded and/or reconfigured air monitoring net-
works. In addition, complexities and problems associated with photochemical
pollutants (e.g., Stasiuk and Coffey, 1975; and Spicer, et al., 1976) which
were unforeseen at the time of the passage of the "Amendments" will require
an expansion of photochemical and photochemical precursor pollution monitor-
ing. The total impact of these issues will require a reallocation of resources
for ambient air monitoring. It will, therefore, be essential for the site
selector to optimize ambient SC>2 monitoring systems in response to these new
monitoring requirements.
Foremost in the discussion of SC>2 monitoring are the National Ambient Air
Quality Standards (NAAQS) which must be attained and maintained in each AQCR
across the country. These standards are summarized in Table 2-1. The primary
standards were set to protect human health and the secondary standard was set
The "Act" resulted in the requirement for the states to prepare, adopt, and
implement air pollution control plans or "state implementation plans" (SIPs)
to attain and maintain air quality standards (Federal Register, 14 August,
1971). These plans included provisions for the design, establishment, and
operation of air monitoring networks.
-------
to protect the public welfare. The primary standards were to be attained in
each AQCR by June of 1975 and the secondary standards attained within a rea-
sonable time thereafter.
TABLE 2-1
NAAQS for S02
Annual Average
2 4 -hour Maximum
3 -hour Maximum
Primary Standards
80 yg/m3
365 yg/m3
Secondary Standard
_
1300 yg/m3
2.2 USES OF SO? MONITORING DATA
The list of SO2 data uses presented below was compiled from a literature
survey (see Appendix F). The order in which the uses are listed does not ne-
cessarily reflect the priority or relative importance of the uses; obviously,
the priority of a given use in- a given area would depend on the nature of the
SO2 problems that characterize that area. However, most of the uses that are
listed are common to most areas of the country and are generally required to
successfully implement those federal and state clean air policies that require
the use of ambient SC>2 data.
1) Judging attainment of SC>2 NAAQS.
2) Evaluating progress in achieving/maintaining the NAAQS or
state/local standards.
3) Developing or revising state implementation plans (SIPs)
to attain/maintain NAAQS; evaluating control strategies.
4) Reviewing new sources.
5) Establishing baseline air quality levels for preventing
significant deterioration and air quality maintenance
planning.
6) Developing or revising national S02 control policies [e.g.,
new source performance standards (NSPS), tall stacks, sup-
plementary control systems (SCS)].
7) Providing data for model development and validation.
8) Providing data to implement the provisions of the Energy
Supply and Environmental Coordination Act (ESECA) of 1974.
8
-------
9) Supporting enforcement actions.
10) Documenting episodes and initiating episode controls.
11) Documenting population exposure and health research.
12) Providing information to
a) public - air pollution indices; and .
b) city/regional planners, air quality policy/decision
makers - for activities related to programs such as
air quality maintenance planning (AQMP) , prevention
of significant deterioration (PSD), and the prepara-
tion of environmental impact statements.
2.3 MONITOR SITING OBJECTIVES
The above data uses are expressed in rather broad terms and are generally
program oriented. For this reason, it was difficult to associate a particular
data use with a specific site selection procedure. To obviate the problem, a
list of siting objectives was developed to provide a link between data uses
and specific site selection procedures. The various siting objectives were
developed such that each could be related to a specific type of monitoring
site that would yield data of a level of quality and spatial and temporal re-
presentativeness appropriate for its intended use. Some of the siting objec-
tives are couched in terms more reflective of the mears by which the appropri-
ate data will be obtained rather than in terms having a broad program connota-
tion. Other siting objectives are worded closely to their related data uses,
since in these cases the intended use is rather specific (e.g., episode moni-
toring) . The monitor siting objectives and their related data uses are listed
and discussed in'the following sections.
2.3.1 Siting Objective 1 - Determination of Peak Concentration in Urban Areas
State and EPA policies and regulations require that SO2 levels be brought
within the primary NAAQS by June of 1975 and the secondary NAAQS within a rea-
sonable time after that date, and that both are maintained thereafter. Maximum
annual, 24- and 3-hour concentrations of S02 are usually found in urban centers
where the use of sulfur-containing fossil fuel for space heating results in ex-
tremely high SO2 emission densities. Subsequently, people living and working
in these areas may be subject to both chronic and acute effects brought about
by exposure to these high concentrations. The problem is exacerbated by S02
emissions from power plants which are often located in the larger urban centers.
SIP control strategies for S02 abatement are usually keyed on achieving
the NAAQS at these points of maximum concentration (therefore, inherently re-
lated to the maximum economic impact of the strategy). Monitoring sites should
be located at or near these points of maximum concentrations as revealed by
modeling, to provide a continuing assessment of the situation. The most rele-
vant uses for which such data are required are as follows:
-------
Judging attainments of SC>2 NAAQS (use 1)*.
Evaluating progress in achieving/maintaining the NAAQS
or state/local standards (use 2).
Developing or revising state implementation plans (SIPs)
to attain/maintain NAAQS; evaluating control strategies
(use 3).
Such data will also be relevant to the implementation of the ESECA of 1974
(use 8} in-those cities where there are power plants subject to the provisions
of the ESECA.** Other uses include the supporting of enforcement actions (use
9) and in providing information to the public, city/regional planners, and air
quality decision makers (use 12).
2.3.2 Siting Objective 2 - Determination of the Impact of Individual Point
Sources in Multi-Source urban Settings
This siting objective is similar to Objective 1 except that the monitor is
placed at or near the maximum ground-level impact point caused by an individual
point source located in an urban area. Because of background "noise" produced
by other urban sources, monitor placement and data interpretation must also be
done in conjunction with diffusion modeling.
This siting objective is related particularly to the ESECA of 1974 (use 8)
which was enacted in response to projected shortages of fuel oil and/or dimin-
ished confidence of availability of supplies of such fuels. Under the Act's
provisions, sourcesmainly power-generating stationsmay be required to con-
vert to coal-burning. The conditions of the conversion will depend on the sta-
tus of the AQCR with respect to the NAAQS. If the NAAQS are not being attained,
a regional limitation (on S02 emissions) applies and all provisions of the SIP
must be met before the conversion. Jf the NAAQS are being attained, then a pri-
mary standard condition applies which results in a variance from SIP emission
limits and still results in attainment of the NAAQS. This siting objective
particularly addresses the situation for such subject sources located in urban
areas.
Another situation applicable to this siting objective is that of a single
scurce located in an urban area that contributes overwhelmingly to SO2 pollu-
tion in that urban area. In such a situation, it would be very desirable to
monitor the maximum ground-level contribution from that source since the attain-
ment and maintenance of the NAAQS in the area would be highly dependent on the
effectiveness of control measures applied to that source. In this connection,
data from monitoring stations so located could be used for:
Developing or revising SIPs to attain/maintain NAAQS; evalu-
ating control strategies (use 3).
* Uses are listed from 1 to 12 in Section 2.2.
** A brief, general summary of the ESECA is presented under Siting Objective 2.
10
-------
Developing or revising national SC>2 control policies; e.g.,
new source performance standards (NSPS), tall stacks, and
supplementary control systems (SCS)(use 6).
Supporting enforcement actions (use 9), including SCS sur-
veillance.
Reviewing new sources (use 4). In this case, the data would
be used to provide urban background concentrations at the
point of maximum concentration contributed by a proposed new
source or at any other point in the major impact area at
which the NAAQS may be threatened.
Providing information to the public, etc. (use 12).
2.3.3 Siting Objective 3 - Determination of the Impact of Isolated Point
Sources
This siting objective is similar to Objective 2. Because there will be
few, if any, interfering sources in rural areas, area diffusion modeling need
not be employed for locating monitoring stations or in data interpretation.
However, because of special problems associated with locating maximum impact
points from individual sources in rural areas, mobile sampling may be required,
particularly in regions of complex terrain. Only the 3-hour and 24-hour aver-
age concentrations need to be considered since the annual standard will not
likely be contravened by an individual isolated point source.
The primary data uses related to this siting objecitve are the same as
those associated with Siting Objective 2, particularly SCS implementation
and surveillance. Other uses include establishing baseline air quality
levels for PSD planning (use 5) and impact assessments associated with the
enforcement of PSD policies.
2.3.4 Siting Objective 4 - Assessment of Interregional SO? Transport
Transport or advection of pollution across state or other jurisdictional
boundaries received considerable attention in the development of some SIP's
(e.g., Ball, et al., 1972). Large urban areas situated near or straddling
state boundaries can result in a considerable exchange of SO2 between the af-
fected statese.g., New York/New Jersey/Connecticut (New York City); Penn-
sylvania/New Jersey (Philadelphia); Missouri/Illinois (St. Louis); and
Illinois/Indiana (Chicago)-. A rather detailed study of interstate transport
of S02 was conducted by the NAPCA in the New York/New Jersey area (DHEW, 1967).
The EPA has acknowledged the existence of these situations and has re-
quired their being taken into account in state SIP's. The main objective of
monitoring interregional transport of S02 is to assess the relative impacts in
adjoining states. This assessment can provide information to the air pollution
control agencies of these states for refining or optimizing control measures
for achieving and maintaining the NAAQS (uses 2 and 3).
11
-------
In certain situations, monitoring sites set up to monitor incoming SO2
may also be considered as sites for measuring background concentrations and
determining base concentrations for environmental impact studies, AQMP and
PSD planning (uses 5 and 12).
2.3.5 Siting Objective 5 - Determination of Base Concentrations in Areas of
Projected Growth
The air quality maintenance provisions of the Clean Air Act require that
once the NAAQS are attained they must be maintained thereafter. To effectuate
this requirement, a series of guideline documents was prepared and issued to
the states (EPA, 1974a) to assist them in establishing Air Quality Maintenance
Areas (AQMAs), and preparing AQMPs. Volume XT of the series ("Air Quality
Monitoring and Data Analysis") addresses rather specifically the air monitor-
ing requirements of AQMPs. The basic requirement involves the design and op-
eration of a monitoring network (or a modification of an existing network) to
establish baseline concentration levels from which air quality levels are pro-
jected into the future. Ongoing air quality measurements are then matched
against projected levels to ascertain AQMP effectiveness. This siting objec-
tive satisfies the air monitoring requirements of AQMP development. Data orig-
inating from monitoring stations satisfying this siting objective will be par-
ticularly relevant to the activities of city/regional planners and air quality
policy/decision makers associated with such programs and the preparing of en-
vironmental impact statements (uses 5 and 12).
2.3.6 Siting Objective 6 - Initiation of Emergency Episode Abatement Actions
States have established (with EPA guidance) air quality levels at which
preplanned abatement strategies must be activitated for precluding air pollution
buildup during air stagnations. These plans are usually "triggered" on the
basis of real-time monitoring information from appropriately located sites.
Episodal concentrations often represent the highest short-term concentra-
tions ever observed during the year in any given area. The highest peaks occur
in the urban core, but are also relatively high and generally uniformly distri-
buted over the areas surrounding the urban core. Since episodes are of rela-
tively short duration (maximum duration of about three days or so) , the acute
effects on human health and public welfare are of greatest concern.
Most emergency episode plans drawn up by the states provide for a four-
stage abatement mechanism. In each successive stage, more stringent emission
limitations are imposed on prespecified sources to deal with the pollution
buildup in a stepwise manner. The air quality situation is continuously moni-
tored and each stage (and the eventual "all clear") is triggered according to
prespecified criteria. Sites established for SO2 monitoring during air stagna-
tions should use continuous type instruments that output directly (via teleme-
tering) to the air pollution control agency office (and computer) to facilitate
rapid data acquisition and evaluation. In most situations, the site should be
located in the very heart of the maximum SC>2 emission density zone of an urban
area, since during air stagnation conditions wind speeds are low and directions
12
-------
are variable so the maximum concentration should occur near to where the emis-
sion density is a maximum. Since it is desirable to maximize monitoring cover-
age during a stagnation episode, other sites can be used to trigger the episode
abatement plan and/or to monitor the progress of each stage. Most often these
will be the peak concentration stations and other stations located in the urban
area. The relevant data uses here are, therefore:
Documenting episodes and initiating episode controls (use 10).
Providing public information via air pollution indices (use 12).
2.3.7 Siting Objective 7 - Assessment of Background Concentrations in Rural
Areas
Background levels of SC>2 in rural areas represent the lowest levels, or
the approximate lowest levels (depending on the degree of interregional SC>2
transport) attainable over a large region. They may be considered as the base-
line concentrations near urban areas that should be known in order to optimize
the degree of control necessary to attain and maintain the NAAQS over the urban
area. This siting objective is also closely related to Siting Objectives 4
and 5; in fact, several of these objectives could be satisfied with one site
strategically located. The data uses relevant to this siting objective include
uses 2, 3, 5 and 12.
2.3.8 Siting Objective 8 - Determination of Population Exposure
Since the primary purpose of the NAAQS is to protect human health, SC>2
monitoring sites should be located in areas characterized by high population
density to ascertain the degree of S02 exposure to large numbers of people.
In most cases, these areas will be the residential areas of cities adjacent to
the central business districts (CBDs) and the peripheral suburbs.
In these areas, S02 concentrations for the three averaging times may be
relatively high. However, the greater spatial variability of the shorter term
peaks shifts the major concern to the annual average concentrations where ef-
fects on people are most likely to be chronic. This siting objective places
the emphasis on the monitoring of S02 where most people live (constant exposure
to relatively high levels) rather than where they work, which is covered by
Siting Objective 1. The relevant data uses are then:
Documenting population exposure and health research (use 11).
Providing information to the public via air pollution indices
(use 12).
2.3.9 Siting Objective 9 - Diffusion Model Calibration and Refinement
»
The calibration and refinement of diffusion models is becoming one of the
most important objectives of air monitoring (use 7). In fact, many of the
13
-------
., ultimate!}
objectives described in this section may well, ultimately, be satisfied by the
operation of appropriate diffusion models.
Many states used diffusion models to develop control strategies (e.g.,
Morgenstem and Hagg, 1972) to satisfy EPA SIP requirements. Diffusion model-
ing by state agencies is expected to continue as an ongoing activity in refin-
ing and/or optimizing control strategies and in providing a development/assess-
ment tool in the design and implementation of AQMPs and PSD plans.
A realistic SO2 model calibration program may require the establishment
of a special, temporary network of SC>2 monitors to facilitate spatial as well
as temporal correlation studies. For a detailed discussion on the problems of
model calibration, see Brier (1973, 1975). Monitoring sites established for
other objectives may also be used to supplement data from the special network.
Diffusion models are of two basic typesGaussian and grid. Gaussian
models simulate individual plumes (continuous or puff) by assuming a Gaussian
distribution of plume material in the crosswind and vertical dimensions. Grid
models, on the other hand, compute mean concentrations for each cell of a three-
dimensional matrix of cells. There are several varieties of grid models, one
of which is the full-airshed Eulerian (fixed-cell) type.
Several problems are associated with each type of model. A major problem
with the Gaussian models, particularly the continuous versions, is their ina-
bility to account for complex air flows in which SC>2 source plumes are imbedded
(e.g., in urban areas and in other regions of complex terrain). Grid models,
however, are difficult to validate because their volume-averaged predictions
must be compared to measurements taken at a point. Neither type of model can
simulate the effects of micro-scale features of complex flows.
The largest air pollution study ever conducted by the EPA is presently
underway in St. Louis, Missouri. The Regional Air Pollution Study (BAPS) has
been referred to as the modeler's model. Models have been developed for simu-
lating emissions, meteorology, photochemical reactions, removal processes, etc.
Twenty-five air monitoring stations have been established in and around St.
Louis for the primary purpose of model validation (Pooler, 1974). All five
primary air pollutants and selected meteorological variables will be measured.
Each site was carefully chosen in order to prevent contamination from small
local sources, dust re-entrainment from the ground, and the measurement of
anomalous winds.
Model calibration and refinement work is very highly specialized. Network
configurations, instrument specification, characteristics, and other factors
all reflect monitoring requirements that are probably unique for any given pro-
ject. It is difficult to anticipate the monitoring requirements of such pro-
jects and impossible to generalize related siting guidelines. An attempt,to
do so was considered beyond the scope of the objectives of routine monitoring
which this report addresses. However, it may be safely stated that ambient
data from any source could probably be utilized, to at least a limited extent,
in model calibration/validation studies if the conditions under which such data
was obtained were known (and documented).
14
-------
3.0 SPECIAL CHARACTERISTICS ASSOCIATED WITH S02 MONITORING
Because of the complex relationships among geographic, topographic, and
climatologic factors; S02 patterns; and the various averaging times of the NAAQS;
the selecting of appropriate sites for S02 monitoring can be a very complex
process. However, the process can be simplified somewhat by first viewing the
various siting objectives in the context of an S02 monitoring "universe". Then,
through an elimination, consolidation and optimization process, one can estab-
lish various site types such that each can be associated with a general siting
approach. Initially, it was expected that each site type could be related to
a specific siting procedure. However, because of the nature of S02 concentra-
tion patterns, the requirements of some of the siting objectives, data uses and
other factors, this was not possible in many cases. As will be seen in Section
4.0, some procedures are more closely related to the siting objective than site
type.
The major objective of this section is to discuss the elements of the uni-
verse. This includes spatial scales of representativeness and how these relate
to the averaging times of the NAAQS and the nature of urban concentration pat-
terns, terrain characteristics, meteorology, land use, and other elements.
Such a discussion will constitute an appropriate introduction to Section 4.0
(which presents the site selection procedures) by providing the site selector
a basis for understanding some of the characteristics and problems associated
with SO2 monitoring.
3.1 MONITORING NETWORK CONCEPTS
It might be appropriate to begin this section with some historical per-
spective of monitoring in general by discussing the two basic types of moni-
toring networks. Many of the networks of the recent past and several existing
ones are typified by these types.
3.1.1 Target Networks
Target networks are source-oriented in that each monitoring site has a
specific and unique objective associated with it (e.g., see Stockton, 1970).
These objectives may include the assessment of the air quality impact of a
specific large source or the combined impacts of many sources in a particular
area (usually where the maximum concentration occurs). The main concept behind
the target network is that if an objective of a control or surveillance strategy
15
-------
is achieved at a maximum concentration point, then they are achieved in all
areas of the affected region. Such a network requires a minimum number of site
locations, and for this reason they are often considered optimum networks.
This optimization also allows for a greater degree of sophistication regarding
instrument types and data acquisition systems.
3.1.2 Area Networks
Prior to the general availability of diffusion models, initial urban air
quality surveys were often conducted via an area network where large numbers
of monitors were uniformly spaced over a region, usually at each point of a
grid. The concepts behind this approach were that the more samples one had in
the field, the more likely the concentration pattern characteristics of inter-
est would be revealed, or the more accurately the regional average concentra-
tion could be computed. The earlier networks of this type were often estab-
lished for purposes of research (e.g., see Keagy, et al., 1961). Because of
the large number of sites, network maintenance was costly, and the use of ex-
pensive, high quality instruments was prohibitive. However, usually after a
year or so of experience, one could drastically reduce the number of stations
and still achieve all monitoring objectives with a reasonable degree of confi-
dence (as discussed by Herrich, 1966). In a sense, the area network was gradu-
ally converted to a quasi-target network.
Area or quasi-target networks have been established in several large metro-
politan areas where large sections are characterized by uniform land use such
as large residential and commercial areas. In these situations., site locations
are often determined on the basis of population and geographical coverage (e.g.,
see Heller and Ferrand, 1969).
There are some interesting variations of the area network type. Some may
be configured on the basis of the orientation of a major topographical feature
such as a river valley; others, on the location of a large emission district
embedded in a larger, more diffuse emission region. In these situations, indi-
vidual sampling sites may be located at points along a series of concentric axes
centered on the high emission district (e.g., see Leavitt, et al., 1957; Rossano,
1956) to "normalize" the distance-concentration factor, or at points along a
series of lines perpendicular to the valley axis to ascertain concentration flux
at each line. Other sites may be located to measure air quality upwind and
downwind of the region.
For the routine uses of SC>2 monitoring data, the characteristics of an
ideal S02 monitoring network should incorporate the desirable characteristics
of both network types.
3.2 SPATIAL SCALES OF REPRESENTATIVENESS
Much of the discussion in this section was stimulated by a recent report
by Ludwig and Kealoha (1975)a counterpart report to this one for carbon mon-
oxide monitoring. Since the scales of measurement as presented in that report
16
-------
are directly applicable to SC>2 (or to any pollutant) measurement scales, they
are presented below but restated in terms applicable to S02 monitoring.
The volume of air sampled by an SC>2 instrument is very small when compared
to the volume of air that the resulting air quality reading is supposed to re-
present (up to tens of thousands of km2). It is not possible for a single
monitor to sample all of the air volume over the area of interest to produce
the number which is the actual average air quality reading for the area. Ideal-
ly, the monitor must be placed such that the air quality of the small sampled
volume is representative of the air quality over the entire area of interest
or reasonably so. (This requirement implies a certain degree of homogeneity
over this area which is not always met, however.) The size of this area of
interest establishes a corresponding spatial scale of representativeness over
which one would like the measurement to apply.
The typical spatial scales of representativeness associated with most SC>2
siting objectives and related data uses are illustrated schematically in Figure
3-1 and discussed below, sequentially, from the smallest scale. In some situa-
tions, there are special problems associated with the representiveness of some
S02 measurements; these problems are discussed in Section 3.2.1.
Microscale. Ambient air volumes with dimensions ranging from
meters up to about 100 meters are associated with this scale.
Studies of the distribution of S02 within plumes either over
flat or complex terrain or within building wake cavities re-
quire measurements of this scale. The development of special
models designed to simulate such small scale S02 distributions
also require microscale measurements for model verification
and refinement.
Middle Scale. This scale represents dimensions of the order
from about 100 meters to 0.5 kilometer and characterizes air
quality in areas up to several city blocks in size. Some data
uses associated with middle scale measurements include assess-
ing the effects of control strategies to reduce urban peak con-
centrations (especially for the 3-hour and 24-hour averaging
times) and monitoring air pollution episodes.
Neighborhood Scale. Neighborhood scale measurements would char-
acterize conditions over areas with dimensions in the 0.5 km
to 4 km range. As will be discussed later, this scale applies
in areas where the SO2 concentration gradient is relatively
flatmainly suburban areas surrounding the urban centeror
to large sections of small cities and towns. In general, these
areas are quite homogeneous in terms of SO2 emission rates and
population density. Neighborhood scale measurements may be
associated with'baseline concentrations in areas of projected
groVth and in studies of population responses to exposure to
S02 (or health effects) . Also, concentration maxima associated
with air pollution episodes may be reasonable uniformly distri-
buted over areas of neighborhood scale, and measurements taken
within such an area would represent neighborhood as well as
middle scale concentrations.
17
-------
,x Micro-Scale
(<0.1 km)
Neighborhood Scales
(0.5 to 5 km)
URBAN COMPLEX
Regional Scales
(>50 km)
Urban Scales
(4 to 50 km)
FIGURE 3-1. Illustration of various spatial scales of representa-
tiveness .
Scale, prban scale measurements woulfl be made to re-
present conditions over areas with dimensions on the order
of 4 to 50 km. Such data could be used for the assessment
of air quality trends, the effect of control strategies on
urban scale air quality.
18
-------
Regional Scale. Conditions over areas with dimensions of as
much as hundreds of kilometers would be represented by re-
gional scale measurements. These measurements would be applic-
able mainly to large homogeneous areas, particularly those
which are sparsely populated. Such measurements could provide
information on background air quality and interregional pollu-
tant transport.
National and Global Scales. These measurement scales repre-
sent concentrations characterizing the nation and the globe
as a whole. Such data would be useful in determining pollu-
tant trends, in studying international and global transport
processes, and in assessing the effects of control policies
on national and global scales.
3.2.1 Measurement Scales Relevant to S02 Monitoring
In SO2 monitoring, a distinction should be made between the spatial scale
desired to be represented by a single measurement and the spatial scale actu-
ally represented by that measurement. The former is determined by the size of
the area of interest which is associated with the intended use of the data and
associated siting objective, while the latter is a function of the spatial vari-
ation of concentration in the horizontal over the area of interest. This vari-
ation results not only from the impacts of local sources within the area, but,
more importantly, from the collective impacts of all sources located outside
of the area of interest. These collective impacts result in background concen-
tration patterns and gradients over the area of interest that essentially dic-
tate the spatial scale that will be represented by a single measurement taken
at a station located anywhere in that area. This dilemma may be stated in an-
other waythe distance from a monitoring station at which measurements become
significantly different from those at the monitoring station determines the
spatial scale represented by measurements at the monitoring station. This
distance is a function of the background concentration gradient. S02 concen-
trations over urban areas generally decrease rapidly outward from a peak near
the urban center, and rather smoothly for annual averaging times (e.g., see
Larsen, et al., 1961; and Figure 2-3, Stern, et al., 1973) as shown in Figure
3-2. Also, superimposed on the relatively smooth concentration pattern are
"bumps"* due to large point sources. Hence, SC>2 concentrations in cities are,
in general, neither uniform over large, homogeneous land use areas within the
city, nor are they contained within numerous individual independent cells or
street canyons as is the case for carbon monoxide (c.f., Figure 3, Ott, 1975).
Because of this nature of SC>2 distributions over urban areas, the middle
scale is the most likely scale to be represented by a single measurement in an
urban area, and only if the undue effects from local sources (minor or major
point sources) can be eliminated. Neighborhood scales would be those most
likely to be represented by single measurements in suburban areas where the
concentration gradients are less steep. Regional scale measurements would be
For shorter averaging times these bumps become large "spikes" superimposed
on a greatly irregular background pattern.
19
-------
associated with rural areas. Micxoscale measurements may be required in cer-
tain situations. For example, in monitoring the impact of an isolated point
source in complex terrain,'initially it may be desirable to use mobile samp-
ling or to establish a dense, area-type network to determine the general loca-
tion of the maximum impact point. This will provide guidance for locating
permanent sites for measurements representing the more relevant middle scale.
Normally, investigators making such microscale measurements have specific sit-
ing requirements that reflect the specific and often unique purposes of .their
projects; these requirements would be difficult to generalize.
Because of the great variation of S02 concentrations in urban areas, it
is unlikely that urban scale concentrations could be measured at a single site.
National and global scale concentrations are not of sufficient interest
to state and local agencies to justify specific treatment. However, concentra-
tions characterizing areas on these scales may be estimated by synthesizing re-
gional, and then national scale measurements.
Figure 3-2 shows relative locations of sites in an urban area for measur-
ing concentrations representing several spatial scales of measurement.
I
c
o
ro
C
0)
o
c
o
o
C>4
O
to
0)
a:
Mi
Md
N
R -
Mi cro Scale Si tes
Middle Scale Sites
Neighborhood Scale
Sites
Regional Scale
Sites
k-
Mi & Md
Peaks Associated with
Major Point Sources
rural areas
suburbs i urban core 'suburbs rural areas
i i
'city limits '
FIGURE 3-2. Relative locations of sites for measuring concentrations
representing several spatial scales of measurement in an urban
complex, with respect to annual averaging times.
20
-------
3.3 MONITORING SITE TYPES AMD ASSOCIATED SITING OBJECTIVES AND DATA USES
Our survey of the literature indicated that SO2 monitoring sites can be
classified as either proximate or general level. Proximate sites are those
associated with siting objectives that require information regarding impacts
from a specific source or a group of specific sources. These sources may be
isolated, such as a smelter complex in a remote area, or a power plant so lo-
cated in a city that it constitutes a large fraction of the total observed SO2-
General-level sites are those located in areas where the total concentration
is important but contributions from individual sources to that concentration
are relatively unimportant.
The siting objectives and related data uses and their associated site
types and spatial scales of representativeness are summarized in Table 3-1.
Blank spaces indicate those scales of measurement that are either inconsistent
with the siting objective, or are simply not very useful. Proximate site types
are indicated by "Pr" and general level by "GL". The underlined Xs indicate
^e desired spatial scale to be represented by a single measurement. The re-
maining Xs indicate other spatial scales that may actually be represented by
a single measurement (because of the conditions imposed by the background con-
centration gradients). The letters (P) and (F) within the site type column
indicate whether the siting objective is concentration pattern oriented or is
associated with a fixed geographical area independent of the SO2 pattern. For
example, an urban peak concentration site (P) will be located as close as pos-
sible to the peak concentration point in the city without regard to the geo-
graphical setting of the siting area, while a site established to determine
base concentrations in areas of projected growth (F) will be located within
the growth area regardless of the characteristics of the prevailing SC>2 con-
centration pattern. It also might be worthwhile here to mention that the less
complicated the source mix and density (i.e., as one approaches rural condi-
tions) the wider the range of spatial scales a reading will represent; for
example, in a homogeneously rural area, an individual reading will represent
all spatial scales ranging from micro to regional and over any averaging time.
3.4 THE SO2 MONITORING UNIVERSE
In the foregoing discussions, we have identified the uses of S02 data and
their relationships to specific monitor siting objectives; we have also related
the individual siting objectives to appropriate spatial scales of representive-
ness (see Table 3-1). However, there are other variables that must be consi-
dered in the site selection processnamely the averaging times of the NAAQS
and the land use and topographical settings. All combinations of the above
variables that must be accounted for in the selecting of monitoring sites, and
to a certain extent, in determining probe exposure and monitoring mode, con-
stitute an S02 monitoring "universe". It is from this universe that specific
site types are selected to which are attached specific site selection proce-
dures.
The five basic variables (the first two have already been discussed) that
constitute the S02 monitoring universe are listed following Table 3-1.
21
-------
TABLE 3-1
Relationships Among Siting Objectives and Related Data Uses, Site
Types, and Scales of Representativeness
(a) If the assumption is made that the peak concentration point will only rarely occur (within middle-scale
limits) at the monitoring site, then the reading will better represent typical maximum values on the neigh-
borhood scale in the maximum impact area.
(b) Microscale measurements may be required to define plume structure via either area network or mobile sanpling
to simulate plume or to estimate permanent middle scale site locations.
(c) Under stagnation conditions, the maximum concentration zone will probably expand in area, in which case the
reading nay represent neighborhood scale averages as well as middle scale averages*
(d) Because of the multitude of scales on which models are designed to simulate air pollution, data on any
scale nay be required in model calibration/refinement work.
* The 'Spatial Scale of Representativeness" is keyed as follows: I - micxoscale; II - middle scalei
III - neighborhood scale; and IV - regional scale.
Siting Objectives/ Data Uses
1. Determination of Peak Concentrations in Urban Areas.
Judging attainment/maintenance of NAAQS.
Evaluating progress in achieving/maintaining of NAAQS.
Developing/revising SIPs/evaluating control strategies.
Providing data to facilitate the ESECA of 1974.
Supporting enforcement actions.
Public information.
2. Determination of the Impact of Individual Point Source in Multi-Source
Urban Setting.
Developing/revising SIPs/evaluating control strategies.
Reviewing new sources.
Developing/revising national S02 control policies(NSPS,SCS, tall stacks).
Providing data to facilitate ESECA of 1974.
Supporting enforcement actions.
3. Determination of the Impact of Isolated Point Sources.
Developing/revising SIPs/evaluating control strategies.
Reviewing new sources.
Developing/revising national S02 control policies(NSPS,SCS, tall stacks).
Providing data to facilitate the ESECA of 1974.
Supporting enforcement actions.
4. Assessment of Interregional S02 Transport
Establishing baseline air quality levels for PSD planning and AQMP.
Evaluating progress in achieving/maintaining NAAQS.
Developing/revising SIPs to attain/maintain NAAQS.
Public Information.
5. Determination of Base Concentration In Areas of Projected Growth.
Establishing baseline air quality levels for PSD planning and AQMP.
Evaluating progress in achieving/maintaining NAAQS.
Developing/revising SIPs to attain/maintain NAAQS.
Public information.
6. Emergency Episode Abatement Initiation and Monitoring.
Documenting episodes and initiating episode controls
Public Information.
7. Assessment of Background Contration in Rural Areas.
Establishing baseline air quality levels for PSD planning and AQMP.
Developing/revising SIPS to attain/maintain NAAQS.
Public Information.
8. Determination of" Population Exposure 1n Populated Areas.
Documenting population exposure and health research.
Public Information.
9. Diffusion Model Calibration and Refinement, (d)
Site
Type
GL
(P)
Pr
(P)
Pr
(P)
GL
(P)
GL
(F)
GL
(P)
GL
(P)
GL
(F)
GL
Pr.
Spatial Scale of
Representativeness*
1
(b)X
X
I
II
(a)X
X
X
X
X
X
I
111
X
X.
(c)X
X
£
IV
X.
X
X.
X
x.
22
-------
1) Site Type
4)' Land Use Setting
Proximate
General Level
2) Spatial Scale of Representativeness
Microscale
Middle Scale
Neighborhood Scale
Regional Scale
3) Averaging Time of MAAQS
3-hour (second highest)
24-hour (second highest)
Annual
Urban
Suburban
Rural
5) Topographical Setting
Coastal
Ridge-valley
Interior Plain
. Rugged, Irregular (interior)
Rugged, Irregular (coastal)
Only a portion of the monitoring universe is presented in Figure 3-3 which
shows only 15 combinations of variables. For the entire universe, the combina-
tions total 360. Each combination could theoretically require a unique set of
siting procedures depending on the siting objective, data use, and the commonali-
ty and availability of meteorological data for the various combinations. How-
ever, it will be seen that the combinations of theso universe elements that
reflect the stated siting objectives can be accommodated by a relatively small
number of site types.
BASIC SITE TYPE
SPATIAL SCALE
AVERAGING TIME
TOPOGRAPHICAL
SETTING
FIGURE 3-3. Portion of SO2 monitoring, universe.
23
-------
3.5 THE *"1VE REuEVANT MONITORING SITE TYPES
The concept of the monitoring universe as presented above can be converted
to more convenient tabular format. Tables 3-2 and 3-3 show the resulting uni-
verse after considering only the elements of Table 3-1, and the desirable spa-
tial scale to be represented by a
single measurement. For example, an
TABLE 3-2 SO2 reading representing a regional
3-hour mean concentration associated
Relationships Among Table 3-1 with an isolated point source either
Elements and Associated does not exist or is irrelevant.
Relevant Averaging Times ' From Table 3-2, all siting objectives
1
2
S
! 3
I
N
G 4
0
B
JE 5
c
T
} «
E
S
7
8
9
Basic
Site
Type
Pr
GL
(P)
Pr
(P)
GL
Pr
(P)
GL
Pr
GL
(P)
Pr
~~ GL~
(F)
Pr
GL
(P)
Pr
GL
(P)
Pr
~ Gl~
(F)
Pr
(P)
GL
(P)
Averaging T me
3-hour
Md
Md
Mi.Md
R
**
N
R
t
Mi ,Md
Mi ,Md,
N,R
24-hour
Md
Md
Mi ,Md
R
N
N
R
N
Md
Md,
' N,R
Annual
Md
Md
JL
R
N
t
R
N
Md
Md,
N,R
can be accommodated by five monitorii
site types :
1) General Level, Regional Scale.
2) General Level, Neighborhood Seal
3) General Level, Middle Scale.
4) Proximate, Middle Scale.
5) Proximate, Microscale.
KEY
Mi - Microscale.
Md - Middle Scale.
N - Neighborhood Scale .
R - Regional Scale .
Pr - Proximate .
GL - General Level .
(P) - Pattern Oriented Site.
(F) - Fixed Geographically Oriented
Site.
* Not likely for an isolated point
source .
** Difficult to estimate since no
specific source is impacting.
t No episodes occur in this time
scale.
t Secondary standard; no signifi-
cant effects .
24
-------
TABLE 3-3
Matrix of Topographical and
Land Use Types
Land
Use
Type
K E
~*- ^»
U -
S -
R -
A -
B -
(U)
(S)
(R)
Topographical Type
(A, B, C, D, E)
A,U B
A,S B
A,R B
Y_
Urban C
Suburban D
Rural
Coastal E
Plain
,U C,U D,U
,S C,S D,S
,R C,R D,R
E,
E,
E,
- Ridge/Valley
- Rugged, Irregular
Interior
- Rugged, Irregular
Coastal
U
S
R
r
r
Each of these site types is associated with a basic procedural siting approach
with variations from the basic approach being functions of the siting objec-
tive, averaging time, and physical setting.
It would be appropriate at this point to summarize the material presen-
ted in this section by showing an example of the process that ties the site
type to the intended use of the data. This can be accomplished in a stepwise
manner as follows: ,
a) Decide the use to which the data will be put.
EXAMPLE: Providing data to implement the provisions of the
ESECA of 1974.
b) Determine all siting objectives that will satisfy the data use.
From Table 3-1, siting objectives 13 2, and 3 will satisfy
the proposed use of the data.
c) From Table 3-2, determine the site type and averaging times
of concern that apply to each siting objective.
SITING OBJECTIVE 1: General-Level, Middle-Scale, all
SITING OBJECTr/E 2:
averaging times.
Proximate, Middle-scale, all averaging
times.
SITING OBJECTIVE 3: Proximate, Middle and/or Miorosoale
3-hour and 24-hour averaging times.
25
-------
d) From Table 3-3, determine the physical setting of the siting
area. There are 15 combinations of physical settings that
are relevant to the SC>2 monitoring site selection process.
The specific siting procedures associated with each site type are presented
in the next section.
26
-------
4.0 SITE SELECTION PROCEDURES AND CRITERIA
Procedures and criteria for selecting specific monitoring site locations
and instrument inlet exposures have been developed for the relevant monitoring
site types and are presented in this section. They are generally uniform for
each siting objective associated with a given site type. It was not possible
to develop specific siting criteria to satisfy some siting objectives (e.g.,
isolated point source monitoring in extremely rough or mountainous terrain).
In these instances, general guidelines addressing "points to consider" are
presented.
The site selection process itself is an elimination process; general sit-
ing areas are selected, then specific prospective sites within the areas are
gradually eliminated in accordance with specific criteria until a small subset
of acceptable sites remains. The final selection is made from this subset.
It should be made clear at this point to state that not every AQCR is re-
quired to have each type .of monitoring site described in this section; the
types of sites required would depend on the nature of the SO2 problems in the
AQCR. These are judgements to be made by the control agency or site selector.
The organization of much of this section is based on that of the report
by Ludwig and Kealoha (1975)essentially flow charts showing the basic struc-
ture and flow of the procedural logic followed by discussion of the elements
of the flow chart. Much of the material is presented without discussion of
the justification or rationale for the various steps of the procedures; this
is reserved for Chapter 5.0 to maintain the clarity and continuity of the pro-
cedures as discussed here.
4.1 DESCRIPTION OF SITE SELECTION AIDS AND BACKGROUND MATERIAL
An integral part of the site selection process is the acquisition and/or
development of background material, data, and, in some situations, the use of
auxiliary equipment (e.g., portable wind station). Such material is needed to
provide the site selector with information mainly regarding the physical char-
acteristics of the siting area. This information may include the terrain and
land-use setting of the prospective monitor siting area, the proximity of large
water bodies, the distribution of SC>2 sources in the area, the location of ap-
propriate National Weather Service (NWS) airport stations from which weather
data may be obtained, etc. Depending on the siting objective, this material
may take the form of:
27
-------
Isopleth maps S02 air quality,
Emissions inventories,
Meteorological data,
Wind roses,
Portable wind equipment,
Topographic/population/land-use maps, and
Mobile sampling equipment.
The purpose of each item will be described briefly below prior to the presen-
tation of the site selection procedures. A more complete discussion of this
material and its sources can be found in the appendices and in Section 5.0.
Isopleth maps, particularly those generated by diffusion models is recom-
mended for use in determining the general location of a prospective monitoring
site, or a prospective siting area within which the final site is to be selec-
ted. For siting monitors in urban areas, multi-source models such as the Air
Quality Display Model (AQDM) are recommended. For isolated point source moni-
toring in relatively uncomplicated terrain, various point source models (PTMPT,
PTDIS, PTMAX; see Appendix E) or graphical solutions of the Gaussian point
source equation are suggested. (It will be seen that the guidelines presented
herein are strongly diffusion-model-output oriented.)
Emission inventory information for point sources is available from the
U.S. Environmental Protection Agency (EPA) for any area of the country for an-
nual and seasonal averaging times. Specific information characterizing the
emissions and large point sources for the shorter averaging times (diurnal
variations, load curves, etc.) can often be obtained from the source. Area
source emission data by season, although not available from the EPA, can be
generated by apportioning annual totals according to degree days. This kind
of information provides some of the input to the diffusion models and are also
important for other reasons that will be discussed later.
The nature of the elements of Table 3-3 in Section 3.0 determine the mete-
orological and diffusion parameter input to the diffusion models. In most
cases, the meteorological data originating from the most appropriate (not ne-
cessarily the nearest) NWS airport station in the vicinity of the prospective
siting area will adequately reflect conditions over the area of interest, at
least for annual and seasonal averaging times. In developing data in complex
meteorological and terrain situations, diffusion meteorologists should be con-
sulted. A complete list of NWS stations that can provide most of the relevant
weather information in support of siting activities anywhere in the country can
be found in Appendix A. Such information includes joint frequency distributions
of winds and atmospheric stability (stability-wind roses). These are provided
by the output of the National Climatic Center "STAR" computer program. For the
shorter averaging times or in complex terrain situations, the use of portable
wind equipment, smoke bombs, time-lapse photography may be necessary. Land use
28
-------
types and topographical characteristics of specific areas of interest can be
determined from U.S. Geological Survey (USGS) and land use maps. Detailed in-
formation on urban physiography (building/street dimensions, etc.) can be ob-
tained from Sanborn maps (see Appendix D). Additional information may have to
be obtained by visual observations, aerial photography, and surveys to supple-
ment that available from the above sources. Such information may be required
to determine the appropriate diffusion coefficients and SC>2 half-life values
to be used by the models as well as determining the locations of local sources
in and around the prospective siting areas.
Finally, after the general location of a site or prospective siting area
has been established, mobile sampling may be required to determine the optimum
site location, particularly in regard to isolated point source monitoring (see
Appendix C).
4.1.1 The Critical Role of Diffusion Modeling in the Site Selection Process
As discussed in Section 3.2.1, the SO2 background concentration gradient
over an area essentially determines the spatial scale represented by measure-
ments taken at a single station located in that area. Also, it was seen that
in Section 3.3 that some siting objectives may be associated with specific fea-
tures of the S02 pattern while others may be associated with fixed geographi-
cal areas that are independent of such features. Since the only objective
means of obtaining such gradients and patterns is by diffusion modeling, the
modeling of the area of interest will usually be a prerequisite for selecting
monitoring sites.
In Sections 4.1.1 and 4.1.2 below, the role of diffusion models in the
site selection process is discussed in general terms to orient the reader.
The model's role in the selecting of sites to satisfy specific siting objec-
tives is discussed in more detail later in the appropriate sections.
4.1.1.1 Siting Objectives Associated with Fixed Geographical Areas
For siting objectives associated with fixed geographical areas (see Table
3-1), measurements from a single monitoring station within such an area can
represent concentrations over any spatial scale. In a given scenario, the
particular spatial scale represented would depend directly on the background
concentration gradient prevailing over the area of interest. In developing
the siting criteria in these situations, we arbitrarily chose specific concen-
tration gradients that we felt appropriately characterized the various spatial
scales (see Section 5.3.3 for rationale) as follows:
1) If the concentration gradient over the area of interest does
not exceed about 0.5 yg/m3-km, the measurements from a single
site will represent concentrations over regional spatial scales.
2) If the concentration extremes over the area of interest are not
within about 25% of the mean value, then more than one site is
required to represent concentrations over the area. To establish
the number of stations, the area is divided into the number of
29
-------
parcels required to bring the extreme concentrations over each
parcel to within 25% of the mean of each parcel. Measurements
from single sites located near the center of each parcel should
adequately characterize the concentration over that parcel. The
spatial scales represented by the measurements will be the same
as the spatial scales of the parcel, namely neighborhood (0.5
to 4.0 tan) or middle (0.1 to 0.5 km) scales.
4.1.1.2 Siting Objectives Associated With Features of the S02 Pattern.
The peak S02 concentration is usually the most important feature consid-
ered for siting objectives associated with features of the SC>2 pattern (see
Table 3-1). The peak concentrations may be due to either single or multiple
sources. The diffusion model is used to determine the approximate location, of
the peak and can consider annual patterns as well as near worst case 3-hour
and 24-hour patterns. Because the concentration gradient in the vicinity of
the peak is often steep and/or irregular, the middle scale is the most likely
scale to be represented by measurements from a single station located near the
peak.
The use of models to aid in determining regional scale site locations in
rural areas is optional. In these situations, the model is used to verify that
the prevailing concentration gradient is relatively flat.
4.2 GENERAL-LEVEL REGIONAL-SCALE STATIONS
Figure 4-1 shows the recommended procedure siting objectives for estab-
lishing general-level regional stations. There are two basic siting objectives
for which regional stations are established: (1) to measure regional mean
background concentrations; and (2) to assess pollutant transport.
The following material should be assembled to provide inputs to the de-
cision-making process;
Wind roses,
Regional uaps of various scales showing topography and
developed areas,
Population data (by town),
Emissions inventory of point and area sources.
Diffusion model output (optional).
Climatological wind data in the form of a statistical table such as that
shown in Table 4-1, or a wind rose shown in Figure 4-2, are the forms most use-
ful in selecting general-level regional stations. These are examples of some
of the kinds of data that are available from the National Climatic Center (NCC)
Asheville, North Carolina. The wind rose is particularly useful in depicting
the wind direction frequency over the area of interest.
30
-------
Assemble Site Selection Aids:
Climatological Wind Data
Maps (various scales) showing
- topography
- developed areas
Population Data by Town
Diffusion Model Ouput (Optional)
Is Purpose of Site to Assess SO?
Transport or Measure Regional Mean
Concentrations?
Interstate, general
Tentative primary site located
as near as possible to state
Tine on out-of-state urban
area side, as close to urban
area as possible, but no closer
to an in-state urban area than
30 km. Additional sites lo-
cated directly downwind of ur-
ban area in prevailing winter
or annual wind direction, or
symmetrically about the pri-
mary/secondary siting areas
(see text).
Tentative Siting
area located as
near as possible
to state line.
If only one site
is considered,
then locate on
winter windward
side; i.e., state
line toward maxi-
mum winter wind
frequency
Eliminate Prospective Specific Sites Within:
30 km of large point sources, e.g., 400 MW power plant.
15 km of medium-sized town, population 25,000.
10 tan of large industrial source, emissions 500 T/yr.
0.6 km of individual home (103 gal/yr of 12 oil).
Avoid low lying areas. (see Table 4-2)
Intercity
Single Site: Tentative
siting area located not
less than 30 km from
large urban area upwind
in most frequent winter
wind direction (also,
see Table 4-2). Addi-
tional sites may be lo-
cated same distances
upwind in the next most
frequent winter wind
direction
Tentative siting area located not
less than 30 km from nearest urban
area. Urban area should be toward
direction of the lowest winter
wind direction frequency and in an
area characterized by flat concen-
tration gradient, as revealed by
diffusion model output, and in an
area of uniform topography
Site Characteristics: Trailer or
existing permanent structure.
Inlet Height: 3-10 m above ground.
FIGURE 4-1.
Flow chart showing procedures for locating
general-level regional-scale stations.
31
-------
TABLE 4-1
Example of a Tabulated Wind Summary
(taken from the National Climatic Center, Asheville, N.C.)
PERCENTAGE FREQUENCIES
OF WIND DIRECTION AND SPEED:
(XMCTION
N
NNE
NE
ENE
E
ESE
SE
SSE
S
SSW
sw
wsw
w
WNW
NW
NNW
CALM
TOTAL
HOURLY OBSERVATIONS OF WIND SPEED
(|N MILES m HOUIJ
0-3
f
f
+
+
+
f
+
+
1
+
+
+
*
4-
-f
+
+
4
4 -7
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
13
8 -i:
2
2
3
2
1
1
2
2
2
1
1
3
3
2
2
1
30
13- II
1
1
2
1
f
+
1
1
3
3
3
A
5
3
2
1
30
19-74
f
+
f
+
+
+
+
2
2
2
4
3
1
1
+
17
33-31
+
+
4-
+
+
1
1
1
1
+
f
f
f
5
37-M
+
+
+
4-
+
+
1
3»- U,
»
+
+
»
47
OVH
+
+
TOT At
4
4
6
4
3
2
4
4
10
8
8
14
12
7
6
4
+
100
AV
ifKO
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
MILES PER HOUR
1-3 6-IS 16-30 >30
10
15
20
3O
PERCENT FREQUENCY
FIGURE 4-2.
A typical wind rose with wind-speed information
(taken from Slade, 1968).
32
-------
Maps showing the physical featuresnatural and man-madeof the region
are important since the monitoring purpose and location of the monitors,are
based on the nature and distribution of these features. Demographic and emis-
sion inventory data will provide additional useful inputs.
The selection process begins by deciding on the siting objective after
which a specific series of steps is followed. This process is presented be-
low.
4.2.1 Regional Mean (Background) Concentration Stations
A tentative siting area should be established no closer than about 30 km
from the jurisdictional boundary of any major urban S02 source area in the re-
gion. (Consider a city with a population of 2xi05 or more as constituting a
major urban area.) The nearest major urban area should be toward the direc-
tion of the lowest winter (Dec, Jan, Feb) wind frequency. If available, a
winter seasonal (or annual concentration map generated by a diffusion model
can be utilized to ensure that the tentative siting area is not located in an
area characterized by a steep concentration gradient (>0.5 yg/m^ - km).
The topography of the region containing the urban SC>2 source areas, the
tentative siting area, and the NWS station from which the wind rose data orig-
inated should be reasonably uniform.
4.2.1.1 Local Characteristics, Interferences and Inlet Placement
Guidelines for considering local physical characteristics, proximity of
interfering sources in the vicinity of the final site and instrument inlet
placement are the same for all regional scale stations. Once the siting areas
have been established, individual prospective sites should be eliminated on
the basis of the proximity of small, local S02 sources that may unduly influ-
ence the measurements. These sources, or source types, and corresponding
"interference" distances* are shown in Table 4-2. Regional scale S02 monitors
should be sited no closer to these sources than the interference distances.
Since low lying areas are associated with relatively higher inversion fre-
quencies, they should be avoided. Open or sparsely forested areas are recom-
mended with the instruments housed in an existing permanent structure or trailer
Since all pollutants are well-mixed in the vertical over outlying areas, exact
inlet height is not important. A height range of from 3 to 10 m above the
ground would be reasonable. In densely forested areas, the inlet tube should
be raised a few meters above the tops of the surrounding trees.
Figure 4-3 is a schematic illustrating the tentative siting area for a
regional mean concentration station.
The interference distances are defined in Section 5.0. They were developed
via solutions to the Gaussian point source formula by assuming certain
"worst case" conditions.
33
-------
TABLE 4-2
Source Types and Related Interference Distances for
Regional Scale Monitoring Stations
Source Type
Interference Distance
Large point source (e.g.., a 400 MW power plant)
Industrial source (500 tons S02 per year)
Towns (various size population)
50,000
25,000
12,500
6,000
Individual home
30 km
10 km
22 km
15 km
10 km
7 km
0.6 km
WIND ROSE
Si te selected from
choices within
shaded area.
FIGURE 4-3.
Schematic illustration showing tentative siting area for
regional mean background concentration stations.
34
-------
4.2.2 S02 Transport Stations
S02 transport stations may satisfy several siting objectives. Three such
objectives and the siting procedures for establishing the stations are dis-
cussed below.
Interstate SO? Urban Transport Sites. These sites are estab-
lished for measuring incoming interstate S02 that originates
from a large urban complex outside of the state (e.g., New
York City/Connecticut; Chicago/Indiana). A primary siting
area should be located as close as possible to the state line
opposite the out-of-state urban area, but no closer than 30
km to an in-state major urban area. If the state or part of
the state is dire'ctly downwind of the out-of-state urban area
in the winter prevailing direction, a secondary siting area
could be established near the state line, as described above,
but directly downwind (winter) of the urban area. These are
illustrated in Figure 4-4. Alternatively, a series of stations
could be placed symmetrically about the primary or secondary
siting areas (to obtain horizontal profiles of incoming pollu-
tant, for example).
SECONDARY
SITING AREA/
OUT-OF-STATE
URBAN CENTER
FIGURE 4-4. Schematic illustration showing primary and secondary siting areas
for interstate urban transport stations.
35
-------
interstate SC>2 Transport-General. If there are no major
out-of-state urban areas contributing significantly to in-
coming S02/ a general SC>2 transport station can be estab-
lished anywhere (but no closer than 30 km to an in-state
major urban area) along the winter windward state line.
Intercity S02 Transport Sites. If S02 flux entering a city
is desired to be measured, a primary siting area may be es-
tablished upwind of the city boundary in the most frequent
winter wind direction, at distances which depend on city
size. These distances arange from about 30 km for cities
of 2*10^ population or more to 15 km for small towns of
25,000 population (see also Table 4-2, "Towns"). Secondary
siting areas can also be established toward the next most
frequent direction, etc. Figure 4-5 shows the location of
the siting areas for intercity S02 transport stations. Al-
ternatively, other sites may be established directly between
two cities, without regard to wind direction, to assess the
exchange of SO2 between the two cities. Similarly, as for
the siting areas for regional mean concentration stations,
the topography of the entire region should be reasonably
uniform.
Guidelines for considering local characteristics, interference distances,
and inlet placement are the same as for regional mean concentration stations
(see Section 4.2.1.1).
SECONDARY
SITING AREA
PRIMARY
SITING AREA
URBAN
CLNTER
FIGURE 4-5.
Schematic illustration showing primary and secondary siting areas
for intercity background stations.
36
-------
4.3 GENERAL-LEVEL, NEIGHBORHOOD-SCALE STATIONS
There are three major siting objectives associated with neighborhood-
scale stationsmonitoring emergency episodes, determining baseline concen-
trations in areas of projected growth, and monitoring SC>2 concentrations to
which certain human populations are exposed. The specific objectives chosen
for which monitoring will be undertaken will determine, in large measure, the"
siting procedure and kind of background information required.
Figure 4-6 shows the recommended procedure for locating the three kinds
of general-level, neighborhood-scale stations. The first step is to decide on
the objective of the monitoring, after which follows the gathering of back-
ground information and the site selection process itself, as discussed below.
4.3.1 Emergency Episode Stations
The background information required for the proper siting of emergency
episode stations includes:
Emissions inventory of point and area sources,
USGS map of urban area, and
Sanborn map of urban area (see Appendix D).
Prospective emergency episode stations should be located near the center of
the maximum low-level emission zone(s) of the urban core. The maximum emis-
sion zone can be found by plotting the S02 emission rates of the area source
fraction of the inventory, in tons per year per UTM grid square*, on a gridded
USGS map of the urban area. Isopleths of constant emission rate may be drawn
as an option to reveal the center(s) of the zone(s). Mobile sampling may also
be undertaken during an actual episode or in near-episode conditions (e.g., in
the morning when winds are light and variable) to better define the general
area of maximum concentration. Figure 4-7 is a schematic illustrating the
general location of the prospective siting area.'
The final site is selected from a list of candidate sites located near
the center of the zone in accordance with the desirable site characteristics
and inlet placement criteria as shown in Table 4-3. In general, because very
little turbulence and unsteady winds usually prevail during atmospheric stagna-
tions, undue influence from an individual nearby source is minimal. However,
if the data from this site is to be used to supplement data from peak concen-
traion stations, then the site characteristics should be consistent with the
criteria shown in Table 4-5 (see Page 45).
Emission inventory grid systems normally utilize the kilometer-based Uni-
versal Transverse Mercator (UTM) system. UTM tick marks are shown in the
margin of most USGS maps.
37
-------
What is the siting objective?
Emergency Episode Monitoring
Population Exposure/Projected Growth Monitoring
Assemble Background Material:
Emission inventory of point and area sources.
USGS map of urban area.
Sanborn maps of urban area.
Assemble Background Material:
Emission inventory" of point and area sources.
Meteorological data.
Other background material (see text).
Plot urban area source emission rates
by grid on USGS map. locate tentative
Siting area near center of maximum
emission zone(s). Conduct mobile
sampling to better define maximum
concentration zone(s).
Perform diffusion model analysis of
region containing areas of interest:
Winter season.
Flna? Site - General Characteristics:
Site Location. Trailer or existing
permanent structure, Ht i 0.8 of
mean building height. No SO?
source on roof.
Inlet Location. Precise location
not critical; little turbulence/
no steady winds, undue local in-
fluence minimal (see Table 4-3).
Population Exposure Stations | Projected growth Area Stations
From population maps, de-
lineate area of interest.
Superimpose winter SOj
concentration pattern over
area of interest.
From growth maps (population,
land use, industry, etc.)
delineate area of interest.
Superimpose winter 502 con"
centration pattern over area
of Interest
Is one station sufficient to represent
concentration over area of interest?
Locate siting area at center of area of
Interest and identify with 0.5 to 1.0 kn
diameter circle.
Determine prevailing winter wind direction
and direction associated with short-term
peak concentration (see text).
Subdivide Area
Depending on size of area, follow:
Regional-scale procedures (Section 4.2.1).
Neighborhood-scale procedures.
Fro" Inventory, Sanborn maps and/or photographs, identify all source points over designated areas
upwind of siting area. Construct 10V20' plumes from each source point in the downwind direction
over designated areas. Consider both wind directions. Then choose prospective sites well within
built-up sections of the siting area.
Eliminate specific sites located within 10°/20° plume sectors and buildings with stacks from con-
sideration. Choose sites such that impacts from SO; sources in other directions are minimized.
Attempt to satisfy prevailing and peak concentration monitoring with one site.
Final Site - General Characteristics:
Rooftop" Ht <. 0.8 of mean building height and 1-2 m above roof. Inlet on windward side of building.
Similar to rooftop. Locate inlet no closer than J-2 m from building.
If possible, avoid lot around which are buildings with stacks.
ntermediate Height.
railer. Locate in non-parking lot.
Inlet height, 3-5 m above ground (see Table 4-5).
FIGURE 4-6.
Flow chart showing procedures for locating general-level
neighborhood-scale stations.
38
-------
City Boundary
UTM Grid Network
I i I .1 t
571 'X f yi-.gzK I "»50'!30//3a 371
J^ ff^CZZH. SITING AREA
FIGURE 4-7 General location of emergency episode siting area.
Numbers indicate relative emission rates. Each
grid square equals 1 square kilometer.
TABLE 4-3
Site Characteristics and Inlet Placement Criteria
for Emergency Episode Stations
Station
Location
Rooftop
Intermediate
Height on
Building
Trailer
Inlet Placement Criteria
Height
Above Ground
Somewhat less than
mean height of
buildings in zone
or lowerC^-O.SFO*
Same as
Rooftop
3-5 meters
Height
Above Roof
Not critical. Be-
tween 1-2 ra, and
away from dirty/
dusty areas.
NA
1-2 meters
Horizontal
Clearance
Eevond Structure
Not Applicalbe
(NA)
> 2 meters
NA
Horizontal of
Inlet Placement
Inlet may be placed
anywhere on roof, but
away from dirty/dusty
areas.
Inlet may be located
on any side of build-
ing, preferable the
side away from near-
est sources.
NA
Remarks
Ho SO? source on
roof
No S02 source on
roof
If possible, avoid
parking lots and
other lots around
which are buildings
with stacks. If lo-
cated in park, avoid
sites under thick
"forest" canopy.
H = mean building height in zone.
39
-------
As a supplementary procedure, population figures could be utilized in a
manner similar to that described above for S02 emission rates, to determine
population densities and characteristics (e.g., age frequencies in the maximum
emission zone.
4.3.2 Population Exposure and Projected Growth Monitoring Stations
Siting procedures for these two siting objectives are very much the same
after the subject population areas and projected growth (residential, indus-
tiral, etc.) areas have been identified and delineated. Since the monitoring
of air quality in regions of projected growth is related to the EPA-mandated
AQMP process, the reader is referred to Vol. IX of the EPA's Maintenance Plan-
ning Guideline series of documents (discussed briefly in Section 2.2.7 of "this
report) for additional information. The siting procedures for projected growth
monitoring sites presented below are consistent with the concepts discusses in
that document.
The background information and other aids that would be useful for selec-
ting population exposure and projected growth stations include:
Emission inventory of point and area sources.
Meteorological data reflecting conditions imposed by topo-
graphical and land use setting;* winter season.
USGS/land use/population mpas of area.
Air Quality Maintenance Plan (AQMP).
Sanborn maps of urban area.
Air Quality Display Model (AQDM) or equivalent.
The emissions inventory and meteorological data will provide the required in-
put to the AQDM for generating an SO2 concentration field over the areas of in-
terest as well as providing information on the location and strengths of all
point sources in the urban area.
After assembling the required background materials, delineate the subject
population area and/or the projected growth area, depending on the monitoring
objective chosen. Simulate an SO2 concentration field over the region contain-
ing these areas using the AQDM with the emission inventory and meteorology re-
flecting winter quarter conditions. If the areas of interest are located in
large urban areas (population >. 10 ), use a half-life of 1 hour, otherwise use
a 3-hour half life. Superimpose the concentration map over the areas of inter-
est. At this point, it must be determined whether measurements from a single
site located within the area of interest will represent the entire area of
interest. The following general procedure (and illustration shown in Figure
4-8) is recommended for making this determination.
* Consult diffusion meteorologist to estimate conditions.
40
-------
ISOPLETHS
S02 CONCENTRATION PATTERN (Winter)
FIGURE 4-8.
Schematic illlustrating
typical concentration pat-
tern over delineated popu-
lation or growth areas of
interest; (a) area in flat
part of gradient, one sta-
tion probably adequate to
represent concentrations
over area; (b) probably two
sites required; and (c) pos-
sibly three sites required:
urban setting, representing
neighborhood or middle spa-
tial scales.
If the concentration gradient over the area of interest is no
more than about 0.5 vg/m km, or if the distance between the
center of the area and the nearest sources are at least equal
to those shown in Table 4-2 (assume monitoring site is near
the center of the area), then measurements from a single sta-
tion located near the center of the area of interest will prob-
ably represent the entire area of interest (see Figure 4-8a).
Use regional-scale station procedures to determine site char-
acteristics and inlet placement (Section 4.2*1).
If the extreme concentrations over the area are not within
about 25 percent of the mean concentration, then a single site
may not be representative of the entire area and more than one
station will be necessary to represent the range of concentra-
tions over the area of interest. Divide the area into sub-
areas (preferably along an isopleth) until the extreme concen-
trations over each sub-area are within 25 percent of the mean
value (see Figure 4-8b,c). Siting areas should be located
near the center of each sub-area. If the sizes of the sub-
areas are in the middle scale range (<0.5 km), then middle-
scale procedures should be followed (see Section 4.4.3).
The tentative siting area within each sub-area (neighborhood) should be
in the vicinity of the mean concentration point of the neighborhood, which is
near the center of the neighborhood. Identify the siting area by drawing a
circle (0.5 to 1.0 km diameter) near the center of the neighborhood as shown
in Figure 4-9. Locate all prospective sites within the siting area and well
41
-------
inside of any built-up area. Eliminate from consideration all buildings with
SO2 source points (stacks). At this point, an elimination process begins, the
result of which is the selection of the final site location; Figures 4-9 through
4-11 illustrate the process. The first step is to establish two "upwind" di-
rections (Figure 4-9) . One direction is toward the prevailing winter wind di-
rection* and the other is the direction toward the center of the maximum emis-
sion zone of the nearest urban area from the tentative siting area. The latter
direction represents the most probable direction associated with the maximum
short-term concentrations (high background from urban center plus undue local
influences). Construct "sector boundaries" upwind of the siting area in both
directions as shown. These boundaries enclose the areas containing the most
important potential "interfering" SO2 sources. An arc is then drawn at a dis-
tance from each prospective site equal to the "point source interference dis-
tance" (PSID).** These distances, for 3 degrees of land use intensity, are
shown in Table 4-4 (see Section 5.1 for discussion). Then, from emissions
inventory data, Sanborn maps and/or photographs, identify and plot the loca-
tions of all S02 point sourcest within the siting area itself and the area
enclosed by the sector boundaries up to the PSID as chosen in Figure 4-10.
Wind direction
associated with
maximum frequency
of impacts from
nearby sources--
most likely the
winter prevailing
direction.
Neighborhood Scale
Area of Interest
FIGURE 4-9.
Schematic illustration of intermediate step by which
neighborhood-scale stations are located; identification
of siting area, and establishment of sector boundaries
within which sources are of concern.
* This wind direction should be associated with the maximum frequency of
occurrence of impacts from nearby sources within the siting area. Direc-
tions other than the winter prevailing may be chosen; appraisal by agency
may be necessary.
** The point source interference distance (PSID) is the distance beyond which
point source impacts are no longer significant at the monitoring site.
t Point sources include all sources identified as such in the emissions in-
ventory .
42
-------
Wind direction <-
associated with \
maximum frequency
of Inpacts from
nearby sources
most likely the
winter prevailing _
direction. ©
0
KEY
© Point sources.
Prospective sites unduly
influenced by point
sources.
O Candidate sites, to be
subjected to minor
source influence
analysis.
PSID = 1000 m.*
* Urban area; see Table 4-4 for
PSIDs associated with other
development intensities.
, Siting Area
t fl(iii^
FIGURE 4-10.
Plan view blowup of siting area of Fig. 4-9 illustrating the tech-
niques by which final candidate sites are selected.
KEY
Minor sources (area
source elements).
O Final monitoring site.
Candidate sites elim-
inated due to minor
source influence.
MSID ='200 m.
FIGURE 4-11.
Blowup of Fig. 4-10 illustrating the technique by which the final
site is selected.
43
-------
TABLE 4--4
Interference Distances for Three Development Intensities*
Interference Distances
Minor Sources (MSID)
Point Sources (PSID)
Urban
Suburban
Rural
200 m
100 m
60 m
1,000 m
2,200 m
3,200 m
The final candidate sites are determined from the following analysis:
Construct 10° plume sectors downwind of each point source
within the PSID of each candidate site.
Use 20° for the nearest sources (within a city block or two).
Eliminate prospective sites that fall within any 10° and 20°
sector (eliminates undue influences from nearby sources).
The remaining sites represent the set from which the final site will be selec-
ted. In a similar manner, identify and plot the locations of all area source
elements** within the area enclosed by the sector boundaries up to the minor
source interference distance (MSID)t as shown in Figure 4-11. Construct 10°
and 20° (nearby sources only) plume sectors downwind of each source and elim-
inate affected sites. From the remaining sites, select a single site that will
satisfy both winter prevailing and short-term peak concentration directions,
if possible. Also, the site should be selected such that effects from sources
in the other directions from the site are minimal, especially if the wind di-
rection frequency distribution is bi-modal (i.e., high frequencies from two
directions, one being the prevailing direction). The procedures used for the
prevailing winter direction analysis may be used for this analysis. Figure
4-12 shows the siting area and site locations in better spatial perspective.
If the environment of the siting area is rural in character, the desir-
able site characteristics and inlet placement are identical to those for re-
gional-scale stations. The desirable site characteristics and inlet placement
criteria for sites in suburban and urban environments are shown in Table 4-5.
* For discussion, see Section 5.1.
** Area source elements are the individual components of an area source
such as an individual home or small office building. They can be
identified on Sanborn maps or photographs.
t The minor source interference distance (MSID) is the analog of the
"PSID" but applicable to the minor sources or individual area source
elements.
44
-------
0 Final Monitoring Site
x Candidate Sites Eliminated
Prevailing winter
wind direction
Direction from
urban center
SITING AREA
FIGURE 4-12.
Oblique view of siting area of Fig. 4-11 showing site locations
and urban structure.
TABLE 4-5
Site Characteristics and Inlet Placement Criteria for Neighborhood Stations
Station
Location
Rooftop
Intermediate
Height on
Building
Trailer
Inlet Placement Criteria
Height
Above
Ground
A little
less than
the mean
height of
buildings
in neigh-
borhood
or lower
(<0.8fl*)t
Same as
Rooftop
3 to 5
meters
Height
Above
Roof
Between
1 - 2 m
and away
from
dirty/
dusty
areas
N.A.
1 to 2
meters
Horizontal
Clearance
Beyond
Structure
Not
Applicable
(N.A.)
1 to 2
meters
Not
Critical
Horizontal of
Inlet Placement
Locate inlet on windward
side of building relative
to the prevailing winter
wind direction, particu-
larly if bluff side of
building is toward pre-
vailing direction.
Same as Rooftop
Not critical
Remarks
No SOj source on roof of building.
Mo 502 source on roof of building.
If possible, avoid parking lots and
lots around which are buildings with
stacks, particularly if nearest
building upwind has a Urge stack.
If located in park. -avoid sites
under thick forest canopy.
* H = mean height of buildings in neighborhood (or in middle-scale area of
interest for middle-scale stations or in zone of maximum emission
densities for emergency episode stations).
t In suburban areas choose a building of low heightpreferably one-story.
45
-------
4.4 GENERAL-LEVEL, MIDDLE-SCALE STATIONS
Middle spatial scales are the smallest practical scales of measurement in
routine SC>2 monitoring. Indeed, in an area characterized by a steep SO2 concen-
tration gradient (in a general-level sense, not an individual plume) measurements
made at any one site within the area may represent concentrations on a scale no
larger than the middle (see Section 4.3.2).
The major siting objective associated with such scales, in a general-level
sense, is to determine peak levels in urban areas. Other siting objectives are
the population exposure and projected growth objectives discussed in Section
4.3.2 (normally associated with neighborhood spatial scales) for such areas lo-
cated in regions of steep concentration gradients.
Figure 4-13 is a schematic illustration of an annual S02 concentration pro-
file and associated ground-level pattern that may be observed over an ideally
configured city and shows the steep S02 gradient that is typically observed.
It is within the area of steep gradients that single sites may be located to
measure concentrations representing middle-spatial scales. Also shown in Figure
4-13 are example relative locations of stations sited for the above objectives.
Actually, most cities are irregularly configured or have industrial complexes
and power plants off to one side resulting in irregular SO2 concentration pat-
terns; nevertheless, the above representation is still relevant.
(a)
(b)
Peak Concentration Station
Population Exposure/Projected
Growth Stations
FIGURE 4-13. Schematic illustration of (a) idealized S02 concentration profile;
and (b) its associated ground-level pattern and example site locations.
46
-------
Peak concentration stations are "pattern oriented"; i.e., the location of
the peak concentration point of the concentration field determines where the
site is established. Diffusion modeling plays the primary role in making this
determination. The population and growth sites, on the other hand, are asso-
ciated with fixed geographical areas and are located within these fixed areas
regardless of the features or characteristics of the S02 pattern over the areas.
The procedures for siting these two kinds of monitoring stations are discussed
separately below.
Figure 4-14 is a flow chart showing the recommended procedure for select-
ing general-level middle-scale stations. If the objective of the monitoring
is to assess population exposures to SC>2/ or areas of projected growth (neigh-
borhood-scale procedure having been deemed inappropriate, Section 4.3.2) the
final steps for determining these sites are discussed later in this section.
Otherwise, the first step is to assemble the required background material re-
lated to the selection of sites to establish peak concentration stations:
Emissions inventory of point and area sources.
Meteorological data (see Appendix A)
USGS map of area.
Sanborn maps.
AQDM or equivalent model.
4.4.1 Peak Concentration Stations
Perform an S02 simulation analysis of the urban area using the AQDM (or
equivalent model) with meteorological data and point and area source emission
rates reflecting winter conditions (Dec, Jan, Feb).* For large urban areas
(population Xl0°) , use an SC>2 half-life of one hour. For other urban areas,
use a three-hour half life. Generate the winter mean, 24-hour "worst case"
and three-hour "worst case" SC>2 concentration patterns over the area. An ap-
proach for generating such short-term worst case patterns is suggested in
Appendix B. In any case, a diffusion meteorologist should be consulted.
4.4.1.1 Winter or Annual Peak Concentration Station
Identify the location of the maximum concentration point on a USGS or de-
tailed city map. (This can be accomplished quite easily if one-kilometer model
output grid spacing and isopleth analysis is used.) Then, draw a 500-meter
If the location of the annual concentration peak would be better estimated
by averaging four seasonal simulations rather than the winter pattern alone,
then this should be done; for example, in cities where there are power
plants, many of which emit peak or near-peak SC>2 rates in summer. For
cities having less than about 1500-2500 heating degree days per year, a
single annual simulation should suffice, unless industrial sources and/or
power plants exhibit seasonal emission patterns.
47
-------
What is the monitoring objective?
Population Exposure/Projected Growth Sites:
(from Section 4.3.2)
Assume that division of area of
Interest into middle-*cale parcels
has taken place.
Peak concentration monitoring.
Assemble background material (see text).
Determine the prevailing winter
wind direction and the direction
toward the maximum emission zone
of the city.
Peak Concentration Stations:
Using AQDM and appropriate meteorological data,
generate winter (or annual) 24-hour and 3-hour
worst-case S0£ patterns over the area of inter-
est (see Appendix B). In large urban regions,
use half-life of 1 hour, otherwise 3 hours.
From AQDM run obtain
the worst-case wind
directions associated
with the worst-case
concentrations. De-
'termine the prevailing
winter direction.
Select siting areas within 250 meters of maximum
concentration points (winter or annual, 24-hour
and 3-hour maximums). Emergency episode slle,
'if established, may serve as heat island peak
site to supplement or replace 3-hour peak site
above (see text).
From emissions inventory data, Sanborn maps or survey identify all $03 source points
in the general upwind directions from each prospective monitoring site up to 200 m
(SID) out from the site (see Sections 4.4.1.1 and 4.4.1.2).
Construct 10° plume sectors from each source point in downwind direction for
all source points identified previously.
Eliminate specific sites located within 10° plume sectors and buildings with stacks from
consideration. Choose sites such that impacts from S02 sources in other directions are
minimized (see text).
Final Site - General Characteristics:
Rooftop. Ht <. 0.8 of mean building height and from 1 - 2m above roof. Inlet on wind-
ward side of building.
Intermediate Height. Similar to rooftop. Locate inlet no closer than 1 to 2 meters
from building.
Trailer. Locate in non-parking lot. If possible, avoid lot around which are buildings
with stacks. Inlet height, 3 to 5 meters above ground (see Table 4-5).
FIGURE 4-14.
Flow chart showing procedures for locating general-
level middle-scale stations.
48
-------
diameter circle centered on that point. This circular area represents the
upper limit of the middle spatial scale and defines the most probable area
within which the maximum winter (and annual) peak concentrations occur.
Locate all prospective monitoring sites within the circular area. Elim-
inate from consideration all buildings with SO2 source points (stacks). Next,
using a procedure similar to that described for neighborhood stations, estab-
lish the prevailing winter (or annual, whichever applies) upwind direction
and draw arcs 200 meters from each prospective monitoring site in the upwind
direction. This 200-meter distance is the "source interference distance"
(SID).* See Figure 4-15 for an illustration of the procedure.
Wind direction
associated with
max frequency of
impacts from nearby
sources probably
the winter prevail-
ing directio'n.
Source points.
o Prospective monitoring
sites unduly influenced
by nearby sources.
0 Candidate sites from
which final selection
will be made.
SID = 200 meters.
Siting Area
Peak Concentration Point
City
Boundary
Winter or Annual SO^ Pattern in Urban Area
FIGURE 4-15. Schematic illustration of middle-scale siting procedure
for peak concentration stations.
Since we want to measure the maximum collective annual impacts from all
point sources, the SID applies only to area source elements. In that
sense, the SID is analogous to the PSID.
49
-------
From the emissions inventory, Sanborn maps and/or survey results identify
and plot all S02 source points (stacks) within the siting area and up to the
SID in the upwind direction. Draw 10° plumes downwind from all source points
located within the SID of each prospective monitoring site as shown. The
final site is selected from among those sites not intersected by a 10° plume.
Since we are now dealing with spatial scales in which we are becoming inter-
ested in impacts from local sources, the use of 20° sectors is not so impor-
tant and their use is optional.
The above analysis should be extended to other wind directions proceed-
ing from the next most frequent until only one prospective site location re-
mains . This will be the final site location and could be considered perman-
ent. The physical characteristics of the site and inlet exposure are the
same as those shown in Table 4-5 (see Page 45).
4.4.1.2 24-Hour and 3-Hour Maximum Concentration Stations
The procedure for locating these stations are the same as for the annual
peak station except for the following points.
1) Assume that the short-term peaks occur in winter. Their loca-
tions will be determined on the basis of winter simulation
analyses.*
2) The wind directions used will be those of the worst case mete-
orologyi.e., those which are associated with the 24-hour and
3-hour concentration peaks.
3) Regarding the peak 3-hour station, prospective stations should
be considered temporary with the final site location refined on
the basis of mobile sampling. Such sampling should be done when
the 3-hour worst case meteorological conditions are forecast.
4) There is an alternative to the 3-hour site location as deter-
mined from the above analysis. It is possible that the 3-hour
peak concentration occurs under an inversion situation with a
general inflow of air toward the urban center ("heat island"
effect). The urban center here can be considered the area of
the maximum SC>2 emission density due to area and point sources,
analogous to that associated with the emergency episode sta-
tions. It is not unlikely that the maximum temperature excess
point, air inflow convergence point, and peak concentration
point will be found near the center of this area. Thus, un-
less the addition of point sources to maximum emission zone
calculations significantly changes the location of the zone
based on area sources alone, or if the city is geographically
complex, the emergency episode station(s) can be considered a
heat-island related 3-hour peak station as well.
* If only major point sources contribute significantly to urban SO2/ see
Section 4.5.
50
-------
It is possible that some or all of the peak'concentrations (3-hour, 24-
hour, and annual) occur near the same location. In these cases, only one site
will be required and located/verified by using the procedures applicable to
each averaging time involved.
If there is a choice between the 3-hour peak site based on diffusion mo-
del results versus the emergency episode station, consider both but make the
final decision on the basis of mobile sampling (as addressed in Item 3 above)
results. See Table 4-5 (Page 45) for site characteristics and inlet exposure
criteria.
4.4.2 Population Exposure and Projected Growth Stations
The siting procedures for these stations are a continuation of the pro-
cedure described in Section 4.3.2 (second item of decision process for deter-
mining number of sites required for characterizing area of interest) and by
Figure 4-8c (see Page 41). Therefore, at this point it can be assumed that
the area of interest has already been divided into an appropriate number of
middle scale parcels. The siting procedure continues by first establishing a
siting area in each parcel. Try to limit the siting area to the central strip
of the parcel as shown in Figure 4-16. Then, use the annual peak station sit-
ing procedure discussed in Section 4.4.1.1, but with one additional wind di-
rectionthat defined by the direction of the siting area and the center of
the maximum emission zone of the city (for example, see Figure 4-9, Page 42).
This wind direction defines the most probable direction associated with the
shorter term peak concentrations resulting from center city sources, which
almost certainly impact essentially uniformly over the entire parcel.
Idealized ground
urban S02 pattern
Siting Area for Parcel 1
Siting Area for Parcel 2
Siting Area for Parcel 3
level
Population
exposure/projected
growth area
of interest
FIGURE 4-16.
Schematic showing population exposure or projected growth area
divided into middle-scale parcels and recommended siting areas.
51
-------
The final site should be chosen such that nearby local sources in this direc-
tion also will not unduly impact at the monitoring site.
4.5 PROXIMATE, MIDDLE-SCALE STATIONS - Urban Sources
There are two siting objectives associated with proximate, middle-scale
stationsassessing the impact of a major point source in a multi-source urban
setting, and assessing the impact of an isolated point source. The procedures
for siting monitors to satisfy the first objective are heavily dependent on
the results of multi-source diffusion model simulations, point source diffu-
sion calculations and "X/Q" type analyses. For the second objective,
knowledge of plume behavior in various terrain environments, special surveys,
* and mobile sampling results may also be required. Although middle-scale
measurements are associated with both objectives, the selection procedures
for siting monitors to achieve the two objectives are totally different. In
this section, only the first objective is addressed. Isolated point source
monitoring is discussed in Section 4.6.
Figure 4-17 is a schematic illustrating the concept of the impact of a
major point source in an urban setting. In this situation, the specific sit-
ing objectives are to:
measure the impact of the point source at the urban peak
concentration point (Figure 4-17a, point X), and
measure the maximum impact of the point source itself (at
point P of Figure 4-17a,b).
Averaging times of 3 hours, 24 hours, and" one year should be considered,
particularly the shorter averaging times.
Figure 4-18 is a flow chart showing the procedure for locating middle-
scale stations for assessing the impact of individual urban point sources.
The first step is to assemble all background information. This will include:
Physical data from point source
- peak and daily mean production rate of SC>2
stack parameters
- exact plant location.
Emission inventory of point and area sources.
Meteorological data
stability wind roses (see Appendix A)
wind persistance tables (see Appendix B, Part I).
USGS/Sanborn maps of urban area.
Frequency statistics of hourly wind speed and direction
(annual data).
52
-------
Prevailing Wind
Direction
Annual or Winter
Wind Rose
Urban Peak
Concentration
Point (x)
Subject Major
Point Source
Location of maximum annual impact
of subject point source alone (P)
Maximum concentration point from
subject point source (P)
Subject Major
Point Source
.' Short-term
worst-case
wind direction
FIGURE 4-17.
Schematic illustration of impact of point source in an urban
setting for two averaging times: (a) annual pattern, and
(b) short-term pattern.
53
-------
the
Assemble Background Information (see text) 1
i
Establish a General-Level Site for assessing
annual urban peak concentration (see Section 4.4.
D
Annual impact point
Using point source data,
annual meteorology, simulate
an annual $03 pattern around
the source
Short-term Impact points
Using procedures for isolated point source
monitoring (see Appendix B, Sec. II)
appropriate~meteorological data and
emission rates, determine the locations
of peak 3- and 24-hour impact points
(see Section 4.5.2).
Select siting areas as close to
peak concentration points as possible
From inventory, Sanborn maps, or survey, identify all source points
1n the upwind directions from each prospective monitoring site up
to 200 meters (SID) out from the site. The upwind
directions are toward the subject point source location from each
monitoring site, plus other directions as discussed
In Section 4.5.1 for the annual impact point station.
Construct 10° plumes from each source in the downwind
direction for all source points previously identified.
Eliminate specific sites located within 10° plume sectors and buildings
with stacks from consideration. Choose sites such that impacts
from S02 sources in other directions are minimized.
NO
T
Annual impact point site?
From wind statistics, determine
frequency of downwash
conditions. Do mobile
sampling either as routine or
to adjust permanent
location.
Final Site - General Characteristics
Rooftop. Ht <_ 0.8 of mean building Ht and l-2m
above roof. Inlet on windward side of building.
Intermediate Ht. Similar to rooftop. Locate
inlet no closer than l-2m from building.
Trailer. Locate in non-parking lot. If possible,
avoid lot around which are buildings with
stacks. Inlet ht, 3-5m above ground
(see Table 4-5)
FIGURE 4-18.
Flow chart showing procedures for locating
proximate middle-scale stations.
54
-------
4.5.1 Annual Peak Concentration Stations
4.5.1.1 General-Level Urban Peak Station
If a monitoring site has not already been established to monitor the
general-level (urban) peak annual concentration, establish such a site (point
X in Fig. 4-17a) by using the procedures presented in Section 4.4.1.1 for the
annual peak stations. However, in the procedure for dealing with undue local
impacts (illustrated in Fig. 4-15) consider the direction toward the subject
point source,f^om the siting area (direction "1", Fig. 4-17a) as well as the
winter (or annual) prevailing direction.
4.5.1.2 Proximate. Station
The remaining steps of this procedure pertain to the siting of an addi-
tional monitor to assess the maximum annual impact from the point source it-
self. Using the AQDM with annual* meteorological data, appropriate half-life
value, annual* average emission rate and stack characteristics of the point
source only, simulate the annual* SO- pattern around the source and establish
a siting area centered on the maximum concentration point, Point "P" in Figure
4-19 (which is point "P" in Fig. 4-17a) . Figure 4-19 shows the annual SC>2
pattern due to the point source only. The next few steps are the same as
those discussed in Section 4.4.1 (illustrated in Fig. 4-15 in that section),
except that an additional direction must be considered for identifying another
critical sector that contains sources which may produce undue influences
the upwind direction between the siting area and the subject point source
(direction "2", Fig. 4-17a); however, this direction is probably the same as
the prevailing wind direction for the time period simulated. Use Table 4-5
for final site characteristics and inlet exposure.
4.5.2 24-Hour and 3-Hour Maximum Concentration Stations
These stations are analogous to the 24-hour and 3-hour maximum concentra-
tion stations associated with the urban peaks except that these assess the
peaks due to the subject point source alone (Point "P" in Fig. 4-17b). In
this case, we can pretend that the point source is located in an isolated
area of rough topography. The location of the peak 24-hour and 3-hour aver-
age concentration points due to the source, and associated "worst case" mete-
orology can be determined by an approach suggested in Appendix B, Section II.
After determining these locations, use the procedure found in Section 4.4.1,1
(regarding annual peak stations) for selecting the siting area and specific
station locations. However, in this situation, only one wind direction is
used for determining undue influence of nearby sourcesthe direction from
the siting area toward the subject point source (direction "3", Fig. 4-17b).
If the emission rates and meteorology vary significantly over the year, it
may be better to use the average of four seasonal simulations; if subject
point source emissions are degree-day dependent, use winter meteorology.
55
-------
Annual average ground level S02 pattern
due to subject urban point source only.
Prevailing Wind
Direction
Subject Urban Point
Source Location
Annual
Wind
Rose
Siting '
Area '
Kilometers
= 200 meters
= Maximum impact point of point source alone.
= Minimum Concentration
= Maximum Concentration (secondary)
Source points.
O Prospective monitoring sites unduly influenced
by nearby sources.
© Candidate sites from which final selection
will be made.
FIGURE 4-19.
Schematic illustration showing annual impact pattern due to
urban point source alone, the siting area, and final candidate
sites for assessing the annual impact from the point source.
Before deciding on a final site location, it might be a good idea to de-
termine the expected frequency of downwash conditions. (It is recognized that
over large urban areas in the daytime, downwash conditions are the rule rather
than the exception, especially for the lower level sources. However, for the
larger (and more elevated) sources such may not be the case; but, because the
sources are large, if and when downwash occurs the ground-level impact of such
occurrences could be substantial.) If the wind statistics (e.g., see Table
4-6) for the area show that the stack exit velocities are less than 1.5 times
the wind speed (Sherlock and Stalker, 1941), or if the height of the stack is
not at least 2.5 times the height of the highest surrounding buildings (Hawkins
56
-------
and Nonhebel, 1955), downwash conditions are likely. Sanborn maps or surveys
will resolve the lattersituation. To resolve the former, the following pro-
cedure is recommended:
From load curve and/or hourly fuel consump-
tion data obtained from the source, determine
the typical hourly exhaust gas flow and velo-
city rates. Then tabulate these velocities
by the hour and compare them to the hourly
wind speed frequencies of Table 4-6 for the
same hour. If downwash conditions will be
frequent, mobile sampling may be necessary,
either as a routine operation to monitor the
3-hour peak or to determine the location of
a final permanent site, or, perhaps, even to
delineate a new siting area within which a
permanent site will be selected. Use Table
4-5 (see Page 45) for final site character-
istics and inlet exposure.
4.5.3 Data Interpretation
TABLE 4-6
Percentage Frequencies of
Sky Cover, Wind, and
Relative Humidity
(Taken from NCC)
HOU*
Of
01
02
03
04
05
06
07
08
09
10
11
12
13
It
It
It
17
18
19
20
21
22
23
AVG
OOUD* ' <
S£AU 0-10
i
': ' i
1 , !
17 7 75 4
17 5 78 51
15 7 77; »
15 9 76 4
!' 7 79 5
10 11 78 51
10, 8. 82, 4i
12 10 78 51
12 12 76 4
11, !» 75 2!
10 1*. 77| 21
10 13 77' 5'
9 15 76| 41
10 14 76, 3
11 15 74 2
14 15 71 3
15 13 72 5,
18 9 73- 5
16 11 T3 4,
18 11 71 4,
18 9 73 4
JO 8 72' 51
14 10 76| 4
MO 1FW
ut ' ni
ovit
44: 44 4
'5, 4 5
48, J 5
47 Z 7
48 1 6
47 2 6
48i > 5
44 * 7
38 52 6
40 52 6
35 5» 9
31 54 11
35 55 7
35 55 3
41 52 5
47 4 i 5
45 4& 4
43 48 5
42 50 5
43 50 4
44 47 5
44 45 6
43 »8 6
Huron HUMB
| i
1 ' 1
O. »- JO- 7O-
W 4f , ** 7f
I 1
i
1! 18 27
ll 16 79
1 16 30
1 15 28
1, 16 28
1! 15 32
15 34
I1 14' 33
1[ 20 34
21 29 28
31 35 25
1 3 36 2«
31 40 25
5 42 22
5 40 22
4 37 25
2' 31 30
2 27 29
I1 24 29
2 21 31
l; 21 31
+i 18 35
l! 16 31
2 24 29
rr r*i
so- vo-
lt 100
35 19
34 20
33 20
3* 22
J4 21
32 20
»2 :o
34 18
28 16
27 14
22 15
18 15
16 15
16, 19
18 13
18 16
21 16
27 16
29 17
29 19
30 18
29 18
32 19
27, 18
The total concentration of S0_ in the sam-
ples taken at the above sites consists of con-
tributions from most of the sources in the urban
area as well as that from the subject point source. The former may be consid-
ered as background noise that cannot be separated from the total. Therefore,
to estimate the percentage of the sampled concentration due to the subject
point source alone, diffusion modeling must be utilized. Figure 4-20 illus-
trates the recommended approach very well for the annual situations. The two
vertical profiles on the right side of the figure are source contribution re-
sults computed by the AQDM model at the indicated points of interest. Dis-
played as seen in the figure, they may be interpreted as "source contribution
profiles" and represent the best estimate of what each source contributes to
the total concentration at those points. In this case, there are two points
of interestthe urban annual peak concentration point, and the corresponding
impact point "P" (pattern, lower left corner), where the maximum annual im-
pact from the point source of interest ("10") occurs. The upper profile re-
presents the highest total concentration in the entire urban area, and in this
particular case, source 10 contributes the least to that concentration. (How-
ever, it is more than likely that in the real world any point source worthy of
individual attention will almost certainly contribute substantially to the
highest urban peaks.) The lower profile represents the total concentration at
source 10"s maximum impact point, with source 10 being the largest contributor
to that concentration.
In a similar approach, for the short-term impacts, the 3-hour and 24-hour
worst case meteorology, and emission data (appropriate for season in which the
worst cases occur) for the entire urban area can be input to the AQDM to de-
termine the short-term background concentrations at the maximum impact points
of the subject point source. Source contribution profiles may also be gener-
57
-------
ated by the model and analyzed in a manner similar to that discussed above.
/Point
I Figure
S02 concentration field over urban
area due to all sources.
Total concentration of urban
concentration peak
Point source
contributions
to the maximum
concentrations.
Contribution from
point source 10
to maximum urban
concentration.
Total concentration
Maximum urban
concentration
point
Point source
contributions
to the maximum
impact point
of source 10.
Contribution of
source 10 to total
concentration at
Point "P"
Maximum impact point
of point source 10 above
Annual pattern due to
source 10 alone
FIGURE 4-20.
Schematic illustration of the concept of the source contribu-
tion profile in an urban area for the annual pattern
(a synthesis of Figures 4-17a and 4-19).
58
-------
These kinds of analyses should be performed at each monitoring site to
estimate the percentage of the measured concentration due to the point source
of interest. The model used can be run either calibrated or uncalibrated,
but appropriate half-life decay factors should be used. A yearly analysis
with updated meteorological and emissions data is recommended in order to ex-
plain trends and to determine relative effects of control strategies.
4.6 PROXIMATE, MIDDLE-SCALE STATIONS - Isolated Sources
Because of the great variety of physical environments in which isolated
point sources* are found, it was not possible to develop a single set of pro-
cedures for selecting sites for monitoring the impacts of such sources applic-
able uniformly, in all environments. Therefore, where possible, it was de-
cided to present examples of monitoring site configurations, each reflecting
an approach to the site selection problem in a given characteristic physical
environment. In other situations where typical settings can be extremely
varied and complex, only a general description of the kinds of siting prob-
lems expected to be encountered in such settings is discussed, mainly in
terms of "points to consider". It is hoped that these examples will serve as
guides for the site selector in developing the steps necessary for the proper
siting of monitors in specific situations. In presenting the examples, spe-
cific points will be addressed wherever possible to help the site selector
in developing and executing the steps.
The material presented is essentially an expansion of existing guidelines
(EPA,1974b) but with more emphasis placed on the use of the diffusion equation
and graphical aides in selecting monitoring sites. Additional points addressed
include problems of plume behavior in various terrain environments and the
role of mobile sampling in the site selection process. In this regard, in
some situations it may be necessary to determine the distribution of the
plume material in order to ascertain the plume's statistical characteristics;
this will require microscale measurements, most easily accomplished via mo-
bile sampling. For a rather comprehensive overview of isolated point source
monitoring, the reader is referred to a paper by Paulus and Rossano (1973).
The situations described in this section will also be applicable to mon-
itoring networks established to satisfy supplementary control system (SCS)
requirements. Since these systems are rather complex and comprehensive
(i.e., they integrate ambient and in-stack monitoring, diffusion modeling,
emission controls, etc. in a predictive scheme) detailed treatment of the
subject was considered beyond the scope of this report. However, a descrip-
tion of the components of a typical supplementary control system can be found
in the Federal Register, Vol. 38, No. 178, Friday, September 14, 1973. See
also Montgomery, et al. (1975) for a description of TVA's SCS system.
Point sources in this context consist of power plants, sulfide smelters,
sulfuric acid plants, coal conversion plants, or refineries located away
from populated or developed areas.
59
-------
The material presented below includes typical examples of isolated point
source monitoring problems expected to be encountered in a variety of physi-
cal terrain settings. The major role expected to be played by mobile samp-
ling would be either in routine monitoring or in the refining of preliminary
site locations for permanent monitoring stations. (To maintain the continuity
of the section, the concept of mobile sampling, per se, is discussed briefly
in Appendix C.)
The specific objectives in monitoring the S02 impact of isolated point
sources, regardless of terrain setting, are to:
determine the short-term maximum concentrations downwind of
the source and where they occur. (Samplers may be placed
where the highest peak is likely to occur and where rela-
tively high peaks are likely to occur very often. It can
be assumed that the annual standard will not be threatened
by emissions from an individual isolated source.)
determine the background concentrations by establishing a
monitoring site in the direction from the source opposite
to those above.
The major problem is to account for the effects imposed by the various terrain
settings in determining where to place the monitors to measure the peak values.
The terrain settings addressed below are: flat, near coastline, ridge-valley,
and irregular-rugged.
4.6.1 Monitoring in Flat Terrain Settings
The recommended procedures in this situation are very similar to those
given in EPA (19743). Plumes behave rather well in this kind of setting and
are amenable to treatment with standard diffusion equations.
The first step is to assemble the background information. It should in-
clude:
USGS maps of the area.
Physical data from point source:
- peak daily mean production rates of SC>2,
stack parameters,
- exact plant location.
Stability wind roses (climatological). \
y See Appendices A and B.
Wind persistence tables. )
The next step is to confirm that the terrain is flat so that the recom-
mended siting techniques are applicable. The terrain is deemed to be flat if:
Terrain elevations more than 2/5 the height of the stack do
not exist within 10 km of the source (EPA, 197-fc).
60
-------
4.6.1.1 Peak Concentration Stations
After determining that the terrain is flat, a determination must then be
made whether the source should be monitored. A screening technique suggested
by the EPA (EPA, 1974b, Appendix C) is also suggested here for this purpose.
If the technique indicates that monitoring need not be undertaken, check for
the possibility of downwash situations occurring. Downwash is likely to occur
if:
the heights of any buildings and other obstructions that
exist within a distance of 10 stack heights of the source
exceed 2/5 of the height of the stack.
Downwash conditions may also result if the ratio between the stack gas velo-
city (75) and the wind velocity (V) is less than about 1.5. In this case, the
effective stack height would be no more (and probably less) than the physical
stack height and ground-level concentrations would increase. To assess this,
one can do an analysis similar to that described in Section 4.5.2 to estimate
the frequency of downwash conditions. Vary the production rates (and Vs )
and then compare V$ to the wind speed frequency (see Table 4-6, Page 57) ex-
pected over the area to estimate an expected frequency of downwash conditions.
Then determine the expected ground-level concentrations by assuming that the
effective stack height equals 1/2 of the physical stack height. If the re-
sulting concentrations exceed the threshold concentration prescribed in the
screening technique, then use mobile sampling when downwash conditions are pre-
dicted. Permanent monitoring sites may be located in "favored" areas if the
mobile sampling results show ground-level peaks consistently occurring in about
the same place. Site characteristics, etc. are discussed in Section 4.6.1.5.
If a need for monitoring has been determined by the screening technique,
but not due to downwash conditions, the estimated locations of the peak 24-hour
and 3-hour maximum concentration points can be determined by a technique sug-
gested in Appendix B, Part II for isolated point sources. From Appendix B,
Part II, and/or from downwash analyses, then, we have determined the approxi-
mate locations of:
the near-worst 24-hour average concentration,
the near-worst 3-hour average concentration,
where a very high concentration occurs very often.
4.6.1.2 Background Stations
The locations of the background stations should be selected to measure
the quality of the incoming air. The difference between this background con-
centration and the peak concentrations measured downwind of the source is
equal to the contribution due to the source alone. Background stations should
be located:
in the direction from the source opposite the peak
concentration stations,
61
-------
These sites should be located within a few kilometers of the source. ^ Figure
4-21 illustrates a possible monitoring site configuration around an isolated
point source in flat terrain.
b
Siting Area
Wind direction:
highest 24-hr peak
Wind direction:
frequent high
3-hr peak
Wind direction:
highest 3-hr peak
Station No. and Purpose
1 - Urban transport station.
2 - Background station.
3 - Highest 3-hr peak station.
4 - Frequent high 3-hr peak station
5 - Highest 24-hr peak station.
Isolated Point Source
FIGURE 4-21.
Illustration of possible monitoring site configuration
around an isolated point source in flat terrain;
(a) relationship to local geography, and (b) blowup
of siting area.
62
-------
4.6.1.3 Fumigation Effects
It is possible that the highest 3-hour concentrations result from "fumi-
gation". This phenomenon usually occurs as a result of inversion "breakup"
after sunrise. Because of the flat terrain and the characteristic light and
variable wind conditions associated with such a situation, it is unlikely that
a single stationary sampling site could be established for the sole purpose of
measuring such concentrations. The recommendation is to use mobile sampling
when fumigation is predicted to occur. It is also likely that at least one
of the other sampling stations will detect the phenomenon occasionally. Sev-
eral analysis techniques for estimating fumigation concentrations and where
they occur are availablee.g., see EPA (1974c), Turner (1974), and Slade
(1968) .
4.6.1.4 Role of Mobile Sampling and Final Site Selection
«
The above procedures can only approximate the location of the peak con-
centration points because:
The available meteorological data may not be exactly repre-
sentative of conditions in the vicinity of the source (main
reason for erecting meteorological towers).
For a given wind direction and stability class (from a sta-
bility wind rose) the frequency of wind speed events is re-
ported as occurring within a range of speeds. This wind
speed range would correspond to a distance range along the
azimuth.
Diffusion equations are only accurate to within a factor of
two or so.
Accordingly, terrain roughness and road accessibility permitting, mobile samp-
ling should be utilized to refine the site locations, particularly those for
the 3-hour peak stations. When the meteorological conditions that produce the
peak concentrations are predicted, the mobile unit can be dispatched. After a
number of occurrences, a plot of observed peak concentration points could prob-
ably be enclosed by a circle of middle-scale dimensions (up to 500 m in di-
ameter) ; the final site should be located near the center of the circle.
4.6.1.5 Site Characteristics and Inlet Placement
The site characteristics and inlet placement for these sites are similar
to those for regional monitoring stations. Since the topography is flat, low
lying areas should not be a problem; in any case, they should be avoided.
Open or sparsely forested areas are recommended with the instruments housed
in either a trailer or other stationary structure. Inlet height should be no
higher than about 3 to 5 meters. If any buildings in the vicinity are heated
by fossil fuels, be sure that they are not between the monitoring site and the
source. Otherwise, such buildings and clumps of trees create little cavity
wakes which tend to increase the effective sampling volume of the instrument.
63
-------
If locating a site in a densely forested area is unavoidable, the inlet tube
should be raised a few meters above the tops of the surrounding trees. Locate
on the lee side of clearings, if possible.
4.6.1.6 Instrument Type and Supplementary Instrumentation
Since we are concerned with the short-term peak concentrations, continu-
ous instrumentation will be required at all stations. Instrumented towers for
measuring pertinent meteorological variables, such as temperature lapse rates
and wind variation with height (also, air -quality as well) , are often con-
structed in the vicinity of a large source or source complex (Munn and Stewart,
1967). They are usually required in situations where the available meteoro-
logical data is not representative of the source area, as is often the case in
conjunction with the preparation of environmental impact statements (EIS) prior
to the construction of large sources. Gill, et al. (1967) and the AEC (1972)
describe optimal design configurations for towers and how resulting data should
be interpreted, respectively.
4.6.2 Monitoring in Near Coastline Settings
When a tall stack is located near a seacoast or other large body of water
such that sea or lake breezes may influence the SC>2 plume, a phenomenon known
as a "sea-breeze fumigation" may occur. It results when the stack plume, ini-
tially embedded in a stable, sea-breeze flow is convectively mixed down to the
ground downwind. The mixing is caused by a vertically growing mixing layer re-
sulting from the stable marine air being heated from below by the land surface
as it moves inland (van der Hoven, 1967).* An effort to model fumigation oc-
curring inland from a large lake has recently been reported by Peters (1975).
Figure 4-22 is a schematic illustrating the phenomenon. Unlike fumigation
resulting from a nocturnal radiation inversion, which might only last for
about a half-hour or so, sea-breeze fumigations may last for several hours
due to the constant replacement of stable air by the on-shore flow. These
short-term sea breeze fumigation concentrations may very well exceed those
observed over flat terrain away from marine influences and, therefore, should
be monitored, either with appropriately placed permanent stations, or via mo-
bile sampling.
£OOL MARINh AIR
AiFI FLOW
MODIFIED MARINE AIR
COLD WATER SURFACE
.On
FUMIGATION
FIGURE 4-22. Schematic illustration of a sea-breeze fumigation situation
(taken from Van der Hoven, 1967).
* In actuality, this phenomena may occur in any on-shore flow if the associ-
ated marine air becomes less stable as it moves inland over a heated ground
surface.
64
-------
A recommended procedure for such monitoring is presented below. It is
based on the results of a study by Collins (1971) in which the occurrence of
sea-breeze fumigations and where they occurred were accurately predicted. The
concept is essentially universally applicable except where the coastal topo-
graphy is extremely rough. The procedure also requires that mobile sampling
be utilized. As an illustration of the procedure, consider a plume embedded
in a layer of cool marine air at an elevation of 100 meters and movinq inJand.
Data from the 100-meter level(the same elevation as the plume) of an instru-
mented meteorological or TV tower (or equivalent) are assumed to be available
as well as sea-surface temperatures (either estimated or specially taken).
For the next series of steps, refer to Figure 4-23 for flat terrain or to
Figure 4-24 for rising terrain (height of terrain is subtracted from height
of plume above MSL). From the tower data and sea-surface temperatures, the
following values are computed:
A9 (potential temperature difference) - the 100-m
temperature (°C) + 0.91°C* minus the sea-surface
temperature (see Figure 4-25).
A9 indicates the stability of the layer between plume height
and the ground with higher values indicating the more stable
conditions. The more stable the atmosphere, the longer it
takes for the mixing layer to grow and farther inland the
plume will advance before intercepting the mixing layer.
U The mean wind speed within the layer of interest
(the wind speed observed at the 50-m level).
U indicates the mean wind speed at which the plume and the
stable air is transported inland. The slope of the upper
boundary of the mixing layer to the terrain is related to
this speed; i.e., at a very low wind speed the mixing layer
would intercept the plume relatively close to its source
(everything else being equal).
Then, from Figure 4-23 or 4-24, the distance to where the plume intercepts the
mixing layer and, therefore, to the initial point of plume touchdown (fumiga-
tion) is ascertained (typically, ^Q.2 to 2.0 km from the source). In a sea-
breeze situation, the direction toward which the plume blows is usually a com-
promise between the direction normal to the mean shoreline orientation and the
flow dictated by the large-scale pressure field. If the wind direction at the
50-meter level on the tower is available (after the sea-breeze has passed),
this could be used. The point defined by the distance from the source to where
the fumigation is predicted to take place and the 50-meter wind direction could
be considered the initial starting point of a search for the maximum fumigation
concentration via mobile sampling. In this instance, a vertical sensing capa-
bility would be quite helpful.
0.91°C is the adiabatic temperature change that a parcel of air will
undergo when brought down to the surface from a 300-ft (100-m) elevation.
65
-------
I
o
UJ
I
H = 88 Vu~A9
u = MEAN WIND
'A9 = POTENTIAL I CMPERATURC PROFILE
PLUME STRATIFIED AT 100 m, MSL, u A8 = 5
LFUMIGATION OCCURRED AT eoo m
X, DISTANCE FROM STACK (m)
tooo
500 -
200
§
100
-§ 50
2
O
10 -;
5
.EXAMPLE
- WIND FROM NORTH
PLUME STRATIFIED AT 100 m, MSL
u AS = 5
FUMIGATION OCCURRED AT 360 m
HEIGHT OF PLUME
ABOVE GROUND
NOTE
RISING TERRAIN CAUSES
INCREASED RATE OF GROWTH
OF MIXING LEVEL
FALLING OR LEVEL TERRAIN
RESULTS IN SAME RATE OF
GROWTH AS FIG 1
FIGURE 4-23.
Mixing depth as
function of sta-
bility, wind
speed, and in-
land travel dis-
tance (taken from
Collins, 1971).
FIGURE 4-24.
Vertical mixing
depth adjusted
for terrain
(taken from
Collins, 1971).
10 20
100 200 500 1000 2000
DISTANCE INLAND (m)
5000 10,000
66
-------
600
500
400
s
o
o
UJ
X
300
200
100
SURFACE
Observed
Temperature
Profile
PLUME HEIGHT
18 19
20 21 22
TEMPERATURE
23
(°C)
24 25
o
H
cn
A6
Regarding a permanent station, it would
be feasible to establish one only if the
plume tended to touch down within "favored"
areas.
Over flat terrain, the monitoring of the
sea-breeze fumigation phenomenon would be
in addition to the monitoring objectives
described in Section 4.6.1. In practical
applications of the procedure, there is a
problem of obtaining temperatures aloft at
plume height (effective height). In the
illustration (Figure 4-25), the temperature
sensor was exactly at plume heightan ideal
situation not likely to be encountered in
the field. Figure 4-23 is universally ap-
plicable in flat coastline topography if
temperatures are taken at effective plume
height. Figure 4-24 would have to be modi-
fied to reflect the slope of the ground in
the coastline area of concern. In all situ-
ations, the services of a diffusion meteor-
ologist is strongly recommended.
4.6.2.1 Site Characteristics and In-
strument Inlet Placement
In the event that a favored area does
existi.e., a small area over which fumi-
gation occursthe site characteristics and
inlet placement should be the same as those
for the flat terrain stations (see Section
4.6.1.5). In irregular, rough terrain
areas, choose well exposed locations.
4.6.3 Monitoring in Ridge/Valley Settings
FIGURE 4-25. Example of com-
putation of A6 (adapted
from Collins, 1971).
This kind of terrain is found mainly in
the Appalachian Mountain area and in parts
of the upland region of several of the west-
ern states. For purposes of this discus-
sion, characteristics of such areas are,
typically, a valley of arbitrary width with parallel walls or ridges and a
more or less definable "up-valley/down-valley" direction.
Because no two ridge/valley configurations are exactly alike, a detailed
treatment of the subject is difficult and the development of siting procedures
uniform for all possible scenarios is impractical. (For a detailed discussion
of the subject of plume behavior in valleys, the reader is referred to the
work of Hewson, et al., 1961; Smith, 1968; and Flemming, 1967.) However, from
the descriptions provided by these references, typical situations were derived
67
-------
from which a general procedural guideline for siting monitors was developed.
The situations include the likely kinds of impacts expected to result from a
large, elevated point SC>2 source located in a valley with steep walls and un-
der a variety of meteorological conditions. The kinds of SC>2 problems asso-
ciated with such a scenario will be briefly summarized below, and followed by
recommended siting procedures.
4.6.3.1 S(>2 Problems
"Fumigation" occurring shortly after sunrise (see Figure 4-26).
A plume may be embedded in a down-valley drainage flow then
brought down to ground level after sunrise via the fumigation
mechanism.
FIGURE 4-26. Inversion aloft-above stack
("fumigation"), (taken from ASME, 1968).
Near intersection of plume with valley wall with a cross-valley
wind flow under stable to unstable conditions (see Figure 4-27).
Distortion and downwash of the plume due to wake effects on lee
side of upwind wall (see Figure 4-28).
68
-------
Side View
End View
Top View
///
(a)
FIGURE 4-27. Plume behavior near a steep bluff when the air is unstable(above)
and when it is very stable(below). (a) Vertical temperature lapse rates in
relation to the critical lapse rate of 5°F/1000 ft, shown by broken line
sloping upward to left: full linelapse rate measured from grade at plant;
dotted linelapse rate measured from top of bluff; both show effect of
ground surface. Corresponding plume
features as observed (b) looking hor-
izontally parallel to steep bluff,
(c) looking horizontally toward steep
bluff, and (d) looking vertically
downward from above. Note that when
air is unstable, effluent moves up
and over the bluff, but when the air
is very stable, as with the inversion
as shown, the bluff acts as a barrier
-X X
VALLEY LOCATION ("WIND CROSS AXIS')
to deflect the plume.
Hewson, et al., 1961.)
(Taken from
Figure 4-28.
Plume dispersion in a deep valley.
With a wind from left to right, as in
Section (a), the plume may be brought
quickly to ground level by aerodynam-
ic eddies. Wind from the opposite
direction may create high concentra-
tions on the plateau. (Taken from
ASME, 1968.)
69
-------
Maximum impact points at ground level "in the valley when wind
direction is parallel to valley (see Figure 4-29).
(A)
VAILCV LOCATIOM CwlMO ALOM Ulil
(B)
PLUME DIMENSIONS
AT DISTANCES
INDICATED IN B- 16o
(k)
//
ZA_
(c't *0»«it« killcitf*
FIGURE 4-29. Plume dispersion in a deep valley. When the wind is parallel to
the valley, dispersion tends to occur fairly normally until confined by
the valley walls. Section (A) is a pictorial representation of the dis-
persion; Section (B) shows the associated concentration patterns. (Taken
from ASME, 1968.)
4.6.3.2 Siting Procedures.
The first step common to any monitor siting study in this kind of terrain
is to acquire supporting data, information, and equipment such as that listed
below:
USGS map of area.
Physical data from the SC>2 source
- peak and daily mean production rates,
- stack parameters,
exact plant location.
Stability wind rose (climatological)
Wind persistence tables
Portable wind measuring system.
Smoke bombs.
See Appendices A and B.
70
-------
Mobile sampling system.
Cameras.
If the meteorological data originates at an observing site located on the high
terrain outside of the valley,* adjustments of the wind data will be necessary
because of the channeling effect of the valley (a meteorologist should be con-
sulted to determine these adjustments)
4.6.3.2.1 Fumigation Concentration Stations. Fumigation situations in
valleys usually occur under inversion break-up conditions (e.g., see Hewson
and Gill, 1944). Winds are often calm or variable at typical airport loca-
tions (higher terrain). However, down in the valley drainage and valley flows
may carry plumes down the valley (see Figure 4-30) . Since there may be little
or no correlation between airport winds and valley winds in these situations,
it is suggested that plume behavior (e.g., typical direction of movement) be
determined through visual observations via photography, smoke bombs, or photo-
graphing the panorama from a ridgetop vantage point. If the plume follows a
similar trajectorye.g., consistently downvalleywhenever an inversion situ-
ation occurs, it may be possible to site a monitor in a permanent location
(see Figure 4-31a). While data for developing the plume "climatology" is be-
ing gathered (photographs, etc.), mobile sampling could be conducted to deter-
mine fumigation concentrations and where the maxima are typically located;
again, a vertical sensing capability would be quite useful. Permanent sites
could be located if ".favored" areas are observed, otherwise mobile sampling
may have to be conducted routinely. As an option, fumigation concentrations
may be calculated (by procedures as discussed, for example, in EPA, 1974c) to
estimate the maximum expected concentrations and the distances downwind where
they theoretically should occur. The 3-hour peak concentration is the averag-
ing time of concern.
aooo
I 3 5 7 9 II 13 13
TIME . HOURS
FIGUKE 4-30. The diurnal variation of valley winds during the summer in the
Columbia River Valley near Trail, B.C. Isopleths give average wind speed
components (mph): hatched areas - downvalley (north); unhatched areas -
upvalley (south). (Taken from Hewson, et al., 1961.)
For purposes of discussion, assume that such an observing site is a first
order NWS airport weather station taking regular observations.
71
-------
(b)
Ca1m or
Var iable
Wind
Drainage
Flow
FIGURE 4-31. Illustration of plume configurations under a variety of meteoro-
logical conditions and relative locations of sampling sites (X); (a) fumi-
gation situation, (b) plume deflected by valley wall (channeled flow in
valley), (c) plume either deflected over wall under unstable conditions or
passing out of valley due to excessive plume rise. Symbol A is location
at point on wall nearest the source (downwind wall).
72
-------
4.6.3.2.2 Valley-Wall Impact Stations. When the large-scale wind blows
in a cross-valley fashion, the valley wind direction is often channeled; i.e.,
the resulting direction on the valley flow is a compromise between the large-
scale flow direction and the valley orientation. Depending on the wind speed,
direction, and stability, the plume may either pass over the valley wall or
interact with it (not impinge upon it) and move downwind along it. Under
stronger winds, aerodynamic downwash conditions may prevail.
A wind station will need to be established on the valley floor; utiliz-
ing the services of a meteorologist, determine the wind climatology on the
valley floor for various speed ranges and stability classes. From wind cli-
matology, several basic valley-wall impact situations can be deduced. These
situations are listed below along with recommended siting procedures, or
points to consider for siting monitors, mainly for measuring 3-hour impacts.
Stable to unstable conditons, light to moderate winds, high
terrain (airport); channeled wind in valley.
In this situation, we are assuming that the plume will not,
in the case of the stable conditions, clear the top of the
valley walls; neither will it intersect it but will approach,
then be deflected by it and move along downwind parallel to
it (see Figure 4-31b). To assess this situation, the first
step is to determine the most frequently occurring wind di-
rection at the airport (from its stability wind rose) for
stable conditions and determine the associated valley resul-
tant direction.* This direction is the vector sum of the
valley flow wind direction and airport wind direction. This
direction defines the azimuth of "intersection" with the
valley wall. A tentative siting area should be established
above the half-way point between the valley floor and top of
the wall where the azimuth intersects the valley wall (see
Figure 4-31b). The final site should be selected on the
basis of visual observations. Caution should also be exer-
cised regarding possible prolonged and continuous dawnwash
conditions; cavity flows on the lee side of the upwind wall
may distort the plume near its source.
Under unstable conditions, the plume will either be de-
flected over the valley wall rather than along it, or will
pass out of the valley without any significant impact as
shown in Figure 4-31c. (See Appendix E for available models
that can deal with such situations.)
Stable conditions, light (not variable) cross-valley wind at
the airport; drainage flow in valley.
This situation is very similar to the one above except for
the following points:
The valley resultant direction as used here is an estimate of the direction
between the source and the closest approach point of the plume to the valley
wall (or "impact" point). Visual confirmation is recommended.
73
-------
- To estimate the azimuth of intersection in this case, the
valley resultant direction is considered a compromise be-
tween the valley orientation and the airport wind direc-
tion.
Also, since the winds are light, the effective height of
the plume may be above the valley wall and pass out of
the valley with reduced impact at the valley wall and be-
yond (see Figure 4-31c) .
Visual observations and the use of standard diffusion equa-
tions, making the proper adjustments for terrain elevation
(see Appendix E for available models), should be made to de-
termine degree of impact and direction of plume movement under
such circumstances. If plume does not clear valley wall, then
the problem is similar to that above (see Figure 4-32a).
Neutral or unstable conditions, moderate to strong winds at the
airport; cross-valley direction.
Under this situation, the plume is expected to be subjected
to downwash conditions due to either wake effects on the lee
side of the upwind wall or to the contravention of the 1.5 VS/U
ratio rule or both. In this instance, siting procedures are
difficult to generalize. However, suffice it to say that the
highest concentration would occur near the stack and measured
most effectively via mobile sampling. Even in this case though,
if favored plume touchdown areas are observed, this would pro-
vide a basis for establishing a permanent station (see Figure
4-32b). Guidance for analyzing the 1.5 ratio rule contraven-
tion problem was discussed in Section 4.5.2.
4.6^3.2.3 Worst-Case Conditions for Along Valley Flow. This situation
can be handled in a manner similar to that for the flat terrain case for both
24-hour and 3-hour average impact assessments. However, a meteorologist
should be consulted for advice. Locate tentative 3-hour and 24-hour peak
monitoring sites using the procedures discussed in Section 4.6.1. Finalize
location via mobile sampling. See Figure 4-32c for illustration.
4.6.3.2.4 Supplementary Monitoring Stations and Concluding Comments. In
all situations, one site should be established at a point nearest the source
on the wall most frequently downwind (based on annual wind rose) and one back-
ground site located a kilometer or so upvalley from the source. Instrument
types, inlet placement, site characteristics and supplementary equipment for
all stations are the same as those discussed in Section 4.6.1. If instru-
mented towers are erected, the elevated point source-in-valley situations and
related monitoring site selection problems can be treated in a manner more
rigorous than that described above.
74
-------
(c)
Valley Flow
(a)
FIGURE 4-32, Illustration of plume configuration under a variety of meteoro-
logical conditions and relative locations of sampling sites; (a) plume
deflected by valley wall (calm or valley wind in valley), (b) plume influ-
enced by wake effects, and (c) maximum concentration configuration in
valley with along-valley flow.
75
-------
4.6.4 Monitoring in Rough, Irregular Terrain Settings
Rough, irregular terrain may range in texture from nearly flat to ex-
tremely severe (e.g., the mountainous areas of Idaho, Utah, etc.). Since this
terrain is "irregular" by definition, no typical setting exists. Thus, we
were considerably hampered by this situation in that it did not permit us to
develop a "typical" scenario from which a site selection rationale or method-
ology could be presented, as, for example, in the previous discussion. Plow-
ever, we dealt with the problem by separating the terrain type into two
"regimes", one in which the setting was characterized by irregular topographic
features of sizes no larger than a typical physical stack height of a point
source (roughly 300 ft), and the other which was characterized by larger fea-
tures, up to and including the extreme mountainous. The former is reasonably
amenable to diffusion model analysis in the more or less traditional sense,
as for example shown by Leahey (1974). In this regard, as will be seen, the
monitor siting approach in this regime can be developed in a manner similar
to that described for flat terrain (see Section 4.6.1). However, plume be-
havior in the latter regime is extremely complex and beyond the simulation
capability of most models. For example, two rather detailed tracer studies
that were conducted by the National Oceanic and Atmospheric Administration
(NOAA) in mountainous terrain in Utah (Start, et al., 1973, 1974) showed the
extremely complex behavior of tracer material under various meteorological
conditions; the preparation of uniform site selection procedures for achiev-
ing specific monitoring objectives in such terrain is clearly impractical.
In view of this, it seems likely that any effort short of an individual dif-
fusion study, uniquely designed for a given situation, to assess the o02 im-
pact of a new or existing source will probably be unsuccessful, at least in
the extreme terrain cases. At the "smooth" end of this rough topographical
regime, tracer and numerical meteorological/diffusion modeling have been con-
ducted (Hinds, 1970; and Hino, 1968, respectively). The reader is urged to
consult these and the other references cited above to gain a better insight
of the problems of monitor siting in rough terrain. The services of a diffu-.
sion meteorologist is also strongly recommended.
4.6.4.1 Monitor Siting Procedures in Terrain of Up to Moderate Roughness
In the context of this discussion, the mean elevations of the terrain are
considered to be reasonably level with the maximum deviations from the mean
not exceeding a value equal to the height of a typical point source stack.
The recommended approach for selecting SC>2 monitoring sites is identical to
that for flat terrain; however, the specifics of the approach differ in the
following respects.
Diffusion Coefficients. Because of the mechanical turbulence
induced by the rough topography, the graphical solutions to
the Gaussian equationviz., Fig. B-2, Appendix B, Part II
must be modified by incorporating diffusion coefficients ap-
propriate for such terrain. The coefficients suggested by
Bowne (1973) for suburban and urban areas seem appropriate.
These would correspond to slightly rough (features up to
sizes of three-story buildings) to moderately rough (up to
stack height) topography. As a more accurate alternative,
76
-------
diffusion coefficients could be derived sepcifically for the
area of interest as shown by Leahey (1974).
Corrections for Terrain Elevation. Because of the scattered
and irregular nature of the terrain features, as opposed to
a solid barrier to the wind, caution should be exercised in
correcting concentration estimates, or in estimating loca-
tions of ground-level concentration maxima. The major effect
of the terrain on the plume will be to increase its rate of
dispersal, which would tend to bring the ground-level con-
centration maxima closer to the stack. Concentrations would
also likely be higher at the top of obstacles. However, a
lower level plume (effective H below the tops of the ob-
stacles) would tend to split and move around the obstacle,
particularly under stable conditions, resulting in lower
concentrations at the top of the obstacle. In more undu-
lating topography, the plume would tend to follow the
terrain. However, height corrections would need to be
made where the terrain elevation changed abruptly. For
example, Figure 4-33 is a numerical model simulation of
the ground-level pattern produced by an elevated plume,
showing increased concentrations over elevated terrain
(Hlno, 1968).
FIGURE 4-33. Distribution at height of 40m from ground surface (£ = Z-h
40m) of concentration of smoke emitted from a source with height 400m
which is derived from the computer experiment (taken from Hino, 1968).
77
-------
Downwash Situations. If obstacles higher than 2/5 of the
height of stack exist within 10 stack heights of the stack,
downwash due to wake effects is very likely. Downwash
analyses, such as those discussed in previous sections,
would be necessary.
Taking into account the above points, the procedure discussed in Section 4.6.1
can be utilized in selecting monitoring stations. However, because of the
heterogeneous nature of the setting, more reliance on mobile sampling and
visual observations of plume behavior may be required.
General Comments on Site Characteristics. The site charac-
teristics should be similar to those discussed in Section
4.6.1.5. However it is recognized that wake disturbances
on the lee sides of obstacles will be the rule. These dis-
turbances, which may extend to twice the height and five
to ten obstacle heights downwind, should not be considered
as things to avoid entirely. Such obstacles close to a
stack may downwash the plume to the ground to complicate
the picture. However, if downwash does not occur near the
stack, any wake effect produced by an obstacle located near
the expected ground-level maximum point would have very
little influence on where the actual maximum concentration
would be found, since the plume has already diffused down
to the ground naturally over an area probably larger in
size than the obstacle itself.
4.6.4.2 Conditions in Extremely Rough Terrain
The nature of this terrain precludes the development of monitoring site
selection procedures that could be uniformly applied to any given mountain-
ous terrain configuration. However, describing some of the gross character-
istics of plume behavior in such terrain may be instructive in terms of
"points to consider" when contemplating establishing SC>2 monitoring sites to
assess the impact of individual point sources so located.
The following summary was abstracted from the two NOAA studies (Start,
et al., 1973, 1974) and from the study by Hinds (1970) cited previously. The
studies described the behavior of plumes over specific sections of California
and Utah characterized by extremely rough terrain. Plume behavior in these
areas may or may not typify such behavior in other similar topographic areas.
Elevated Plume. Centerline concentrations are reasonably
well predicted by the standard Pasquill-Gifford diffusion
curves when the plume does not pass over mountainous ter-
rain. However, over mountainous terrain elevated center-
line concentrations average from 3 to 4 times more dilute.
Lateral Plume. Spreading is almost twice that expected
for over flat terrain. Several physical processes con-
tribute to this increased spreading.
- Plumes tend to be deflected around obstacles.
78
-------
- The descending portions of looping plumes spread
laterally as they approach steeply sloped canyon
floors.
Increased mechanical turbulence enhances lateral
spreading.
- Vertical shearing of wind direction with height en-
hances la,teral spreading.
When a low, strongly stable layer aloft combines with flow-
blockage effects of the terrain, a quasi-stagnant air pocket
can develop that may contain an elevated plume layer. Pro-
longed ground-surface contact with this layer is probable.
>
A higher stable layer will allow the plume to flow over the
ridge-tops and the plume tends to become uniformly distribu-
ted in the vertical. Because of ground-reflection effects,
ground-level concentrations may be twice as large as those
aloft.
With no stable layer aloft, plumes are deflected aloft over
the ridges and follow a path similar to the shape of the
underlying topography. The lateral distribution of pollu-
tants from the centerline is generally Gaussian.
The locations of the maximum ground-level concentrations
were at the ridgetops. Specific impact areas were identified
best via pilot balloon (pibal) wind observations near the ef-
fective plume height.
Figure 4-34 is a schematic illustrating the dilution of an
airborn plume as it interacts with elevated, rough topography.
In unstable conditions rates of dilution in mountainous ter-
rain are about the same as those over flat terrain; the rates
increase by a factor of 5 in neutral conditions and a factor
of 15 in stable conditions.
Peak to mean concentration ratios in mountainous terrain are
lower than those over flat terrain.
In canyon settings within mountainous terrain, mechanical
turbulence is enhanced by:
Turbulence generated near the mountain tops and the
upper confines of the canyon.
- Airflows originating within side canyons.
- Wake effects of airflows over and around canyon topo-
graphic variations.
Because of strong diurnal wind cycles characteristic of
canyon topography, synoptic stagnation conditions are not
the worst diffusion conditions.
79
-------
ZONE 1
ZONE 2
ZONE 3
ZONE 4
ZONE 1: "Simple" elevated plume with buoyant rise, becoming the bent-over form. Near Gaussian
vertical distribution.
ZONE 2: Deflection zone with plume tending to parallel ground surface. Near Gaussian vertical
distribution.
ZONE 3: Mixing or transitional zone affected by turbulence about the toooaraohv. Quasi-
Gaussian vertical distribution.
ZONE 4: Well-mixed zone. Ouasi-uniform vertical distribution of plume mass.
Plume effluent concentrations are greatest where the shading is the most dense.
FIGURE 4-34. Schematic illustration of the dilution of an airborne plume as
it approaches and flows over nearby elevated terrain. Four zones of plume
behavior and the postulated vertical mass distributions are depicted.
(Taken from Start, et al., 1974,.)
Figure 4-35 illustrates the turbulent wake effects of obsta-
cles characteristic of canyon topography.
Diffusion over a ridge-canyon system often results in sub-
stantially lower concentrations on the canyon floor than
would occur at the same distance over flat terrain.
4.6.4 .3 Implications for SC>2 Monitoring
Based on the above observations, the following general guidelines for
selecting S02 monitoring sites in extremely rough terrain are suggested.
In regions subject to at least occasional periods of low
mixing depths, locate monitors in basins that have inlets
for S02 source plumes. Very high concentrations could re-
sult from stagnant air pockets that could develop in such
areas. The rough terrain, upland areas of the west coast
of the United States would seam to be particularly liable.
80
-------
BOUNDARY
Wake
Cavity
Two Dimensional Canyon Flow
Primary
Secondary
FIGURE 4-35.
Schematic illustration of turbulent wake effects caused by
obstacles protruding into the primary flow pattern.
(Adapted from Start, et al, 1973).
81
-------
Site monitors at rigetop locations in the general downwind
directions from the source, or perhaps at ridgetop loca-
tions surrounding the source, particularly those nearest
the source at near effective height (H) elevations.
Site monitors in passes that may receive the plume advected
either by drainage or channeled winds.
A complete survey of the entire area influenced by the SC>2
source would almost certainly be required in all situations.
Visual observations, aerial photography, mobile sampling,
remote sensing, etc. would probably be the most important
means for conducting such surveys.
82
-------
5.0 RATIONALE AND SUPPORT DOCUMENTATION FOR SITING CRITERIA
The site selection and inlet placement procedures and criteria discussed
in Section 4.Q are quite specific, particularly those regarding location pa-
rameters such as height of the inlet, proximities of interfering sources (un-
due influence) and horizontal positioning of the inlet for rooftop sties, etc.
The rationale for some of these procedures and siting approaches was included
to explain certain points of the procedural ^pgic. However, it was felt that
justifying certain other elements of the siting procedures and criteria would
have muddled the continuity. Therefore, we have reserved this section for
their presentation.
The logic underlying the procedures of Section 4.0 can be considered em-
bodied in three basic elements:
1) Determining the general location of the monitoring site,
mainly via simulation modeling.
2) Refining the location to minimize undue influences from
nearby sources, including meteorological effects.
3) Placing the instrument inlet in such a location to avoid
local contamination.
The first element, we believe, has been adequately covered in previous sec-
tions and in the appendices and requires no further discussion here. There-
fore, much of the material presented in this section will pertain to elements
2 and 3. Several miscellaneous items that are relevant to all three elements
will also be discussed.
5.1 UNDUE INFLUENCE EFFECTS
Regarding the problem of establishing a site location such that undue
influences from nearby sources are minimized,* we had to first define what
constituted undue influence. We wanted to use a fairly stable, maximum SC>2
concentration as a level of undue influence and then establish a separation
distance between the monitoring site and all sources such that any one
source's contribution at the monitoring site would not exceed the undue
In just about every reference cited in this study the problem of undue
influence of nearby sources was mentioned, but no objective procedures
or approaches for dealing with such influences were ever suggested.
83
-------
influence level. Typical rural background levels over all parts of the country
seemed to be excessive, 10-30 yg/m3 (viz., Figure 4-3, EPA,1974d), to be used
as an undue influence level for regional-scale stations. Also, these levels
were decreasing due to the effectiveness of S02 emission controls. Thus, it
was decided to use the natural background level of 2.6 yg/m3 (1 ppb) as re-
ported by Robinson and Rabbins (1970), a very low level and probably quite
stable as well. Using this value and typical emission rates for various
classes and configurations of sources, we determined sets of distances beyond
which impacts from any source did not exceed the undue influence level. Exam-
ples of these distances, described as "interference distances" (IDs), were shown
in Table 4-2 (see Page 34).
The ID of a major urban area was determined by using the normalized con-
centration pattern resulting from a circular area source as shown in Figure 5-1
(from Ludwig and Kealoha, 1975) Slong with a speed of 1 m/sec and a half-life
value of 3 hours.* Typical maximum emission rates for a major urban area were
assumed to be represented by the city of Philadelphia, Q = 0.86 x 10~5 g/sec/m
(from EPA,1973b). This gave an ID of about 30 km.
RELATIVE SURFACE CONCENTRATIONS (XO2 decay characteristics.
84
-------
TABLE 5-1
Configurations and Emissions for Typical Source Types Assumed in
Determining Interference Distances for Regional-Scale Stations
Source
Type
Power Plant
(400 MW)
Industrial
Space Heat
(500 T S02/yr)
Small Town
(25,000 pop..
6000 homes)
Individual
Home
Characteristic
Emission Period
365 days/yr
Winter Quarter
(Dec, Jan, Feb)
Winter Quarter
(Dec. Jan, Feb)
Winter Quarter
(Dec, Jan, Feb)
Fuel Rate
S Content (%)
280 x lo6 gal
16 oil 9 IX S
14 x 10s gal
16 oil 9 0.5X S
TO3 gal/home
f2 oil 9 0.2X S
103 gal
12 oil * 0.21 S
Source
Configuration
Point, Uniform
Wind over 22.5°
Sector
Point
Area Source
4 ml2
Point
Emission
Rate
(g/sec)
575
58
10
.0016
Meteorology
Wind Speed
5 m/sec
5 m/sec
1 m/sec
1 m/sec
Stability
Class
D
D
0
F
Effective
Ht. (m)
300
200
0
0
During this phase of the study, we concluded that a major urban area, in
an ID sense, may be considered as having a population of about 2 x io5 or
more. This contention was based on the observation that the ID varied more
closely with emission intensity rather than with total emissions; large cities
emit more SC>2 than small cities, but it is emitted over a larger area. For
example, the ID for a 25 x io3 population town was 15 km (Table 4-2) versus
an ID of 30 km for a 1 x io6 population city; the 2 x io5 figure seemed an
appropriate cut-off point to separate the large urban areas from smaller towns.
Analogous to the IDs for regional-scale stations is the concept of point,
minor, and source IDs (PSID, MSID, and SID, respectively), as presented in
Table 4-4 (see Page 44). These values were obtained by considering an undue
influence level of 10 yg/m3 (instead of 2.6 yg/m3 ), which was the cleanest
rural SC>2 background level observed. We felt that using this higher undue in-
fluence level was justified since the associated ID values are meant to apply
in urban and suburban areas where existing SO2 levels are much higher than in
rural areas where regional-scale station IDs apply. Table 5-2 shows how the
IDs of Table 4-4 were obtained. In developing the concentrations shown, zero
effective height, a wind speed of 1 m/sec and stability class D were assumed;
this would tend to produce a maximum impact at the site to provide a modest
safety factor. The diffusion coefficients used in the calculations were those
suggested by Bowne (1973) for rural, suburban, and urban areas. Large point
sources were considered as those using IO6 gal/yr of fuel oil. Minor sources
used IO3 gallons in rural areas (home) , 101* gallons in suburban areas (small
office building) and 10s gallons in urban areas (large office building). All
fuel was burned during the winter quarter of the year. The concentrations as
shown in Table 5-2 are unadjustedthat is, the concentrations have not been
modified to account for effects due to decay and averaging time; the actual
3-hr mean concentrations were estimated by multiplying the given concentra-
tions by an appropriate half-life factor (considering corresponding travel
time) and the correction factor of 0.51 to account for additional dilution due
to wind direction variability. Multiplying the circled numbers in Table 5-2
by these factors will result in an actual concentration estimate of 10 yg/m3
85
-------
Downwind distances associated with these concentrations that are shown in Table
5-2 are the IDs of Table 4-4.
TABLE 5-2
Rationale for PSIDs and MSIDs of Table 4-4 (see Page 44)
Development
Intensity
""^^^ Fuel Use
^\ (ga.l/yr)
Downward ^^^^^
Distance (m) ^^-v^
10
30
100
300
600
1000
2000
3000
Unadjusted Concentration (yg/m3)
Urban
103* 10" 10s 10et
5.1 51 510 -
3.4 34 340
0.9 9 90 900
0.23 2.3 mj 230
.078 .78 7.8 78
.034 .34 3.4 M4J
.011 .11 1.1 11
.006 .06 .58 5.8
Suburban
103 101* 10s 106
15.9 159
Rural
103 101* 10s 10s
84.9 849
9.8 98 980 K2.a 525
3.07 GOj) 307
.68 6.8 68 680
.26 2.6 26 260
.117 1.17 11.7 117
.04 .4 4.0 MOJ
.021 .213 2.13 21.3
m.j) 113
1.6 16 160
.5 5.0 50 500
.20 2.0 20 200
.073 .73 7.3 73
.038 .38 3.8 MSj
* Individual Home
t Large Point Source
A 10° plume sector roughly corresponds to +_ la around the centerline of
the plume. This sector width is suggested as a guide for determining those
upwind sources that may unduly influence the measurements at the site. The
schematic shown in Figure 5-2 illustrates the rationale for this criterion.
For neighborhood-scale stations, the sector sizes were increased to 20° for
the very nearby point sources.
5.2 METEOROLOGICAL PROCESSES PERTINENT TO SITE LOCATION REFINEMENT AMD INLET
PLACEMENT
Wind direction establishes the general transport direction and determines
which sector of the area surrounding the source will receive the pollutant.
86
-------
Site
Number
FIGURE 5-2. Schematic illustrating undue influence of nearby sources on
measurements at three sampling sites: (1) within 10° plume
sector; (2) at a minimum impact point within area; and (3) at
a point beyond the MSTD but within zone of concentration
characteristic of the area as a whole.
87
-------
The location of impact points within the sector are determined by the trajectory
of the polluted air stream or parcel. The trajectory is only rarely a straight
linethe parcel being subject to effects of obstructions that can change a
given direction to another. These obstructions include mountains, valleys,,
buildings, and other parcels or masses of air. Even in the absence of physical
obstructions, the wind varies in space and time due to thermal effects, shear
effects, and turbulence advected from upwind (e.g., Fig. 6, Anderson, 1971) .
The parcel may also be deformed; being a fluid, the air parcel will be
subject to changes in shape and separation. However, from mass continuity con-
siderations, there must also be corresponding changes in air flow. For example,
parcels passing between two obstructions (viz., two buildings) or over a moun-
tain will be squeezed horizontally (transverse) and vertically, respectively.
In either situation, stretching of the parcel longitudinally will take place to
compensate, resulting in a faster air flow. The reverse will occur for parcels
moving across a valley (vertical stretching and longitudinal squeezing)(see
Fig. 5-3). However, from mass continuity considerations, the concentration of
the pollutant within the parcel must remain essentially the same throughout the
deformation process.
PLAN VIEW
VERTICAL SECTION
VERTICAL SECTION
FIGURE 5-3. Topography effects on
wind. Length of arrow is propor
tional to the wind speed.
In illustrating the concepts of wind
speed and direction, and deformation,
we have assumed that the pollutant re-
mains contained within the parcel as
the parcel proceeds downwind. This is
not actually the case because turbulent
mixing and diffusion processes are con-
stantly at. work dispersing the pollu-
tant from the initial parcel to adja-
cent ones as the pollutant plume moves
downwind. [Note in this regard, a
point apparently missed in some of the
papers reviewed that no physical mech-
anism exists in the atmosphere, includ-
ing deformation, which can reverse the
process and "unmix" the atmosphere to
create higher concentrations of S02-
Therefore, turbulence in the atmosphere
can only lead to dilution ox- dispersion
of a polluted air mass that it affects;
"cavity" flows cannot accumulate pollu-:
tantthey can only partically contain
it; nor can "channeling", i.e., the
squeezing of streamlines, squeeze to-
gether the pollutant and increase its
concentration. Indeed, no flow nor
even a stagnant air mass can contain
a higher pollutant concentration than
that of its most intense inlet.] The
rate of diffusion is a function of at-
mospheric stability, which is associ-
ated with varying degress of thermal
88
-------
or convective turbulence and the degree and nature of the roughness of the sur-
face over which the parcels are transported, which is associated with varying
degrees of mechanical turbulence.
Mechanical turbulence is produced when air moves over a rough surface,
which tends to interrupt an otherwise smooth air flow. Air swirling about
buildings, rough ground, and clumps of irregular sized vegetation are examples
of mechanical turbulence. The degree of mechanical turbulence is directly pro-
portional to the wind speed.
5.2.1 Effects of Natural Topography; Wakes, Channeling, and Turbulence
The transport and diffusion of a pollutant plume is complicated by the ef-
fects of natural terrain features on the flow of air in which the plume is
transported.
It is suggested on the basis of work described by Anderson (1973), that
topographic effects on the windfield are scaled by the ratio of topographic
slopes to the depth of the mixing layer. That is, if the ground rises or falls
a significant fraction of the mixing layer depth, the wind speed components
should change a comparable fraction; for mixing depths on the order of 1000 ft,
100-ft elevations would usually be significant. Topographic features consider-
ably larger than this can produce even more dramatic changes in the flow field.
For example, wind flowing towards a very steep hill face is merely guided by
the topography,, but wind blowing off the top of a similar face will typically
break up into severe turbulence and may even form a "cavity wake" as shown in
Figure 5-4. (For a good review of the basic mechanism that causes wakes and
wake cavities, the reader is referred to Halitsky, 1962.)-
FIGURE 5-4.
Assymmetry of flow approaching and leaving steep
topography.
If topographic slopes exceed 10 percent, much increased turbulence can
be expected with downslope winds. If topographic slopes exceed 20 percent,
cavity flows are g_uite possible.
Inasmuch as the air entering a cavity may turn over many times before
leaving, the cavity tends to average the concentration from the pollution
89
-------
sources that feed it. If -the pollutant enters from the flow passing the ob-
stacle, it may be thought of as continuously "sampling" the passing air, mix-
ing it up, and passing it on much delayed. Thus, the cavity averages on both
time (due to the delay) and space (due to its size) scales. The cavity can-
not "collect" pollutant because it also "collects" the air which carries the
pollutant, each in direct ratio to its concentration.
When the general wind direction is oblique to a ridge-valley axis, the
channeling of the wind often occurs as shown in Figure 5-5. The surface wind
speed in the valley is usually diminished because of friction (Flemming, 1967)
If the valley wall is bluff, wake cavities (mechanical turbulence) on the lee
side of the upwind wall may be produced. Wind blowing perpendicular to the
valley axis will not be'significantly channeled, but surface speeds can be
considerably diminished (frictional drag and vertical stretching) and the pro-
bability and size of wake cavities will increase.
(a)
(b)
FIGURE 5-5.
Distortions of the wind flow by topographic obstacles.
(a) Channeling of the wind by a valley, and
(b) The effect of a mountain pass on the wind flow.
(Taken from Slade, 1968.)
At night under clear skies and light winds, the air adjacent to the
ground along the valley floor and slope are cooled through radiational cool-
ing. As this cooling progresses, a density differential between air at the
same elevation (relative to a horizontal plane) develops and results in a
flow of air down the slope toward the valley floor. This flow is called a
slope or drainage wind. The same mechanism also causes a flow of air down-
hill along the valley axis (valley wind). In the daytime under light wind
conditions, the drainage flow mechanism is reversed and causes upslope and
upvalley flows. Downslope and upslope air flows can be complicated in com-
plex valley systems where several valleys merge at various angles or where
slopes vary. Also, the differential heating of valley slopes can further
complicate the already complex flow pattern.
The only significant modification of winds and turbulence (and hence SC^
concentrations) due to irregular topography is that it increases mechanical
turbulence. The scales of this turbulence depend on the strength of the wind
90
-------
and the size and size distribution of the individual features. If the features
are large, wake cavities may form; otherwise the turbulence elements are of
small scale. If the area is generally level, flows produced by drainage and
valley flow mechanisms will be insignificant. In general, smooth hills alter
the flow least. Under unstable conditions, air parcels tend to move over ob-
stacles, while under stable conditions, air parcels tend to move around obsta-
cles. In strong winds, cavity wakes can form on the lee side of large bluff
hills.
Trees can obstruct an otherwise smooth wind flow and increase mechanical
turbulence. Dense clumps of large trees can have the same effect on wind flow
as a small, bluff hill and produce wake cavities. Similarly, wake cavities may
exist on the windward side of a forest clearing. Below a forest canopy, wind
speeds may be very low and diurnal temperature variations tempered. Irregular-
ly spaced trees or small clumps of trees of varying height, lines of trees,
and low vegetation will have the same effect as rough topography and increase
mechanical turbulence.
5.2.1.1 Effects of Above Considerations on SOg Distribution
The effects of ridge-valley topography on SC>2 concentrations and patterns
depend on several factors. The major factors are the time of day the pollu-
tants are being emitted, where they are being emitted, the height of release,
and the prevailing meteorology. An S02 plume emitted from the top of the val-
ley wall or on an adjoining plateau may be caught up in a cavity wake (downwash)
and brought down into the valley. At night, an 302 plume released at a high
level at a high exit velocity may escape the valley and surrounding high ter-
rain entirely. On the other hand, lower level releases may become imbedded in
the drainage flow and move down the valley, or, emerge above the drainage flow
upper boundary and impact (not intersect) on the valley wall or slope. Emis-
sions from intense area sources (low-level) located in valleys and released in-
to a very stable drainage flow or a deep, intense inversion layer will be se-
verely restricted both horizontally and vertically.
Unless large obstacles ars present, moderately rough natural topography
will decrease the pollutant concentration by increasing mixing because of mech-
anical turbulence; therefore, the concentration levels measured will be less
sensitive to the location of the site or placement of the instrument inlet.
Wake cavities formed on the lee sides of the largest bluff obstacles may cause
the downwash of a passing plume.
The effects on SO2 plume behavior induced by vegetation are similar to
those caused by irregular, rough terrain. However, an individual SO2 plume
passing over a clearing in a forest at a low level (viz., a low-level release
from a nearby source) may be downwashed to the ground via a wake cavity formed
on the windward side of the clearing.
5.2.2 Effects of Urbanization: General
The effect of urbanization on meteorological elements is described very
well by Pooler (1963) and summarized by Peterson (1969). In their work, they
91
-------
discuss urban effects on the horizontal and vertical distributions of tempera-
ture, humidity, visibility, radiation, wind, and precipitation. The urban
"heat island" phenomena has been well documented in studies by DeMarrais (1961),
Mitchell (1961), Bornstein (1968), and Oke (1975). In these studies, charac-r
teristic vertical and horizontal variations, wind flows, and stability changes
are discussed and compared to adjacent rural areas. Oke (1973) related, city
size and urban heat island intensity. Hutcheon, et al. (1967) show that even
small cities can produce urban heat islands. From these studies, the effects
of the various meteorological elements relevant to SC>2 behavior (and, there-
fore, monitoring site exposure criteria) are summarized here.
Nighttime temperatures in cities are higher than those observed over ad-
jacent rural areas. Generally, the larger the city and more intense the noc-
turnal inversion, the larger the temperature differential between the urban
core area and the outlying rural area (as great as 20°F). Heat island inten-
sity has been shown to be related to the logarithm of population according to
a study by Oke (1973) for a group of North American cities. Daytime tempera-
ture differentials are generally much less apparent than at night and occa-
sionally are reversed.
Higher surface temperatures in urban areas reduce atmospheric stability.
In fact, in large cities surface-based inversions are quite rare. Decreased
stability, increased mixing depths, and increased mechanical turbulence, due
to the rough urban topography, all tend to enhance the mixing and dispersion
of pollutants.
Frictional drag and the urban heat island effect modify the urban wind
direction. During the daytime, particularly under unstable conditions, wind
directions over the cities and rural areas are reasonably homogeneous. How-
ever, at night the relatively warmer air of the city rises, causing low-level
convergence and an inflow of air toward the urban center as observed by Pooler
(1963). The inflow is strongest at night when the urban heat island is well
developed. Of course, the magnitude of the effect is dependent on city size.
In many cases, this inflow toward the urban center is observed whenever re-
gional winds are weak. Often in extreme conditions, an outflow of urban air
aloft is observed and results in a closed circulation.
5.2.2.1 Building-Induced Turbulence and Wakes
The material in this section is presented as support documentation for
the criteria contained mainly in Tables 4-3 (see Page 39) and 4-5 (see Page 45).
The major area of concern here is the representativeness of urban monitor-
ing sites in view of the complication in the wind and turbulence fields, par-
ticularly in the daytime, due to urban structures. Even the idealized situa-
tion of a single building is quite complex as shown in Figure 5-6. In cities,
complex street canyon flows and cavity wakes on the lee side of buildings dom-
inate the flow pattern. Since high SO2 concentrations are often observed in
urban areas (and is the rule in the north), a major effort was made to reflect
in the urban site selection guidelines an accounting of such complex urban flow
characteristics. In addressing the phenomena, some studies trea~ting the prob-
lem by two basic techniques were reviewedmathematical modeling (Hotchkiss
92
-------
and Harlow, 1973) and physical modeling (Halitsky, 1962). Neither technique
provided comprehensive quantitative answers. However, each provided some in-
sight into the problem. In general, features of complex urban flows are dif-
ficult to generalize via both mathematical and physical modeling because the
dominant flow features near buildings cannot be represented as turbulence or
as "potential flow" (i.e., a solution to the potential function equation).
r . lioloting furtoct
FIGURE 5-6. General arrangement of flow zones near a sharp-edged
building (taken from Slade, 1968).
Cavity flows, as shown in Figure 5-7, and wake flows, as shown in Figure 5-8,
contain features as large or larger than the obstacles creating them and yet
are in no way random. They may, however, be embedded in random turbulence,
and are, in general, dependent on very fine scale features of the flow such as
the precise angle of incidence and the smoothness and precise shapes of the
surface contours. Thus, numerical models which only introduce turbulence of
a fixed intensity in the incident flow cannot reproduce the effect of the in-
cident flow containing a regular train of eddies as large as the building.
FIGURE 5-7. Cavity flows.
93
-------
If models contain upstream structures to generate such eddies, they still can-
not represent the formation or release rate of such eddies which depend on mi-
croscale surface details of the building.
FIGURE 5-8. Wake region in flow past bluff obstacle (viewed
from above).
A virtue shared by both numerical and physical modeling is that gross,
qualitative features are quickly, cheaply, and/or distinctly shown, and the
existence of large, coherent disturbances and their general nature can often
be determined by such modeling programs. However, their location, size,
typical residence time when trapped, or the pollutant concentrations result-
ing from plume interaction cannot usually be determined. For example, the
wind tunnel studies described by Halitsky (1962), show that in some cases
highest pollution flux is against the mean wind; whereas on superficially
comparable building parts, no such reverse flow is evidenced. On other
building roofs, pollutant transport is transverse to the mean wind. All of
the observations might lead to conclusions that might be invalidated by
slight changes in incident wind direction, or variability. The experiment
does legitimately warn the selector of monitoring sites that one cannot
easily or casually choose a site and be sure that it will be upwind (statis-
tically) of a nearby source in a complex structural environment. However,
in a later study, Drivas and Shair (1974) showed that a reverse circulation
in the wake downwind of a building generally exists and that tracer experi-
ments indicated that the extent of the recirculation back onto the roof was,
in general, systematically confined to less than one-half the width of the
building from the downwind edge.
94
-------
5.2.3 Relevance of Above Considerations on Siting Criteria
The urban modifications to the regional meteorology that have a major im-
pact on S02 concentrations and patterns are the air inflow characteristics un-
der stable conditions when regional winds are weak, particularly at night, and
the effects of wake cavities on the lee sides of buildings under stronger re-
gional wind conditions.
When a heat island circulation exists, individual pollutant plumes may
tend to converge toward the center of the city where they will rise, then
return aloft to the periphery of the urban area and return again, completing
the circulation (convergence at low levels, divergence aloft (see Figure 5-9).
With a large number of SC>2 plumes tending to converge toward one central
point and the return from ,aloft of already polluted air, a pollutant maximum
may be located near the urban center, From heat island circulation dynamics,
Chandler (1968) deduced that urban-rural pollution gradients would be very
sharp with the strongest gradients on the lee side of the city. Ball (1969)
observed that pollutant peaks roughly coincided with the thermal maximum of
New York City's heat island and that the thermal pattern shifted and elon-
gated in response to the regional wind flow.
300 -
. 200 -
til
o
t 100 -
WIND
CITY
COUNTRY
FIGURE 5-9. Urban circulation and dispersion before sunrise.
Under stronger wind speed conditions, particularly during the day, heat
island circulations break down and lose their identity. Complex building wake
patterns then distort the wind flow resulting in a very turbulent urban atmo-
sphere, and become a major factor in influencing SO2 concentrations and dis-
tributions. Figures 5-10 and 5-11 show mathematical representations of the
effects of building wakes and cavity flows on pollutant distribution. In a
large urban area, these complex flows would typically produce an averaging
effect in both the horizontal and vertical. This statement is suppprted by
Pelletier (1963) who measured SO2 distributions in Paris; at specific locations
he found no appreciable difference in 24-hour mean SC>2 concentrations measured
at 13 m and 53 m above the ground. The same conclusions were reached by Clif-
ton, et al. (1959) in a Sheffield, England study, particularly for locations
not too near the upwind edge of the city (to allow sufficient time and dis-
tance for the wake/cavity-induced averaging effect to take place). Simon (1969)
described a similar situation, in considerable detail, in his discussion of New
York City's meteorology program. From the above considerations, what seems to
emerge regarding the vertical distribution of SC>2 in urban areas is illustrated
in Figure 5-12 and described as follows:
95
-------
A parcel of air entering the city is characterized by very low con-
centrations uniformly distributed in the vertical. At it passes
through the suburbs, it picks up contributions from relatively
small sources at low-to-moderate heights. As it passed through
the central business district (CBD) , very large amounts of SC>2 are
picked up; but because of plume reflection and the mixing/averaging
effect produced by building wakes, the S02 distribution is substan-
tially uniform up to at least the mean building height. However,
it is likely that a maximum concentration level may be observed
above the mean building height near the mean effective height of
the stronger sources which are emitted at higher levels. This
statement is substantiated by Simon (1969, Fig.6). As the air
leaves the city, the upper profile "fills in" due to the upward
dispersion from lower levels. This effect plus horizontal disper-
sion continues until a vertically uniform, general low-to-moderate
concentration level results just downwind of the city.
FIGURE 5-10.
The dispersal of a narrow
plume passing over a single
building. Recirculation in
wake region is clearly evi-
dent. (Taken from Hotchkiss
and Harlow, 1973.)
Figure 5-11.
The dispersal of pollutant
from a flush vent on the top
of a complex building struc-
ture. (Taken from Hotchkiss
and Harlow, 1973.)
96
-------
Uniform concentration in vertical.
Upper profile filling, lower profile decaying.
Upper profile begins filling in due to
vertical dispersion from lower levels.
Distribution essentially uniform
up to mean building height.
Quasi-instantaneous
concentration maxima at
effective plume height.
Incoming S02
profile, dis
tribution is
uni form wi th
height.
FIGURE 5-12.
Schematic illustration of vertical distribution of SG>2 within
a vertical column of air passing through an urban area.
97
-------
From the above discussions, the following conclusions can be drawn:
Mixing produced by mechanical turbulence and wake effects of
larger obstacles over moderately rough, natural terrain aver-
ages the pollutant over space which lessens the concern of
exact site location and inlet placement in rural areas.
Micro-scale urban features substantially increase mixing and
promote uniformity of pollutant levels from mid and far dis-
tant sources. This mixing between source and monitoring site,
reduces the monitoring site selection problem to the consider-
ation of only near sources (less than the interference dis-
tance) .
A single building wake can mix, over its volume, pollutant
from a source that enters it. Thus, standard analyses com-
pute a "virtual point source" upwind of a building if the
building wake is thought to catch its own plume.
The uniform mixing principle is not absolute and cavity flows
often build, get swept away, and reform, leading to large
"puff" type releases.
Except for near the windward edge of a city, a vertically uni-
form SO2 distribution up to at least the mean building height
over the area of interest in the city can be assumed. The
choosing of 0.8 H (or lower) for inlet location above the
ground (see Table 4-3, Page 39; and Table 4-5, Page 45) is
somewhat arbitrary, but was meant to insure that the instru-
ment (or inlet) would be placed at a point in the vertical
where the measured levels would approximate those existing
near the breathing zone5-6 ft above the ground.
If pollutant release is known to be well within a cavity
(e.g., emissions from a vehicle in a deep, street canyon)
averaging will not be complete and concentration fluctuations
and gradients are apt to be found within the flow. Minimum
velocities and maximum concentrations should be found near
the ground on the leeward side of the obstacle. This is the
justification for avoiding trailer locations just downwind
of buildings with large stacks (see Table 4-5). This situ-
ation is precluded, however, if the interference distance
criteria are satisfied.
Pollutants from sources located downwind of a building may
be emitted into the wake cavity behind the building. The
reverse flow of the wake may advect pollutants up to the
roof of the building to at least one-half of the width of
the building from the downwind edge. This observation is
the rationale for recommending that inlet placement loca-
tions be on the windward side of the building (see Table 4-5).
It also justifies the recommendation of not having SO2 sources
98
-------
on the roof of the building chosen for the monitoring site
or inlet location.
Emergency episode stations should be located in the very
heart of the maximum SC>2 emission density zone of an urban
area; during air stagnations wind speeds are low and direc-
tions are variable so the maximum concentration should occur
where the emission density is a maximum. However, even in
this case, caution should be exercised in locating monitors.
S02 emissions from high stacks (which are often the largest
sources) may not reach the ground (if at all) until several
miles away during a low wind-speed/stable situation. Appro-
priate site locations can best be found by using gridded
emission inventory data with most of the weight being given
to the area source fraction of the inventory.
The heat island mechanism may produce maximum concentrations
near the wind inflow convergence point which may be located
near the center of maximum SC>2 emission zone of the city. This
justifies considering the emergency episode station as an al-
ternative site for measuring the 3-hour peak concentration.
5.3 MISCELLANEOUS CONSIDERATIONS
In this section, additional justification and rationale regarding certain
siting criteria, modeling approaches and miscellaneous considerations along
with support documentation is presented.
5.3.1 Tempera tur e
The temperature at a point has little direct effect on the concentration
of S02. Only temperature gradients, mainly in the vertical, have a major in-
fluence. However, temperature may influence the rate of emission of S02; for
example, the amount of fuel burned for space heating is directly proportional
to heating degree days, a number which is equal to the average temperature for
the day minus 65°F. Turner (1968) and Roberts, et al. (1970) related S02 emis-
sion rate response to changes in temperature on an hourly (diurnal variation)
as well as a daily mean basis. Power plant load (and SO2 emissions) may vary
seasonally and diurnally. In the northern part of the United States, power-
plant emission maxima occur in both summer and winter in response to power
demand to run air conditioners (related to cooling degree days), and in re-
sponse to demand to run electric and oil heating systems (related to heating
degree days)(Federal Power Commission, 1971).
5.3.2 Chemical-Physical Interactions
S02, being soluble in water, interacts both chemically and physically
with atmospheric moisture. S02 is also photochemically and catalytically
99
-------
reactive with other atmospheric constituents. The reaction kinetics of such
interactions are very complex, all aspects of which are not yet fully under-
stood. Some of the end products 'of atmospheric interactions involving SO2
are sulfur trioxide, sulfuric acid, and sulfates. SC>2 also interacts with
ground, vegetative, and water surfaces. Since it is beyond the scope of this
report to present a detailed discussion of this topic, only a brief summary
of the more pertinent aspects is presented.
_5_.3.2.1 Reactions of SO^ With Atmospheric Liquid Water
Precipitation scavenging consists of three basic components as described
by Slade (1968):
1) transport of the SC>2 to the scavenging site,
2) in-cloud scavenging by precipitation and cloud elements
(rainout), and
3) below-cloud scavenging by falling raindrops (washout).
The rate of scavenging of SO2 is based on the molecular diffusion of S02 to
the droplets in accordance with the vapor pressures and solubility of SO2
Laboratory tests by Bracewell and Gall (1967) indicated that the occurrence
of H2SO4 in urban fogs could be accounted for by the catalytic oxidation of
SO2 dissolved in fog droplets in the presence of certain metallic ions.
5".3.2.2 Catalytic and Photochemical Oxidation Reactions
SC>2 may be catalytically oxidized to 863 in the presence of oxides of
nitrogen. The SOj then readily converts basic oxides to sulfates (NAPCA,1970).
Liberti and Devitofranesco (1967) reported that S02 may be catalytically oxidized
to sulfate after being adsorbed by suspended particles. in a report to the
U.S. Senate, the National Academy of Sciences (NAS, 1975) reported that SO;2
oxidation rates (to sulfates) varies from 0.17 percent/hour to 50 percent/hour,
depending on the relative humidity and the presence and relative concentra-
tions of other pollutants. The rate is typically more rapid in urban air.
Urone, et al. (1968) reported that SO2 can be photo-oxidized to H2SC>4 aerosol
in the presence of water vapor and at a faster rate when hydrocarbons and ni-
trogen dioxide are present. In the same NAS report cited above, it was esti-
mated that in the northea-stern United States, roughly one-third of SC>2 emis-
sions are returned to the earth as sulfates.
5.3.2.3 Reactions With Ground and Water Surfaces
In a study by Spedding (1972), the ocean was found to be a major sink for
SO2. He concluded that SC>2 deposition velocities (Vg = deposition/surface
area/time of exposure/atmospheric concentration) were proportional to the flow
rate of the S02~air mixture. Under calm conditions, he estimated a value of
Vg of 0.28 cm/sec. Owers and Powell (1974) estimated an SC>2 deposition velo-
city of 0.8 cm/sec over land and water surfaces. Similar values were found
by Shepard (1974) for grass in summer but were much less in the autumn (0.3
100
-------
cm/sec). Over water he found that Vg was proportional to windspeed. Gar-
land, et al. (1974) reported a deposition velocity of 0.55 cm/sec onto short
grass. Individual values varied widely and were independent of the weather.
He also estimated that 25 percent of the SC>2 emitted within Great Britain was
deposited by dry deposition.
5.3.2.4 Residence Times and Half-Life
Residence times and half-lives of S02 are extremely variable because of
the very complex interactions of SO2 with other reactive pollutants, atmo-
spheric liquid and gaseous water, surface water, land and vegetative surfaces,
sunshine, and weather. Eliassen and Saltbones (1975) reported a residence
time for S02 of about one-half day, or a decay rate of about .002 percent/sec.
In their work, they considered dry deposition and oxidation to sulfates.
In general, the various oxidation/deposition rates for SC>2 as summarized
above correspond to half-lives ranging from about one hour to several days.
The shorter half-lives are probably characteristic of urban SO2 where suffi-
cient quantities of reacting pollutants exist to hasten the transformation
process.
Figure 5-13 is a schematic showing SC>2 transport, diffusion, and various
removal processes.
Stack Gas
Plume
Distance from Source
FIGURE 5-13.
Processes involved in the relationship
of sulfur oxide emissions to air quality
(taken from HAS, 1975).
101
-------
5.3.2.5 SO2 Reactivity and Monitoring Device Inlet Tubes
Because of the adsorption-desorption characteristics of SC>2, care must
be taken in choosing the kinds of inlet tubes for SC>2 monitoring devices.
Stainless steel, glass, and teflon have been shown (Wohlers, et al., 1967) to
be nonreactive for either high flow (28.3 1/min) or low flow (1.4 1/min) sam-
pling through 30.5 m lengths (1.27 cm i.d.) of tubing made from such material.
5.3.3 Relevance of Above Considerations to Siting Criteria
Ambient temperature levels dictate the amount of fuel to be consumed for
space heat. Since the coldest temperatures occur in winter, the prevailing
wind direction for the three core winter months (December, January, and Febru-
ary) was considered as the "upwind" direction for determining the sources most
frequently upwind of prospective monitoring sites.
The very complex problem regarding the chemical/physical interactions of
SC>2 was addressed by assuming that S02 decays exponentially with a half-life
of three hours, generally, but one hour for cities of over 105 population.
The population figure is somewhat arbitrary, but the objective was to address
the fact that the chemical conversion of SC>2 in the largest urban areas pro-
ceeds at a faster rate than in rural areas. Using an appropriate half-life
value in all modeling exercises is important, particularly for assessing point
sources in urban settings and in the process of determining the sizes of pro-
jected growth or population areas to be represented by either neighborhood or
middle-scale stations (see Section 4.3.2). In the former, the contribution
of S02 at the monitoring site due to the source is a function of transit time
(and half-life) as well as distance; in the latter, the concentration gradient
is a function of half-life. In the same section (4.3.2), the 0.5 yg/m3 - km
gradient value is arbitrary; it was considered a realistic threshold value
separating steep and flat concentration gradients and was chosen via inspec-
tion of SC>2 concentration maps. Also, in the same section, the extreme con-
centration value span of +_ 25 percent of the mean concentration over an area,
for determining whether one monitoring site will represent concentrations over
the area, is also arbitrary. However, this value, considered as reasonable
for most purposes as given, could be adjusted to satisfy any purpose at the
discretion of the site selector.
The remaining siting and inlet placement criteria of Table 4-3 (see Page
39) and Table 4-5 (see Page 45) not specifically addressed above are consis-
tent with those found in existing guidelines (e.g., EPA, 1971, 1974b.) . They
are recommended mainly to prevent contamination of the instrument from dust
sources on the roof of the site building or to insure a proper exposure to the
open ambient air stream.
102
-------
6.0 REFERENCES
AEC, 1972; Safety Guide 23, On-Site Meteorological Programs, Atomic Energy
Commission (AEC), Washington, D.C., 23.1-23.6.
Anderson, G.E., 1971: Mesoscale influences on windfields. J. Appl. Meteor.,
10, 377-386.
, 1973: A Mesoscale Windfield Analysis of the Los Angeles Basin, EPA
Contract No. 68-02-0223, EPA, National Environmental Research Center,
Meteorological Laboratory, Research Triangle Park, N.C., 42 pp.
ASME, 1968: Recommended Guide for the Prediction of the Dispersion of Airborne
Effluents, ASME Committee on Air Poll. Cont., United Engineering Center,
New York, N.Y., 85 pp.
Ball, R.J., 1969: Interstate Transport of Air Pollution in Southwestern
Connecticut, in-house report, Department of Health, State of Connecticut,
Hartford, Connecticut, 94 pp.
, et al., 1972: Air Quality Implementation Plan, in-house report, De-
partment of Environmental Protection, State of Connecticut, Hartford,
Connecticut, 320 pp.
Bornstein, R.D., 1968: Observations of the urban heat island effect in New
York City. J. Appl. Meteor., 7, 575-582.
Bowne, N.E., 1973: Diffusion Rates, Paper #73-130 presented at the 66th Annual
Meeting of the Air Poll. Control Assoc., Chicago, Illinois, June 1973,
18 pp.
Bracewell, J.M. and D. Gall, 1967: The Catalytic Oxidation of Sulfur Dioxide
in Solution at. Concentrations Occurring in Fog Droplets. Proc,3 Symp.
Physiao-Chemical Trans, of Sulfur Compounds in the Atmos. and the Forma-
tion of Acid Smogs, Organization for Economic Cooperation and Development,
Mainz, Germany, June, 1967.
Brier, G.w,, 1973: Validity of the Air Quality Display Model Calibration Pro-
cedure, EPA-R4-73-017, Office of Research and Monitoring, EPA, Research
Triangle Park, N.C., 28 pp.
, 1975: Statistical Questions Relating to the Validation of Air Quality
Simulation Models, EPA 650/4-75-010, National Environmental Research Cen-
ter, Meteorology, Lab., EPA, Research Triangle Park, N.C., 21 pp.
103
-------
Cavender, J.H., D.S. Kircher, and A.J. Hoffman, 1973: Nationwide Air Pollu-
tion Emission Trends^ 1940-1970, in-house report, Pub., AP-115, Office
of Air and Water Prog., EPA, Research Triangle Park, Jfl.C., 52 pp.
Chandler, T.J., 1968: The bearing of the urban temperature field upon urban
pollution patterns. Atmos. Environ., 2, 619-620.
Clifton, M., D. Kerridge, W. Moulds, J. Pemberton, and J.K. Donoghue, 1959:
The reliability of air pollution measurements in the rleation to the
siting of instruments. Int'l. J. Air and Water Poll., 2, 188-196.
Collins, G.F., 1971: Predicting sea breeze fumigation from tall stacks at
coastal stations. Nuclear Safety, 12, 110-114.
DeMarrais, G.A., 1961: Vertical temperature difference observed over an urban
area. Bull. AMS, 42, 548-554.
DHEW, 1967: Technical Report: New York - New Jersey Air Pollution Abatement
Activity; Sulfur Compounds and Carbon Monoxide, in-house report, Dept.
Health, Education, and Welfare, DHEW, PHS, National Center for Air Pollu-
tion Control, Cincinnatti, Ohio, 59-75.
Drivas, P.J. and Shair, F.H., 1974: Probing the air flow within the wake down-
wind of a building by means of a tracer technique. Atmos. Environ., 8,
1165-1175.
Eliassen, A. and J. Saltbones, 1975: Decay and transformation rates of SO2 as
estimated from emission data, trajectories and measured air concentra-
tions. Atmos. Environ., 9, 425-430.
EPA, 1971: Guidelines: Air Quality Surveillance Networks, in-house report,
Publ. AP-98, Office of Air Programs, EPA, Research Triangle Park, N.C.,
16 pp.
, 1973a: Monitoring and Air Quality Trends Eeport3 1972, in-house report,
Publ. EPA 450/1-73-004, Office of Air Quality Planning and Standards,
EPA, Research Triangle Park, N.C., 210 pp.
, 1973b: The National Air Monitoring Program: Air Quality and Emissions
Trends. Annual Eeport3 Vol. II, in-house Report, Publ. EPA 450/1-73-
0016, Office of Air Quality Planning and Standards, EPA, Research Tri-
angle Park, N.C., 350 pp.
, 1974a: Guidelines for Air Quality Maintenance Planning and Analysis,
Vols. 1-12, Office of Air Quality Planning and Standards, Monitoring
and Data Analysis Div., EPA, Research Triangle Park, N.C.
, 1974b: Guidance for Air Quality Monitoring Network Design and Instru-
ment Siting, in-house Report, OAQPS No. 1.2-012, Office of Air Quality
Planning and Standards, EPA, Research Triangle Park, N.C., 33 pp.
104
-------
, 1974c: Guidelines for Air Quality Maintenance Planning and Analysis:
Vol. 10, Reviewing flew Stationary Sources,' EPA Contract 68-02-1094,
Office of Air Quality Planning and Standards, EPA, Research Triangle
Park, N.C., 78 pp.
, 1974d: Monitoring and Air Quality Trends Report, 1973, in-house report,
Publ. EPA 450/1-74-007, Office of Air Quality Planning and Standards,
EPA, Research Triangle Park, N.C., 312 pp.
, 1974e: Guidelines for Air Quality Maintenance Planning and Analysis:
Vol. 12, Applying Atmospheric Simulation Models to Air Quality Mainten-
ance Areas, in-house report, Publ. OAQPS No. 1.2-031, Office of Air
Quality Planning and Standards, EPA, Research Triangle Park, N.C., 42 pp.
Federal Power Commission, 1971: The 1970 National Power Survey, Part 4, U.S.
Government Printing Office, Washington, D.C.
Federal Register, 1971: Requirements for Preparation, Adoption, and Submittal
of Implementation Plans, 36(158), Saturday, August 14, 1971.
, 1973a: Maintenance of National Ambient Air Quality Standards, 35(116),
Monday, June 18, 1973.
, 1973b: Submission of Transportation and/or Land Use Plans, 35(52),
Tuesday, March 20, 1973.
, 1973c: Use of Supplementary Control Systems and Implementation of Se-
condary Standards, 35(178), Friday, September 14, 1973.
, 1974: Prevention of Significant Air Quality Deterioration, 35(235),
Thursday, December 5, 1974.
Flemming, G., 1967: Concerning the effect of terrain configuration on smoke
dispersal. Atmos. Environ., 1, 239-252.
Garland, J.A., D.H.F. Atkins, C.J. Readings, and S.J. Caughey, 1974: Deposi-
tion of gaseous sulphur dioxide to the ground. Atmos. Environ., 8,
75-80.
Gill, G.C., L.E. Olsson, J. Sela, and M. Suda, 1967: Accuracy of wind measure-
ments on towers or stacks. Bull. AMS, 48, 665-674.
Halitsky, J,, 1962: Diffusion of vented gas around buildings. J. Air Poll.
Control Assoc., 12, 74-80.
Hawkins, J.E. and G. Nonhebel, 1955: Chimneys and the dispersal of smoke. J.
Inst. Fuel, 28, 530-545.
Heller, A.N. and E.F. Ferrand, 1969: The Aerometric Network of the City of
New York, in-house report, Dept. of Air Resources, The City of New York,
N.Y., 25 pp.
105
-------
Herrick, R.A., 1966: Recommended standard method for continuing dust fall
survey (APM-1, Revision 1). J. Air Poll. Control Assoe,, 16, 372-377.
Hewson, E.W. and G.C. Gill, 1944: Meteorological investigation in Columbia
River Valley near Trail, B.C. U.S. Bureau of Mines Bull., 453, 23-228.
, E.W. Bierly, and G.C. Gill, 1961: Topographic influences on the be-
havior of stack effluents. Proa, of the Am. Power Conf., XXIII, 358-
370.
Hinds, W.T., 1970: Diffusion over coastal mountains of southern California.
Atmos. Environ., 4, 107-124.
Hino, M., 1968: Computer experiment on smoke diffusion over complicated topo-
graphy. Atmos. Environ., 2, 541-558.
Hosier, C.R., 1975: The meteorology program of the Environmental Protection
Agency. Bull. AMSt 56, 1261-1270.
Hotchkiss, P.S. and F.H. Harlow, 1973: Air Pollution Transport in Street Can-
yons, EPA Interagency Agreement No. EPA-IAG-0122(D), U. of California,
Los Alamos Scientific Lab., Los Alamos, N.M., 100 pp.
Hutcheon, R.J., R.H. Johnson, W.P. Loury, C.H. Black and D. Hadley, 1967: Ob-
servations of the urban heat island in a small city. Bull. AMS, 48, 7-9.
Jepson, A.F., and Weil, J.C., 1973: Maryland Power Plant Air Monitoring Pro-
gram Preliminary Results, Paper #73-157 presented at the 66th Annual
Meeting of the Air Poll. Control Assoc., Chicago, Illinois, June 1973,
19 pp.
Jutze, G. and E. Tabor, 1963: The continuous air monitoring program. J. Air
Poll. Control Assoc., 13, 278-280.
Keagy, D.M., W.W. Stalker, C.E. Zimmer, and R.C. Dickerson, 1961: Sampling
station and time requirements for urban air pollution survey. Part I:
Lead Perioxide Candles and Dustfall Collectors. J. Air Poll. Control
ASSOG., 11, 270-280.
Earsen, R.I., W.W. Stalker, and C.R. Claydon, 1961: The radial distribution
of sulfur dioxide source strength and concentration in Nashville. J. Air
Poll. Control Assoc., 11, 529-534.
Leahey, D.M., 1974: A study of air flow over irregular terrain. Atmos.
Environ., 8, 783-791.
Leavitt, J.M., F. Pooler, Jr., and R.C. Wanta, 1957: Design and interim Mete-
orological evaluation of a community network for meteorological and air
quality measurements. J. Air Poll. Control Assoc., 7, 211-215.
106
-------
Liberti, A. and G. Devitofrancesco, 1967: Evaluation of sulfur compounds in
Atmospheric Dust. Proa. Symp. Fhysioo-~chemical Trans, of Sulfur Compounds
in the Atmos. and the Formation of Aaid Smogs, organization for Economic
Cooperation and Development, Mainz, Germany, June, 1967.
Ludwig, F.L. and J.H.S. Kealoha, 1975: Selecting Sites for Carbon Monoxide
Monitoring, EPA Contract 68-02-1471, Environmental Protection Agency,
Research Triangle Park, N.C., 149 pp.
Martin, D.O., P.A. Humphrey, and J.L. Dicke, 1967: Interstate Air Pollution
Study, Phase II Project Report, V. Meteorology and Topography, in-house
report, U.S. Dept. of Health, Education and Welfare, PHS, National Center
for Air Pollution Control, Cincinnati, Ohio, 42 pp.
Mitchell, M.J., 1961: The temperature of cities. Weatherwise, 14, 224-229.
Montgomery, T.L., J.W. Frey, and W.B. Morris, 1975: Intennittant control sys-
tems. Environ. Sci. and Tech,, 9, 528~533.
Morgenstern, P. and K.A. Hagg, 1972: A system for abatement control strategy
evaluation. J. Air Poll. Control Assoc., 22, 774-778.
Munn, R.E. and I.M. Stewart, 1967: The use of meteorological towers in urban
air pollution programs. J. Air Poll. Control Assoc., 17, 98-101.
NAPCA, 1970: Air Quality Criteria for Sulfur Oxides, in-house report, AP-50,
National Air Pollution Control Administration, U.S. Dept. of Health, Edu-
cation and Welfare, Washington, D.C., 178 pp.
NAS, 1975: Air Quality and Stationary Source Emission Control, prepared for
the Committee on Public Works pursuant to S. Rev. 135; Serial No. 94-4,
U.S. Gov't. Printing Off., Washington, D.C., 909 pp.
Oke, T.R., 1973: City size and the urban heat island. Atmos. Environ., 7,
769-779.
, 1975: Urban heat island dynamics in Montreal and Vancouver. Atmos.
Environ., 9, 191-200.
Ott, w.R., 1975: Development of Criteria for Siting Air Monitoring Stations,
Paper # 75-14.2, presented at the 68th Annual Meeting of the Air Pollu-
tion Control Assoc., Boston, Mass, June 1975.
Owers, M.J. and A.W. Powell, 1974: Deposition velocity of sulpher dioxide on
land and water surfaces using a ^S tracer method. Atmos. Environ., 8,
67-68.
Pasquill, P., 1961: The estimation of the dispersion of windborne material.
Meteorol. Mag. (London), 90, 33-49.
Paulus, J.J. and A.T. Rassano, 1973: Siting of Air Quality and Meteorological
Monitoring Stations to Investigate Air Quality Effects of a Point Source.
3rd Int'l Ocean Air Cong., Duesseldorf, W. Germany, 133-136.
107
-------
Pelletier, J., 1963: Difficulties encountered in the measurement of air pollu-
tion and in the interpretation of results. Int'1. J, Air & Water Poll.,
7, 973-978.
Peters, L.K., 1975: On the criteria for the occurrence of fumigation inland
from a large lake. Atmos, Environ,, 9, 809-816.
Petersen, J.P., 1969: 'The Climate of Cities: A Survey of Recent Literature,
in-house report, Publ. No. AP-59, National Air Pollution Control Admin.,
U.S. Dept. of Health, Education and Welfare, Raleigh, N.C., 48 pp.
Pooler, Jr., F., 1963: Airflow over a city in terrain of moderate relief. J
Appl. Meteor., 2, 446-456.
, 1974: Network requirements for the St. Louis regional air polltuion
study. J. Air Poll. Control Assoc., 24, 228-231.
Roberts, J.J., E.J. Croke, A,S. Kennedy, J.E. Norco, and L.A. Conley, 1970:
Chicago Air Pollution System Analysis Program: A Multiple-Source Urban
Atmospheric Dispersion Model, in-house report, Publ. # ANL/ES-CC-007,
Argonne National Laboratory, Argonne, Illinois, 148 pp.
Robinson, E. and R.C. Robbins, 1968: Sources3 Abundance and Fate of Gaseous
Atmospheric Pollutants, SRI Report No. PR-6755, Stanford Research Insti-
tute, Menlo Park, Calif., 127 pp.
and , 1970: Gaseous sulfur pollutants from urban and natural sources.
J. Air Poll. Control Assoc., 20, 233-235.
Rossano, A.T., 1956: The joint city, county, state, and federal study of air
pollution in Louisville. J'. Air Poll. Control Assoc. , 6, 176-181.
Shepard, J.G., 1974: Measurements of the direct deposition of sulphur dioxide
onto grass and water by the profile method. Atmos. Environ., 8, 69-74.
Sherlock, R.H. and E.A. stalker, 1941: A Study of Flow Phenomena in the Wake
of Smoke Stacks. Bull. #29, Dept. of Engineering Res., U. of Michigan,
Ann Arbor, Michigan.
Simon, C., 1969: New York City's Meteorological Program, presented at the Mid-
Atlantic States Section, APCA, Semi-Annual Tech. Conf., Philadelphia, Pa.,
March 21, 1969, 9 pp.
Slade, D.H.(editor), 1968: Meteorology and Atomic Energy3 1968, Publ. No. TID-
24190, Air Resources Labs., Environ. Sci. Serv., Admin. (ESSA), U.S. Dept.
of Commerce, Washington, D.C., 445 pp.
Smith, D.B., 1968: Tracer study in an urban valley. J. Air Poll. Control
Assoc. , 18, 600-604.
Spedding, D.J., 1972: Surphur dioxide absorption by sea water. Atmos. Environ.,
6, 583-586.
108
-------
Spicer, C.W., J.L. Gemma, D.W. Joseph, P.R. Stricksel, and G.F. Ward, 1976:
The Transport of Oxidant Beyond Urban Areas, EPA Contract 68-02-1714,
Off. of Res. and Dev., ESRL, U.S. Environmental Protection Agency,
Research Triangle Park, N.C., 235 pp.
Start, G.E., C.R. Dickson, and L.L. Wendell, 1973: Diffusion in a Canyon with-
in Rough Mountainous Terrain, in-house report, Publ. # NOAA-TM-ERL-ARL-
38, Environ. Res. Labs,, Air Resources Lab., NOAA, Idaho Falls, Id.,
43 pp.
, N.R. Ricks, C.R. Dickson, 1974: Effluent Dilutions Over Mountainous
Terrain, NOAA Tech. Memo., ERL ARL-51, Environ. Res. Labs., Air Resources
Lab., NOAA, Idaho Falls, Id.,. 162 pp.
Stasiuk, N.W. and P.E. Coffey, 1975: Evidence of atmopsheric transport of
ozone into urban areas. Environ. Sci. and Tech., 9, 59-62.
Stern, A.C., H.C. Wohlers, R.W. Boubel, and W.P. Lowry, 1973: Fundamentals of
Air Pollution, Academic Press, New York, N.Y., 492 pp.
Stockton, E.L., 1970: Experience with a computer oriented air monitoring pro-
gram. J. Air Poll. Control Assoo., 20, 456-460.
Turner, D.B., 1968: The diurnal and day-to-day variations of fuel usage for
space heating in St. Louis, Missouri. Atmos. Environ., 2, 339-351.
, 1974: Workbook of Atmospheric Dispersion Estimates, in-house report,
7th printing, Publ... # AP-26, Office of Air Programs, U.S. Environmental
Protection Agency, Research Triangle Park, N.C., 84 pp.
Urone, P., H. Lutsep, C.M. Noyes, and J.F. Parcher, 1968: Static studies of
sulfur dioxide reactions in air. Environ. Sci. Tech., 2, 611-618.
U.S. Geological Survey, 1969: Topographic Maps, USGS, Washingtpn, D.C., 22 pp.
, 1974: Orthophotoquad Index, USGS, Washington, D.C., (chart).
Van der Hoven, I., 1967: Atmospheric transport and diffusion at coastal sites.
Nuclear Safety, 8, 490-499.
Wohlers, H.C., N.M. Trieff, H. Newstein, and W. Stevens, 1967: Sulfur dioxide
adsorption on - and desorption from teflon, tygon, glass, stainless steel,
and aluminum tubings. Atmos. Environ., 1, 121-130.
Yamada, V.M., 1970: Current practices in the siting and physical design of
continuous air monitoring stations. J. Air Poll, Control Assoc., 20 ,
209-213.
109
-------
APPENDIX*
SOURCES OF CLIMATOLOGICAL AND
METEOROLOGICAL INFORMATION
(adapted from Ludwig and Kealoha, 1975)
* All references found in this appendix are listed in Section 6.0 (References)
of the main text.
-------
A. SOURCES OF CLIMATOLOGICAL AND METEOROLOGICAL INFORMATION
One of the most helpful publications specifically designed to assist po-
tential users of climatological data is called "Selective Guide to Climatic
Data Sources," Key to Meteorological Records Documentation Number 4.11, pre-
pared by the staff at the National Climatic Center, Ashville, N.C., for sale
by the Government Printing Office, Washington, D.C. Its format indicates the
publication(s) in which various climatological categories (temperature, pre-
cipitation, wind, humidity, and so on) may be found. Although this publica-
tion refers primarily to published climatological data, a wealth of unpublished
climatological summaries are also available on special order from the files of
the National Climatic Center. An index to the summaries that can be ordered
is given in the "Guide to Standard Weather Summaries," NAVAIR 50-IC-534, U.S.
Navy, March 1968.
The National Climatic Center makes every effort to obtain a copy of all
meteorological records collected in the United States. These data are avail-
able and can be ordered on microfilm, magnetic tape, hard copies, or as copies
of raw data. The address and phone number are:
Director, National Climatic Center
Federal Building
Ashville, North Carolina 28801
Telephone: (704) 258-2850
The Center answers inquiries and analyzes, evaluates, and interprets data.
Routine letters or telephone inquiries are usually answered without charge;
other services are provided at cost.
The bulk of the data at the Climatic Center are meteorological observa-
tions made at airfields by the National Weather Service, the Federal Aviation
Administration, and the Defense Department. Table A-l shows an example of the
kind of information to be found on a three-hourly tabulation for one month at
one station. Climatic information is seldom available to the extent that it
will be desired, but ingenuity can often be used to ferret out sources not
generally available from the usual public data repositories.
At the State and regional level, fire stations, highway and transportation
departments, environmental studies groups, air pollution districts, and utility
districts may have continuing meteorological records or special weather studies
available. A call directly to these agencies may result in a data source not
available anywhere else.
A-l
------- An error occurred while trying to OCR this image.
-------
Schools, colleges, industrial complexes (such as oil refineries), agricul-
tural research stations, radio-TV stations, and electrical power plants may co-
operate with a data collection program if asked.
The following publications provide important information concerning use-
ful data sources.
1) Air Pollution Control Association (1973-1974): Directory,
Government Air Pollution Agencies, published in cooperation
with the Office of Air Programs, EPA. This directory lists
federal, state, regional, and county agencies conducting air
pollution programs. Names of officials, titles, addresses,
and telephone numbers are given. Wirte to W.T. Beery, Editor,
Directory Governmental Air Pollution Agencies, Air Pollution
Control Association, 4400 5th Ave., Pittsburg, Pa. 15213.
2) World Weather Records, Smithsonian Misc. Collections, Vol. 79,
Publication 2913, Assembled and arranged for publication by
H.H. Clayton, published by the Smithsonian Institution, August
1927. This reference book contains monthly and annual means
of pressure, temperature, and totals of rainfall.
A more extensive collection consisting of climatological data for selected
airfields and for the climatic areas in which they are located has been com-
piled by the USAF Environmental Technical Application Center (ETAC), Building
159, Navy Yard Annex, Washington, D.c. 20333. This series is also being pub-
lished by the U.S. Naval Weather Serice, Navy Yard, Washington, D.C. 20390,
under the title, U.S. Naval Weather Service World-Wide Airfield Summaries.
Table A-2 lists the available volumes in this series. Volume VIII contains
summaries for the United States. Information requests should be addressed to:
The National Technical Information Service (NTIS)
Springfield, Virginia 22151.
3) The Climatic Atlas of the United States, 1968, is a comprehen-
sive series of monthly and annual analyses showing the national
distirbution of mean, normal, and/or extreme values of tempera-
ture, precipitation, wind, pressure, relative humidity, dewpoint,
sunshine, sky cover, heating degree days, solar radiation and
evaporation. It was prepared by the Environmental Data Service,
NOAA, U.S. Department of Commerce, for sale by the Superintendent
of Documents, Washington, D.C.
4) Mixing Heights, Wind Speedst and Potential for Urban Air Pollu-
tion Throughout the Contiguous United States, by George C.
Holzworth, illustrates seasonal and annual, morning and after-
noon mean mixing heights, wind speeds, and normalized pollutant
concentrations that were exceeded 10, 25, and 50 percent of the
time for specified city sizes. Copies of this report Office of
Air Programs Pub. No. AP-101) may be ordered from the Office of
Tech. Information & Publs., Off. of Air Programs, EPA, Research
Triangle Park, N.C. 27711.
A-3
-------
TABLE A-2
Published Volumes of World-Wide Airfield Summaries
Volume
I
II
III
IV
V
VI
VII
VIII
IX
X
NTIS
Name Accession No.
Southeast Asia (revised) AD-706-355
Middle
Part 1
Part 2
East
AD-662-425
AD-622-427
Far East AD-662-426
Canada -Greenland-Iceland AD-662-424
Australia-Antarctica (including S. Pacific Is.) AD-662-648
South America
Part 1
Part 2
AD-664-828
AD-664-829
Central America AD-671-845
United
Part 1
Part 2
Part 3
Part 4
Part 5
Part 6
Part 7
Part 8
Africa
Part 1
Part 2
Europe
Part 1
Part 2
Part 3
Part 4
States of America
W. Coast, Western Mtns., & Great Basin AD-688-472
. Rocky Mtns. and Northwest Basin AD-689-792
. Central Plains AD-693-491
. Great Lakes AD-696-971
. Mississippi Valley AD-699-917
. Southeastern Region AD-701-719
East Coast and Appalachis^i Region AD-703-606
. Alaska and Hawaii AD-704-607
. Northern Half AD-682-915
. Southern Half AD-682-915
Scandinavia and Northern Europe
Low Countries and British Isles
Alps and Southwest Europe
Mediterranean
The National Climatic Center will prepare special data summaries. They
also have standard computer programs available for special summaries. One of
the most useful summaries for air pollution studies is that prepared by the
STAR program. It is a joint frequency distribution of atmospheric stability
and wind speed and direction. The atmospheric stability is calculated objec-
tively from the cloud cover and wind data. This stability algorithm is based
A-4
-------
on the work of Pasquill (1961). The summaries can be based on any extended
period of record with separate outputs for the months or seasons, as well as
an annual summary. There are some pollution models that use the output of
STAR program as part of their input requirements. The National climatic Cen-
ter has computed these summaries for over 250 weather stations in the United
States. These stations are listed in Table A-3,
TABLE A-3
List of Stations for Which Stability-Wind-Rose
Tables have been Prepared*
abc
abc
abe
abc
abe
abe
abc
abc
abc
ab
ab
abc
abc
ac
abc
abc
abc
abe
abc
abc
abc
ab
abc
abe
abe
abc
abc
ab
abc
abe
abc
abc
abc
abc
abc
abc
abe
Scat Ion* for
ftTABAtiV
Birmingham
Huntavllle
Mobile
Montgomery
ALASKA
Anchorage
Anne tee
Barrow
Barter Island
Bethel
Bettles
Big Delta
Cold Bay
Fairbanks
Farewell
Gulkana
Homer
Ulanou
Juneau
King Salmon
Kotzebue
McGrath
Nenana
None
St. Paul Island
Shemya
Sum'.'.
Talkeetna
Tanana
Onalakleet
Yakutat
ARIZONA
Flagstaff
Phoenix
Tucson
Vlnslov
Yunta
ARKANSAS
Fort Smith
Little Rock
C O N T I N
which Local Cllnatologieal Data are issued, aa of January 1, 1969
abc
ac
ac
ac
abc
abc
abc
ac
ae
abc
abc
abc
ac
be
abc
ac
ac
abe
ac
abc
abc
abc
abc
abc
abc
ac
abe
CALIFORNIA
Bakersfield
Bishop
Blue Canyon
Eureka
Fresno
Long Beach
Los Angeles
Airport
Civic Center
Mt. Shasta
Oakland
Red Bluff
Sacramento
Sand berg
San Diego
San Francisco
Airport
City
Santa Maria
Stockton
COLORADO
Alamo sa
Colorado Springs
Denver
Grand Junction
Pueblo
CONNECTICUT
Bridgeport
Hartford
New Haven
DELAWARE
Wilmington
DISTRICT OF COLUMBIA
abc Washington
U E D
ae
abe
abc
abc
abe
ac
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
ac
abe
k<»
ADC
abc
abc
abc
abc
ac
abe
ac
abe
ab
abc
abe
abc
abc
abc
abc
abc
abc
FLORIDA
Apalachicola
Daytona Beach
Fort Myers
Jacksonville
Key West
Lakeland
Miami
Orlando
Pensacola
Tallahassee
Tampa
West Palm Beach
GEORGIA
Athens
Atlanta
Augusta
Columbus
Macon
Rome
Savannah
HAWAII
Hilo
Honolulu
Kahulul
Lihue
IDAHO
Boise
Lewlston
Pocatello
ILLINOIS
Cairo
Chicago
Midway Airport
O'Hare Airport
Moline
Peoria
Rockford
Springfield
INDIANA
Evansville
Fort Wayne
Indianapolis
South Bend
abc
abc
ac
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
abe
abc
ac
abc
abc
abc
ac
abc
abc
abc
abc
abc
abc
abe
abc
abe
ac
abc
abc
IOWA
Burlington
Des Moines
Dubuque
Sioux City
Waterloo
KANSAS
Concordia
Dodge City
Good land
Topeka
Wichita
KENTUCKY
Lexington
Louisville
LOUISIANA
Alexandria
Baton Rouge
Lake Charles
New Orleans
Shreveport
MAINE
Caribou
Portland
MARYLAND
Baltimore
MASSACHUSETTS
Boston
Airport
Blue Hill Obs.
Nantucket
Worcester
MICHIGAN
Alpena
Detroit
City Airport
Detroit Metro A?
Flint
Grand Rapids
Houghton Lake
Lansing
Marquette
Muskegon
Sault Ste. Marie
* Only those stations having at least the "To" indicated.
A-5
-------
Table A-3. Continued.
abc
abc
abc
abc
ac
t
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
e
abc
abe
ac
ac
abc
abc
abc
ac
abc
abc
abc
abc
abc
abc
ac
abc
a
abc
ac
abc
ac
abc
KWKESOTA
Duluth
International
Fa 11 a
Min'p'1's-St. Paul
Rochester
St. Cloud
MISSISSIPPI
Jackson
Meridian
MISSOURI
Columbia
Kansas City
St . Joseph
St. Louis
Springfield
MONTANA
Billings
Glasgow
Great Falls
Havre
Helena
Kalispell
Miles City
Missoula
NEBRASKA
Grand Island
Lincoln
Norfolk
North Platte
Omaha
Scottsbluff
Valentine
NEVADA
Elko
Ely
Las Vegas
Reno
Winnemucca
NEW HAMPSHIRE
Concord
Mt, Washington
NEW JERSEY
Atlantic City
Airport
State Marina
Newark
Trenton
NEW MEXICO
Albuquerque
Clayton
Roawell
abc
abc
abc
abc
abc
abc
abc
abe
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
ac
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
ac
abc
abc
abc
abc
ac
ac
abc
be
abc
abc
abc
abc
abc
abc
abc
HEW YORK
Albany
Blnghamton
Airport
Buffalo
New York
Central Park
Int'l. Airport
LaCuardia Field
Rochester
Syracuse
NORTH CAROLINA
Asheville
Cape Hatterai
Charlotte
Greensboro
Raleigh
Wilmington
NORTH DAKOTA
Bismarck
Fargo
Willlston
OHIO
Akron-Canton
Cincinnati
Abbe Obs.
Airport
Cleveland
Columbus
Dayton
Mansfield
Toledo
Youngstown
OKLAHOMA
Oklahoma City
Tulsa
OREGON
Astoria
Burns
Eugene
Meachan
Medford
Pendleton
Portland
Salem
Sexton Summit
PACIFIC ISLANDS
Guam
Johnston
Koror
Kwajalein
Majuro
Pago Pago
Ponape
Truk (Moen)
Wake
Yap
abc
abc
abc
abc
abc
ac
ac
abc
abc
ac
abc
abc
a
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
abc
a
ac
abc
abc
abc
abc
abc
abc
abc
abc
abc
ac
abc
abc
abc
abc
abc
abc
abc
abc
abc
PENNSYLVANIA
Allentovn
Erie
Harrisburg
Philadelphia
Pittsburgh
Airport
City
Reading
Scranton
Wllliamsport
RHODE ISLAND
Block Island
Providence
SOUTH CAROLINA
Charleston
Airport
City
Columbia
Greenville-
Spartanburg
SOUTH DAKOTA
Aberdeen
Huron
Rapid City
Sioux Falls
TENNESSEE
Bristol
Chattanooga
Knoxville
Memphis
Nashville
Oak Ridge
Area Stations
City
TEXAS
Abilene
Amarillo
Austin
Brownsville
Corpus Christ!
Dallas
Del Rio
El Paso
Fort Worth
Calves ton
Houston
Lubbock
Midland
Port Arthur
San Angelo
San Antonio
Vic tor ia
Waco
Wichita Falls
ac
abc
ac
abc
ac
abc
abc
abc
a
abc
abc
abc
abc
ac
ac
abc
abc
ac
abc
abc
ac
abc
ac
abc
abc
abc
abc
abc
abc
abc
abc
UTAH
Mil ford
Salt Laie City
Wend over
VERMONT
Burlington
VIRGINIA
Lynchburg
Norfolk
Richmond
Roanoke
Wallops Island
WASHINGTON
Olumpia
Quillayute Airport
Seattle-Tacoma AP
Spokane
Stampede Pass
Walla Walla
Yakima
WEST INDIES
San Juan, P. R.
Swan Island
WEST VIRGINIA
Beck ley
Charleston
Elk ins
Huntington
Parkersburg
WISCONSIN
Green Bay
La Crosse
Madison
Milwaukee
WYOMING
Casper
Cheyenne
Lander
Sheridan
a. Monthly summary issued.
b. Monthly summary includes 3-hourly observations.
c. Annual Summary issued.
A-6
-------
APPENDIX
B
Suggested Approaches for Determining Worst Case
SO Patterns and Associated Meteorology
PART I. Multi-Source Urban Settings
PART II. Isolated Point Source Monitoring
All references found in this appendix are listed in Section 6.0 (References)
of the main text.
-------
B. SUGGESTED APPROACHES FOR DETERMINING WORST CASE
S02 PATTERNS AND ASSOCIATED METEOROLOGY
This Appendix was prepared to serve as guidance for utilizing short-term
diffusion model simulations in selecting monitoring sites. Such simulations
may be required to determine approximate locations of monitoring sites for
measuring short-term peak concentrations under near worst case conditions.
Part I addresses the situation in multi-source urban settings. Part II ad-
dresses isolated point source monitoring.
It is strongly recommended that a diffusion meteorologist be consulted
in developing procedures for specific applications. Presented below are ap-
proaches that address the problem areas; they should not be construed as con-
stituting the only approaches.
B.I MULTI-SOURCE URBAN SETTINGS
This suggested approach utilizes the programming logic incorporated in
the AQDM computer program. What is presented here is not a detailed listing
of a program modification but a description of a suggested modification and
how such a modification, when used in conjunction with wind direction persis-
tence tables, can estimate the approximate locations where the short-term con-
centration peaks should occur. The program modifications themselves could be
accomplished quite easily with the aid of a diffusion meteorologist.
The AQDM utilizes emission rates from a set of pollutant sources and a
joint frequency distribution of wind speed, wind direction, and atmospheric
stability (stability-wind-rose, SWR). The total concentration at a specific
receptor is obtained by utilizing each element of the SWR to calculate the
partial concentration that a source contributes to the receptor and summing
over all sources. There are six wind speed classes each representing a wind
speed range, five stability classes, and sixteen wind directions for a total
of 480 elements or "cells" that comprise the SWR. However, normally only
about one-half of the cells are "filled" and only those that establish the
source as being upwind of a receptor are utilized.
Utilizing the winter quarter SWR (Dec-Jan-Feb) and making appropriate
changes in "print" or "write" statements, it would be possible to print 480
(or a number equal to the number of filled cells) maps of pollutant concentra-
tions, one map for each combination of wind speed range, wind direction, and
stability class. All 480 maps could be stored on tape and only those five or
so maps having the highest concentration peaks at any receptor need be printed
B-l
-------
out and analyzed. It is recommended that the SWR be "normalized" by changing
the indicated frequency of occurrence of each meteorological combination (cell)
to a constant value for all cells. This value could be "1" or the inverse of
the number of cells filled, etc. The purpose of this procedure is to elimin-
ate the bias due to relative frequencies of occurrence. For example, a high
frequency of occurrence'of a given combination could actually be composed of
many short periods, no individual period being characterized by a high concen-
tration; on the other hand, a low frequency of occurrence may consist of a sin-
gle long period of persistent wind direction resulting in a high concentration
peak. Therefore, the use of normalized meteorology along with wind direction
persistence information will more likely reveal the location of the peak con-
centrations. Another important point is that the distribution of sources,
source strengths and emission heights in combination with a unique meteoro-
logical condition results in the highest ground-level concentrations (the near
worst case meteorological condition).
The persistence of the wind direction over the worst case condition de-
termines the averaging time of the peak. Three basic time periods are recom-
mended to be considered. For the three-hour peak analyses, the daytime period
from 1000 to 1900 LST and the nighttime period (2000-0900LST) should be con-
sidered separately. For the 24-hour peak analysis, consider the 24-hour period
0000 to 2400 LST.
Wind direction persistence tables for the weather station of interest for
each of the three basic time periods may be requested from the National Clima-
tic Center or generated by a computer program using LCD data (wind), A sug-
gested format is shown in Table B-l, or graphically, in Figure B-l. Since ob-
servations are taken at 3-hour intervals, two consecutive observations having
the same wind direction would constitute a 3-hour persistence case; three con-
secutive observations, a 6-hour case; etc.
TABLE B-l
Tabulated Persistence Data for NW Wind Situation
(192 obs/yr, 7% of total wind obs)
NW Wind
Frequency (#)
Frequency (%)
(192 base)
Median Speed
Modal 2 Speed
(%)
Persistence (hours)
3
40
21
15
17
(50)
6
12
6
11
9
(30)
9
7
4
9
11
(60)
12 15 1
5 4
3 3
12
10
(25)
8 21 24
321
211
1 Wind speed that divides sample in half.
2 Most frequently occurring wind speed.
B-2
-------
100
90
80
Q. 70
T 60
r i
50
40
30
20
10
0
CD
W
sw
15/17(50)
NW
24 21 18 15 12 9 6 3
Wind Direction Persistence (hrs)
FIGURE B-l. Graphical presentation of persistence data, NW wind case from
Table B-l. Graphs may be prepared by season, day/night, etc.
B.II ISOLATED POINT SOURCE MONITORING
This approach also utilizes standard Gaussian diffusion concepts. It is
assumed that the reader is familiar with the various kinds of graphical solu-
tions to the Gaussian point source diffusion equation such as that shown in
Figure B-2. In any case, "Turner's Workbook" (Turner, 1974) contains most of
the support information and should be consulted (also, see Appendix E, Sec.E.3)
- FIGURE B-2.
Distance of maximum
concentration & maxi-
mum xu/Q as a func-
tion of stability
(curves) and effective
height (m) of emission
(numbers).
B-3
-------
The approach presented here is generally applicable to terrain settings
ranging from flat to moderately rough and irregular. Its application to rough
and irregular terrain can be accomplished by substituting appropriate diffu-
sion coefficients, such as those presented by Bowne (1973) for those valid
only for flat terrain in the Gaussian point source equation. In this manner,
graphical solutions can be prepared for various terrain settings. Suggested
approaches for dealing with the 3-hour and 24-hour averaging times are given
below.
B.II.l Three-Hour Peak Concentrations, Their Locations, and Associated
Meteorology
The critical wind speeds for each stability class (A, B, and C) are first
determined; these winds produce the highest ground-level concentrations down-
wind of the plant. The wind directions associated with the critical wind
speeds (see Table B-2 for example) establish the downwind azimuths along which
candidate siting areas are established. Graphical solutions to the Gaussian
equation in a form similar to Figure B-2 is used to determine their approxi-
mate distances downwind. Next, a decision has to be made regarding selecting
the final site or sites. Since stability Class A will usually be associated
with the highest peak but may not have the highest frequency of occurrence,
two sites may be appropriate:
a) one site associated with the most unstable stability class
to measure the highest peak;
b) the other site associated with a high peak that occurs very
often (see Table B-2, second column).
The meteorology associated with each of the above situations is the worst case
meteorological condition for those situations.
B.II.2 Twentyfour-Hour Peak Concentrations, Their Locations, and Associated
Meteorology
The situation associated with the 24-hour average impact is somewhat dif-
ferent than that associated with the 3-hour impact. The plant will probably
not be operating at its peak production rate for 24 hours and the atmospheric
stabilities may involve the full range of classes.
The following procedure is recommended:
Determine the critical wind speed for stability class D (repre-
sents 24-hour average stability); in the calculalation of ef-
fective stack height, stack parameters should reflect the
average daily emission rate. From wind persistence tables
(e.g., Figure B-l) representative of the area of interest,
ascertain the most persistent wind directions (e.g., the five
or so of longest duration. The median and modal speeds asso-
ciated with the directions are then noted. These speeds can
B-4
-------
TABLE B-2*
Illustration of Use of Stability-Wind-Rose for Determining Site Locations
Monitoring Isolated Point Sources
[Assume a critical wind speed of 2 knots; NNE direction from source: azimuth
for monitoring the highest peak concentration; NE direction from sources az-
imuth associated with frequently occurring high peaks. Use Fig. B-2 for de-
termining downwind distances. Verify via mobile sampling.]
Direction I
1-3
N 19
NNE 15
NE 6
ENE 2
E 7
ESE 2
SE 9
SSE 2
S 16
ssw - [29]
SW »- 22
WSW 15
W 12
WNW 15
NW 4
NNW 7
Average 2 . 6
Total 182
Frequency Distribution
Speed (knots)
t
4-6 7-10
35 15
29 8
11 3
9 4
10 4
6 3
4 1
3 0
14 4
34 22
J41J 13
13 7
22 5
18 13
18 15
12 6
5.2 7.3
250 123
11-16
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
17-21
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
>21
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Average
Speed
5.0
5.0
4.8
5.6
4.8
5.2
3.6
4.0
4.2
4.9
4.8
4.6
4.7
5.1
5.9
4.8
3.5
Total
69
52
21
15
21
11
14
5
34
85
76
35
39
46
37
25
Number of Occurrences of A Stability = 805
Number of Calms with A Stability =
221
* Station = 14703 Chicopee Falls, Mass 60-64 240B (adapted from NCC printout)
be used to characterize each persistence period; the modal speed should be used
if it occurs about one-third of the time or moreotherwise the median speed
may be more representative. Next, a judgemental decision must be made regard-
ing the selection of the probable direction associated with the highest concen-
tration. For this purpose, it can be assumed that the maximum concentrations
are associated with the longest persistence periods with characteristic speeds
nearest the critical wind speed. Therefore, if there are several persistence
periods exceeding 24 hours in duration, the direction whose characteristics
speed is closest to the critical wind speed may be deemed the worst case con-
dition. Otherwise, the most persistent direction, regardless of its character-
istic wind speed, will best represent such conditions, Graphical solutions to
the Gaussian equation similar to Figure B-2 may be used to determine the re-
quired distance downwind to the siting area.
B-5
-------
APPENDIX *
C
MOBILE SAMPLING
* All references found in this appendix are listed in Section 6.0 (References)
of the main text.
-------
c. MOBILE SAMPLING
Mobile sampling can satisfy several important needs in SO2 monitoring,
particularly those related to isolated point source impacts. Among them are:
to reveal ground-level SO2 patterns, especially over complex
terrain;
to determine the location of the maximum ground-level impact
point of a plume emitted by an elevated source; and
to define the spatial distribution of plume material.
Several kinds of instruments are available for mobile sampling of SC>2 -
Those based on Flame Photometric Detection (FPD) are usually sensitive and
fast-responding instruments. Also, remote sensing devices such as a correla-
tion spectrometer may be used to measure the total vertical burden (or inte-
gral of the pollutant) over a point. A small vehicle may be used to transport
the equipment.
Several sampling techniques can be utilized in mobile monitoring. Jepsen
and Weil (1973) describe a rather successful approach in the State of Maryland's
power plant impact studies. From their work, Figure C-la shows an idealized
mobile lab traverse route setup and Figure C-lb shows a typical plume disper-
sion pattern at ground-level and total burden in the vertical.
Martin, et al. (1967) describe a tracer study in which the spatial dis-
tribution of plume material is revealed by sampling at 30 to 50 sites along
several crosswind arcs downwind of the release point (see Figure C-2). The
crosswind dispersion parameter, ay can be computed from the data obtained
and used in conjunction with values of other key variables (plume height, wind
speed, etc.) to derive the vertical parameter, a . Mobile sampling tech-
niques can be utilized to accomplish the same thing by traversing along the
arcs instead of measuring at 30 to 50 stationary points along the arcs.
In complicated terrain, plume patterns become distorted and a regular
pattern may not be observed. Crosswind traversing is probably not possible
either. Therefore, for additional guidance in these situations, the vertical
sensing capability becomes more important; the total burden aloft over a point
can indicate whether high concentrations will be found farther downwind.
C-l
-------
Qi Plume Dispersal Area
(plan view)
stopped for ,
timed data ,'
source
0 km 2
Pfc Transportable lab
f"l Mobile lab
Mobile lab routes
Chalk Point Power Plant
15 Sept. 1972
FIGURE C-l. (a) Ideal plume and measurement
pattern. (b) Plume dispersion pattern at
Chalk Point power plant. (Adapted from
Jepsen and Weil, 1973.)
ground level SC>2
ill total burden SC^
route traversed
0 200
ppm-M
ppb
0 km 2
SEPTEMBER 12, 1963,1215-1315 C.D.T.
TRACER RELEASE
3XIO~'° Lower limit of significant data
FOREST
PARK
^
N
\
3X10
FIGURE C-2.
Ground-level pattern down-
wind of point source: Pat-
tern of relative concentra-
tion, in grams per cubic
meter (indicated by dots),
for a daytime tracer experi-
ment with steady winds
(concentrations based on
emission rate of 1 gram per
second).
(Taken from Martin, et al.,
1967.)
C-2
-------
APPENDIX *
SOURCES OF LAND-USE AND
TOPOGRAPHICAL INFORMATION
(adapted from Ludwig and Kealoha, 1975)
All references found in this appendix are listed in Section 6.0 (References)
of the main text.
-------
D. SOURCES OF LAND-USE AND TOPOGRAPHICAL INFORMATION
The extent and availability of land use data is dependent on the specific
area under study and what one chooses to call "land-use information". The
more formal information can be obtained at different levels of government.
Some states have developed sizable data bases as an aid in generalized state
planning (i.e., Connecticut, Florida, Hawaii, Maine, and Vermont). The major-
ity of states, however, are just beginning the information gathering process.
Land-use information for nonurban areas is best obtained from State Planning
and regional agencies.
Regional planning agencies (e.g., the Denver Council of Governments in
Colorado, Southeastern Virginia, Planning District Commission, and the Com-
prehensive Planning Organization in San Diego) can be excellent sources of
information. These regional agencies gather socioeconomic, existing land-use,
and transportation data. Comprehensive regional plans can then be prepared
to provide projections of long range demographic growth and land use.
Cities and counties will usually have current, readily available data on
population, employment, existing and projected land use, general development
plans, and zoning regulations. Also, they will be able to provide basic
transportation information and maps of major arterials and proposed thorough-
fares .
In cities with schools of urban and regional planning, planning profes-
sors can help the researcher meet specific needs. Also, their libraries can
be researched for applicable graduate and doctoral theses which are frequently
case studies of the immediate vicinity.
There are other sources of land-use information that are not specifically
directed to the topic. Good maps or aerial photographs can provide a lot of
useful information that may not be available from conventional land use sources.
Useful sources of information for the United States are discussed next.
D.I U.S. BUREAU OF THE CENSUS
Demographic and socioeconomic information of use to planners is available
from the Department of the Census. Data developed by census tracts can be
used to answer questions regarding a neighborhood's population and character-
istics. Census tract information can be outdated, so it should be supplemented
by material developed by the city, county, or regional planning bodies.
D-l
-------
D.2 U.S. GEOLOGICAL SURVEY (USGS)
D . 2 . 1 Topographic Maps
Topographic maps portray man-made and natural features, and the shape and
elevation of the terrain. The usefulness of topographic maps is revealed in
their accuracy, availability, economy, and wealth of detail. All maps are
classified according to scale. The map scale expresses a ratio between the
features shown on the map and the same features on the earth's surface. A
scale of 1:24,000 states that one unit on the map represents 24,000 units on
the ground. Figure D-l is an example of three maps scales of the same area
showing the type of information that is available in large, medium, and small
scale maps . Table D-l is a summary of the principal maps and their essential
characteristics . A booklet describing topographic maps and symbols is avail-
able free upon request from the Geological survey. To order maps of a speci-
fic area, first obtain a state index map by asking for the "Index to Topo-
graphic Maps of (state) . " An order form is included with each Index as well
as a list of local merchants that may stock topographic maps. Map reference
facilities are also maintained in many public libraries. All maps for areas
west of the Mississippi may be purchased by mail or over the counter from:
Distribution Section
U.S. Geological Survey
Federal Center
Denver, Colorado 80225
And for areas east of the Mississippi:
Distribution Section USGS
1200 S. Eads Street
Arlington, Virginia 33303
D.2.2 Photoimage Maps
Photoimage maps are available in the 1:24,000 scale. These are a ne.w
standard product called the orthophotoquad maps. An orthophotoquad portrays
by aerial photoimagery a wealth of detail not found in conventional line maps.
Yet there is the same positional accuracy as in standard topographic maps.
Orthophotoquads are reproduced in black and white as photographic, diazo, or
lithographic copies. Diazo or lithographic products are comparable in price
with 7.5 minute topographic maps. To obtain an index of orthophotoquad avail-
ability, contact the:
National Cartographic Information Center (703) 860-6045
U.S. Geological Survey
National Center, Stop 507
Reston, Virginia 22092
Figure D-2 shows a portion of the orthophotquad index, legend, and a
coded portion of the state of Florida (USGS, 1974).
D-2
-------
(a) LARGE
SCALE
1:24,000 scale,
1 inch = 2000 feet
Area shown,
1 sauare mile
(b) MEDIUM
SCALE
1:62,500 scale,
1 inch nearly 1 mile
Area shown,
6-3/4 square miles
(c) SMALL
SCALE
1:250.000 scale,
1 inch = nearly 4 miles
Area shown,
107 square miles
FIGURE D-l.
An example of the informational content of the large (a),
medium (b), and small scale (c) topographic map.
(Taken from USGS, 1969.)
D-3
-------
-ORTHOPHOTOQUAO INOCX
0«e*mb«r 1974
STATUS OF ORTHOPHOTOQUADS
Monocolor orthographic photograph or
photoao«aic In quadrangle fomac with
Btlninal cartographic enh«ne*n«nt*
So product or copy available.
Aerial Photography
Advance
High Altitude. noecly quad centered.
final copy (or approximately 90 percent
of orthophotoquads. (For ordering
in»truction* refer to state advance
Material index, available fro* appropriate
Happing Center at indicated below or froa
the National Cartographic Information
Center, I'. S. Geological Survey (507),
Rittoo, Virginia 22042.)
(For ordering inetructlona refer to state
index to topographic pup», available froa
the Branch of Distribution, U. S. Geological
Survey, 1200 S. Eeda Street, Arlington,
Virginia 22202 or Federal Ctntir, Denver,
Colorado 80225.)
FIGURE D-2.
A portion of the orthophotoquad index showing the legend and.
a portion of the state of Florida.
D-4
-------
TABLE D-l
National Topographic Maps
(taken from USGC, 1969)
Series
7'4-minute
Puerto Rico 7'i-minute
15-minute
Alaska 1:63.360
U.S. 1:250,000
U.S. 1:1.000.000
Scale
1:24.000
1:20.000
1:62.500
1:63.360
1:250.000
1:1,000.000
1 inch represents
2,000 feet
about 1,667 feet
nearly 1 mile
1 mile
nearly 4 miles
nearly 16 miles
Standard
quadrangle size
(latitude-longitude)
rixT'imin.
7'.ix7'4min.
15x15 min.
15x20 to 36 min.
M'x2'
34*x6*
Quadrangle
area
(square miles)
49 to 70
71
197 to 282
207 to 281
4.580 to 8,669
73,73410
102,759
Paper size
E-W N-S
width length
(inches)
'22x27
29'.ix32H
1 17x21
* 18x21
4 34x22
27x27
'South of latitude 31' "'^minute sheets are 23 x 27 inches: 15-minute sheets are 18 x 21 inches.
1 South of latitude 62* sheets are 17 x 21 inches.
1 Maps of Alaska and Hawaii vary from these standards,
4 North of latitude 42' sheets are 29 x 22 inches. Alaska sheets are 30 x 23 inches.
D.2.3 Earth Resources Technology Satellite
The Earth Resources Technology Satellite (ERTS) has the multi-spectral
sensors on board that "photograph" the earth's surface in the visible through
near-infrared range. The potential of such a capability for land-use mapping,
updating, and projection is currently a subject of extensive study. The im-
ages received from the satellite are available for sale as individual frames
each covering an area approximately 1000 x 100 nautical miles with a 10 per-
cent overlap along the spacecraft track. Table D-2 lists the picture products
available from ERTS.
TABLE D-2
Picture Products Available from ERTS
(adapted from EDCDM Form 6)
70
lh
15
30
Image Size
mm (contact size)
x 7^ inches
x 15 inches
x 30 inches
Scale
1:3,369,000
1:1,000,000
1:500,000
1:250,000
Material
Resin coated paper, film positive
or negative
Resin coated paper
Resin coated paper
Resin coated paper
For more information, contact the ERTS Data Center, Sioux Falls, S.D., tel.
(605) 339-2270. The ERTS Data Center has substantial holdings of images ac-
quired by aircraft throughout the United States. They invite inquire regarding
availability of suitable coverage of your area of interest (USGSEDC).
D-5
-------
S3
*< wr tut i xn
COUCH'S AOO'M
._..,-»* t_. N w HO-TT
(asp;
i-r:*T
51
g *: _3/
rt .,
'~i?J*.
N W CLISAN
lai
i
102
i; ./
* '
N.W EVESETT
i i_.
r- * I8!
*' r *i A !-s%
n-
[* i'
58
fitt.*
I '
1 ,»
I I f>
i '. I
i i *
! 2
*
e \
54
,?.'!.': N W FLANDERS«"-
<
*
-------
D.2.4 Sanborn Maps
Sanborn maps are land use maps prepared by the Sanborn Map Company. They
are plotted to a scale of 1 inch to either 50 or 100 feet showing details such
as streets, railroad tracks, lot lines, building dimensions, nature of the
building material, number of stories, height of the building, and use of the
building. Sanborn maps are used primarily for fire insurance purposes. Local
sources may be fire insurance offices, realtors offices, city planning, and
the county courthouse. For information, contact:
The Sanborn Map Company, Inc.
629 5th Avenue
Pelham, New York 10803
ATTN: G. Greeley Wells
Tel. (914) 738-1649
Figure D-3 is an example of a Sanborn map for a section of Portland,
Oregon.
D-7
-------
APPENDIX*
AVAILABLE AIR QUALITY SIMULATION
MODELS APPROPRIATE FOR S02
MONITOR SITING ACTIVITIES
* All references found in this appendix are listed in Section 6.0 (References)
of the main text.
-------
E. AVAILABLE AIR QUALITY SIMULATION MODELS APPROPRIATE
FOR S02 MONITOR SITING ACTIVITIES
This Appendix briefly describes several air quality simulation models
that are appropriate for use in determining approximate locations for S02 mon-
itors. Most have been widely used by government agencies and private groups
for routine source impact assessments and control strategy evaluations. The
model descriptions presented below were abstracted from user manuals or from
other sources as indicated.
E.I AIR QUALITY DISPLAY MODEL (abstracted from EPA, 1974e)
The Air Quality Display Model (AQDM) is a Gaussian plume urban dispersion
model which is best used to determine the impact of a wide variety of station-
ary source classes on annual average concentrations of SC>2 and TSP. Short-
term averages and single source assessments may also be obtained. The model
has been applied.to many multi-source urban areas with a high degree of suc-
cess.
The model is based on the standard long-term concentration equation:
x(x,e) = E
S N
exp
2Q f (e.S.N)
(E-l)
where f(0,S,N) is frequency during the period of interest that the wind is
from direction 9 , for the stability condition S, and the wind speed class
N ; and
'zS
u,
is the vertical dispersion parameter evaluated at the
distance X for the stability condition S ;
is the representative wind speed for class N ;
H is the effective height of release for the wind speed
UN .
E-l
-------
The model is used to determine the impact of all sources at a given receptor
for a given set of meteorological conditions. It then weights this concentra-
tion by the frequency with which that particular set of meteorological condi-
tions occurs and then sums over all meteorological conditions, thus producing
a long-term average concentration. Basic inputs to the model are a comprehen-
sive emissions inventory including both point sources and area sources. Mete-
orological input is a joint frequency distribution of wind speed (6 classes),
wind direction (16 cardinal points), and stability class (Pasquill classes
A-E) along with an annual average mixing height. The dispersion model can be
used to estimate concentrations at any point downwind that is specified. Up
to 1000 sources may be specified and concentrations may be calculated for up
to 237 receptors. Although the original model did not contain a decay term,
such a term can be easily incorporated. The AQDM produces a source contribu-
tion file which allows the impact of each individual source on air quality to
be obtained. The AQDM is available from the National Technical Information
Service (NTIS), NTIS PB 189 194.
E. 2 EPA SINGLE SOURCE MODEL (unpublished EPA computer program; Meteorological
Laboratory, 1972)
E.2.1 Basic Model
The model used to estimate ambient concentrations is one developed by the
Meteorology Laboratory, EPA. This model is designed to estimate concentra-
tions due to sources at a single location for averaging times from one hour
to one year.
This model is a Gaussian plume model using diffusion coefficients sug-
gested by Turner (1974). Concentrations are calculated for each hour of the
year, from observations of wind direction (in increments of 10 degrees), wind
speed, mixing height, and atmospheric stability. The atmospheric stability-
is derived by the Pasquill classification method as described by Turner (1974).
In the application of this model, all pollutants are considered to display the
dispersion behavior of non-reactive gases.
Meteorological data for 1964 are used as input to the model. The reasons
for this choice are: (a) data from earlier years did not have sufficient re-
solution in the wind direction; and (b) data from subsequent years are readily
available on magnetic tape only for every third hour.
Mixing height data are obtained from the twice-a-day upper air observa-
tions made at the most representative upper air station. Hourly mixing heights
are estimated by the model using an objective interpolation scheme.
Calculations are made for 180 receptors (at 36 azimuths and five select-
able distances from the source). The model used here can consider both diurnal
and seasonal variations in the source. Separate variation factors can be ap-
plied on a monthly basis to account for seasonal fluctuations and on a hourly
basis to account for diurnal variations. Another feature of the model is the
E-2
-------
ability to compute frequency distributions for concentrations of any averag-
ing period over the course of a year. Percentages of various ranges in pollu-
tant concentrations are calculated.
E.2.2 Complex Terrain Options
E.2.2.1 Terrain Elevations Below Plume Height
To simulate the effect of moderately elevated terrain in the vicinity of
a plant, an optional form of the basic model may be used. This optional mo-
deling program uses a terrain adjustment procedure which considers the differ-
ence between the plant elevation and the elevation at each receptor. Ground
elevations on 10 degree radials as well as points of maximum elevation are
determined from USGS quadrangle maps. The diffusion model then uses the dif-
ference between the plant elevation and receptor elevation to modify the ef-
fective stack height and thereby adjust the predicted concentrations.
E.2.2.2 Terrain Elevation Above Plume Height
In higher relief areas, the topography at certain plant sites is above
the calculated plume height for at least one stack at the plant. In this case,
an alternate model is used. The model used to estimate short-term concentra-
tions is these situations is one previously developed by EPA for application
to sources located in complex terrain (Valley Model). Elevations of receptor
sites are derived from the USGS quadrangle maps of the area. The model calcu-
lates a daily average concentration at these receptor locations based on a 10
meter nearest-approach point of the plume and an assured persistence of mete-
orological conditions for 6 hours out of the 24 hours. During this period,
the wind direction is assumed to be confined to a 22.5 degree sector. The
model assumes stability class "E" and a wind speed of 2.5 m/sec. The plume
is uniformly distributed horizontally over the 22.5 degree sector.
E.3 UNAMAP* MODELS PTMAX, PTDIS, AND PTMPT (abstracted from Hosier, 1975)
These models calculate hourly averaged concentrations resulting from
point source emissions. Each model provides a different set of options for
the user.
E.3.1 PTMAX
PTMAX is a computer program that performs an analysis of the maximum
short-term concentration from a single point source as a function of stability
and wind speed. A separate analysis is made for each stack. Required inputs
Users Network for Applied Models of Air Pollution, a library of air quality
simulation models stored at EPA's Computer Center at Research Triangle Park,
N.C.; also available on tape via OTIS (Accession Number PB 229 771).
E-3
-------
to the program include ambient air temperature, emission rate, physical stack
height, and stack temperature; either stack gas volume flow or both the stack
gas velocity and inside diameter at the top are required. The program com-
putes effective height of emission, maximum ground level concentration, and
distances of maximum concentration for each condition of stability and wind
speed.
E.3.2 PTDIS
PTDIS is a computer program that calculates downwind ground-level concen-
trations for various downwind distances for specified meteorology. Input re-
quirements include both source and meteorological conditions. The primary
output of the program consists of a table with height of emission, concentra-
tion for each downwind distance, and a relative concentration normalized for
wind speed and source strength. An optional feature of the program allows
the user to enter a value of concentration to be used for the determination
of half-width isopleths; for each distance, if the concentration exceeds the
stated isopleth value, the half-width of an isopleth will be determined. The>
half-width will be compared in the form of a ratio to the half-width of a sec-
tor of given angular size in terms of degrees. The user is given the option
of either specifying effective height of emission or having it calculated us-
ing Briggs' plume rise methods.
E.3.3 PTMPT
PTMTP calculates hourly concentrations at up to 30 receptors whose loca-
tions are specified from up to 25 point sources. Required inputs to the pro-
gram consist of the number of sources to be considered, the emission rate,
physical height, stack gas temperature, volume flow of stack gas velocity and
diameter, and the stack locations, in coordinates. The number of receptors,
the coordinates of each, and the height above ground are required. Concentra-
tions for a number of hours up to 24 can be estimated, and an average concen-
tration over this time period is calculated. The hourly meteorological infor-
mation required consists of wind direction and speed, stability class, mixing
height, and ambient air temperature.
E-4
-------
APPENDIX
F
BIBLIOGRAPHY
-------
APPENDIX F
BIBLIOGRAPHY
The literature search resulted in a collection of articles and reports on
most of the topics discussed in this report, as well as other related topics.
The bibliography is arranged alphabetically by author and numbered consecu-
tively. This arrangement provides a convenient means for relating each ele-
ment of the bibliography to a subject area, as shown in Table F-l. Reports
addressing more than one subject are are tallied accordingly. The biblio-
graphy is by no means exhaustive? it provides only a sampling of the relevant
citations used in the report and other related information sources.
F-l
-------
TABLE F-l
Bibliography Subject Reference
Subject Area
S02 sources: types and emission characteristics.
Air pollution regional climatology,
special problems.
Heat island phenomena and effects on
air pollution.
Ambient SO2 patterns from major source types
in a variety of physical sectir.gs and
weather conditions.
Diffusion/modeling studies of S02 ar.d turbulence
in a variety of physical and meteorological
environments .
SOj monitoring methods.
Supplementary control systems (SCS) and other con-
trol strategies: relationships to S02 monitoring.
Data handling/statistics including relationships
to continuous versus integrated S02 sampling.
Air monitoring networks and
instrument siting guidelines.
Information on environmental impact statements
(EIS) ; monitoring data requirements, nature,
and use of existing data.
Studies of topographical and building waJce
effects on point source pollutant plumes,
especially S02-
Sea-lake breeze influences on 802 plumes from
tall stacks and other configurations.
Bsergency episodes, especially SC>2-
Model calibration studies; relevance of
monitoring data.
Use of instrumented towers: nature and
interpretation of data resulting therefrom.
Environmental data sources.
Mobile sampling; particularly with respect to SC^-
General: characterization of S02 sources,
problems, land use planning, etc.
Reactions of SOj with other pollutants, natural
atmospheric constituents, land, vegetation,
water surfaces.
Number Associated with Bibliographic Element
24 40 59 103 156 160 167 185 191 196
13 59 36 S3 190 200 201
5 8 15 25 27 39 56 62 66 93 122 136 137 148 176
197
1 4 8 9 12 14 16 32 36 53 64 66 69 72 77 87 101
103 109 116 117 120 123 126 132 146 147 149 151
152 153 159 160 163 169 176 183 185 190 197 199
4 11 17 22 29 33 36 38 53 66 69 71 77 79 82 83
84 87 88 89 91 99 101 104 106 109 110 113 118
119 120 123 127 133 134 153 163 164 169 170 171
172 173 176 178 179 183 186 187 188 193 194 197
189 199 201
58 108 141 143 177
10 50 99 114 125 127
2 46 53 84 92 100 102 115 124 135 139
162 171
3 14 19 26 28 30 35 43 44 45 48 49 53 70 73 74
75 76 85 90 94 96 97 98 100 107 111 112 125 138
142 144 150 154 157 166 169 178 184 192 197 204
31 33 54 80 180
5 6 8 11 12 31 33 34 37 41 57 57 68 77 78 81 82
83 89 104 1C6 118 129 133 148 153 171 174 176
181 186 187 193 203
5 6 29 52 59 60 79 91 113 128 145 194
38 62 131 158 174
20 21 130 158 161
7 63 134
23 55 165
1 43 96
3 19 47 49 51 65 94 95 98 105 112 121 155 157
175 189 195 196
18 42 61 108 140 163 182 202
F-2
-------
1 AHLQUIST, N.C. and R.J. CHARLSQN, 1968: "Measurement of the Vertical and
Horizontal Profile of Aerosol Concentration in Urban Air with Inter-
grating Nephelometer," Env. Sci & Teoh., Vol. 5, pp. 363-366.
2 AKLAND, G.G., 1972: "Design of Sampling Schedules," J. Air Poll. Control
Assoo., Vol. 22, pp. 264-266.
3 ALLEN, P.W., 1973: "Regional Air Pollution Study. An Overview," present-
ed at the Air Poll. Control Assoc. Annual Meeting, 66th Chicago,
Illinois, Paper 73-72, 23 pp.
4 ANDERSON, G.E., R.R. HIPPLER, and G.D. ROBINSON, 1969: An Evaluation of
Dispersion Formulas, Final Report, prepared for The American Petroleum
Institute, The Travelers Research Corp., Hartford, Connecticut, 96 pp.
5 ANDERSON, G.E., 1971: "Mesoscale Influences on Wind Fields," J. Appl.
Meteor., Vol. 10, pp. 377-386.
6 ANDERSON, G.E., 1973: A Mesosoale Windfield Analysis of the Los Angeles
Basin, CEM Final Report No. 4121-01-490C to EPA, National Environ-
mental Research Center, Meteorological Labs., 42 pp.
7 ATOMIC ENERGY COMMISSION (AEC), 1972: Safety Guide 22, On-Site Meteoro-
logical Programs, pp. 23.1-23.6.
8 BALL, R.J., 1967: Air Quality and Meteorological Variables in Summer at
Johnstown, Pennsylvania, Master's Thesis, Department of Meteorology,
Penn. State U., University Park, Perm.
9 BALL, R.J., 1969: Interstate Transport of Air Pollution in Southwestern
Connecticut, APTIC, EPA, Research Trinagle Park, N.C., 94 pp.
10 BALL, R.J., et al., 1972: Air Quality Implementation Plan, Conn. Dept.
of Environmental Protection, Hartford, Conn., 320 pp.
11 BIERLY, E.W. and W. HEWSON, 1962: " Some Restrictive Meteorological Con-
ditions to be Considered in the Design of Stacks," J. Appl. Meteor.,
Vol. 1, pp. 383-390.
12 BIGGS, W.G., R.G. ALLEN, L.W. CROW, E.L. HOVEND, and R.G. LARSON, 1972:
"An Assessment of the Combined Effects of Air Pollution by the Major
Power Plants in the Southwest United States," Paper No. 72-130, pre-
sented at the 65th Annual Meeting of the Air Poll. Control Assoc.,
Miami Beach, Florida, 20 pp.
13 BOETTGER, C.M., 1961: " Air Pollution Potential East of the Rocky Moun-
tains: Fall 1959," Bull. AMS.', Vol. 9, pp. 615-620.
14 BOORAS, S.G., 1966: "Air Quality Telemetering System," for presentation
during June, 1966, Great Lakes Regional Meeting of the American
Chemical Society in Chicago, 111., 3 pp.
F-3
-------
15 BORNSTEIN, R.D., 1968: "Observations of the Urban Heat Island Effect in
New York City," -I. Appl. Meteor., Vol. 7, pp. 575-582.
16 BOWNE, N.E., 1971: Space and Time Variability of SC>2 and Partieulate
Concentrations in Connecticut, prepared for EPA, Research Triangle
Park, North Carolina, Contract No. CPA 70-155, 62 pp.
17 BOWNE, N.E., 1973: "Diffusion Rates," Paper 73-130, presented at the 66th
Annual Meeting of the Air Pollution Control Asso., Chicago, 111., The
Research Corp. of New England, Weathersfield, Conn., 18 pp.
18 BRACEWELL, J.M. and D. GALL, 1967: "The Catalytic Oxidation of Sulfur
Dioxide in Solution at Concentrations Occurring in Fog Droplets,"
Proa., Symp. on the Physico-chemical Transformation of Sulfur Com-
pounds in the Atmosphere and the Formation of Acid Smogs, organiza-
tion for Economic Cooperation and Development, Mainz, Germany.
19 BRASSER, L.J., 1974: "Questions Related to the Selection of Air Pollution
Measuring Systems," Design of Environmental Information Systems,
R.A. Deininger (ed.), Ann Arbor Sci. Pub., Ann Arbor, Michigan,
Chapter 6, pp. 135-153.
20 BRIER, G.W., 1973: Validity of the Air Quality Display Model Calibration
Procedure, EPA-R4-73-017, prepared for the Office of Research and
Monitoring, U.S. Environmental Protection Agency, Research Triangle
Park, N.C., 28 pp.
21 BRIER, G.W., 1975: Statistical Questions Relating to the Validation of
the Air Quality Simulation Models, EPA 650/4-75-010, prepared for
the National Environmental Research Center, Meteorology Lab., U.S.
Environmental Protection Agency, Research Triangle Park, N.C., 21 pp.
22 CARPENTER, S.B., T.L. MONTGOMERY, J.M. LEAVETT, W.C. COLBAUGH, and F.W.
THOMAS, 1971: "Principal Plume Dispersion Models: TVA Power Plants,"
J. Air Poll. Control Assoc., Vol. 21, 491-495.
23 CASWELL, V.D., 1974: "Information Systems Enter the Space Age," Env. Sci.
& Tech., Vol. 8, pp. 990-992.
24 CAVENDER, J.H., D.S. KIRCHER, and A.J. HOFFMAN, 1973: Nationwide Air
Pollutant Emission Trends, 1940-1970, EPA, Office of Air and Water
Programs Publ. No. AP-115, 52 pp.
25 CHANDLER, T.J., 1968: "The Bearing of the Urban Temperature Field Upon
Urban Pollution Patterns," Atmos. Env., Vol. 2, No. 6, pp. 619-620.
26 CHARLSON, R.J., 1969: "Note on the Design and Location of Air Sampling
Devices," J. Air Poll. Control Assoc., Vol. 19, p. 802.
27 CLARKE, J.F., 1969: "Nocturnal Urban Boundary Layer Over Cincinnati,
Ohio," Mo. Wea. Eev., Vol. 97, pp 582-589.
F-4
-------
28 CLIFTON, M., D. KERRIDGE, W. MOULDS, J. PEMBERTON, and J.K. Donoghue,
1959: "The Reliability of Air Pollution Measurements in the Relation
to the Siting of Instruments," Int'l. J. Air & Water Poll., Vol. 2,
pp. 188-196.
29 COLLINS, G.F,, 1971: "Predicting Sea-Breeze Fumigation from Tall Stacks
at Coastal Locations," Nuclear Safety, Vol. 12, pp. 110-114.
30 COPLY, C.M., 1973: The St. Louis Regional Telementered Air Pollution
Monitoring System, Air Pollution Control Program, St. Louis, Mo.,
10 pp.
31 COTE, W.A., J.E. YOCOM and N.E. BOWNE, 1973: "Collection and Utilization
of Environmental Quality Data," AIchE Symp. Series, Vol. 70, pp. 31-39.
32 CROW, L.W., 1964: "Airflow Related to Denver Air Pollution," J. Air Poll.
Control Assoo. , Vol. 14, pp. 56-65.
33 CROW, L.W., 1974: "Collection and Analysis of Meteorological Data for
Environmental Impact Assessment at Remote Sites," paper presented at
the 67th Annual Meeting of the Air Pollution Control Association,
Denver, Colo., 14 pp.
34 CULKOWSKI, W.M., 1967: "Estimating the Effect of Buildings on Plumes from
Short Stacks," Nuclear Safety, Vol. 8, pp. 257-259.
35 DARBY, W.P., P.J. OSSENBRUGGEN and C. GREGORY, 1974: "Placement of Samp-
lers in an Air Monitoring Network," Proc.j 20th Annual Meeting3 In-
stitute of Env. Sci.j Washington3D.C., pp 326-335.
36 DAVIDSON, B., 1967: "A Summary of the New York Urban Air Pollution Dy-
namics Research Program," J. Air Poll. Control Assoc., Vol. 17,
pp. 154-158.
37 DAVIES, P.O.A.L. and D.J. Moore, 1964: "Experiments on the Behavior of
Effluent Emitted from Stacks at or Near the Roof Level of Tall Re-
actor Buildings," Int'l. J. Water Poll., Vol. 8, pp. 515-533.
38 DAVIS, F.K. and H. NEWSTEIN, 1968: "The Meteorology and Vertical Distri-
bution of Pollutants in Air Pollution Episodes in Philadelphia,"
Atmos. Env., Vol. 2, pp. 559-574.
39 DE MARRIS, G.A., 1961: "Vertical Temperature Difference Observed Over an
Urban Area," Bull., AMS, Vol. 42, pp. 548-554.
40 DEPARTMENT OF HEALTH, EDUCATION AND WELFARE (DHEW), 1967: Technical Re-
port: New lork - New Jersey Air Pollution Abatement Activity; Sul-
fur Compounds and Carbon Monoxide, DHEW, PHS, National Center for
Air Pollution Control, Cincinnatti, Ohio, pp. 59-75.
41 DRIVAS, P.J. and F.H. SHAIR, 1974: "Probing the Air Flow Within the Wake
Down Wind of a Building by Means of a Tracer Technique," Atmos. Env.,
Vol. 8, pp. 1165-1175.
F-5
-------
42 ELIASSEN, A. and J. SALTBONES, 1975: "Decay and Transformation Rates of
S02 as Estimated from Emission Data, Trajectories and Measured Air
Concentrations," Atmos. 'Env., Vol. 9, No. 4, pp. 425-430.
43 ENVIRONMENTAL MEASUREMENTS, INC., 1971: SO2 and NO2 Measurements - Metro-
politan Los Angeles, California, EPA Contract 6H-02-0124, Task B,
APTD 1148, 42 pp.
44 ENVIRONMENTAL PROTECTION AGENCY, 1971: Guidelines: Air Quality Surveil-
ance Networks, Office of Air Programs, NO-AP-98, Research Triangle
Park, N.C., 16 pp.
45 ENVIRONMENTAL PROTECTION AGENCY, 1973: Air Quality Monitoring Interim
Guidance, OAQPS No. 1.2-007, Monitoring and Data Analysis Division,
Research Triangle Park, N.C.,. 36 pp.
46 ENVIRONMENTAL PROTECTION AGENCY, 1974: Guidelines for the Evaluation of
Air Quality Data, OAQPS No. 1.2-015, Office of Air Quality Planning
and Standards, Research Triangle Park, N.C., 32 pp.
47 ENVIRONMENTAL PROTECTION AGENCY, 1974: Health Consequences of Sulfur
Oxides: A Report from CHESS, 1970-1971, EPA 650/1-74-004, office of
Research and Devel.r Research Triangle Park, N.C., 332 pp.
48 ENVIRONMENTAL PROTECTION AGENCY, 1974: Guidance for Air Quality Monitor-
ing Network Design and Instrument Siting, OAQPS NO. 1.2-012,, off ice;
of Air Quality"Planning and Standards, Research Triangle Park, N.C.,
33 pp.
49 ENVIRONMENTAL PROTECTION AGENCY, 1974: Guidelines for Air Quality Main-
tenance Planning and Analysis, Vol. II, Air Quality Monitoring and
Data Analysis, Office of Air Quality Planning and Standards, Research
Triangle Park, N.C., 104 pp.
50 ENVIRONMENTAL PROTECTION AGENCY, 1975: . Report to Congress on Control of
Sulfur'Oxides, EPA 450/1-75-001, Office of Air and Waste Management,
Research Triangle Park, N.C., 63 pp.
51 EPSTEIN, A.M., C.A. LEARY, S.T. McCANDLES, B.J. GOLDSMITH, J.C. GOODRICH,
and B.H. WILLIS, 1974: A Guide for Considering Air Quality in Urban
Planning, EPA 450/3-74-020, Env. Res. and Tech., Inc., Lexington,
Mass., 103 pp.
52 ESTOQUE, M.A., 1962: "The Sea Breeze as a Function of the Prevailing
Synoptic Situation," J. Atmos. Sci., Vol. 19, pp. 244-250.
53 FARMER, J.R. and J.D. WILLIAMS, 1966: Interstate Air Pollution Study,
Phase II, Project Report; III, Air quality Measurements, U.S. DHEW,
PHS, National Center for Air Pollution Control, Cincinnati, Ohio,
182 pp.
54 FEDERAL REGISTER, 1973: Preparation of Environmental Impact Statements,
Guidelines, Vol. 38, No. 147.
F-6
-------
55 FEDERAL REGISTER, 1975: National Oceania and Atmospheric Administration,
Dept. of Commerce, General Regulations of the Env. Data Sen)., Env.
Data, Vol. 40, No. 234, Thursday, December 4, 1975.
56 FINDLAY, B.F. and M.S. HIRT, 1969: "An. Urban-Induced Meso-Circulation,"
Atmos. Env., Vol. 3, pp. 537-542.
57 FLEMMING, G., 1967: "Concerning the Effect of Terrain Configuration on
Smoke Dispersal," Atmos. Env., Vol. 1, pp. 239-252.
58 FOREST, J. and L. NEWMAN, 1973: "Ambient Air Monitoring for Sulfur Com-
pounds, A Critical Review," J. Air Poll. Control. Assoc., Vol. 23,
pp. 761-768.
59 FRENZEL, C.W., 1962: "Diurnal Wind Variations in Central California,"
J. Appl. Meteor. , Vol. 1, pp. 405-412.
60 FRIZZOLA, J.A. and E.L. FISHER, 1963: "A Series of Sea Breeze Observa-
tions in the New York City Area," J. Appl. Meteor., Vol. 2, pp. 722-
739.
61 GARLAND, J.A., D.H.F. ATKINS, C.J. READINGS, and S.J. CAUGHEY, 1974:
"Deposition of Gaseous Sulphur Dioxide to the Ground," Atmos. Env.,
Vol. 8, No. 1, pp. 75-80.
62 GATZ, D.F. and D.W. KLAPPENBACH, 1969: "City of Chicago Pollution Inci-
dents: Case Studies," paper presented at the Air Pollution Control
Assoc. Annual Meeting, New York, N.Y., 13 pp.
63 GILL, G.C., L.E. OLSSON, J. SELA, and M. SUDA, 1967: "Accuracy of Wind
Measurements on Towers or Stacks," Bull., AMS, Vol. 48, pp. 665-674.
64 GOLDSTEIN, I.F., L. LANDOVITZ, and G. BLOCK, 1974: "Air Pollution Pat-
terns in New York City," J. Air Poll. Control Assoc., Vol. 24,
pp..148-152.
65 GOODRJCH, J.C. and B.H. WILLIS, 1972: . "A Methodology for Determining
Emissions from Land Use Planning Data," Paper No. 72-122 presented
at the 65th Annual Meeting of the Air Poll. Control. Assoc., Miami
Beach, Florida, 22 pp.
66 GORSHKO, B.B., 1973: The Optimizing Sampling Stations and the Sampling-
Frequency for Surveying Urban Air Pollution, Canada, Dept. of Env.,
Atmos. Env. Serv., Meteorological Translation No. 20, 10 pp.
67 HALITSKY, J., 1975: "Diffusion of Vented Gas Around Buildings," J. Air
Poll. Control. Assoc., Vol. 12, pp. 74-80.-
68 HAWKINS, J.E. and G. NONHEBEL, 1955: "Chimneys and the Dispersal of
Smoke," J. Inst. Fuel, Vol. 28, pp. 530-545.
F-7
-------
69 HALPERN, P., C. SIMON and L. RANDALL, 1971: "Source-Emissions and the
Vertically Integrated Mass Flux of Sulfur Dioxide Across the New
York City Area," J. Appl. Meteor., Vol. 10, pp. 715-724.
70 HAMBURG, F.C., 1971: "Some Basic Considerations in the Design of an Air
Pollution Monitoring System," J. Air Poll. Control Asoa. , Vol. 21,,
pp. 609-613.
71 HARRIS, E.K. and R.A. McCORMICK, 1963: "NOTES AND CORRESPONDENCE, A
Simple Procedure for Estimating the Standard Deviation of Wind Fluc-
tuations," J. Appl. Meteor., Vol. 2, pp. 804-805.
72 HARRIS, D.N., J.R. HUFFMAN and J.H. WEILAND, 1968: "Another Look at New
York City's Air Pollution Problem," J. Air Poll. Control Assoo.,
Vol. 18, pp. 406-410.
73 HARRISON, P.R., 1972: "Considerations for Siting Air Quality Monitors in
Urban Areas," Paper No. 73-161 presented at the 65th Annual Meeting
of Air Poll. Control Assoc., Miami Beach, Florida, 16 pp.
74 HEINS, C.F., F.D. JOHNSON, and E.G. MANGOLD, 1975: "Monitoring the En-
vironment," Env. Sai & Tech., Vol. 9, pp. 720-725.
75 HELLER, A.N. and E.F. FERRAND, 1969: The Aerometrio Network of the City
of New York, Dept. of Air Resources, The City of New York, 25 pp.
76 HERRICK, R.A., 1966: "Recommended Standard Method for Continuing Dust
Fall Survey (APM-1, Revisionl)," J. Air Poll. Control Assoo., Vol.
7, pp. 372-377.
77 HEWSON, E.W. and G.C. GILL, 1944: "Meteorological Investigations in
Columbia Rvier Valley near Trail, B.C.," U.S. Bureau of Mines Bull.,
Vol. 453, pp. 23-228.
78 HEWSON, E.W., E.W. BIERLY, G.C. GILL, 1961: "Topographic Influences on
the Behavior of Stack Effluents," Proa.} Am. Power Conf., Vol. XXIII,
pp. 358-370.
79 HEWSON, E.W. and L.E. OLSSON, 1967: "Lake Effects on Air Pollution Dis-
persion," J. Air Poll. Control Assoo., Vol. 17, pp..757-761.
80 HIGH, M.D., 1974: "A review of the Background, Preparation and Use of
Environmental Impact Statements," J. Air Poll. Control Assoc. , Vol..
24, pp. 111-114.
81 HINDS, W.T., 1969: "Peak to Mean Concentration Ratios from Ground Level
Sources in Building Wakes," Atmos. Env., Vol. 3, pp. 145-156.
82 HINDS, W.T., 1970: "Diffusion over Coastal Mountains of Southern Cali-
fornia," Atmos. Env., Vol. 4, pp. 107-124.
83 HINO, M., 1968: "Computer Experiment on Smoke Diffusion over a Compliccited
Topography," Atmos. Env., Vol. 2, pp. 541-558.
F-8
-------
84 HINO, M., 1968: "Maximum Ground-Level Concentration and Sampling Time,"
Atmos. Env., Vol. 2, pp. 149-165.
85 HOFFMAN, D.W., J.H. PAULUS, and D.P. LINGO, 1975: "An Evaluation of Air
Quality Monitoring for Fossil-Fuel Power Plants in a Regional Power
System," Paper No. 75-45.4 presented at the 68th Annual Meeting of
the Air Poll. Control Assoc., Boston, Mass., 14 pp.
86 HOLZWORTH, G.C., 1962: "A Study of Air Pollution Potential for the West-
ern United Statex," J. Appl. Meteor., Vol. 1, pp. 366-382.
87 HOTZWORTH, G.C., 1969: "Large-Scale Weather Influences on Community Air
Pollution Potential in the United States," J. Air Poll. Control Assoc.,
Vol. 19, pp. 248-254.
88 HOLTZWORTH, G.C., 1972: Mixing Heights, Wind Speeds, and Potential for
Urban Air Pollution Throughout the Contiguous United States, EPA,
Office of Air Programs, 118 pp.
89 HOTCHKISS, R.S. and F.H. HARLOW, 1973: Air Pollution Transport in Street
Canyons, U. Calif., Los Alamos Sci. Lab., Los Alamos, N.M., EPA
Contract EPA-1AG-0122(D), 79 pp.
90 HOUGLAND, E.S. and N.T. STEPHENS, 1974: "Air Pollutant Monitor Siting by
Analytical Techniques," Paper No. 74-49 presented at the 67th Annual
Meeting of the Air Poll. Control Assoc., Denver, Col.
91 HOYDYSH, W.G., 1973: A Scale Model Wind Tunnel Study of Dispersion in the
Cleveland Area - Laboratory Simulation of Lake Breezes Effects on
Diffusion from Around Level Emissions+. Final Report for NASA, Grant
NGR-33-016-197, 24 pp.
92 HUNT, Jr., W.F., 1972: "The Precision Associated with the Sampling Fre-
quency of Log-Normally Distributed Air Pollutant Measurements," J.
Air Poll. Control Assoc., Vol. 22, pp. 687-691.
93 HUTCHEON, R.J., R.H. JOHNSON, W.P. LOURY, C.H. BLACK, and D. HADLEY, 1967:
"Observations of the Urban Heat Island in a Small City," Bull., AMS,
Vol. 48, pp. 7-9.
94 INHABER, H., 1975: "A Canadian View of Monitoring Activities," Env. Sci.
& Tech., Vol. 9, pp. 206-209.
95 INTERNATIONAL JOINT COMMISSION, OTTAWA (Ontario), ST. CLAIR-DETROIT AIR
POLLUTION BOARD, 1971: Joint Air Pollution Study of St. Clair-
Detroit River Areas, for Int'l. Joint Commission, Canada and U.S.,
APTD 13-5, 266 pp.
96 JEPSEN, A.F. and J.C. WEIL, 1973: "Maryland Power Plant Air Monitoring
Program Preliminary Results," Paper 73-157 presented at the 66th
Annual Meeting of the Air Poll. Control Assoc., Chicago, 111, 19 pp.
F-9
-------
97 JUDA, J., 1974: "Design of Air Quality Monitoring Networks," Design of
Environmental Information Systems, Ralph Deininger (ed.), Science
Publs., Ann Arbor, Michigan; Chapter 18, 347-356.
98 JUTZ, G.A., 1972: Guidelines for Technical Services of a State Air Pollu-
tion Control Agency, APTD-1374, EPA Contract No. 68-02-041, Research
Triangle Park, N.C., 291 pp.
99 KAUPER, E.K. and C.J. HOPPER, 1966: "Control of Emissions Through Effi-
cient Use of Atmospheric Dispersiona Forcast System," J. Air Poll.
Control Assoc,, Vol. 16, pp. 606-608.
100 KEAGY, D.M., W.W. STALKER, C.E. ZIMMER, and R.C. DICKERSON, 1961: "Samp-
ling Station and Time Requirements for Urban Air Pollution Survey.
Part I: Lead Peroxide Candles and Dust Fall collectors," J. Air
Poll. Control Assoc., Vol. 11, pp. 270-280.
101 KNOX, J.B., 1974: "Numerical Modeling of the Transport Diffusion and De-
position of Pollutants for Regions and Extended Scales," J. Air Poll.
Control Assoc., Vol. 24, pp. 660-664.
102 KORNREICH, L.D.(ed.), 1974: Proc.3 Symp. on Statistical Aspects of Air
Quality Data, EPA 650/4-74-038, U. of N.C., Chapel Hill, prepared
for EPA, Office of Research and Development, NERC, Research Triangle
Park, N.C., 247 pp.
103 LARSEN, R.I., W.W. STALKER and C.R. CLAYDON, 1961: "The Radial Distribu-
tion of Sulfur Dioxide Source Strengh and Concentration in Nashville,"
J. Air Poll. Control Assoc. , Vol. 11, pp. 529-534.
104 LAVERY, T.F., B.A. EGAN and R.M. IWANCHUK, 1974: "A Numerical Simulation
of the Advection and Diffusion of a Plume under Aerodynamic Downwash
Conditions," APCA paper #74-215, Environmental Research and Technology
Inc., Lexington, Mass., 16 pp.
105 LEADERER, B.P. and G.H. SOVAS, 1972: "Allocation and Projection of Resi-
dential and Commerical Emissions through Use of the LUNR Inventory,"
paper # 72-57, presented'at the 65th Annual Meeting of the Air Poll.
Cont. Assoc., Miami Beach, Fa., 18 pp.
106 LEAKEY, D.M., 1974: "A Study of Air Flow over Irregular Terrain," Atmos.
EnV., Vol. 8, 783-791.
107 LEAVETT, 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," J. Air Poll. Control Assoc., Vol. 7,
pp. 211-214.
108 LOGSDON, O.J. and M.J. CARTER, 1975: "Comparison of Manual and Automated
Analysis Methods for Sulfur Dioxide in Manually Impinged Ambient Air
Samples," EnV. Sci. and Tech., Vol. 1, pp. 71-86.
F-10
-------
109 LUCAS, D.H., 1958; "The Atmospheric Pollution of Cities," Int'l. J. Air
Poll., Vol. 1, pp. 71-86.
110 LUCAS, D.H., 1967: "Paper'VI: Application and Evaluation of Results of
the Tilbury Plume Rise and Dispersion Experiment," Atmos. Em)., Vol.
1, pp. 421-424.
Ill LUDWIG, F.L. and J.H.S. KEALOHA, 1975: Selecting Sites for Carbon Monox-
ide Monitoring, EPA Contract No. 68-02-1471, Stanford Research In-
stitute, Menlo Park, Calif., Final Report, 149 pp.
112 LYNN, D.A. and T.B. McMullen, 1965: "Air Pollution in Six Major U.S.
Cities as Measured by the Continuous Air Monitoring Program (CAMP),"
paper presented at the Annual Meeting of the Air Poll. Control.
Assoc., Toronto, Ontario, Canada.
113 LYONS, W.A. and L.E. OLSSON, 1972: "Mesoscale Air Pollution Transport
in the Chicago Lake Breeze," J. Air Poll. Control Assoc., Vol. 22,
pp. 876-881.
114 MacDONALD, B.I., 1975: "Alternative Strategies for Control of Sulfur
Dioxide Emission," J. Air Poll. Control Assoc., Vol. 25, pp. 525-
528.
115 MAGE, D.T., 1975: "An Improved Statistical Model for Analyzing Air Pol-
lution Concentration Data," paper No. 75-514, presented at the 68th
Annual Meeting of the Air Poll. Control Assoc., Boston, Mass., 27 pp.
116 MARSH, K.J. and M.D. FOSTER, 1967: "An Experimental Study of the Dis-
persion of the Emissions from Chimneys in Reading - I: The Study
of Long-Term Average Concentrations of Sulfur Dioxide," Atmos. Env.,
Vol. 1, pp. 527-550.
117 MARTIN, A. and F.R. BARKER, 1967: "Sulfur Dioxide Concentrations Mea-
sured at Various Distances from a Modern Power Station," Atmos. Env.,
Vol. 1, pp. 655-677.
118 MARTIN, D.O., P.A. HUMPHREY, and J.L. DICKE, 1967: Interstate Air Pollu-
tion Studyj Phase II Project Report3 V. Meteorology and Topography,
U.S. DHEW, PHS, National Center for Air Poll. Control, Cincinnati,
Ohio, 42 pp.
119 McELROY, J.L. and F. POOLER, Jr., 1968: St Louis Dispersion Study, Vol.
II - Analyses, U.S. Dept. of HEW, Bureau of Engineering and Physical
Sciences, Div. of Meteorology, Arlington, Va., 51 pp.
120 McKEE, H.C. and R.E. CHILDERS, 1972: "Estimation of Additive Effects
from Multiple Sources," J. Air Poll. Control Assoc., Vol. 22, pp.
785-789.
121 MEGONNELL, W.H., 1975: "Atmospheric Sulfur Dioxide in the United States:
Can The Standards be Justified or Afforded?", J. Air Poll. Control
Assoc., Vol. 25, pp. 9-15.
F-ll
-------
122 MITCHELL, Jr., J.M., 1961: "The Temperature of Cities," Weatherwise,
Vol. 14, pp. 224-258.
123 MONTGOMERY, T.L. and M. CORN, 1967: "Adherence of Sulfur Dioxide Concen-
trations in the Vicinity of a Steam Plant to Plume Dispersion Models,"
J. Air Poll. Control Assoa., Vol. 17, pp. 512-517.
124 MONTGOMERY, T.L. and J.H. COLEMAN, 1975: "Empirical Relationships Between
Time-Averaged S02 Concentrations," Env. Sai. and Tech., Vol. 9, pp.
953-957.
125 MONTGOMERY, T.L. J.W. FREY, and W.B. NORRIS, 1975: "Intermittent Control
Systems," Era). Sai. and Tech., Vol. 9, pp. 528-533.
126 MOORE, D.J., 1967: "Paper IV: S02 Concentration Measurements Near Til-
bury Power Station," Atmos, Env., Vol. 1, pp. 289-410.
127 MORGENSTEEN, P. and K.A. HAGG, 1972: "A System for Abatement Control
Strategy Evaluation," J. Air Poll. Control, Vol. 12, pp. 774-778.
128 MOROZ, W.J., 1967: "A Lake Breeze on the Eastern Shore of Lake Michigan:
Observations and Model," J. Atmos. Sai., Vol. 24, pp. 337-355.
129 MOSES, H., G.H. STORM and J.E. CARSON, 1964: "Effects of Meteorological
and Engineering Factors on Stack Plume Rise," Nucl. Safety, Vol. 6,
pp. 1-18.
130 MOSES, H., 1969: "Mathematical Urban Air Pollution Models," paper pre-
sented at the 62nd Annual Meeting of the Air Poll. Control Assoc.,
New York, N.Y., 46 pp.
131 MOSES, H., D.F. GATZ, J.E. CARSON, F.C. KULHANEC, and G.A. ZERBE, 1972:
"The Anatomy of an Air Pollution Episode," paper No. 72-139, pre-
sented at the 65th Annual Meeting of the Air Poll. Control Assoc.,
Miami Beach, Fa.
132 MOSHER, J.C., M.J. LEONARD, T.P. MULLINS, and M.F. BRUNELLE, 1970: "The
Distribution of Contaminants in the Los Angeles Basin Resulting from
Atmospheric Reactions and Transport," J. Air Poll. Control Assoa.,
Vol. 20, pp. 35-42.
133 MUNN, R.E. and A.F.W. COLE, 1967: "Turbulence and Diffusion in the Wake
of a Building," Atmos. Env. , Vol. 1, pp. 33-34.
134 MUNN, R.E. and I.M. STEWART, 1967: "The Use of Meteorological Towers in
Urban Air Pollution Programs," J. Air Poll. Control Assoa., Vol.. 17,
pp. 98-101.
135 NEHLS, G.J. and G.G. AKLAND, 1973: "Procedures for Handling Aerometric
Data," J. Air Poll. Control Assoa., Vol. 23, pp. 180-184.
136 OKE, T.R., 1973: "City Size and the Urban Heat Island," Atmos. EnV. ,
Vol. 7, pp. 769-779.
F-12
-------
137 OKE, T.R. and G.B. MAXWELL, 1975: "Urban Heat Island Dynamics in Mon-
treal and Vancouver," Atmos. Env. , Vol. 9, pp. 191-200.
138 OTT, W.R., 1975: "Development of Criteria for Siting Air Monitoring
Stations," paper No. 75-14.2, presented at the 68th Annual Meeting
of the Air Poll. Control Assoc., Boston, Ma., 17 pp.
139 OTT, W.R. and D.T. Mage, 1975: "Random Sampling as an Inexpensive Means
for Measuring Average Annual Air Pollutant Concentrations in Urban
Areas," paper No. 75-14.3, presented at the 68th Annual Meeting of
the Air Poll. Control Assoc., Boston, Ma., 25 pp.
140 OWERS, M.J. and A.W. POWELL, 1974: "Deposition Velocity of Sulphur Di-
oxide on Land and Water Surfaces Using a 35S Tracer Method," Atmos.
., Vol. 8, pp. 67-68.
141 OZOLINS, G., W.R. OTT, and T.W. STANLEY, 1974: "A National Program for
Quality Assurance in Environmental Monitoring," paper No. 74-15,
presented at the 67th Annual Meeting of the Air Pollution Control
Assoc., Denver, Coloardo, 19 pp.
142 PAULUS, J.J. and A.T. ROSSANO, 1973: "Siting of Air Quality and Meteor-
ological Monitoring Stations to Investigate Air Quality Effects of
a Point Source," paper presented at Int'l. Clean Air Congr. 3d,
Duesseldort, West Germany, pp. 133-136.
143 PEDCO-Environmental Specialists, Inc., 1972: Field Operations Guide for
Automatic Air Monitoring Equipment, EPA, Office of Air Programs,
Research Triangle Park, N.C., EPA Contract No. CPA-70-124, 154 pp.
144 PELLETIER, J., 1963: "Difficulties Encountered in the Measurement of
Air Pollution and in the Interpretation of Results," Int'l. J. Air
Wat. Poll., Vol. 1, pp. 973-978.
145 PETERS, L.K., 1975: "On the Criteria for the Occurrence of Fumigation
Inland from a Large Lake," Atmos. Env., Vol. 9, pp. 809-816.
146 PETERSON, J.T., 1970: "Distribution of Sulfur Dioxide Over Metropolitan
St. Louis as Described by Empirical Eigenvectors and its Relation
to Meteorological Parameters," Atmos. Env., Vol. 4, pp. 501-518.
147 PETERSON, J.T., 1972: "Calculations of Sulfur Dioxide Concentrations
Over Metropolitan St. Louis," Atmos. EnV., Vol. 6, pp. 433-442.
148 POOLER, F,, Jr., 1963: "Airflow Over a City in Terrain of Moderate Re-
lief," J. Appl. Meteor., Vol. 2, pp. 446-456.
149 POOLER, F.,Jr., 1966: "A Tracer Study of Dispersion Over a City," J.
Air Poll. Control Assoo. , Vol. 11, pp. 677-681.
150 POOLER, F.,Jr., 1974: "Network Requirements for the St. Louis Regional
Air Pollution Study," J. Air Poll. Control Assoo., Vol. 24, pp. 228-
231.
F-13
-------
151 RANDERSON, D., 1970: "A Numerical Experiment in Simulating the Trans-
port of Sulfur Dioxide Through the Atmosphere," Atmos. Env., Vol.
4, pp. 615-632.
152 RAYNOR, G.S., M.E. SiMITH, and I.A. SINGER, 1974: "Meteorological Effects
on Sulfur Dioxide Concentrations on Suburban Long Island, New York,"
Atmos. Env., Vol. 8, pp. 1305-1320.
153 KEIQUAM, H., 1970: "An Atmospheric Transport and Accumulation Model for
Airsheds," Atmos. Env., Vol. 4, pp. 233-247.
154 RHOADES, B.J., 1973: "A Methodology for Minimizing and Optimizing Sta-
tion Location in a Two Parametered Monthly Sampling Network," paper
no. 73-159, presented at the 66th Annual Meeting of the Air Poll.
Control Assoc., Chicago, 111.
155 ROBINSON, E. and R.c. ROBBINS, 1968: Sources, Abundance and Fate of Gas-
eous Atmospheric Pollutants, Stanford Research Institute, Final
Report PR-6755, for Am. Petroleum Inst., 123 pp.
156 ROBINSON, E. and R.C. ROBBINS, 1970: "Gaseous Sulfur Pollutants from
Urban and Natural Sources," J. Air Poll. Control Assoc., Vol. 20,
pp. 233-235.
157 ROSSANO, A.T., 1956: "The Joint City, County, State and Federal Study
of Air Pollution in Louisville," J. Air Poll. Control Assoc., Vol.
6, pp. 176-181.
158 ROTE, D.M. and J.W. GUDENAS, 1971: "A Steady State Dispersion Model
Suitable for Air Pollution Episodes," paper presented at the 64th
Annual Meeting of the Air Poll. Control Assoc., Atlantic City, N.J.
159 RUBIN, E.S., 1974: "The Influence of Annual Meteorological Variations
on Regional Air Pollution Modeling: A Case Study of Allegheny
County, Pennsylvania," J. Air Poll. Control Assoc., Vol. 24, pp.
349-356.
160 RUBIN, E.S., and F.C. McMICHAEL, 1975: "Impact of Regulations on Coal
Conversion Plants," Env., Sci. and Tech., Fol. 9, pp. 112-117.
161 RUFF, R.E. and D.G. FOX, 1974: "Evolution of Air Quality Models Through
the Use of the RAPS Data Base," paper presented at the 67th Annual
meeting of the Air Poll. Control Assoc., Denver, Colo., 24 pp.
162 SALTZMAN,B.E., 1972: "Simplified Methods for Statistical Interpretation
of Monitoring Data," J. Air Poll. Control Assoc., Vol. 22, pp. 90-95.
163 SANDBURG, J.S., W.J. WALKER, and R.H. THUILLER, 1970: "Fluorescent Tracer
Studies of Pollutant Transport in the San Francisco Bay Area," <7.
Air Poll. Control Assoc., Vol. 20, pp. 593-598.
164 SCRIVEN, R.A., 1967: "Paper V: Properties of the Maximum Ground Level
Concentration from an Elevation Source," Atmos. Env., Vol. 1, pp.
411-419.
F-14
-------
165 SEARLE, V.C., 1969: "Technical Information Resources in the Air Pollu-
tion Field," J. Air Poll. Control Assoc., Vol. 19, pp. 137-141.
166 SEINFELD, J.H., 1972: "Optimal Location of Pollution Monitoring Sta-
tions in an Airshed," Atmos. Env., Vol. 19, pp. 137-141.
167 SEMRAU, K.T., 1971:. "Control of Sulfur Oxide Emissions from Primary
Copper, Lead, and Zinc SmeltersA Critical Review," J, Air Poll.
Control Assoa., Vol. 21., pp. 185-194.
168 SHEPARD, J.G., 1974: "Measurements of the Direct Deposition of Sulfur
Dioxide Onto Grass and Water by the Profile Method," Atmos. Env.,
Vol. 8, pp. 69-74.
169 SIMON, C., 1969: "New York City's Meteorological Prggram," paper pre-
sented at Mid Atlantic State Section APCA Semi-annual Technical
Conf., Philadelphia, Pa., 9 pp.
170 SINGER, I.A., and M.E. SMITH, 1953: "Relation of Gustiness to Other
Meteorological Parameters," J. Meteor., Vol. 10, pp. 121-126.
171 SINGER, I.A., K. IMAI, and R. Del CAMPO, 1963: "Peak to Mean Pollutant
Concentration Ratios for Various Terrain and Vegetation Cover,"
J. Air Poll. Control Assoc., Vol. 13, pp. 40-42.
172 SINGER, I.A., J.A. FRIZZOLA, and M.E. SMITH, 1966: "A Simplified Method
of Estimating Atmospheric Diffusion Parameters," J. Air Poll. Control
Assoo., Vol. 16, pp. 594-596.
173 SINGER, I.A. and M.E. SMITH, 1966: "Atmospheric Dispersion at Brook-
haven National Laboratory," Air and Wat. Poll. Int'l. J., Vol. 10,
pp. 125-135.
174 SKEPKA, K.J. and T.T. Frankenberg, 1974: "SO2 Concentrations During Air
Stagnation Episodes in the Ohio River Valley," paper No. 44-68, pre-
sented at the 67th Annual Meeting of the Air Poll. Control Assoc.,
Denver, Colo.
175 SH.T, C.M., W.G. RIGGAN, J.G. FRENCH, W.C. NELSON, R.C. DICKERSON, F.B.
BENSON, J.F. FINKLEA, A.V. COLUCCI, D.I. HAMMER, and V.A. NEWELL,
1974: An Overview of Chess. Health Consequences of Sulfur Oxides,
EPA, Contract No. 650/1-74-004, Research Triangle Park, N.C., pp.
1-1 to 1-9.
176 SMITH, D.B., 1968: "Tracer Study in an Urban Valley," J. Air Poll. Con-
trol Assoc., Vol. 18, 600-604.
177 SMITH, F., D.E. WAGONER, and A.C. NELSON, Jr., 1974: "Quality Assurance
Procedures Applicable to Ambient Air Monitoring Methods," paper No.
74-18, presented at the 67 Annual Meeting of the Air Poll. Control
Assoc., Denver, Colo., 28 pp.
F-15
-------
178 SMITH, J.H., 1961: "Reconnoitering the Air: Procedures for Analyzing
Local Air Pollution Concentrations and Contributions," J. Air Poll,
Control Assoc., Vol. 11, pp. 57-59.
179 SMITH, M.E. (ed.), 1968: Recommended Guide for the Prediction of the
Dispersion of Airborne Effluents, ASME Committee on Air Pollution
Controls, New York, N.Y., 85 pp.
180 SMITH, M.E., 1974: "Deficiencies in Data and Analyses for Environmental
Impact Statements," paper No. 74-126, presented at the 67th Annual
Meeting of the Air Pollution Control Association, Denver, Colo.,
18 pp.
181 SNYDER, W.H., 1972: "Similarity Criteria for the Applcation of Fluid
Models to the Study of Air Pollution Meteorology," Bound.-Layer
Meteor., Vol. 3, pp. 113-134.
182 SPEDDINGjD.J., 1972: "Sulphur Dioxide Absorption by Sea Water," Atmos.
Env., Vol. 6, pp. 583-586.
183 SPIEGLER, D.B., E. NEWMAN, and R. E. BAILEY, 1974: "Ambient Pollutant
Concentrations Around Power Plants as a Function of Meteorological
Conditions," paper No. 74-7, presented at the 67th Annual Meeting
of the Air Poll. Control Assoc., 17 pp.
184 STALKER, W.W., R.C. DICKERSON, and G.D. KRAMER, 1962: "Sampling Station
and Time Requirements for Urban Air Pollution Surveys. Part IV: 2-
and 24-hour Sulfur Dioxide and Summary of Other Pollutants," J. Air
Poll. Control Assoc., Vol. 12, pp. 361-375.
185 STALKER, W.W., P.A. KENLZNE, and H.J. PAULUS, 1964: "Nashville Sulfur
Dioxide Emission Inventory and the Relationship of Emission to Mea-
sured Sulfur Dioxide," J. Air. Poll. Control Assoc. , Vol. 14, pp.
469-474.
186 START, G.E., C.R. DICKSON, and L.L. WENDELL, 1973: Diffusion in a Can-
yon Within Rough Mountainous Terrain, NOAA, Env. Res. Labs, Air
Resources Labs, Idaho Fall, Id., Report No. NOAA IM ERL ARL-38,
43 pp.
187 START, G.E., N.R. RICKS, and C.R. DICKSON, 1974: Effluent Dilutions
Over Mountainous Terrain, NOAA Tech. Memo., ERL ARL-51, Air Resources
Lab., Idaho Falls, Id., 162 pp.
188 START, G.E. and L.L. WENDELL, 1974: Regional Effluent Dispersion Calcu-
lations Considering Spatial and Temporal Meteorological Variations,
NOAA Tech. Report COM-74-11627, 61 pp.
189 THUILLER, R.H., 1972: Air Monitoring and its Application to Land Use
Planning, Bay Area Air Poll. Control District, San Fransicso, Calif.,
Tech. Dif., Conf. on Methods in Air Poll, and Industrial Hygiene
Studies, 7 pp.
F-16
-------
190 TURNER, D.B. , 1961: "Relationships Between 24-Hour Mean Air Quality
Measurements and Meteorological Factors in Nashville, Tennessee,"
J. Air Poll. Control Assoc., vol. , pp. 183-489.
191 TURNER, D.B., 1968: "The Diurnal and Day-to-Day Variations of Fuel Usage
for Space Heating in St. Louis, Missouri," Atmos. EnV., Vol. 2, pp.
339-351.
192 U.S. DEPT. OF HEW, 1959: "Air Monitoring and Sampling Networks," Pro-
ceedings of a seminar held at the Robert A. Taft Sanitary Engineer-
ing Center, Cincinnati, Ohio, 4 pp.
193 U.S. DEPT. OF HEW, 1967: Lewiston, Idaho, Clarkston, Washington Air
Pollution Abatement Activity, Tech. Report, 7 pp.
194 VAN DER HOVEN, I., 1967: "Atmospheric Transport and Diffusion at Coastal
Sites," Nuclear Safety, Vol. 8, pp. 490-499.
195 VAUGHN, D.J. and E.J. STANEK, II, 1973: "Sulfur Dioxide Standards:
Primary More Restrictive than Secondary?", J. Air Poll. Control
ASSOC., Vol. 23, pp. 1039-1041.
196 VENEZIA, R., G. OZOLINS, 1966: Air Pollutant Emission Inventory_, Inter-
state Air Pollution Study, Phase II, Project Report, Dept. of HEW,
Cincinatti, Ohio, 50 pp.
197 VUKOVICH, P.M., 1974: On Optimum Meteorological- and Air Pollution Samp-
ling Site Selection in Urban Areas: Phase I, Final Progress Report,
NSF Grant No. GI-34345, Research Triangle Park, N.C., 64 pp.
198 WHITE, F.D. and D.H. PACK, 1956: "Meteorology as Related to Reactor-Site
Selection," J. Air Poll. Control Assoc., Vol. 6, pp. 151-170.
199 WIEDERSUM, G.C. and S. BARR, 1973: "The Effect of Power Plant Stack
Emissions on the Ground Level S02 Concentrations in an Urban Area,"
paper No. 73-123, presented at the 66th Annual Meeting of the Air
Poll. Control Assoc., Chicago, 111., 27 pp.
200 WEEDFALL, R.O. and B. LXNSKY, 1969: "A Mesoclimatological Classification
System for Air Pollution Engineers," J. Air Poll. Control Assoc.,
Vol. 19, pp. 511-513.
201 WILLIAMS, P., Jr., 1964: "Air Pollution Potential Over the Salt Lake
Valley of Utah as Related to Stability and Wind Speed," J. Appl.
Meteor., Vol. 3, pp. 92-97.
202 WOHLERS, H.C., N.M. TRIEFF, H. NEWSTEIN, and W. STEVENS, 1967: "Sulfur
Dioxide Adsorption On - and Desorption from Teflon, Tygon, Glass,
Stainless Steel and Aluminum Tubing ," Atmos. Env., Vol. 1, pp. 121-
130.
203 WU, Th. Y-T., 1972: "Cavity and Wake Flows," Annual Review of Fluid
Mech., Vol. 4, Annual Reviews Inc., Pasadena, California.
F-17
-------
TECHNICAL REPORT DATA
(Please read Instructions on the n r -rsv before completing)
1 REPORT NO
3. RECIPIENT'S ACCESSIOt*NO.
4 TITLE AND SUBTITLE
Optimum Site Exposure Criteria for
Monitoring
5. REPORT.
6. PERFORMING ORGANIZATION CODE
7 AUTHORIS
;)
Robert Ball and Gerald Anderson
8. PERFORMING ORGANIZATION REPORT NO.
PERFORMING ORGANIZATION NAME AND ADDRESS
Center for Environment and Man
275 Windsor Street
Hartford, Connecticut 06120
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-2045
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE
OAQPS
Monitoring & Data Analysis Division
Research Triangle Park, N.C. 27711
EPQRT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
IB. SUPPLEMENTARY NOTES
16. ABSTRACT
This report presents procedures and exposure criteria for selecting S02
monitoring sites. Data uses are first reviewed and summarized; from this summary a
list of specific siting objectives is developed. Site selection procedures were then
prepared for specific site types each of which was associated with either a grouping
of siting objectives or with an individual objective.
Detailed procedures are provided for selecting sites to measure regional
mean concentrations, interregional S02 transport, representative concentrations
for areas of various sizes, peak concentrations in urban areas, and emergency episode
levels. Recommendations and the rationale for inlet height and orientation, and for
minimizing undue influence from nearby sources are presented.
Sources of special information and data relevant to selecting specific sites
and guidelines for determining locations of sites for satisfying specific objectives
are provided in a series of appendices. A bibliography, conveniently arranged
according to specific subject areas, is included.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Meld/Group
Sulfur Dioxide
Site Exposure Criteria
Site Selection
Monitoring Objectives
Air Pollution
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report/
Unclassified
21. NO. OF PAGES
180
20 SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
P-19
------- An error occurred while trying to OCR this image.
------- |