\ <3
Pt5
EPA-450/3-76-027
March 1976
ANALYSIS
POPULATION EXPOSURI
TO AIR POLLUTION
NEW YORK-NEW JERSEY
CONNECTICUT
TRI-STATE REGION
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
>' REPRODUCED BY ~" """
NATIONAL TECHNICAL !
INFORMATION SERVICE
U. S. DEPARTMENT OF COMMERCE
SPRINGFIELD, VA. 22161
-------
j TECHNICAL REPORT DATA
(Please read Inttnietioiu an the reverse before completing)
1. REPORT NO.
EPA-450/3-76-027
2.
4. TITLE AND SUBTITLE
Analysis of Population Exposure to Air Pol
the New York-New Jersey-Connecticut Tri-S
7. AUTHOR(S)
Yuji Horie and Arthur C. Stern
9. PERFORMING ORGANIZATION NAME AND ADDRESS
University of North Carolina
Chapel Hill, N.C, 27514
12. SPONSORING AGENCY NAME AND ADDRESS
Monitoring and Data Analysis Division (MD-
Office of Air Quality Planning and Standar
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
15. SUPPLEMENTARY NOTES
3. RECIPIENT'S ACCESSION" NO.
6. REPORT DATE
. . . March 1976
lutlOn in 6. PERFORMING ORGANIZATION CODE
tate Region
8. PERFORMING ORGANIZATION REPORT HO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
R803461-01-0
13. TYPE OF REPORT AND PERIOD COVERED
14) Final
(JS 14. SPONSORING AGENCY CODE
EPA-OAQPS
IB. ABSTRACT /\ population exposure methodology has been developed and applied to total
suspended particulate (TSP) in the NY-NJ-Conn Tri -State Region. Ambient TSP data pro-
duced by 72 monitoring stations, 1971 to 1973, were used for the analysis of popula-
tion exposure to TSP. Census data are aggregated into 215 points to form a demographic
network. The monitored air quality data are spatially interpolated to each demographic
network point to calculate a local population exposure.
Annual and quarterly geometric mean concentrations are used to estimate long-term
population exposure to TSP. Long-term exposure is characterized by a population dosage
spectrum that indicates a population distribution of exposures at various mean concen-
trations. Population average air quality is computed to indicate representative air
quality levels. A health risk index indicates a percentage of the population exposed
to air pollution above the annual standard.
Percenti le concentrations are used to estimate short-term population exposure.
Short-term exposure is characterized by a population-at-risk spectrum that indicates
a population distribution for various exposures to air pollution above the 24-hour
standard. A population-at-risk index indicates a percentage of time that an average
person in the region is exposed to air pollution above the 24-hour standard.
Methods of forming the optimal subnetwork out of an existing monitoring network are
also explored with respect to the objective of minimizing the error in estimating
exposure of the population to air pollution.
17.
KEY WORDS AND DOCUMENT ANALYSIS
1. DESCRIPTORS
Air Pollution
Air Quality Monitoring
Interpolation
Exposure
Optimization
18. DISTRIBUTION STATEMENT
Unlimited
EPA Form 2220-1 (9-73)
b. IDENTIFIERS/OPEN ENDED TERMS
19. SECURITY CLASS (This Report)'
Unclassified
20. SECURITY CLASS (THbpage)
Unclassified
c. COSATI Field/Group
21. NO. OF PAGES
22. PF
PRICES SUBJECT TO CHANGE
-------
EPA-450/3-76-027
ANALYSIS
OF POPULATION EXPOSURE
TO AIR POLLUTION
IN NEW YORK-NEW JERSEY-
CONNECTICUT
TRI-STATE REGION
by
Yuji Horie* and Arthur C. Stern
Department of Environmental Sciences and Engineering
University of North Carolina at Chapel Hill
Chapel Hill, North Carolina 27514
Grant No. R803461-01
F,PA Project Officer: Neil H. Frank
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Kesearch Triangle Park, North Carolina 27711
March 1976
, Research Scientist, Technology Service Corporation, Santa Monica, California.
-------
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 (MD35) , 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 Department of Environmental Sciences and Engineering, University
of North Carolina at Chapel Hill, North Carolina 27514, in fulfillment
of Grant No. R803461-01. The contents of this report are reproduced
herein as received from the Department of Environmental Sciences and
Engineering, University of North Carolina at Chapel Hill. The opinions,
findings, and conclusions expressed are those of the author and not
necessarily those of the Environmental Protection Agency. Mention of
company or product names is not to be considered as an endorsement
by the Environmental Protection Agency.
Publication No. EPA-450/3-76-027
-------
Table of Contents
Page
I. INTRODUCTION 1
II. AIR QUALITY AND POPULATION DATA 3
2.1 Population Data 4
2,2 Air Quality Data 5
2.3 Interfacing Population and Air Quality Data Sets 7
III. FORMULATION OF POPULATION EXPOSURE 10
3.1 Parameters Based on Mean Concentrations 11
3.2 Parameters Based on Percentile Concentrations 13
IV. LONG-TERM POPULATION EXPOSURE 14
4.1 Air Quality Indices 15
4.2 Dosage and Population Dosage Spectrum 16
V, SHORT-TERM POPULATION EXPOSURE 17
5.1 Risk Probability Mapping 18
5.2 Population-At-Risk Spectrum 20
VI. ANNUAL POPULATION EXPOSURE 22
6.1 Trend in Mr Quality Indices 23
6.2 Changes in Dosage Spectra 25
VII. EMPIRICAL AIR MONITORING OPTIMIZATION 26
7.1 Rank-Order of Monitoring Stations 27
7.2 Performance of Sub-Networks 29
VIII. RESULTS AMD DISCUSSION 31
REFERENCES 33
APPENDIX A - Analysis of Interpolation formulae 34
APPENDIX B - Regional Risk and Population-at-risk indices 39
APPENDIX C - Mathematical Formulation of Schemes I, II, and III 41
-------
I. INTRODUCTION
A Population Exposure Approach , as contrasted with an Air Quality
Approach, for reporting ambient air quality in a concise and comprehensive
manner is proposed and developed in this work. Presently, ambient air
quality is reported in terms of mean concentrations and/or percentlle
concentrations which are derived from air monitoring data through standard
statistical manipulation. However, these quantities, of themselves, do
not describe accurately the state of air quality to which a population is
exposed. The population at risk could be any population, such as the tree
population or the cattle population, but in this report "population"
always means "human population".
The purpose of a Population Exposure Approach is to provide a better
method to describe the state of air quality representative of the population
at risk. For this purpose, air quality data'^) are merged with demographic
data^. The New York-New Jersey-Connecticut Tri-State Region was chosen
for the analysis of population exposure, and the pollutant to which they
were exposed chosen was total suspended particulates (TSP). Of 164 air
monitoring stations scattered over the Tri-State Region, 72 stations reported
statistically valid air quality data during the study period, the second
quarter of 1971 (designated 71/2) and the second quarter of 1973 (designated
73/2).
1970 census summaries for county subdivisions and 1970 population
traits prepared by the Tri-State Regional Planning Commission' ' were used
to generate the population data set. These population data were tabulated
for contiguous regional statistical areas. In order to interface the
population data with the air quality data, the population data were aggregated
into 215 standard network points. Each point is supposed to indicate the
-------
- 2 -
local population size, local subpopulations (school-age, elderly, and non-
white populations), and the area in which the local population resides.
TSP air quality at each standard network point was estimated by
interpolating to the point the air qualities observed at the nearest f
three monitoring stations to the network point. The data set of air
quality and population at the 215 standard network points was then used "*
to compute local aid regional values of the various population exposure
Indices and variables discussed below. Geometric mean concentrations are
used to estimate the long-term average exposure of the population to TSP
air pollution„ whereas percentile concentrations are used to describe a
cumulative distribution of the short-term population exposure to various
levels of TSP air pollution.
Long-term population exposure is summarized in: (a) a population
average air quality which indicates the air quality level to which the
population was exposed; (b) a health index which indicates the percentage
of tbe population exposed to air pollution exceeding the United States
federal primary air quality standard; (c) a welfare index which indicates
the percentage of the population exposed to air pollution exceeding the
United States federal secondary air quality standard; and (d) a population
dosage spectrum which indicates the distribution of the population asso-
ciated with various air pollution dose levels.
Short-term -exposure is summarized in: (a) a risk probability which indicates
the time percentage of local population exposure in excess of a given concentration
threshold (tue United States federal 24 hour or annual air quality standards ;u ?!
used for the threshold values), and (b) a population-at-risk spectrum which indi- »
cates the distribution of the population associated with various risk probabilitie
Air pollwrlon effects on health vary among sub-populations such as the
-------
- 3 -
child population and the elderly population. Therefore, the population
exposure parameters described above are determined not only for the whole
population but also for three sub-populations: (a) school-age, (b)
elderly, and (c) non-white. The working-age population was intentionally
dropped from this analysis because of its great daily mobility. A popu-
lation exposure analysis for the working-age population would require use
of a measure of a dynamic population whose size and composition vary with
time, instead of the static population exemplified by the resident
population, on which this report is based.
The air quality trend on an annual base is investigated in Section VI.
The annual geometric mean concentrations observed at 69 air monitoring
stations during 1971 and 1973 were used for the analysis of trends in the
air quality indices noted above. Of the 69 stations, 14 stations failed
to report valid air quality data in one or two quarters during 1971 and
1973. Where there was missing quarterly data in one year, the air quality
data in the corresponding quarters of the other year, i.e. either 1971 or
1973, was substituted.
An empirical method of upgrading an existing air monitoring network
is discussed in Section VII. 130 of the 164 Tri-State Region air monitoring
stations reported statistically valid air quality data during the second
and third quarter of 1973. These 130 stations were used to explore optimal
sub-networks.
II. AIR QUALITY AND POPULATION DATA
The area under study is the New York-New Jersey-Connecticut Tri-State
air quality control region (AQCR) less the eastern half of Suffolk County.
This area is a little smaller than the Tri-State Region t-5' but a little
larger than the New York-North Eastern New Jersey Standard Metropolitan
-------
Statistical Area (Fig. 1). It is comprised of parts of Connecticut, New York, and
New Jersey, i.e. the 19 counties listed in Table I. According to the 1970
census summaries^3', the study area Includes approximately 12,000 square
kilometers (4,600 square miles) and 17.0 million people in and surrounding
New York City. There are 164 air monitoring stations in the area operated
by federal, state or city governments. Of these, 72 stations report to
the Nation Aerometric Data Bank statistically valid air quality data for
TSP measured by Hi-Vol Sampler during the two quarters, 71/2 and 73/2.
The individual statistical areas for the county sub-divisions vary
in area and population and have complex geographic boundaries. Air moni-
toring stations are not distributed uniformly over the area but tend to be
concentrated in heavily populated areas. As a result, there is a very
poor geographical agreement between the distribution of air monitoring
stations and that of census statistical areas. To solve this problem, the
standard network shown in Figure 2 was devised. The standard network
consists of 215 standard network points located by considering the geo-
graphical distribution of the population. The boundaries between one point
and the neighboring points were determined rather arbitrarily but
geographical boundaries were considered in the partitioning process. Air
quality at a standard network point was estimated by interpolating to that
point the observed air qualities at neighboring monitoring stations.
2.1 Population Data
The 1970 census results have been summarized In many ways In print
and on magnetic tape. Of these summaries, the "Population of County Sub-
divisions" wan used as the population data base for this study. However,
they do not include the statistics of the schooJ-age, elderly, and non-white
sub-populations. It is very time consuming to aggregate such subpopulatit.n
-------
- 5 -
statistics of each census tract into subdivisions for each county. In this
study, this impediment was circumvented by using "Population by Age Group"
data previously aggregated by the Tri-State Regional Planning Commission .
The commission also issues a series of computer-produced maps called
(4)
Regional Profilesv '. The Regional Profile computer display of 1970 popu-
lation distribution over the Tri-State Region was used to produce the
standard network by assigning to each sub-region the number of network
points approximately proportional to the population density of that sub-
region. The 215 standard network points thus selected are shown on the
regional map of Figure 2. Each standard network point is intended to
represent its spatial location, its assigned area, and the average
population density In that assigned area. These data are listed in Table
II. The code number indicates the county in which the network point is
located. The corresponding county name can be found from Table I. Table
III lists for each network point the percentage of each sub-population in
the total population.
2.2 Air Quality Data
There are 164 air monitoring stations operated by federal, state,
or city governments in the study area. These stations measure 24 hour
average air quality of TSP with Hi-Vol samplers. The frequency of sampling
is 61 samples per year, or once every 6 days. EPA's National Aerometric
Data Bank (NADB) receives the observed air quality data from the local air
pollution agency and stores them for retrieval and use by a variety of
purposes, such as for a study like this one. NADB's quarterly summaries
of TSP air quality data during the period 1970 to 1974 were examined for
use in this study. Judging from the data retrieved, the performance of
the Tri-State Regional air quality monitoring network is disappointingly
-------
- 6 -
poor. The number of stations consistently reporting statistically valid
air quality data during the five year period is less than 30% of the total
number of stations.
The present analysis of population exposure to TSP air pollution
Includes a comparison of trend in air quality and that in population ex-
posure. Such analysis needs statistically valid air quality data fro« a station
at two points? in time. Among the many possible combinations of quarters
during the thr*« year period examined the number of monitoring stations
satisfying this condition is greatest for the combination of the second
quarters or 1971 Ul/2) and of 1973 (73/2). There were 72 monitoring
stations whose data were valid for both these periods. The geometric mean
concentrations and spatial coordinates of all the 164 stations operating
during the two quarters, 71/2 and 73/2 are listed in Table IV. The
locations of the 72 monitoring stations valid for both periods are shown
in Figure 3 by open circles, the invalid stations by solid circles. The
data set presented in Table IV was used for the analysis of the long-term
population exposure, I.e. for a season, discussed in Section IV. The
percentile concentrations at these 72 valid stations during the same period
are presenteel in Table V. These percentile data were used for the analysis
of ffhort-t.ens« population exposure discussed in Section V.
There are 45 monitoring stations that report statistically valid air
quality dst*» for the entire years of both 1971 and 1973. For these two
years, 14 stations failed to report statistical valid air quality data for
one quarter and 10 stations failed to do so for two quarters. For each of
these 24 imperfect stations, valid data from the corresponding quarter of
the other year was used to replace the data for the invalid quarters. In
this way, th«- air quality data for 69 monitoring stations was generated
-------
- 7 -
and used for the analysis of trend in air quality and population exposure
discussed in Section VI. These data are shown in Table VI.
The annual geometric mean concentration of each monitoring station
was computed by taking a geometric mean of the geometric mean concentrations
during each of the four individual quarters (Table VII). The resulting
value is a little different from the reported value for the annual geometric
mean concentration. However, the difference between the computed and the
reported value is generally less than 0.1 ug/nP. The cause of the difference
is that the computed value is based on the assumption of an equal number
of samples during each quarter while the reported value is based on the
actual number of samples measured during each quarter.
The two consecutive quarters, second and third quarters of 1973 have
the largest number of valid monitoring stations among many combinations of
two quarter periods. There are 130 valid monitoring stations during these
periods (Table VIII). The spatial locations of these stations are given
in Table VIII and also shown in Figure A. The data set of Table VIII was
used for the sensitivity analysis of monitoring networks discussed in
Section VII.
2^3 Interfacing Population and Air Quality Data Sets
To know the exposure of a person to air pollution, the spatial location
of the person and the air quality at his location must be known as a
function of time. In the present study, however, we are not interested in
the actual exposures of individual persons to air pollution, but rather
interested in the ensemble of potential exposures of a large population,
say, a million people. For this purpose, an appropriate estimate of air
quality at each standard network point should be sufficient to make an esti-
mate of population exposure at that particular locale, if the assumption is
-------
- 8 -
made that the population size and sub-population composition will be
approximately stationary over the study period. This assumption should be
good for the analysis of exposure of elderly and school-age populations
because these subpopulations tend to be locationally fixed, i.e. most
school-age children and elderly people stay close to their residence loca-
tions most of the time.
However, the above assumption would not hold for the working-age
population because a substantial percentage of that population spends a
substantial part of their time at their working places where the air
environment may be quite different from that of their residential location.
Thus, the working-age population has been intentionally omitted from this
population exposure analysis. A substantial percentage of the whole popu-
lation and of the non-white population may also spend a substantial part of
their time at locations with an air environment different from that of
their residence. However, such percentages of the total population and of
the non-white population would certainly be smaller than that for the working-
age population.
A look at: the census data^3' for the Tri-State Region Indicates that
although the daytime population of New York City is significantly greater
than the nighttime population, regionwide the population which commutes to
work from their residential location is small compared to the total population.
Nassau County is a case in point. The census data show that, of the total
population (1,428,077), the working population that commutes to working places
outside of Nassau County is 263,592 (i.e. all workers [558,931] less those
working in Nassau County [295,3391) or 18% of the total population. The
percentage of such a population to the total population of the Tri-State
region has been assumed to be somewhat lower than 18%, e.g. 10%. A 10% error
-------
- 9 -
in the estimate of population should not invalidate analysis of exposure
of the total population or the non-white population, when compared with the
error in the estimate of air quality, which may be estimated as around
10 * 20%.
As mentioned earlier, the spatially distributed population over the
study area was aggregated at 215 standard network points. Therefore, all
the information on air quality necessary for the population exposure
analysis was assumed to be contained in the air quality at the 215 network
points. Their air quality was estimated from the air quality observed at
the 72 valid air monitoring stations. The air quality at a network point
was estimated by interpolating the observed air quality at the three near-
est neighboring stations to that point as
3 -2 3 -2
C,| » £ Cj dj_ It d^ for di $ 0
C.j » Ci for d£ » 0
where Cj is the concentration estimated at j-th network point (X.t, Y.I),
GI (i ™ 1, 2, 3) are the concentrations observed at the three nearest
neighboring stations, l~th (i = 1, 2, 3) air monitoring stations (Xj,
around the j-th network point, and d± is the distance between the i-th
monitoring station and the j-th network point, i.e.
X.j)2 + (Y± - Yj)2 (2)
The interpolation formula given by Equation (1) was arrived at after a
careful and detailed analysis was made of various interpolation formulae
and their performance, A discussion of this analysis of interpolation
formulae is given in Appendix A.
-------
- 10 -
III. FORMULATION OF POPULATION EXPOSURE
There are national air quality standards for partlculate air pollution
both for short-terra (24 hour average concentration) and long-term (1 year
average concentration). The reasons of having an air quality standard for
two different averaging times is that adverse effects can occur from either
a short exposure to higher concentrations, or longer exposure to lower
concentrations. These relations are explained by the dose-response curve.
For each averaging time, there are two air quality standards, a primary
standard to protect public health and a secondary standard to protect
public welfare.
The present analysis of population exposure to air pollution is based
on these same basic dose and threshold concepts. An air pollution dose of
a person at r_ during a time period T is given as
T
DOSE (_r) « /0 C (r, t) dt (3)
When the wean concentration of air quality at r_ over T is estimated from
air monitoring data. Equation (3) can be expressed as:
DOSE (r) « T €„,<£) (3-a)
Equation (3-t) indicates that when exposure time Is given, the air pollution
doss of a person is estimated from the mean concentration at his location.
Tti this report Che mean of pollution concentrations at the location of a
population over a given time is called "dose D(r)" i.e.
(A)
Equation (4) says that because a population which resides at r; stays close
to their resident location for most of time, the ensemble average of air
pollution doses of its individual members is given by the mean of concentre
-------
- 11 -
tions over the exposure time.
3.1 Parameters Based on Mean Concentrations
A spatial average concentration may be computed by
AQS - /£ D(r) dr/Aj, (5)
where AQ is the total area under study. The spatial average air quality
has been used to indicate the air quality representative of a given region.
However, the air quality representative of the populace which resides in
the region may be better expressed by the population average air quality,
AQp - /s D(r) p(r) dr/Po (6)
where p(r)ia the population density at r_ and Po is the total population
of the region.
The air quality indices AQS and AQp described above Indicate the air
quality for the entire region or population under study. Another kind of
air quality index may be defined by the percentage of population exposed
to air pollution exceeding a given level. A health index, HI, is defined by the
percentage of the population exposed to air pollution exceeding the primary
air quality standard D^ (the health standard).
HI - /£ H(r, D) p(r) dr/Po (7)
where H(r^, D) is a discriminant function defined as'')
H(r, D) - 1 if D(r) > Dh
(8)
H(r, D) « 0 otherwise
~*~ ^r
Similarly, a welfare index, HI, may be defined by the percentage of the populatior
exposed to air pollution exceeding the secondary air quality standard 1^,
(the welfare standard).
-------
- 12 -
WI - /r W(r, D) p(r) djr/P0 (9)
where W(jr, D) - 1 if D(r) > I\, )
( (10)
W(jr, D) » 0 otherwise I
The discretized forms of Equations (5), (6), (7) and (9) are written
as
AQS - E Di AAiMo (5-a)
i
AQp » E DI pi AAi/PQ (6-a)
HI - E H(DA) PjL AAt/P0 (7-a)
WI - I WCD-t) P1 AAi/P0 (9-a)
where AAj Is the area represented by the i-th standard network point, Dj
is the annual mean concentration at the 1-th point, and Pi is the population
density averaged over the area AAi.
A dosage isopleth map, similar to an air quality isopleth map, can be
obtained by using a threshold function such that
N(r, D) - 1 if D(r> > D*
(ID
N(£, D) - 0 otherwise
where D* is the dosage threshold. For a given threshold, one can draw a
'iosage isopleth by plotting all the points with N(ir, D) - 1. Using the same
threshold function, one can also compute the dosage spectrum S(D*) and the
population dosage spectrum P(D*) that arc defined as
" /r N(r, D) dr/A,, (12)
P(D*) - /r N(r, D) p(£) dr/P0 (13)
-------
- 13 -
Equation (12) says that a fraction of the total area, S(D*), is polluted
more than D*, or io receiving a pollution dosage greater than D*T where
T is 8760 hours. Equation (13) says that a fraction of the total population,
P(D*), is exposed to air pollution exceeding D*, or is receiving a pollution
dosage greater than D*T.
3.2 Parameters Based on Percentile Concentrations
The individual values of 24 hour Hi-Vol measurements of TSP can be
sorted in descending order of magnitude and normalized in percentile con-
centrations. These percentile concentrations can be used to evaluate the
short-term (24 hour) population exposure to air pollution in relation to
national air quality standards.
At a given location, the percentile concentrations may exceed a level
of, say, the 24 hour primary standard. Then, the percentage of time during
which local air pollution exceeds the prescribed level may be associated
with the risks to the local populace incurred by the adverse effects of
such excess. The risk probability is defined as the percentage of time
during which local pollution concentration exceeds the level prescribed by
each of the four air quality standards; the annual primary, annual secondary,
24 hour primary, and 24 hour secondary standards.
Pr(r) - fg(r) (14)
where fg(r) is the percentile that corresponds to C(_r, f) « Cs, a concentration
threshold designated by one of the four air quality standards. By plotting
fs(r) against _r, one can draw a risk probability map that shows a spatial
distribution of risk probability.
Let us define another threshold function such that
M(r, f.) - 1 if fs(r) > f*l
~ ( (15)
M(_r, fs) « 0 otherwise j
-------
- 14 -
where f* is a frequency threshold. Using this threshold function, one can
compute the risk spectrum R(f*|cs) and the population-at-risk spectrum
PR(f*|C8), defined as
R(f*|€8) - /r M(r, fg) dr/A0 (16)
PR(f*|ca) - /Mf seasonal exposure of the population whose compositions are
presented in Tables IT and III. Although there is no air quality standard
for seasonal mean concentration, the same numerical values as the annual
air quality standards (primary and secondary) were assumed as hypothetical
air quality standards for quarterly geometric mean concentrations.
The air quality at each of the 215 standard network points was estimated
trom interpolations of the observed air qualities at the 72 valid air
-------
- 15 -
monitoring stations during 71/2 and 73/2 by using the interpolation formula,
Equation (1). The resulting concentration isopleths are shown in Figures
5 and 6 for 71/2 and 73/2, respectively. A comparison of the two figures
shows that the air quality for 73/2 was, in most parts of the Tri-State
Region, much better than that for 71/2.
4.1 Air Quality Indices
The interpolated concentration at a standard network point is assumed
to represent the average air quality over the area represented by AA^ whose
value for each network point is presented in Table II. Under this assump-
tion, the spatial average concentration AQS, population average concentration
AQp, health index HI, and welfare index WI were computed for the total
population using Equations (5-a), (6-a), (7--a) and (9-a), respectively
(Fig. 7). Indices AQp5 HI and WI can be used to differentiate air qualities
to which individual sub-populations such as school-age, elderly, and non-
white population of the study area are exposed. The sub-population data
for the Tri-State Region are presented in Table III. The spatial
average air quality AQS is constant among the sub-populations, but the
population average air quality AQp reveals that the non-white and the elderly
population are exposed to poorer air quality than the total population and
the school-age population exposed.
Air quality improvement may be quantified better by the health index
HI and the welfare index WI than by the average air qualities AQ8 and AQp.
The health index indicates a percentage of the population exposed to air
pollution exceeding the national primary air quality standard. Figure 7
shows that during the study period such percentages of the total population,
the school-age population, the elderly population, and the non-white popula-
tion all decreased, between 1971 and 1973, from 49% to 37%, 45% to 33%, 54%
to 42%, and 69% to 54%, respectively. Knowing that the sizes of these four
-------
- 16 -
populations in the study area are, respectively, 17.0, 4.0, 1.8, and 2.7
million persons, the number of persons exposed to a below-standard air
quality were reduced from 1971 to 1973 from 8.3 millions to 6.3 millions
for the total population, from 1.8 millions to 1.3 millions for the school-
age population, from 1.0 million to 0.8 million for the elderly population,
and from 1.9 millions to 1.5 millions for the non-white population. A
similar Interpretation can be made for the welfare index.
4.2 Dosage and Population Dosage Spectrum
Ositig the threshold* function N(jr, D) defined by Equation (11), the
interpolated air quality and the size of the population at each standard
network point were stratified according to a level of air pollution dose
D*. Then, the dosage spectrum S(D*) and the population dosage spectrum
P(D*) defined, respectively, by Equations (12) and (13) were computed by
taking the sums, I N(r, D) AA^ and I N(_r, D) p^ AAj_, over the entire study
area.
The dosage spectrum is plotted in Figure 8. It shows the fraction of
the study area in which the TSP seasonal geometric mean concentration ex-
ceeds any stated dose level. Prom the figure we can see that in the second
quarter of 1971 the primary air quality standard (annual mean 75
level was exceeded in 15. 5% of the area, while in the same quarter of 1973
it was exceeded in 10.3% of the area. Similarly, the secondary air quality
standard (annual mean 60 yg/m^) level was exceeded in 56. 0% of the area in
71/2, but only in 29.4% of the area in 73/2. The air quality improvement
was «ore pronounced in the higher dosage range. For instance, the percentage
of area exposed to mean concentrations equal to or greater than 100 yg/nT
was reduced from 5.7% in 71/2 to 0.5% in 73/2.
The population dosage spectrum for the total population is plotted in
Tlgure 9. It shows the fraction of the total population exposed to air
-------
- 17 -
*
pollution exceeding any stated dose level. From the figure we can see that
the number of persons exposed to air pollution exceeding any stated level
was reduced substantially from 71/2 to 73/2. The air quality improvements
over the higher concentration range are again emphasized in this figure.
The percentage of population exposed to a mean concentration equal to or
greater than 100 pg/m3 dropped from 22.5% in 71/2 to 1.3% in 73/2. Knowing
the total population of the study area, 17.0 millions, the population exposed
*j
to above 100 yg/m mean concentrations dropped from 3.8 millions in 71/2 to
0.2 millions in 73/2* Similar interpretations can be made for other dose
levels.
Distribution of the population dosage spectrum for sub-populations
are plotted in Figures 10 and 11, which reveal that, in the Tri-State Region,
the non-white population and elderly population were exposed to a dirtier
air than the total population. The school-age population benefitted most
by cleaner air. Population dosage spectra of 71/2 and 73/2 are plotted in
Figures 10and 11, respectively. A comparison of these figures shows that
all four populations benefitted by air quality improvement from 71/2 to
73/2. In particular, the improvement for the non-white population was the
greatest among the four populations. It should be noted that the non-white
population which was highest of the four population groups exposed to air
pollution exceeding 100 yg/tn^ in 71/2 dropped to the lowest among the four
populations in 73/2,
V. SHORT-TERM POPULATION EXPOSURE
The adverse effects of particulate air pollution is caused not only
by long-term exposure of persons or material to air pollution but also by
short-term exposure to more severe pollution. The dose threshold above
which there should be a noticable adverse effect from short-term exposure
-------
- 18 -
to partlculate air pollution should be defined by the 24 hour air quality
standards (primary and secondary standard). The percentile concentrations
of 24 hour Hi-Vol measurements of total suspended particulate matter can
be used to describe the short-term exposure of the population at each
locale. In the present study, short—term exposure of the population to
TSP air pollution Is summarized by the risk probability which Indicates the
percentage of time exposure to air pollution exceeding any given level of
concentration, and the population-at-risk spectrum which indicates the
distribution of the population exposed to air pollution exceeding any given
level for various percentages of time.
5.1 Risk Probability Mapping
The TSP concentrations observed at each air monitoring station were
rank-ordered and tabulated in a percentile form as shown in Table V. At
each percentile the concentrations observed at the 72 valid monitoring
stations were Interpolated to each of the 215 standard network points using
thf Interpolation formula of Equation (1). In this way, percentile concen-
trations were computed for each of the 215 standard network points. Then,
the risk probability defined by Equation (14) was determined at each network
ooiot from these Interpolated percentile data.
Risk probabilities of total suspended particulate concentrations
3
exceeding the level of the primary 24 hour air quality standard (260 yg/m )
are plotted in Figure 12 for the air quality data of 71/2. It can be seen
that people residing in some areas of Bergen, Hudson, New York, and Kings
counties were exposed for about 5 to 10% of time to air pollution exceeding
the level of the primary 24 hour air quality standard. People residing in
the rest of t:he areas were not so exposed.
Similar risk probabilities In excess of the primary 24 hour air
qaality standard in 73/2 are plotted in Figure 13. A comparison of Figures
-------
- 19 -
12 and 13 indicates that areas having a risk probability of greater than
5% were reduced substantially from 71/2 to 73/2. In 73/2, such areas were
limited to only parts of Richmond and Hudson counties.
The risk probabilities of TSP daily concentrations exceeding the level
of the secondary 24 hour air quality standard (150 yg/m-*) are plotted in
Figures 14 and 15 for 71/2 and 73/2, respectively. The areas having a risk
probability exceeding the secondary air quality standard for more than 5%
of time were much more widespread than the corresponding areas for the
primary air quality standard. The areas at such risk in 71/2 extended over
most of New York City and a part each of Hudson, Passaic, Bergen, Rockland,
Westchester, Nassau and Fairfield counties. In 73/2, the risk areas shrunk
substantially from those of 71/2.
The risk probabilities of daily concentrations exceeding the level of
75 yg/nH are plotted in Figures 16 and 17 for 71/2 and 73/2, respectively.
These risk probability maps are to be compared with the isopleth maps of
annual geometric mean concentrations in 71/2 and 73/2 shown, respectively,
in Figures 5 and 6. As seen from Figure 5, the areas exceeding 75 yg/m in
24 hour average concentrations in 71/2 were limited to around New York City.
However, Figure 16 shows that almost all the people in the Tri-State Region
were exposed, for at least 25% of time, to TSP daily concentrations exceeding
75 yg/m . The greatly reduced risk probabilities in 73/2 as seen from Figure
17 reflect the air quality improvement from 71/2 to 73/2.
The risk probabilities of TSP daily concentrations exceeding the level of
o
60 yg/m are mapped in Figures 18 and 19 for 71/2 and 73/2, respectively. The
majority of the Tri-State Region experienced TSP concentrations in excess of 60
/m-* at least 50% of time in 71/2, whereas in 73/2 the areas experienced such higV
-------
- 20 -
population exposure shrunk to about one third of the total area. However,
in 73/2 the entire Tri-State Region still experienced concentrations in
excess of 60 yg/m3 at least 5% of time.
5.2 Population-At-Rlsk Spectrum
The risk probability fg(r) is numerically integrated with respect to
an incremental area d£ or AA, after which the integrals are stratified
according to the frequency threshold f*. The results are summarized in
the risk spectrum R(fft|cg) defined by Equation (16) and in the population-
at-rlsk spectrum PR(f*|cs).
Figure 20 shows the risk spectrum distributions for 71/2 air quality
data while Figure 21 is for 73/2 air quality data. The abcissa R(f*|cs)
indicates the fraction of the study area which experiences 24 hour average
concentrations exceeding the concentration threshold Cs given by either the
24 hour primary standard, the 24 hour secondary standard, the annual primary
standard, or the annual secondary standard. The ordlnate f* indicates the
percent of time during which exposure to 24 hour average concentrations
exceeds the concentration threshold C8 given by one of these four standards.
The graphic area below each curve quantifies the status of the ambient air
quality under study as to the extent the ambient air quality as measured
by the 24 hour average concentrations meets the air quality goal Cs designated
by the 24 hour primary, the 24 hour secondary, the annual primary, or the
annual secondary air quality standard.
It can be seen from Figures 20 and 21 that the higher the concentration
threshold Cs, the lower the corresponding curve of the risk spectrum R(f*!cs).
This means that the higher the concentration threshold, the smaller the area
and time in excess of the threshold. In these figures, the air quality
improvement from 71/2 to 73/2 can be visualized by the smaller area below
the curve of 73/2 as compared to that below the corresponding curve of 71/2.
-------
- 21 -
The air quality Improvement during this period is more pronounced in the
risk spectrum curves for the annual standards (C8 - 75 and 60 ug/m ) than
in the curves for the 24 hour standards (Cg - 250 and 150 ug/m^).
Let us define a regional risk index RI such that
RI(CS) - /J R(f*|cs) df* (18)
As more fully described in Appendix B, the regional risk index indicates an average
percentage of time at which a typical locale within the region is exposed to air
pollution exceeding the air quality standard Cs. There are two extreme situations;
"total compliance"' corresponding to RI = 0, and "total violation" corre-
sponding to RI « 1, RI » 0, which is given by the horizontal axis of
Figures 20 and 21, indicates that the air quality of a given air-shed meets
the air quality standard C§ everywhere all the time. On the other hand,
RI » 1, which is givejs by a horizontal line through R(f*JG8) - 1 in the
above figures, indicates that the air quality of the air-shed exceeds the
standard everywhere all the time.
The regional risk index RI enables us to quantify the degree of excess
of the standard over an entire air-shed or Air Quality Control Region (AQCR)
based on percentile concentration statistics. The area below each curve of
Figures 20 and 21, shows the improvement in short-term air quality from 71/2
to 73/2 to be RI =• 0.025 to 0.025 (no change) for the 24 hour primary stan-
dard, RI = 0.042 to 0,044 (slight deteriolation) for the 24 hour secondary
standard, RI » 0,328 to 0.192 for the annual primary standard, and RI - 0.492
to 0.379 for the annual secondary standard (Fig. 24).
The population-at-risk spectrum distribution is shown in Figures 22 and
23 for 71/2 and 73/2 air quality, respectively. The abscissa PR(f*|cs)
indicates the fraction of the population exposed to air pollution exceeding
the concentration threshold Cg in daily average concentrations for a given
-------
- 22 -
>'
percentage of time f*. The improvement in terms of population exposure
to
during the period is more pronounced in the populatlon-at-rlsk spectrum
curves for the annual standards (C8 - 75 and 60 yg/m^) than in the curves
for the 24 hour standards (C8 » 250 and 150 yg/m^).
Similar to the regional risk index RI, a population-at-risk index
PRI is defined as
PRI - /J PR(f*|C8) df* (19)
As more fully described in Appendix B, the population-at-rlsk index Indicates thc-
average peteentage of time during which an average (or typical) person in an air-
shed or AQCR is exposed to air pollution exceeding the air quality standard
C8. Again PRI » 0 corresponds to "total compliance" and PRI - 1 to "total
violation". The improvement In population exposure over an entire air-shed
or AOC1 can be quantified by PRI from the available percentile concentration
statistics of air monitoring stations in the area.
The improvement in population exposure over the study area from 71/2
to 73/2 is quantified as PRI - 0.025 to 0.025 (no change) for the 24 hour
primary standard, PRI - 0.073 to 0.053 for the 24 hour secondary standard,
PRI - 0.466 to 0.292 for the annual primary standard, and PRI - 0.643 to
0,473 for the annual secondary standard (Fig. 24).
VI. ANNUAL POPULATION EXPOSURE
There was no good data base of annual geometric mean concentrations for
conducting analysis of long-term exposure of the population to TSP air pollution
1x> the T*"'--State Region. Since the long-term air quality standards are given
for annual geometric mean concentrations, the analysis of long-term population
exposure made in Section IV based on the quarterly geometric mean concentrations
may be misleading. Therefore, annual air quality data sets were created from
-------
- 23 -
45 stations with complete data and 24 stations with missing data in one or
, ^r's
two quarters during the entire year of 1971 and 1973. The quarterly -**
geometric mean concentrations of these 69 stations are presented in Table
VI. Each missing quarterly concentration in one year was replaced by that
of the corresponding quarter of the other year.
The annual geometric mean concentration for each air monitoring station
is computed as
-------
- 24 -
Using Equations (5-a), (6-a), (7-a) and (9-a), the four long-term
air quality indices described earlier were computed for the annual air
quality data. The spatial average concentration, AQ8, population average
concentration AQp, health index HI, and welfare index WI are all plotted
in Figure 28 for the four different populations; total population, school-
age population, elderly population, and non-white population.
All the indices show a substantial improvement of air quality from
1971 to 1973. The spatial average air quality AQs remains the same among
the four populations, while the population average air quality AQp reveals
that the non-white and elderly populations were exposed to dirtier air
than that to which the total population and the school-age population
were exposed.
Air quality improvement is conveniently measured by the health index
HI and the welfare Index WI. The health index tells the percentage of the
population exposed to air pollution exceeding the annual primary standard,
whereas the welfare index tells the percentage exposed to air pollution
exceeding the annual secondary standard. Figure 28 shows that during the
period from 1971 to 1973 the percentage of the population exposed to below
primary standard air quality decreased from 49% to 32% for the total popu-
lation, 44% to 27% for the school-age population, 54% to 38% for the elderly
population, and 64% to 50% for the non-white population. Similarly, the
percentage of the population exposed to below secondary standard air quality
In 1971 and 1973 were, respectively, 82% and 56% for the total population,
80% and 51% for the school-age population, 83% and 61% for the elderly
population, and 88% and 71% for the non-white population.
The sizes of the four populations within the study area (which is a
little smaller than the Tri-State Region) are 17.0 millions for the total
population, 4,0 millions for the school-age population, 1.8 millions for the
-------
- 25 -
elderly population, and 2.7 millions for the non-white population. There-"
fore, the number of people exposed to a below primary standard air quality
decreased from 1971 to 1973 from 8.3 millions to 5.4 millions for the total
population, 1.8 millions to 1.1 million for the school-age population, 1.0
million to 0.7 millions for the elderly population, and 1.7 millions to 1.4
million for the non-white population. Similarly, the number of people exposed to
below secondary standard air quality decreased from 1971 to 1973 from 13.9 mil-
lions to 9.5 millions for the total population, 3.2 millions to 2.0 millions for
the school-age population, 1.5 millions to 1.1 million for the elderly popula-
tion, and 2,4 millions to 1.9 millions for the non-white population.
6.2 Changes in Dosage Spectra
The dosage spectrum is plotted in Figure 29. It shows the fraction
of the area under study ±n which the TSP annual geometric mean concentration
exceeds any stated dose level. It can be seen from the figure that In 1971
the primary standard (75 pg/mr) was exceeded by 14.7% of the area, while in
1973 by 6.7% of the area. The secondary standard (60 yg/m-*) was exceeded
by 67% of the area in 1971 and in 1973 only by 27% of the area. The air
quality improvement during the same period is more pronounced in the higher
dosage range. For instance, the percentage of the area polluted by at least
an annual geometric mean of 90 ug/m-* dropped from 5.5% in 1971 to a. mere
0.2% in 1973.
The population dosage spectrum for the total population is plotted in
Figure 30. It shows the fraction of the population exposed to air pollution
exceeding any stated dosage level. It can be seen from the figure that the
number of people exposed to air pollution exceeding ar>v stated dose level
dropped substantially from 1971 to 1973. The percentage of the total popu-
lation exposed to below primary standard air quality decreased from 49% in
1971 to 35% in 1973. The percentage of the population exposed to below
-------
- 26 -
secondary standard air quality decreased from 82% to 56% during the same
period. The Improvements are again most pronounced in the higher exposure
range. For Instance, the percentage of the population exposed to an annual
geometric mean concentration of at least 90 ug/m3 dropped from 29% in 1971
to a mere 0,6% In 1973. This means that the number of people exposed to
an annual geometric mean of at least 90 yg/m3 dropped from 4.9 millions in
1971 to 102,000 in 1973.
Figures 31 and 3? show the distributions of population dosage spectra
for various sub-populations. It can be seen from these figures that the
school-agf? population benefitted more by being exposed to a cleaner air in
1973 than in 1971 than the other populations exposed. In both figures the
population dosage spectrum of the non-white population behaved differently
from those of the other populations. Although the non-white population is
generally exposed to dirtier air than the other populations, the percentage
of the non-white population exposed to the dirtiest air was less than for
any other population. The air quality improvement, from 1971 to 1973
benefitted all the four populations by reduced exposure to air pollution.
VII. EMPIRICAL AIR MONITORING OPTIMIZATION
Thers have been many studies on selection of proper air monitoring
sites and on th* number of monitoring stations required for accurately
monitoring air quality in a given area. In this section we explore an
empirical method of updating an existing air monitoring network. In particular
we seek a sub-network whose number of monitoring stations are smaller than
the existing, number, but whose monitoring performance is as good as that of
the existing total network.
The 130 air monitoring stations that reported valid air quality data
during the second and third quarters of 1973 were chosen as the test network:
-------
- 27 -
The air quality data and spatial coordinates of these 130 stations are presented
in Table VIII. The spatial locations of the individual stations are shown in
Figure 3. The data set Riven by Table VIII was used for exploring empirical
methods to improve the monitoring performance of an existing network. The con-
centration isopleth maps for the second and the third quarter of 1973 are
shown in Figures 33 and 34, respectively.
7.1 Rank-Order of Monitoring Stations
If each monitoring station is rank-ordered according to the impact of
its monitored concentration on the performance of the entire monitoring
network, this should tell us of which stations can be removed from the
existing network without significantly impairing network performance, or
where additional stations should be located to improve the performance of
the enlarged network to the maximum extent. The importance of each station
is evaluated first, by the difference between its measured concentration and
the concentration Interpolated from the three nearest neighboring stations
to that station site, second, by the sum of differences between the receptor
concentrations interpolated by using all N stations and those by using (N-l)
stations, and third, by the Jack Knife method (Appendix C).
The first scheme of rank-ordering monitoring stations is based on the
error induced at the site of a station when that station is removed from the
network. The error in concentration is measured by the difference between
the concentration observed by the station and that interpolated from the three
nearest neighboring stations to that station site. The maximum error among
the ten stations at every ten rank interval is plotted in Figure 35. The
error grows nearly linearly up to the 90-th rank and thereafter more rapidly.
The 10 highest rank stations and the 10 lowest among the 130 monitoring
stations are shown in Figure 36. The highest rank station may be said to be
least important to the monitoring network because the concentration at that
station can be estimated correctly from the readings at the neighboring
-------
- 28 -
stations. On the other hand, the lowest rank station nay be said to be
most important to the network because its concentration readings bring to
the network information unavailable from any of the other monitoring stations.
The second scheme of rank-ordering monitoring stations is based on
the sum of errors induced at the 215 standard network points when a station
is removed from the monitoring network. The error at each network point
is measured by the difference between the receptor concentration interpolated
from all N monitoring stations and that from (N-l) stations. The maximum
error among the ten stations at every ten rank interval is plotted in Figure
37. The sum of errors in receptor concentrations initially grows nearly
linearly and thereafter grows exponentially. The 10 highest rank stations
and the 10 lowest among the 130 monitoring stations are shown in Figure 38.
The highest rank station may again be said to be least important to the
monitoring network. However, the reasoning in Scheme II is different from
that in Scheme I. Receptor concentrations around the highest rank station
can be estimated correctly from the readings of neighboring monitoring stations.
Thus, loss of that station would have the least Impact on the performance of
the monitoring network. On the other hand, loss of the lowest rank station
vould bring a significant deterioration in network performance because receptor
concentrations around that station are estimated erroneously from the readings
at the neighboring stations.
The third scheme of rank-ordering the monitoring stations is also based on
tba sum of errors induced at the 215 standard network points when a station is
removed from the monitoring network. The third scheme is different from the
second in that for the K-th rank station the error at each network point Is
measured by the difference between the receptor concentration interpolated from
the (N-K-fl) stations (N stations less the first (K-l) stations) and that from
the (N-K) stations (N stations less the first K stations). The 10 highest:
rank stations among the 130 monitoring stations are shown in Figure 39.
-------
- 29 -
Because Che rank-ordering of monitoring stations according to Scheme III
is quite computational, only the first 68 stations were rank-ordered. The
sum of errors in receptor concentrations using Scheme III grows a little
faster than the linear growth of Scheme II (Figures 37 and 40).
7.2 Performance of Sub-Networks
The entire 164 monitoring stations, of which 130 stations reported
valid air quality data during the 2nd and 3rd quarters of 1973, were first
divided into two classes, one with odd numbered stations and the other with
even numbered stations. This division resulted in 68 valid monitoring stations
for the one class and 62 valid monitoring stations for the other class. The
resulting concentration isopleth maps from the sub-network with odd numbered
stations and that from the one with even numbered stations are strikingly
different from each other as seen from Figures 41 and 42.
Three half size sub-networks were formed by removing the 68 least
important stations from the 130 station network according to Scheme I, Scheme
II, and Scheme III, respectively. The performances of these sub-networks
and those of the sub-networks with odd numbered and even numbered stations
were then compared with that of the total network, by determining the space
average air qualities AQS for the entire study area, each state and each
county from the total network and from each of the five half size sub-networks.
The results are shown in Figures 44 through 47. The meanings of the symbols used
In Figure 44 and the following three figures are found from Figure 43. For each
averaging area, there are two vertical bars Indicating the second quarter
values by the left bar and the third quarter values by the right bar. The
distance between the longer horizontal bar and each symbol on a vertical bar
indicates the relative error in the estimate of AOS by that particular sub-
network.
It can be seen from Figure 44 that the performances of sub-networks with
-------
- 30 -
odd numbered and even numbered stations are poorer than those of sub-networks
formed by Schemes I, II, and III. Among the sub-networks by the three
schemes, the sub-network formed by Scheme III out-performs those by Scheme
I and Scheme II. The space average air qualities estimated from the sub-
network by Scheme III is close to that from the total network in every
averaging area. This is in contrast to the fact that the values estimated
from the other sub-networks deviate from the true values progresively as
the averaging; area becomes smaller. Figure AS shows the performances of
the five sub-networks in estimating population average air quality AQ_. The
performance of each sub-network is similar to that revealed In estimating AQS.
The health indices HI and the welfare indices WI estimated from the
total network and from each of the five sub-networks are plotted in Figures
46 and 47, respectively. A comparison between these figures and the previous
two figures indicates that the health index and the welfare index are more
sensitive to monitoring network size than is space average air quality and
population average air quality. The values estimated from the sub-networks
with odd numbered and even numbered stations deviate wildly from the true
values at the county level. In contrast to such wild mis-estimates by the
odd and even number station sub-networks, the values estimated from the sub-
network by Scheme III. stay consistently close to the true values. The sub-network
by Scheme II out-performed that by Scheme I in the estimates of AQS and AQp
but under-performed it in the estimates of HI and WI.
Scheme III, the Jack-Knife Method, is useful to identify stations which
contribute minimally among the existing stations and to form an optimal
sub-network from the existing network. Further, the optimal half-size sub-
network does estimate the four indices, AQS, AQp, HI, and WI down to each
county level with an acceptable accuracy.
-------
- 31 -
VIII. RESULTS AND DISCUSSION
A population exposure approach as contrasted to an air quality approach
has been explored in order to report the state of ambient air quality more
meaningfully for the population exposed to such air quality. Although the
final products of this work turned out to be quite different from those
initially anticipated, they appear to be useful for reporting air monitoring
data in a more comprehensive manner than that currently used by control
agencies.
Ways have been demonstrated for merging air quality data with demographic
data to estimate the degree of exposure of a population and its components
to air pollution. A methodology for quantifying population exposure using
both mean concentration and percentile concentration statistics has been
developed. Three indices for reporting long-term population exposure; popu-
lation average air quality, a health index, and a welfare index have been
proposed. New dosage spectrum and population dosage spectrum concepts have
been proposed to describe long-term population exposure comprehensively but
yet concisely.
A method for utilizing percentile concentration statistics for estimating
short-term population exposure has been developed. A risk probability con-
cept is proposed to describe spatial distribution of excess exposure of a
population to air pollution. Risk spectrum and population-at-risk spectrum
are proposed to describe short-term population exposure. A regional
risk index and a population-at-risk index are developed to report the
improvement or deterioration of short-term air quality over a large area.
An empirical approach to improving an existing air monitoring network
has been explored. Rank-ordering of monitoring stations according to their
impact on network performance has proven to be useful to identify those stations
-------
- 32 -
which contribute maximally and those which contribute minimally among the
existing stations. A Jack-Knife method, based on receptor concentrations,
appears to be useful for forming an optimal sub-network from an existing
network.
-------
- 33 -
REFERENCES
1. Zupan, J. M., "The Distribution of Air Quality in the New York Region,"
Resources for the Future, Inc., Washington, D. C., 1973.
2. "Computer Printouts of SAROAD Air Quality Data in the Tri-State Region,"
Private Communication with Neil Frank, USEPA, OAQPS, January 1975.
3. "Census Tracts," Bureaus of the Census, U. S. Dept. of Commerce,
PHC(1)-30, 96, 145, 146, 149, 206, May 1972.
4. "Computer Printouts of Population Data in the Tri-State Region,"
Private Communication with personnel of the Tri-State Regional Commission,
February 1975.
5. "Regional Profile - 1970 Population Traits," the Tri-State Regional
Commission, Vol. II, No. 1, January 1973.
6. "Directory of Air Quality Monitoring Sites - Active in 1973," USEPA,
OAQPS, EPA-450/2-75-006, March 1975.
7. Csanadys G. T. , "The Dosage-Area Problems in Turbulent Diffusion,"
Atmospheric Environment, Vol. 1, pp. 451-459, 1967.
-------
- 34 -
APPENDIX A Analysis of Interpolation Formulae
There are a number of numerical schemes to obtain a smooth continuous
map from a set of discrete sampling points. One well known scheme is a
(psuedo) linear Interpolation formula as represented by the SYMAP computer
algorithm,''-'' More sophisticated is the g-spllne method that is used for
meteorological mapping.'^) This research explores algebraic interpolation
formulae that do not contain a derivative term. The nearest three neighboring
stations are used for all the interpolation schemes discussed herein.
The psuedo™linear Interpolation formula can be written as:
3 3
imi t, 82 " ± 1 respectively for T2 > i»
-------
- 35 -
and I is the interdistance /(xj - X2)2 + (yi - ya)2 • Referring to Figure
Al, Equation (A2) may be generalized for two dimensional case as:
3
C - E (ej Ci/ri) /J^ (.a±/r±) for r± * 0
(A3)
C . Ct for r± - 0
where sj » ± 1 respectively for h^ > J-j (Fig. Al) . As seen from Figure
Al, there are seven combinations of the signs of s^, i - 1, 2, 3. For
example, the signs in region II are sj_ » -1, 82 * +1. and 83 » +1 because
h^ > fc^, h2 < *2 and h^ < £3.
The straight line that goes through the points (x2, 72) and (x3,
in Figure Al is given bys
sl y + ^1 s "*" cl * 0 (A4)
where aj - - (»2 - X3) , bj = y£ - 73, and GI » x2 Y3 - X3 y2- The line
parallel to the above through the receptor point (xr, yr) is given by
ar y H- br x 4- cr « 0 (A5)
where ar « a1$ br » bi, and cr « (x2 - X3> yr 4- (ya - y2) xr. The heights
&1 and hi are given by
bl xl-»
(A6)
and
71 + bi xx + cr
- - - (A7)
/&l2
The heights &2 and h2, and £3 and 113 can be computed by equations similar
-------
- 36 -
to Equations (AA) through (A7).
The performance of the (true) linear interpolation formulae [Eqn. (A2)
or (A3)] is compared with that of the pseudo-linear interpolation formula
[Eqn. (Al)] in Figures A2, A3 and AA. Figure A2 is the comparison of the one-
dimensional case, while Figures A3 and AA provide a comparison of the
•»
two-dimensional case. In the one-dimensional example CA - 80 yg/m and
Cjj - 20 yg/m* are placed A length units apart on the x-axls, while, in the
two-dimensional example, CA - 85 yg/m3, CR - 15 yg/m , and Cc - 60 yg/m3
are placed in the x-y plane at (5.0, 8.0), (2.0, 2.0), and (8.0, 2.0),
respectively. It can be seen from Figures A2 and A3 that the (true) linear
Interpolation formula would either overestimate or underestimate receptor
concentrations when applied to extrapolation. On the other hand, Figure AA
shows that the pseudo-linear Interpolation formula does not retain the
monitored concentrations C^, Cg, and Cg in the interpolated concentration
field except at those exact points A, B, and C. Thus, neither formula
appears to be proper for Interpolating air monitoring data to get a smooth
concentration map.
Next, parabolic interpolation was examined. The pseudo-parabolic
formula used is:
33
C - z (C±/ri2) / E (1/r,2) for r., *t 0
1-1 1-1 x
(A8)
C « C± for rj - 0
where C, C-i and ri are the same as those for Equation (Al). The (true)
parabolic formula used is:
-------
- 37 -
The performances of the (true) parabolic and pseudo-parabolic inter-
polation formulae are shown in Figure A5 for the one dimensional case. It
can be seen from the figure that the concentrations interpolated by both
formulae vary much more naturally around the monitored concentrations C^
and Cg than do those generated by the (true) linear and the pseudo-linear
interpolation formulae (Fig. A2), The concentrations interpolated in two
dimensional space are shown in Figures A6 and A7 for the (true) parabolic
and the pseudo-parabolic interpolation formulae, respectively. Both
interpolation formulae yield reasonable concentration isopleth maps. How-
ever, Figure A7 shows a better isopleth map than Figure A6 in the sense that
the same numerical values as the monitored concentration values are
distributed over the confined area around each monitoring station whereas
this is not so in Figure A6. The concentrations interpolated by the
pseudo-parabolic interpolation formula exhibit both a representative area
around each monitoring station with numerical values the same as the monitored
concentrations with smooth continuous concentration variation elsewhere.
Thus, this is the formula used in this report as the "geographic inter-
polation formula".
From previous experience with air monitoring networks, it has been
recognized that monitoring stations are more densely distributed in a high
air pollution area than in a low pollution area and, thus, a monitoring
station in a low pollution area covers air quality over a wider area than
that in a high pollution area. As seen from Figures A2 and A5, both the
linear and the parabolic interpolation formulae do not reflect this feature
in their interpolated concentrations. Representation of area can be
incorporated into the interpolated concentrations by introducing a weighting
function into the interpolation formula. For example, a weighted psuedo-
parabolic Interpolation formula may be expressed as:
-------
- 38 -
L2) /,E, U/w, r,2) (A10)
where WA is a weight that is a function of the monitored concentration C^.
The performances of various weighted psuedo-parabolic interpolation
formulae are shown in Figure A8. The weighting functions WA - In C^, wi « C^,
2
and w^ » GI , 1 - 1, 2, 3, are used in the computations. It should be noted
that the intersections between the line A-B and the curves of Interpolated
concentrations are progressively closer to point A as the weight w^ becomes
a stronger function of monitored concentration, i.e. wj » In C^ to w^ « C^
«\
to W£ « C± . These results should be compared with those of the unweighted
interpolation formulae shown in Figure A5 in which the line A-B and the
curves of Interpolated concentrations Intersected at the mid-point between
points A and B. The representative areas Inversely proportional to the
relative magnitudes of monitored concentrations are reproduced in the two
dimensional examples of the weighted interpolation formulae (Figures A9,
A10 and All). Although the representativeness of each monitored concentration
is reconstructed reasonably well in these figures, a serious flaw is found
in the results In that the higher polluted area A and C are now divided by
a narrow strip of lower concentrations. This is obviously an artifact
caused by the weighted interpolation formulae, and is contradicted by our
comraon Interpretation of monitored concentrations. Because this drawback
outweighs its merit of area representation of each monitoring station, weighted
Interpolation formulae were not employed in this study.
-------
- 39 -
APPENDIX B Regional Risk and Population-at-Risk Indices
{1 - R(f*|cs)} and {1 - PR(f*|cs)} are cumulative distribution functions
Of the random variable fs, i.e., P(fs - f*) where f* is a particular
frequency threshold. Let us use the common notation for random variables.
X random variable, corresponding to fs
x particular value of X, corresponding to f*
f(x) probability density function, i.e., f(x) - P(X - x)
F(x) cumulative distribution function, i.e., F(x) - P(X 2 x),
corresponding to {1 - R(f*|cs)} or {1 - PR(f*|C8)}.
As x varies in the range 0 to 1, the well-known statistical relations can be
written as
F(x) - / f (x) dx or f(x) - -^- F(x) (Bl)
o dx
E(X) - /i xf(x) dx « / xdF(x) (B2)
0 o
where E(X) is the mean of a random variable X.
Using the above notations, the regional risk index RI or the population-
at-risk index PRI can be written as
RI or PRI » / (1 - F(x)} dx
o
(B3)
- 1 - f1 F(x) dx
o
Using the "integration by part" method, we can transform
f1 F(x) dx - F(x) x| - f1 ^El xdx
o o o dx
- F(l)«l - /* xdF(x) (BA)
• 1 - E(X)
Substitution of Equation (BA) into (B3) yields
-------
- 40 -
RI or PRI - E(X) (B5)
Therefore, we can interpret RI or PRI as the mean percentage of time at
which a typical locale within the region or an average person in the popu-
lation is exposed to air pollution exceeding the air quality standard C8.
-------
- 41 -
APPENDIX C Mathematical Formulation of Schemes I, II and III
Rank-ordering of monitoring stations according to the impact of their
monitored concentrations on the performance of the air monitoring network was
conducted by introducing the following performance index * ' appropriate for
each of the three schemes.
Scheme I
The first scheme of rank-ordering monitoring stations is based on the
magnitude of errors induced at the station location when that station is re-
moved from the monitoring network. The performance index, P, for the first
scheme may be written as
Pi - [|CU - DU| + |C2i - D21|]/2 (Cl)
where Pj_ is the value of P for the i-th station, C-^ and €2^ are the concen-
trations observed at the i-th station in the second and the third quarter of
1973, respectively, and D^ and D2i are the concentrations interpolated
to the station location from the three nearest neighboring stations to that
station in the two quarters, respectively.
The first rank station is the one having the smallest P. among {P}, the
collection of Pf. The second rank station is the one having the second
smallest PI, and so forth. In general, the K-th rank station is given by
Sv - K-th min {P} K - 1, 2, . . ., N (C2)
* i
where N is the total number of monitoring stations in the network. The rank-
order of each station, the second quarter error, the third quarter error, and
the value of its performance index are all listed in Table Cl.
Scheme II
The second scheme of rank-ordering the stations is based on the sum of
errors induced at all the receptor points when a station is removed from the
-------
- 42 -
monitoring network. The performance index for the second scheme may be
written as
S [|D1:J - D1:j(i)| + |D2j - D2j(±)|]/2
(C3)
where Dj. and D2. are the second and third quarter concentrations interpolated
to the point from the three nearest neighboring stations to the j-th receptor
point, and D^ and D2^ are also the second and third quarter concentra-
tions interpolated to that point. The three nearest stations to compute DJJ
and 02j are selected among the entire N stations, whereas the three stations
to compute Dip ' and D2- are selected among the (N-l) stations, i.e., N
stations less the 1-th station.
The K-th rank station is given by
SK - K-th rain (P) K - 1, 2, . . ., N (CA)
where fP) is the collection of P^ whose value is computed by Equation (C3).
Talle C2 lists the rank-ordered stations, and the mean error in concentration
induced at the effected receptor points, which are defined by a receptor
having an Induced error greater than or equal to 0.01 yg/m^.
Scheme^ JLltl
A
-------
- 43 -
and the first rank station is given by
S - min {P1} (C6)
Equations (C5) and (C6) are essentially the same as Equations (C3) and (C4)
used for the second scheme. However, the difference between the second and
the third scheme appears in the following rank-ordering process.
The performance index to find the second rank station is written as
and the second rank station is given by
STT - min {P11} (C8)
LL I
where D, . and ®2\ are fc^e concentrations interpolated from the three
nearest stations among the (N-l) stations, i.e., N stations less the first
rank station, and DJJ ^ ' ' and D2.t ' are the concentrations Interpolated
from the three nearest stations among the (N-2) stations, i.e., N stations
less the first rank station and the i-th station.
In general, the K-th rank station is given by
ID,/1'" ..... M)-D2J<1'11 ..... "'"llrt (C9)
S,, - min {PK} K - 1, 2, . . . , N-3 (CIO)
^ i
_ (I, II ..... K-l) (I,II,...,K-1) . . fc
where D.. and D21 are the concentrations inter-
polated from the three nearest stations among the (N-[K-1]) stations, i.e.,
N stations less the first (K-l) stations, and D (I»n'' • • »K~1»1) and
D-.^1' '""'K~ ' are the concentrations interpolated from the three nearest
stations among the (N-K) stations, i.e., N stations less the first (K-l)
-------
stations and the i-th station.
The rank-order of monitoring stations by Scheme III can proceed only to
K - N-3 while those by Schemes I and II can complete the entire N stations.
Table C3 lists the first 68 stations and the effected receptors, mean error
and performance index associated with each of the 68 stations.
Network subsets were formed by removing the first 25, 43, and 68 stations
among the rank-ordered stations by Schemes I, II, and III from the 130 station
network. The performance of these network subsets was measured by the mean
error in concentration Induced at the effected receptor points, and the number
of the effected receptors. The results are plotted in Figure Cl. The network
subsets formed by Scheme I causes a greater number of effected receptors and
a greater mean error than those by Schemes II and III. The network subsets
formed by Scheme II causes a lesser number of effected receptors than those
by Scheme III but causes a greater mean error than those by Scheme III. As
a result, the sum of errors induced at all the receptor points is greater for
the network subsets formed by Scheme II than those by Scheme III. Therefore,
Scheme III can be said to be the best to form a network subset among the three
ec'
-------
- 45 -
REFERENCES TO APPENDICES
1. Harvard Laboratory for Computer Graphics, "User's Reference Manual for
Synagraphic Computer Mapping 'SYMAP* Version V". Harvard University,
Cambridge, Massachusetts, 1968.
2. M. J. Munteanu and L, L. Schumaker, "On a Method of Carasso and Laurent
for Constructing Interpolating Splines", Mathematics of Computation,
Vol. 2£, No. 122, April 1973.
3. E. Parzen, "Stochastic Processes", Chapter 1, Holden Day, San Francisco,
California, 1967.
A. W. P. Darby, P. J. Oasenbruggen and C. J. Gregory, "Optimization of
Urban Air Monitoring Networks", Journal of the Environmental Engineering
Division, ASCE, Vol. 100, No. EE3, PP577-591, June 197A.
-------
- 46 -
TABLE I
Code representing county and state
Code
11
21
22
23
24
?,5
26
27
28
29
31
32
33
34
35
36
37
38
?9
County
Fairfield
West Chester
Rockland
Bronx
New York
Queens
Kings
Richmond
Nassau
Suffolk
Bergen
Hudson
Essex
Union
Pas sale
Morris
Somerset
Middlesex
Monmouth
State
Connecticut
New York
New York
New York
New York
New York
New York
New York
New York
New York
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
New Jersey
-------
- 47 -
TABLE II Population data of the Tri-State Region
(population density in persons/Km2)
No,
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Code
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
Tract
New Fairfield
Danbury
Newt own
Newtown
Ridgefield
Redding
She It on
Stratford
Bridgeport
Fairfield
Fairfield
Western
Wilton
New Canaan
Wilton
Westport
Fairfield
Fairfield
Bridgeport
Greenwich
Greenwich
Stamford
X-Coord.
105.0
105.0
115.0
125.0
105.0
115.0
125.0
137.5
122.5
117.5
112.5
107.5
102.5
97.5
102.5
107.5
112.5
117.5
122.5
87.5
92.5
97.5
Y-Coord.
145.0
135.0
135.0
135.0
125.0
125.0
125.0
117.5
117.5
117.5
117.5
117.5
117.5
112.5
112.5
112.5
112.5
112.5
112.5
107.5
107.5
107.5
Area Wt.
152.7
113.5
103.2
68.1
101.1
103.2
125.9
37.2
25.8
25.8
25.8
25.8
18.6
39.2
25.8
25.8
25.8
21.7
17.5
18.6
25.8
25.8
Population
Density
107.1
4,454.5
116.7
116.7
200.5
70.8
367.1
989.6
4,152.3
750.2
750.2
145.7
198.1
315.6
198.1
573.5
750.2
750.2
4,152.3
476.5
476.5
1,079.3
-------
- 48 -
TABLE II (continued) Population data of the Tri-State Region
(population density in persons/Km^)
No.
23
24
25
26
27
28
29
30
31
32
33
34
35
36
17
38
39
40
41
42
43
44
Code
11
11
11
11
11
21
21
21
21
21
21
Tract
Darien
Norwalk
Greenwich
Stamford
Darien
Cor t land
Yorktown
Somers
Ossining
New Castle
Bedford
21
21
21
21
21
21
21
21
21
21
21
Tarrytown
Mt. Pleasant
Irvington
Wbite Plains
Harrison
Hastings
Greenburgh
Scarsdale
Rye
Yonkers
East Chester
X-Coord.
102.5
107.5
92.5
97.5
102.5
75.0
85.0
95.0
75.0
85.0
95.0
76.5
82.5
76.5
82.5
86.5
73.5
77.5
82.5
87.5
73.0
77.5
Y-Coord .
107.5
107.5
102.5
102.5
102.5
125.0
125.0
125.0
115.0
115.0
115.0
107.5
107.5
102.5
102.5
102.5
97.5
97.5
97.5
97.5
92.5
92.5
Area Wt.
25.8
24.8
40.2
25.8
15.5
123.3
114.6
127.6
75.7
97.1
61.2
34.0
31.1
34.0
24.3
14.6
12.6
24.3
24.3
24.3
19.4
24.3
Population
Density
602.1
1,428.0
476.5
1,079.3
602 . 1
328.0
294.8
124.3
1,065.8
310.4
191.2
1,345.8
553.3
778.0
2,039.0
742.5
1,746.4
1,147.1
954.0
1,064.4
4,468.0
3,006.0
-------
- 49 -
TABLE II (continued) Population data of the Tri-State Region
(population density in persons/Km^)
No.
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Code
21
21
23
23
23
24
24
24
24
25
25
25
25
25
25
25
26
26
26
26
26
26
Tract
Mamaroneck
New Roche lie
Bronx
Bronx
Bronx
Manhattan
Manhattan
Manhattan
Manhattan
Queens
Queens
Queens
Queens
Queens
Queens
Queens
Kings
Kings
Kings
Kings
Kings
Kings
X-Coord.
82.5
79.0
75.0
72.5
77.5
67.5
65*0
63.5
69.0
71.5
72.5
77.5
68.5
83.0
82.0
77.0
72.0
67.0
63.5
67.5
72.0
67.0
Y- Coord.
92.5
88.5
87.0
82.5
82.5
77.5
73.5
70.0
83.0
76.0
72.5
72.5
76.5
72.5
66.5
68.0
66.5
68.0
62.5
62.5
63.0
59.0
Area Ut.
28.2
14.6
35.7
29.4
18.9
15.1
9.8
10.6
12.1
12.4
33.8
28.1
21.4
31.5
34.9
39.4
16.2
41.0
18.3
27.0
17.2
23.7
Population
Density
1,545.2
2,781.2
13,857.8
13,857.8
13,857.8
25,826.1
25,826.1
25,826.1
25,826.1
7,102.2
7,102.2
7,102.2
7,102.2
7,102.2
7,102.2
7,102.2
14,352.0
14,352.0
14,352.0
14,352.0
14,352.0
14,352.0
-------
- 50 -
TABLE II (continued)
Population data of the Trl-State Region
(population density in persons/Km^)
t -• -
No.
67
68
69
70
71
72
73
74
75
.
76
77
78
79
BO
81
82
83
W
85
•?-
87
88
Code
27
27
27
27
27
28
28
28
28
28
28
28
28
28
28
28
28
28
28
28
28
28
Tract
Richmond
Richmond
Richmond
Richmond
Ri chmond
Port Washington
Old Brookville
Muttontown
Mill Neck
Oyster Bay
Plainview
Jericho
Old Westbury
Manhasset
Gardesi City
Hemps tead
Levittown
Old Bethpage
Franklin Square
noosevelt
North Bellmore
Wantagh
X- Coord.
57.0
52.0
52.0
57.0
48.0
87.0
92.5
97.5
95.0
102.5
102.5
97.5
92.5
87.0
88.0
92.5
97.5
103.0
87.0
92.5
97.5
99.0
Y-Coord.
62.5
62.0
57.5
58.0
52.5
83.0
82.5
82.5
87.0
82.5
77.5
77.5
77.5
77.5
72.5
72.5
72.5
72.5
67.5
67.5
67.5
64.0
Area Wt.
22.7
24.9
34.6
16.2
20.5
30.3
27.1
27.1
36.8
29.2
28.2
27.1
27.1
39.0
23.8
27.1
27.1
32.5
32.5
27.1
27,1
21.7
Population
Density
1,967.0
1,967.0
1,967.0
1,967.0
1,967.0
1,573.7
272.3
131.9
i
131.6
532.3
2,273.9
1,339.2
129.1
1,341.7
2,006.1
2,278.6
3,704.5
695.5
4,412,7
i
3,324.3
3,308.9
2,239.6
-------
- 51 -
TABLE II (continued) Population data of the Tri-State Region
(population density in persons/Km^)
No.
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
Code
28
28
28
29
29
29
29
29
29
29
29
29
29
29
29
29
22
22
22
22
Tract
Massapequa
Oceanside
Freeport
Lloyd Harbor
Huntington
Station
Half Hollow
Hills
Babylon
Copiague
East Northport
Isliptown
Smithtown
Isliptown
Selden
East Patchogue
Brookhaven
Brookhaven
Stony Point
Ram apo town
Clarkstown
Congers
X- Coord.
102.5
87.0
92.5
106.0
107.5
107.5
108.0
108.0
115.0
115.0
125.0
125.0
135.0
135.0
145.0
145.0
64.0
56.0
65.0
72.0
Y-Coord .
67.5
63.0
63.0
87.0
82.5
77.5
72.5
67.0
85.0
73.5
85.0
75.0
85.0
75.0
85 oO
77.0
122.0
113.0
115.0
113.0
Area Wt.
26.0
33.6
23.8
42.7
24.5
21.8
18.2
21.8
89.0
121.7
95.3
95.3
90,8
77.2
90.8
64.5
66.1
71.6
90.6
19.9
Population
Density
2,675.1
2,756.5
3,762.8
134.3
2,062.6
522.6
553.7
2,386.8
1,847.0
1,053.7
982.2
1,053.7
786.0
826.4
465.6
465.6
190.1
591.6
628.0
629.1
-------
- 52 -
TABLE II (continued)
Population data of the Tri-State Region
(population density in persons/Km^)
No.
! —
109
110
111
112
113
|
114
115
116
11?
118
119
i
,
120
121
i
j.22
123
124
125
126
12?
1.78
129
130
Code
22
22
22
22
22
31
31
31
31
31
31
11
33
31
31
31
31
31
31
31
32
32
Tract
Ramapotown
West Nyack
Upper Nyack
Pearl River
Orangetown
Ramsey
Montvale
Wyckoff
Westwood
Norwood
Fair Lawn
New Mil ford
Ten a fly
Lodi
Teaneck
Englewood Cliffs
Rutherford
Carlstadt
Ridgefield
Lyndhurst
North Bergen
Jersey City
X-Coord.
62.0
67.5
71.5
67.0
71.5
52.0
62.0
55.0
62.5
68.5
56.5
62.5
67.5
67.5
72.5
77.5
58.0
62.5
66.0
57.0
62.5
58.0
Y-Coord.
108.0
107.5
107.5
103.0
102.0
105.0
102.0
97.5
97.5
97.5
92.5
92.5
93.0
87.5
87,5
87.5
82.5
82.5
82.5
78.0
77.5
72.5
Area Wt.
35.3
22.6
13.6
23.6
18.1
122.1
23.3
50. A
24.2
31.0
28.1
24.2
31.0
22.3
24.2
24.2
20.3
22.3
13.6
17.4
30.9
19.9
Population
Density
591.6
449.3
635.3
1,043.3
898.5
891.8
1,659.3
976.0
302.2
649.8
2,904.4
3,567.7
1,286.5
4,192.4
3,001.0
1,189.5
3,006.7
756.0
1,717.6
1,999.7
3,503.9
7,118.1
-------
- 53
TABLE II (continued)
Population data of the Tri-State Region
(population density in persons/Km2)
No.
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
Code
32
32
32
35
35
35
35
35
35
35
35
33
33
33
33
33
33
33
33
33
33
36
Tract
Jersey City
Bayonne
Kearny
West Milford
Ringwood
West Milford
Bloomingdale
Wayne
Hawthorne
West Paters on
Cliffton
Fairfield
Roseland
Monte lair
Nut ley
Livingston
West Orange
Newark
Millbum
Newark
Newark
Rockaway
X-Coord.
61.5
57.0
55.0
35.0
44.0
37.0
43.0
47.0
52.0
49.0
53.5
43.0
42.5
47.5
52.0
42.0
47.5
51.5
42.0
47.5
52.0
25.0
Y-Coord.
72.5
68.0
75.0
110.0
110.0
103.0
102.0
95.0
93.5
88.0
86.0
37.0
82.5
83.0
81.0
77.5
77.5
77.5
78.5
78.0
72.0
95.0
Area Wt.
14.0
15.0
16.9
116.4
83.4
29.1
27.2
53.3
27.2
21.3
35.9
23.7
25.8
35.5
15.1
39.8
26.9
19.4
23.7
25.8
31.2
167.7
Population
Density
7,118.1
5,650.8
1,688.2
83.8
137.4
83.8
322.5
771.0
2, 427. A
1,367.2
3,081.4
252.3
523.8
2,999.6
4,379.7
888.1
1,492.0
6,708.2
893.2
6,708.2
6,708.2
172.2
-------
- 54 -
TABLE II (continued)
Population data of the Trl-State Region
(population density in persons/Km^)
No.
153
154
155
156
157
158
159
160
161
162
163
164
163
166
167
168
169
170
171
171
173
174
Code
36
36
36
36
36
36
36
36
37
37
37
37
37
34
34
34
34
34
34
34
34
38
Tract
Klnnelon
Roxbury
Lincoln Park
Randolph
Hanover
Chester
Mendham
Chatham
Bedminster
Warren
Hillsborough
Franklin
Montgomery
Summit
Berkley Heights
Cranferd
Elizabeth
Plainfield
Scotch Plains
Clark
Linden
Fiscataway
X-Coord.
35.0
15.0
42.0
25.0
35.0
15.0
27.0
33.5
15.5
23.0
15.0
23.0
17.0
39.0
36.0
42.5
47.5
33.0
37.5
42.5
47.5
28.0
Y- Coord.
95.0
85.0
95.0
85.0
85.0
77.0
75.0
75.0
65.0
64.0
54.0
55.0
46.0
71.0
68.0
67.5
67.5
62.0
63.0
62.5
62.5
56,0
Area Wt.
121.4
106.0
39.1
102.9
105.0
63.8
95.7
73.1
91.2
99.4
117.9
71.7
166.0
27.1
40.7
26.1
32.3
13.6
21.9
26.1
24.0
30.2
Population
Density
155,6
294.6
512.6
254.3
420.6
73.8
130.2
589.2
40.7
185.8
87.2
268.2
i
79.8
(
1,629.1
851.3
2,411.6
4,048.7 1
3,234.0
999.6
1,680.1 |
1,536,3
765_9
-------
- 55 -
TABLE II (continued)
Population data of the Tri-State Region
(population density in persons/Kn»2)
No.
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
Code
38
38
38
38
38
38
38
38
38
38
38
38
38
38
38
38
39
39
39
39
39
39
Tract
South Plainfield
Metuchen
Woodb ridge
Carteret
New Brunswick
Edison
Perth Amboy
North Brunswick
South River
Madison
South Brunswick
East Brunswick
Old Bridge
Madison
Plainsboro
Monroe
Matawan
Hazlet
Middletown
Marlboro
Colts Neck
Rums on
X-Coord.
32.5
37.5
42.5
47.0
32.5
37.5
43.0
31.0
37.5
42.0
25.0
32.5
37.5
42.0
27.0
33.0
47.5
52.5
57.5
44.0
55.0
63.0
Y-Coord.
57.5
57.5
57.5
57.5
52.5
52.5
52.5
46.0
47.5
47.5
40.0
40.0
40.0
42.0
33.0
32.0
42.5
42.5
42.0
34.0
35.0
38.0
Area Wt.
25.1
29.2
25.1
18.1
25.1
25.1
28.2
41.2
25.1
23.1
106.6
50.3
50.3
36.2
58.3
43.3
23.9
28.9
19.9
106.7
99.7
39.9
Population
Density
1,015.7
2,551.9
1,707.3
2,110.2
2,920.9
915.7
3,230.2
575.3
2,291.0
518.2
141.6
644.0
1,631.0
518.2
57.0
91.0
914.6
1,523.2
596.6
171.7
74.0
564.7
-------
- 56 -
TABLE II (continued)
Population data of the Tri-State Region
(population density in persons/Km2)
fj_ -ra— 1-»,_ _
No.
197
198
199
200
| 201
202
203
j
204
205
206
207
208
209
210
211
212
!
213
214
215
Code
39
39
39
39
39
39
39
39
39
37
36
39
25
28
28
29
29
29
29
Tract
Oceanport
Ocean
Spring Lake
New Shrewsbury
Howe 11
Freehold
Millstone
Howell
Wall
Bernardsville
Chester
Sea Bright
Queens
Long Beach
Oyster Bay
Is lip
Brookhaven
Brookhaven
Brookhaven
, ,
X-Coord.
63.0
62.5
61.5
55.0
45.0
35.0
25.0
47.0
55.0
18.5
7.5
64.0
75.0
88.0
105.0
120.0
132.0
135.0
145.0
Y-Coord.
32.5
27.5
21.0
25.0
25.0
24.0
18.0
18.0
17.0
71.5
82.0
46.0
56.5
57.5
59.5
61.5
64.5
92.0
92.0
Area Wt.
29.9
22.9
33.9
99.7
99.7
111.7
118.6
53.8
79.8
84.0
85.4
8.0
20.3
5.4
14.1
10.9
16.3
36.3
36.3
Population
Density
888.4
831.8
1,071.7
148.4
139.6
102 .,4
27.7
139,6
I
217.4
207.4
73.8
153.9
7,102.2
5,723.6
1,395.9
1,053.7
465.6
465.6
465.6
. _ .
-------
- 57 -
TABLE III Sub-population data of the Trl-State Region
(population density in persons/Km^,
sub-population in % of population density)
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Code
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
Tract
New Fairfield
Danbury
Newt own
Newtown
Bidgefield
Redding
Shelton
Stratford
Bridgeport
fairfield
Fairfield
West on
Wilton
New Caaasa
Wilton
Westport
Fairfield
Fairfield
Bridgeport
Greenwich
Greenwich
Stamford
Population
Density
107.1
4,454.5
116.7
116.7
200.5
70.8
367.1
989.6
4,152.3
750.2
750.2
145.7
198.1
315.6
198.1
573.5
750.2
750.2
4,152.3
476.5
476.5
1,079.3
School-Age
27.0
25.0
29.0
27.0
31.0
29.0
31.0
27.0
29.0
29.0
31.0
31.0
31.0
25.0
31.0
29.0
27.0
22.0
25.0
25.0
25.0
25.0
Elderly
9.0
9.5
11.0
8.0
6.5
9.0
8.0
9.5
10.5
8.0
7.5
6.5
8.0
9.0
8.0
8.0
8.0
11.0
9.5
11.5
10.0
8.0
Non-White
1.5
8.0
1.0
1.5
1.5
1.5
1.5
0.8
1.0
1.0
1.0
1.0
1.0
7.0
3.5
5.0
1.0
10.5
7.0
3.5
9.0
10.0
-------
- 58 -
TABLE III (continued)
Sub-population data of the Tri-State Region
(population density in persons/Km2,
sub-population in Z of population density)
No.
23
i
24
25
26
27
28
29
i
30
31
3?
33
34
!
35
i
I 36
•
i 3?
38
!
39
40
&l
42
43
44
Code
11
11
11
11
11
21
21
21
21
21
21
21
21
21
21
21
21
21
21
21
21
21
Tract
Darien
Norwalk
Greenwich
Stamford
Darien
Cortland
Yorktmm
Sowers
Ossining
New Castle
Bedford
Tarrytown
Mt. Pleasant
Imfington
Whit*» Plains
Harrison
Hastings on
Hudson
Greenburgh
Scarsdale
Rye
Yonkers
Eastchester
Population
Density
602.1
1,428.0
476.5
1,079.3
602.1
328.0
294.8
124.3
1,065.8
310.4
191.2
1,345.8
553.3
778.0
2,039.0
742.5
1,746.4
1,147.1
954.0
1,064.4
4,468.0
3,006.0
School-Age
27.0
25.0
25.0
25.0
27.0
26.0
31.0
30.0
26.0
30.0
31.0
27.0
28.0
25.0
25.0
25.0
25.0
27.0
25.0
27.0
22.0
24.0
Elderly
8.0
8.0
11.0
10.0
7.0
11.0
7.0
10.5
11.0
9.0
7.0
9.5
8.5
8.0
12.0
8.0
8.0
9.0
12.0
9.5
11.0
11.5
1
Non-White i
H.O
1
2.0
4.0
11.0
5.0
9.0
1.5
2.0
4.0
7.0
1.7
7.0
3.5
12.0
10.0
5.0
10.0
4.5
2.0
4.5 ;
7.0 j
9.0
-------
- 59 -
TABLE III (continued)
Sub-population data of the Tri-State Region
(population density in persons/Km^,
sub-population in % of population density)
No.
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Code
21
21
23
23
23
24
24
24
24
25
25
25
25
25
25
25
26
26
26
26
26
26
Tract
Mamaroneck
New Rochelle
Bronx
Bronx
Bronx
Manhattan
Manhattan
Manhattan
Manhattan
Queens
Queens
Queens
Queens
Queens
Queens
Queens
Kings
Kings
Kings
Kings
Kings
Kings
Population
Density
1,545.2
2,781.2
13,857.8
13,857.8
13,857.8
25,826.1
25,826.1
25,826.1
25,826.1
7,102.2
7,102.2
7,102.2
7,102.2
7,102.2
7,102.2
7,102.2
14,352.0
14,352.0
14,352.0
14,352.0
14,352.0
14,352.0
School-Age
27.0
24.0
22.5
22.5
22.5
15.6
15.6
15.6
15.6
19.2
19.2
19.2
19.2
19.2
19.2
19.2
22.7
22.7
22.7
22.7
22.7
22.7
Elderly
11.0
12.5
11.6
11.6
11.6
14.0
14.0
14.0
14.0
12.4
12.4
12.4
12.4
12.4
12.4
12.4
11.1
11.1
11.1
11.1
11.1
11.1
Non-White
6.0
11.0
26.6
26.6
26.6
29.2
29.2
29.2
29.2
14.7
14.7
14.7
14.7
14.7
14.7
14.7
26.8
26.8
26.8
26.8
26.8
26.8
-------
- 60 -
TABLE III (continued)
Sub-population data of the Trl-State Region
(population density in persons/Km2,
sub-population in % of population density)
No.
67
68
69
70
71
72
73
74
75
76
77
78
79
80
8),
82
83
84
85
86
87
88
Code
27
27
27
27
27
28
28
28
28
28
2ft
28
28
28
28
28
28
28
28
28
28
28
Tract
Richmond
Richmond
Richmond
Richmond
Richmond
Port Washington
Old Brookville
Muttontown
Mill Neck
Oyster Bay
Plalnview
Jericho
Old Westbury
Man basset
Garden City
Hempstead
Levittovn
Old Bethpage
Franklin Square
Roosevelt
North Bellmore
Wantagh
Population
Density
1,967.0
1,967.0
1,967.0
1,967.0
1,967.0
1,573.7
272.3
131.9
131.6
532.3
2,273.9
1,339.2
129.1
1,341.7
2,006.1
2,278.6
3,704.5
695.5
4,412.7
3,324.3
3,308.9
2,239.6
School-Age
25.4
25.4
25.4
25.4
25.4
25.0
25.0
31.0
30.0
31.0
31.0
31.0
28.0
25.0
26.0
26.0
28.0
30.0
27.0
27.0
27.0
28.0
Elderly
8.7
8.7
8.7
8.7
8.7
8.0
7.0
6.0
9.0
6.0
6.0
6.5
7.5
8.5
8.0
8.0
7.5
6.5
8.0
8.0
8.0
7.5
Non-White
6.0
6.0
6.0
6.0
6.0
7.0
3.0
1.8
1.5
1.5
1.5
1.5
3.0
8.0
8.0
7.0
6.0
1.8
8.0
8.0
!
7.0
7.0
i
-------
- 61 -
TABLE III (continued)
Sub-population data of the Tri-State Region
(population density in persons/Km2,
sub-population in Z of population density)
No.
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
Code
28
28
28
29
29
29
20
29
29
29
29
29
29
29
29
29
22
22
22
22
22
22
Tract
Massapequa
Oceanside
Freeport
Lloyd Harbor
Huntiugton
Station
Half Hollcw
Hills
Babylon
Ccpiagua
East Northport
Isliptown
Smith!: own
Isliptcrcm
Selden
East Patchgue
Brookhaven
Brookhaven
Stony Point
Ramapoto^m
Clarkstown
Congers
Ramapotcwri
West Nyack
Population
Density
2,675.1
2,756.5
3,762.8
134.3
2,062.6
522.6
553.7
2,386.8
1,847.0
1,053.7
982.2
1,053.7
786.0
826.4
465.6
465.6
190.1
591.6
628.0
629.1
591.6
449.3
School- Age
30.0
27.0
27.0
31.0
31.0
31.0
31.0
31.0
31.0
30.0
30.0
29.0
29.0
29.0
29.0
29.0
27.0
31.0
29.0
31.0
31.0
31.0
Elderly
6.0
12.0
8.0
6.0
6.0
6.0
6.0
6.0
6.0
7.0
7.0
8.0
8.0
8.0
8.5
8.0
8.0
6.0
7.0
6.0
6.0
6.0
Non-White
1.9
7.5
7.5
3.5
3.5
4.0
10.0
11.0
2.0
5.0
1.5
4.0
4.0
4.0
4.0
4.0
1.5
7.5
7.0
7.5
7.5
7.5
-------
- 62 -
TABLE III (continued)
Sub-population data of the Tri-State Region
(population density in persons/Km2,
sub-population in Z of population density)
So.
Ill
112
113
114
115
116
117
118
119
120
121
122
123
; 124
1
125
126
12?
128
11-
130
131
132
Code
22
22
22
31
31
31
31
31
31
31
31
31
31
31
31
31
31
31
32
32
32
32
Tract
Upper Nyack
Pearl River
Orangetown
Ramsey
Mont vale
Wyckoff
Keatwood
Norwood
Fair Lawn
New Milford
Tenafly
Lodi
Teaneck
Englewood Cliffs
Rutherford
Car 1st ad t
Ridgefleld
Lyndharut
North Bergen
Jexsey City
Jersey City
Bayotme
Population
Density
635.1
1,043.3
898.5
891.8
1,659.3
976.0
302.2
649.8
2,904.4
3,567.7
1,286.5
4,192.4
3,001.0
1,189.5
3,006.7
756.0
1,717.6
1,999.7
3,503.9
7,118.1
7,118.1
5,650.8
School-Age
39.0
29.0
29.0
28.0
30.0
30.0
30.0
31.0
26.0
30.0
28.0
25.0
20.0
22.0
23.0
22.0
19.0
22.0
19.0
22.0
22.0
22.0
Elderly
7.0
9.5
9.5
6.0
5.0
7.5
6.0
7.0
9.5
8.0
8.0
10.0
10.5
11.5
11.5
11.5
11.5
10.5
12.5
11.5
11.0
11.0
Non-White
7.0
6.0 |
6.0
2.5
0.5
0.5
0.8
1.0
1.5
2.0
2.0
0.5
11.0
10.0
3.5
0.9
0.5
3.5
1.5
20. U
20,0
5.0
-------
- 63 -
TABLE III (continued)
Sub-population data of the Tri-State Region
(population density in persons/Km^,
sub-population in % of population density)
No.
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
Code
32
35
35
35
35
35
35
35
35
33
33
33
33
33
33
33
33
33
33
36
36
36
Tract
Kearny
West Milford
Ringwood
West Milford
Bloomingdale
Wayne
Hawthorne
West Paterson
Cliff ton
Fail-field
Roseland
Hontclair
Nut ley
Livingston
West Orange
Newark
Millburn
Newark
Newark
Rockaway
Kinnelon
Roxbury
Population
Density
1,688.2
83.8
137.4
83.8
322.5
771.0
2,427.4
1,367.2
3,081.4
252.3
523.8
2,999.6
4,379.7
888.1
1,492.0
6,708.2
893.2
6,708.2
6,708.2
172.2
155.6
294.6
School- Age
20.0
30.0
29.0
30.0
29.0
29.0
25.0
28.0
21.0
30.0
28.0
23.0
23.0
29.0
21.0
25.0
24
20
26.5
28
29
29
Elderly
11.0
6.0
4.5
6.0
8.0
6.0
11.5
7.5
11.5
6.0
11.0
11.5
12.0
6.0
11.0
10.0
12.0
10.5
8.0
5.0
4.5
6.0
Non-White
1.5
1.5
4.0
1.5
0.5
0.9
13.0
2.0
12.0
0.9
2.0
7.5
10.0
2.0
20.0
40.0
2.0
20.0
50.0
0.5
2.0
1.0
-------
- 64 -
TABLE III (continued)
Sub-population data of the Tri-State Region
(population density in persons/Ksr ,
sub-population in Z of population density)
No.
155
156
157
158
159
160
161
162
163
164
165
Code
36
36
36
36
36
36
37
37
37
37
17
t
i
166
16?
i
1 168
169
170
171
172
173
174
175
176
34
34
34
34
34
34
34
34
38
38
38
Tract
Lincoln Park
Randolph
Hanover
Chester
Mendhan
Chatham
Bednlnster
Warren
Hillsborough
Franklin
Montgomery
Summit
Berkley Heights
Cranford
Elizabeth
Plainfield
Scotch Plains
Clark
Linden
Pia cat away
South Plainfield
Metuchen
Population
Density
512.6
254.3
420.6
73.8
130.2
589.2
40.7
185.8
87.2
268.2
79.8
1,629.1
851.3
2,411.6
4,048.7
3,234.0
999.6
1,680.1
1,536.3
765.9
1,015.7
2,551.9
School-Age
30
28
23
30
27
26
28
30
30
28
28.5
24
24
23
23.0
23.0
29.0
26.0
23.0
27.0
30.0
28.0
Elderly
8.0
8.0
6.5
7.5
9.0
8.5
10.5
7.5
6.5
6.0
6.0
12.5
10.5
11.0
11.0
11.0
7.0
7.5
8.0
4.0
4.5
5.0
Non-White
1.0
7.0
5.0
0.9
3.0
0.5
1.5
1.0
1.0
2.0
7.0
7.5
2,0
5.0
20.0
15.0
9.0
10.0
20.0
20.0 ;
i
10.0 !
4.0 |
-------
- 65 -
TABLE III (continued)
Sub-population data of the Trl-State Region
(population density in persons/Km^,
sub-population in Z of population density)
No.
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
Code
38
38
38
38
38
38
38
38
38
38
38
38
39
38
39
39
39
39
39
39
39
39
Tract
Woodridge
Carteret
New Brunswick
Edison
Perth Aaifaoy
Nortb Brunswick
South River
Madison
South Brunswick
East Brunswick
Old Bi-idge
Madison
Plainsboro
Monroe
Matawaa
Hazlet
Middletoim
Marlboro
Colts Neck
RUE»8O»
Oeeanport
Ocean
Population
Density
1,707.3
2,110.3
2,920.9
915.7
3,230.2
575.3
2,291.0
518.2
141.6
644.0
1,631.0
518.2
57.0
91.0
914.6
1,523.2
596.6
171.7
74.0
564.7
888.4
831.4
School- Age
29.0
28.0
26.0
27.0
25.0
25.0
29.0
29.0
30.0
29.0
28.0
31.0
26.0
25.0
30.0
30.0
31.0
26.0
30.0
22.0
28.0
24.0
Elderly
6.0
7.5
9.0
6.0
10.0
9.0
6.0
5.0
6.0
10.0
9.5
4.5
10.0
11.0
4.5
6.0
6.0
7.5
6.5
12.0
11.0
11.5
Non-White
4.0
5.0
10.0
4.0
5.0
2.0
0.5
1.5
3.5
5.0
5.0
2.0
15.0
10.0
10.0
1.0
3.5
8.0
3.5
5.0
10.0
10.0
-------
- 66 -
TABLE III (continued)
Sub-population data of the Trl-State Region
(population density in persons/Km2,
sub-population in 2 of population density)
No.
199
200
201
202
203
204
205
206
207
208
209
210
j 211
212
213
214
'"
Code
39
39
39
39
39
39
39
37
36
39
25
28
28
29
29
29
29
Tract
Spicing Lake
'W«w Shrewsbury
Howell
Freehold
Millstone
Howell
Wall
Bernardsville
Chester
Sea Bright
Queens
Long Beach
Oyster Bay
Islip
Brookhaven
Brookhaven
Brookhaven
Population
Density
1,071.7
148.4
139.6
102.4
27.7
139.6
217.4
207.4
73.8
153.9
7,102.2
5,723.6
1,395.9
1,053.7
465.6
465.6
465.6
School-Age
22.0
27.0
31.0
30.0
27.0
30.0
24.0
26.0
29.0
30.0
19.2
27.0
30.0
29.0
29.0
29.0
29.0
Elderly
12.5
8.0
7.0
9.0
9.5
8.0
12.5
12.0
11.0
6.0
12.4
12.0
8.0
8.0
8.0
8.0
8.0
Non-White
10 ,,0
10 ,,0
5.0
15.0
10.0
8.0
7.5
0.4
0.7
3.5
14.7
7.5
10.0
4,0
i
i
4,,0
4.0
10.0
>
-------
- 67 -
TABLE IV TSP air quality data, 24 hr. Hi-Vol.
(geometric mean CQ in yg/nr)
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Code
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
32
34
32
SAROAD #
070060001
070060001
070060002
070175001
070260002
070330001
070330002
070330003
070330004
070330007
070330008
070820001
070820005
071080001
071080003
071080004
071080010
071110001
071110005
310180001
311300002
312320001
X- Coord.
124.0
124.0
123.5
106.0
118.5
92.0
94.5
91.0
87.5
91.5
94.5
108.5
109.0
99.5
98.5
99.5
96.5
129.0
129.0
56.5
49.5
60.5
Y-Coord.
116.5
116.5
114.5
135.5
116.5
100.5
103.5
99.0
107.0
105.5
102.5
111.0
112.0
104.5
106.5
109.0
110.0
119.0
116.0
67.5
65.5
71.5
Cn, 71/2
—
52.63
—
—
59.18
45.06
61.40
56.69
—
44.62
86.48
53.23
65.98
—
—
—
—
47.51
—
—
83.25
117.35
V 73/2
—
42.16
53.29
73.30
40.94
46.28
63.09
51.20
41.40
33.03
62.45
51.13
60.53
—
—
129.90
42.76
—
53.99
—
73.90
—
-------
- 68 -
TABLE IV (continued)
air quality data, 24 hr. Hi-Vol.
metric mean CH, in vg/n )
TSP
(geometric
No;
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Code
33
35
38
39
32
33
37
39
38
36
36
33
33
31
31
36
31
31
31
32
33
32
SARD AD f
313480001
314140001
314220001
310060002
310180003
310400002
310500001
310560001
310820001
311100001
311100002
311160002
311380001
311440001
311460001
311540001
311560001
311560002
311820001
312180001
312280001
312320003
X-Coord.
52.0
53.0
43.5
64.0
57.5
51.5
24.5
61.0
48.5
24.0
23.5
43.0
45.0
57.0
65.0
36.5
67.0
67.5
62.0
68.5
47.5
60.5
Y-Coord.
70.0
90.5
51.0
28.0
66.5
83.0
58.5
13.5
58.5
88.0
89.0
75.0
75.0
93.5
81.0
77.5
84.0
85.0
89.5
74.0
72.0
73.0
C,,,, 71/2
—
82.44
—
76.38
102.24
—
76.34
—
67.22
—
—
83.38
48.77
—
125.96
—
—
—
214.00
—
68.39
145.36
Cm, 73/2
—
—
—
53.52
83.10
—
—
34.10
70.52
__
37.08
132.39
39.07
42.80
79.48
33.26
—
46.14
—
116.93
55.49
108.96
1
-------
- 69 -
TABLE IV (continued)
TSP air quality data, 24 hr. Hi-Vol.
(geometric mean CQ In
No.
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Code
32
34
38
38
38
38
39
39
36
33
33
38
33
35
35
38
34
39
39
34
38
38
SAROAD $
312320004
312580001
313020002
313060001
313060002
313060003
313180001
313180002
313300002
313480002
313480003
313500001
313980001
314100001
314140001
314220002
314440001
314500001
314500002
314760001
314920001
314920002
X-Coord.
58.5
47.0
38.0
39.0
44.5
23.5
41.5
45.0
29.5
54.5
53.5
31.0
47.5
55.5
54.0
44.0
44.0
59.5
58.5
45.5
37.5
40.5
Y-Coord.
71.5
61.5
54.5
52.5
46.5
39.0
21.0
17.5
79.5
72.5
71.5
49.5
76,5
134.0
139.5
52.0
61.0
35.5
34.5
65.5
47.5
48.5
Cm. 71/2
95.89
76.10
—
—
—
—
45.52
—
55.60
—
45.49
—
72.95
54.63
82.44
—
67.20
—
—
89.22
—
—
C,,, 73/2
81.07
67.91
44.61
—
39.79
41.73
32.57
30.46
—
151.24
—
59.28
—
—
53.44
53.67
—
50.52
72.95
—
56.19
-------
- 70 -
TABLE IV (continued)
TSP air quality data, 24 hr. Hi-Vol.
(geonetrlc mean CB In tig/v )
No.
67
68
69
70
71
72
73
74
75
76
77
78
79
80
31
82
83
i
84
85
96
87
88
Code
32
37
38
32
34
31
33
31
38
38
23
21
29
21
28
28
28
28
28
23
28
28
SAROAD 1
314960001
315060002
315080001
315420001
315440001
315500001
315860001
315920001
316040001
316040002
334680001
337620001
330280001
331560001
332300001
332300002
332360001
332460001
332900001
332900003
332900004
332900005
X-Coord .
60.5
19.5
43.0
61.5
47.5
57.5
46.5
62.5
42.5
45.5
70.5
73.5
107.0
75.0
94.5
95.5
92.0
92.0
91.0
86.5
87.0
95.0
Y-Coord.
78.0
57.5
49.5
75.5
67.0
105.0
79.0
98.0
57.0
56.0
80.0
92.0
71.5
100.0
64.0
65.0
71.0
86.0
62.5
67.5
63.0
71.0
Cm, 71/2
—
57.08
68.27
109.22
—
—
—
61.79
73.39
—
120.69
—
—
—
—
74.34
—
81.82
98.53
85.57
66.48
69.87
Cm, 73/2
70.40
48.54
52.21
80.24
46.97
35.15
57.03
46.03
68.89
50.65
—
—
52.94
50.52
—
50.85
52.63
—
53.98
59,04
50.06
71.31
-------
71
TABLE IV (continued)
TSP air quality data, 24 hr. Hi-Vol.
(geometric mean Cm in iig/m->)
No.
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
Code
28
28
21
21
21
21
28
28
28
28
28
21
24
24
21
21
21
21
29
29
22
22
SAROAD #
332900007
333480001
334100001
334100002
334480001
334480003
334520001
334520002
334520004
334520005
334520006
334620002
334680050
334680057
334880001
335200001
335360001
335520001
33550001
335550002
335780001
335780002
X-Coord.
84.0
83.5
84.5
84.0
77.5
77.0
96.5
103.5
101.0
99.5
104.0
81.5
68.0
68.0
76.0
75.5
71.5
89.5
134.5
135.0
67.0
71.0
Y-Coord.
62.5
80.0
94.0
93.0
89.5
91.0
72.5
65.5
84.5
85.0
79.0
90.5
75.5
75.5
107.0
114.5
127.0
99.0
93.5
92.5
107.0
101,0
Cm, 71/2
—
74.22
64.02
49.38
—
—
56.34
60.37
—
51.91
72.45
77.71
90.95
87.12
59.69
54.89
91.11
71.94
—
—
53.71
—
CB, 73/2
74.57
40.10
54.46
73.64
—
76.08
53.17
60.09
—
42.58
46.30
58.19
75.97
83.92
40.74
48.04
73.29
50.46
—
—
53.08
50.36
-------
- 72 -
TABLE IV (continued)
TSP air quality data, 24 hr. Hl-Vol.
(geometric mean Cn In
No.
Ill
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
12?
128
129
130
131
132
Code
28
21
29
22
29
29
29
29
21
21
21
21
22
21
21
24
23
25
24
23
26
25
SAROAD *
335800001
335910001
336340001
336560001
336580001
336580002
336580011
336580023
337320003
337320004
337320005
337320006
337400001
337480001
337620001
334680002
334680003
334680004
334680005
334680006
334680007
33468C008
X-Coord.
90.5
88.5
160.0
54.5
129.5
139.0
119.0
118.0
78.5
82.0
85.0
77.5
66.0
82.5
73.0
68.0
72.0
78.0
65.5
72.5
69.0
77.0
Y-Coord .
65.0
97.5
89.0
110.5
91.0
79.0
73.0
83.5
108.5
125.5
129.0
103.5
119.0
101.0
93.0
81.5
84.0
73.0
75.0
87.0
58.5
77.5
Cm, 71/2
93.18
72.18
37.25
52.78
64.87
65.26
52.06
37.08
44.98
40.74
53.51
98.99
—
83.98
117.81
—
—
—
—
—
—
—
Cm, 73/2
67.60
52.96
43.54
65.23
45.43
38.28
48.18
48.84
35.81
30.78
32.45
52.04
53.02
52.72
—
84.99
84.20
55.25
80.44
71.22
63.35
102.09
-------
- 73 -
TABLE IV (continued) TSP air quality data, 24 hr. Hi-Vol.
(geometric mean Cm in iig/m )
No.
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
Code
23
26
26
24
25
25
24
26
26
25
26
23
26
25
25
27
27
27
27
27
27
24
SAROAD #
334680009
334680010
334680011
334680014
334680015
334680016
33A680017
334680018
334680019
334680020
334680021
334680022
334680025
334680029
334680030
334680031
334680032
334680033
334680034
334680035
334680036
334680037
X- Coord.
78.5
64.0
68.0
69.0
80.5
85,5
66.0
67.0
70.5
73.5
73.5
73.0
68.0
76.5
82.0
51.5
58.0
50.0
54.5
57.5
46.5
64.0
Y-Coord .
80.5
65.5
72.5
79.0
76.5
73.5
73.0
68.0
67.0
69.5
64.5
81.0
64.0
64.5
66.0
62.5
63.5
57.5
59.5
58.0
51.5
72.0
Co, 71/2
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
_
Cm, 73/2
53.89
85.52
87.03
78.53
57.08
60.13
85.45
50.94
69.49
74.62
—
84.76
—
56.81
53.12
75,46
75.52
79.58
60.50
68.10
53.65
62.63
-------
- 74 -
TABLE IV (continued)
TSP air quality data, 24 hr. Hi-Vol.
(geometric mean CB in pg/m-*)
No.
155
156
157
158
159
160
161
162
163
164
Code
23
23
25
25
25
25
26
26
25
26
SAROAD *
334680038
334680039
334680040
334680041
334680042
334680044
334680045
334680046
334680047
334680064
X-Coord.
77.0
77.0
71.5
74.0
70.5
83.0
62.5
72.5
77.0
66.0
Y-Coprd.
83.0
87.0
77.0
75.5
75.5
69.5
63.0
62.5
57.0
62.0
CB, 71/2
—
—
—
—
—
—
—
—
—
—
Cm. 73/2
61.87
71.61
86.41
65.43
83.56
73.72
67.70
63.90
103.32
60.00
-------
- 75 -
TABLE V TSP percentlle concentrations In 71/2 and 73/2
(71/73 values in yg/m3)
No.
2
5
6
7
8
10
11
12
13
18
21
22
24
26
27
29
31
34
35
37
41
43
SAROAD t
070060001
070260002
070330001
070330002
070330003
070330007
070330008
070820001
070820005
071110001
311300002
312320001
314140001
310060002
310180003
310500003
310820001
311160002
311380001
311460001
311820001
312280001
Min.
19/23
23/30
16/29
19/36
14/25
12/20
60/25
21/26
31/28
29/22
59/56
73/46
49/56
47/28
25/26
37/28
47/36
54/81
20/18
55/43
10%
37/25
23/34
17/31
38/39
41/28
28/20
60/27
36/34
50/33
29/22
59/56
73/46
49/56
56/36
45/35
46/29
47/41
56/89
30/23
86/55
110/77| 110/89
31/30
43/31
30%
46/35
41/38
43/36
52/48
50/45
39/29
67/56
47/41
55/47
38/48
64/70
81/62
75/63
64/46
89/66
64/48
48/56
68/107
46/35
114/62
151/118
61/47
50%
50/41
44/39
49/42
61/52
60/50
44/30
69/60
54/43
64/59
40/52
67/70
109/73
81/66
71/52
104/71
79/57
68/68
80/120
52/36
121/74
214/121
68/51
70%
66/42
88/45
55/50
83/88
65/71
52/38
100/79
61/76
78/71
48/58
91/73
150/120
87/91
84/72
127/107
89/80
77/77
95/174
64/55
148/109
232/127
81/77
90%
95/79
107/49
80/79
115/122
95/88
92/58
172/88
92/98
92/111
70/75
183/110
227/161
147/109
134/78
179/293
129/90
107/136
126/196
75/67
211/138
339/226
103/68
95%
106/94
176/59
96/83
129/122
101/89
95/79
172/248
100/110
125/139
107/75
183/110
227/161
147/109
147/88
239/331
144/94
107/154
134/240
78/68
218/155
339/258
121/87
Max.
106/94
176/59
96/83
129/12!
101/89
95/79
172/24:
100/lH
125/13
107/75
183/11
227/16
147/10
147/88
239/33
144/94
107/15
134/24
78/68
218/15
339/25
121/87
-------
- 76 -
TABLE V (continued) TSP percentlie concentrations in 71/2 and 73/2
(71/73 values in ug/«3)
No.
44
45
46
I 51
53
56
i
57
58
59
61
64
68
69
70
74
75
77
8?.
85
86
87
88
SAROAD i
312320003
312320004
312580001
313180001
313300002
313500001
313980001
314100001
314140001
314440001
314760001
315060002
315080001
315420001
315920001
316040001
334680001
332300002
332900001
332900003
332900004
332900005
Min.
62/60
35/41
43/32
18/14
20/23
32/19
28/39
28/31
49/56
30/31
55/34
23/21
36/28
56/44
35/23
29/40
94/59
33/24
64/38
39/29
28/25
25/21
10%
87/60
64/48
43/33
20/20
41/26
32/40
55/41
28/37
49/56
46/34
67/37
31/25
51/29
83/48
40/23
43/40
94/59
51/27
66/39
59/29
41/29
40/21
30%
98/80
85/72
58/63
36/24
48/37
35/49
67/49
53/48
75/63
57/45
74/60
50/41
57/39
92/63
51/40
63/58
107/89
58/46
80/44
68/52
59/41
53/63
502
156/101
91/83
85/69
43/38
57/43
45/54
69/56
59/56
81/66
68/52
88/77
59/52
62/51
112/78
62/44
70/65
111/98
67/55
89/48
74/62
61/63
66/64
702
168/160
126/97
98/92
52/47
61/60
56/69
93/76
65/73
87/91
79/65
99/83
79/60
83/71
128/105
73/66
92/72
124/107
101/64
121/63
119/75
86/64
100/78
902
323/177
132/118
130/111
98/54
100/74
71/71
104/82
88/116
147/109
104/85
158/112
91/72
91/89
158/142
98/84
123/126
186/144
113/78
142/87
147/102
119/74
123/332
95%
372/177
196/134
130/114
107/54
101/75
72/75
115/98
88/125
I" 7/109
127/88
159/124
104/112
146/107
200/150
103/95
139/146
186/144
233/85
235/92
171/110
124/77
149/332
Max.
327/177
196/134
130/114
107/54
101/75
72/75
115/98
88/125
147/109
127/88
159/124
104/112
146/107
200/150
103/95
139/146
186/144
233/85
235/92
171/110 ]
124/7?
149/332
-------
- 77 -
TABLE V (continued) TSP percentile concentrations in 71/2 and 73/2
(71/73 values in UB/m3)
No.
90
91
92
95
96
98
99
100
101
102
103
104
105
106
109
111
112
113
114
115
116
117
SAROAD t
333480001
334100001
334100002
334520001
334520002
334520005
334520006
334620002
334680050
334680057
334880001
335200001
335360001
335520001
335780001
335800001
335910001
336340001
336560001
336580001
336580002
336580011
Min.
24/16
19/26
17/39
21/23
33/25
17/20
26/23
57/24
37/44
53/52
18/10
18/23
31/29
32/25
19/27
48/31
31/22
11/15
23/30
19/16
20/14
20/18
10%
51/18
19/30
36/51
28/30
37/26
34/26
1
50/23
.
57/38
71/48
62/60
37/20
36/26
61/43
35/35
35/36
66/31
53/30
22/23
33/34
38/26
36/22
48/30
30%
60/34
57/44
45/65
49/45
54/53
44/33
61/33
67/39
85/53
31/67
47/33
44/41
80/56
58/41
40/39
81/59
63/47
31/36
46/52
53/40
58/32
60/36
50%
67/42
64/55
45/74
57/51
59/70
55/45
72/51
73/60
86/67
91/71
68/38
63/47
94/67
77/46
56/50
90/73
76/51
41/49
54/68
66/42
69/43
64/37
70%
99/47
66/78
55/83
71/60
75/81
69/55
86/58
82/90
97/112
100/119
76/45
69/67
116/109
81/59
72/78
115/79
86/80
51/65
68/91
86/49
102/48
75/67
90%
132/84
109/97
70/120
95/89
92/92
84/69
114/62
114/108
148/134
118/130
105/100
86/82
149/125
124/89
78/88
135/96
99/83
58/71
91/110
110/74
107/56
88/84
95%
166/88
144/101
131/124
117/107
98/96
109/74
143/108
119/124
168/143
119/156
118/134
97/86
178/134
192/93
111/94
181/112
104/105
79/76
92/133
115/75
109/67
94/118
Max.
166/88
144/101
131/124
117/107
98/96
109/74
143/108
119/124
168/143
119/156
118/134
97/86
178/134
192/93
111/94
181/112
104/105
79/76
92/133
115/75
109/67
94/118
-------
- 78 -
TABLE V (continued) TSP percentlie concentrations in 71/2 and 73/2
(71/73 values in yg/m3)
Ho.
118
119
120
121
122
124
SAROAD #
336580023
337320003
337320004
337320005
337320006
337480001
Min.
15/22
13/16
9/15
13/16
33/26
31/31
10%
16/22
24/17
16/17
21/16
66/31
60/31
30%
31/41
40/27
42/26
47/21
88/39
60/38
50%
37/51
53/33
46/28
59/31
100/48
84/54
70*
54/58
56/58
57/39
64/52
125/69
115/65
90%
60/78
85/70
70/61
112/65
152/92
133/92
95%
82/84
87/89
84/68
112/71
178/98
171/93
Max,
82 /S4 ^
87/89
84/68
112/71
178/98
171/93
-------
- 79 -
TABLE VI TSP air quality data
(quarterly geometric mean in
SAROAD
070060001
070260002
070330001
070330002
070330003
070330007
070330008
070820001
070820005
071080004
311300002
312320001
310060002
310180003
310500001
310820001
311160002
311380001
311460001
312280001
312320003
312320004
1971
1
62.29
69.61
67.39
69.80
70.64
43.38
94.32
66.97
86.14
49.05
95.10
96.23
90.94
91.38
75.44
81.40
80.38
38.65
127.16
60.50
126.93
106.34
2
52.63
59.18
45.06
61.40
56.69
44.62
86.48
53.23
65.98
30.85
83.25
117.35
76.38
102.24
76.34
67.22
83.38
48.77
125.96
68.39
145.36
95.89
3
53.46
50.74
53.93
53.37
52.47
43.78
70.36
54.69
63.17
61.04
71.76
93.85
70.88
97.91
73.42
90.16
—
48.07
114.23
65.37
109.07
95.46
4
64.01
129.51
56.88
61.97
56.48
50.31
62.94
66.84
81.11
49.38
102.58
93.37
67.16
73.49
77.07
73.36
127.92
44.86
91.35
63.16
96.09
96.04
1973
1
—
—
—
—
—
—
—
—
—
—
70.67
74.47
48.87
53.41
31.06
60.52
123.99
28.00
78.73
46.12
85.39
75.22
2
42.16
40.94
46.28
63.09
51.20
33.03
62.45
51.13
60.53
129.90
73.90
87.94
53.52
83.10
34.10
70.52
132.39
39.07
79.49
55.49
108.96
81.07
3
50.00
48.43
49.95
63.80
53.69
44.53
70.40
72.37
63.94
83.98
—
—
61.22
88.04
43.96
77.92
137.70
48.83
—
58.67
98.01
110. 30
4
45.78
47.48
39.77
44.62
45.37
27.34
56.37
47.91
55.90
73.31
—
—
44.05
64.08
33.39
64.33
55.78
30.30
—
38.44
73.17
80.12
-------
- 80 -
TABLE VI (continued)
TSP air quality data
(quarterly geometric mean in yg/nr)
SARD AD
312580001
313180001
313300002
313500001
313980001
314220002
314440001
314760001
315080001
315420001
315920001
316040001
332300002
332360001
332460001
332900001
332900003
332900004
332900005
333480001
334100001
334520001
1971
1
103.62
56.44
60.00
39.36
70.59
81.48
67.79
95.85
86.23
109.13
61.80
75.80
66.21
73.62
101.33
195.26
107.79
87.63
89.95
73.46
69.50
62.75
2
76.10
45.52
55.60
45.49
72.95
67.59
67.20
89.22
68.27
109.22
61.79
73.39
74.34
56.16
81.82
98.53
85.57
66.48
69.87
74.22
64.02
56.34
3
79.21
36.06
53.99
39.43
64.84
62.54
61.80
79.12
61.14
91.49
61.83
68.58
79.69
82.99
80.33
83.34
73.79
71.04
74.24
79.03
60.56
67.67
4
88.60
34.99
58.34
38.41
75.56
71.85
—
80.84
73.13
96.06
64.63
78.33
70.53
61.90
81.78
66.78
67.83
66.53
72.93
45.04
57.67
57.54
1973
1
76.35
23.34
51.43
46.74
54.24
53.02
53.14
70.79
57.34
86.61
41.46
62.49
50.22
51.43
59.75
92.07
56.93
53.95
58.20
39.56
51.86
51.45
2
67.91
32.57
43.97
52.12
59.28
53.44
53.67
72.95
52.21
80.24
46.03
68.89
50.85
52.63
—
53.98
59.04
50.06
71.31
40.10
54.46
53.17
3
88.59
—
—
—
62.11
66.26
66.33
76.94
63.78
—
—
81.42
62.99
64.38
—
73.08
61.37
72.99
72.29
57.47
49.02
62.87
4
69.60
—
—
—
50.74
53.88
48.41
60.93
60.13
—
—
89.11
60.74
52.04
72.11
61.42
61.49
57.59
78.30
43.66
50.21
54.24
-------
- 81 -
TABLE VI (continued)
TSP air quality data
(quarterly geometric mean in
SAROAD
334520002
334520005
334520006
334620002
334680050
334680057
334880001
335200001
335360001
335520001
335780001
335800001
335910001
336340001
336560001
336580001
336580002
336580011
336580023
337320003
337320004
337320005
1971
1
72.20
58.40
80.55
97.19
87.76
109,40
61.81
54.15
83.50
77.14
50.25
108.65
82.37
37.78
57.23
96.28
83.06
62.12
53.52
44.48
38.58
61.82
2
60.37
51.91
72.45
77.71
90.95
87.12
59.69
54.89
91.11
71.94
53.71
93.18
72.18
37.25
52.78
64.87
65.26
52.06
37.08
44.98
40.74
53.51
3
107.60
60.47
68.06
76.25
81.04
87.14
50.36
49.14
58.54
66.25
57.98
84.71
64.81
36.91
57.59
63.12
125.81
54.99
44.10
45.75
38.13
38.22
4
132.95
43.26
57.03
72.22
74.75
80.12
51.09
42.79
63.29
54.37
46.36
82.59
71.89
30.42
47.08
80.13
40.79
48.91
38.68
38.09
26.97
32.91
1973
1
52.81
38.05
43.20
57.48
69.40
75.65
45.88
51.83
72.77
51.42
59.24
75.83
58.06
54.63
59.86
58.21
33.93
45.84
46.67
39.99
33.49
39.44
2
60.09
42.58
45.30
58.19
75.97
83.92
40.74
48.04
73.29
50.46
53.08
67.60
52.96
43.54
65.23
45.43
38.28
48.18
48.84
35.81
30.78
32.45
3
57.57
50.44
58.60
90.75
83.95
80.76
51.81
41.07
63.85
52.69
54.08
77.12
55.44
42.80
68.83
55.24
49.10
53.79
52.12
55.05
38,65
41.54
4
60.72
38.08
49.24
58.54
64.32
70.43
47.10
37.37
53.38
50.07
38.81
72.92
68.10
33.96
41.34
52.05
43.50
58.30
45.93
32.30
27.25
33.36
-------
- 82 -
TABLE VI (continued)
TSP air quality data .
(quarterly geometric mean in yg/m )
SARD AD
337320006
337480001
337620001
1971
1
103.30
108.56
113.21
2
98.99
83.98
117.81
3
79.47
68.55
104.37
4
65.88
72.40
65.77
1973
1
72.77
59.77
—
2
52.04
52.72
—
3
57.35
65.05
83.60
4
57.54
50.40
60.95
-------
- 83 -
TABLE VII TSP air quality data, 24 hr. Hi-Vol.
(geometric mean C^ in
No.
2
5
6
7
8
10
11
12
13
16
21
22
26
27
29
31
34
35
37
43
44
45
Code
11
11
11
11
11
11
11
11
11
11
34
32
39
32
37
38
33
33
31
33
32
32
SAROAD #
070060001
070260002
070330001
070330002
070330003
070330007
070330008
070820001
070820005
071080004
311300002
312320001
310060002
310180003
310500001
310820001
311160002
311380001
311460001
312280001
312320003
312320004
X-Coord.
124.0
118.5
92.0
94.5
91.0
91.5
94.5
108.5
109.0
99.5
49.5
60.5
64.0
57.5
24.5
48.5
43.0
45.0
65.0
47.5
60.5
58.5
Y-Coord.
116.5
116.5
100.5
103.5
99.0
105.5
102.5
111.0
112.0
109.0
65.5
71.5
28.0
66.5
58.5
58.5
75.0
75.0
81.0
72.0
73.0
71.5
Cm, 71
57.83
72.13
55.30
61.51
58.70
45.47
77.52
59.92
73.62
46.17
87.15
99.90
75.94
90.79
75.65
77.58
104.36
44.85
114.00
64.32
117.78
98.37
Cm, 73
49.45
50.53
49.81
59.56
54.53
36.40
69.44
58.55
65.75
79.14
78.65
87.14
51.49
70.81
35.32
68.10
106.20
35.70
89.92
49.02
90.47
85.70
-------
- 84 -
TABLE VII (continued)
TSP air quality data, 24 hr. Hi-Vol.
(geometric mean Cm in
No.
46
51
53
56
57
60
61
64
69
70
74
75
82
83
84
85
86
87
88
90
91
95
Code
34
39
36
38
33
38
34
34
38
32
31
38
28
28
28
28
28
28
28
28
21
28
SAROAD #
312580001
313180001
313300002
313500001
313980001
314220002
314440001
314760001
315080001
315420001
315920001
316040001
332300002
332360001
332460001
332900001
332900003
332900004
332900005
333480001
334100001
334520001
X- Coord.
47.0
41.5
29.5
31.0
47.5
44.0
44.0
45.5
43.0
61.5
62.5
42.5
95.5
92.0
92.0
91.0
86.5
87.0
95.0
83.5
84.5
96.5
Y-Coord.
61.5
21.0
79.5
49.5
76.5
52.0
61.0
65.5
49.5
75.5
98.0
57.0
65.0
71.0
86.0
62.5
67.5
63.0
71.0
80.0
94.0
72.5
Cm» 71
86.15
42.47
56.97
40.53
70.90
70.46
60.79
85.89
71.52
100.99
62.33
74.07
72.60
68.03
85,84
101.49
82.48
72,42
76.52
66.52
62.65
60.79
Cm' 73
75.19
31.34
51.68
43.71
56.36
56.36
54.87
70.10
58.41
88.46
52.46
74.81
55.98
54.73
72.97
68.72
59.74
57.97
69.76
44.70
51.42
55.15
-------
- 85 -
TABLE VII (continued)
TSP air quality data, 24 hr. Hi-Vol.
(geometric mean Cm in
No.
96
98
99
100
101
102
103
104
105
106
109
111
112
113
114
115
116
117
118
119
120
121
Code
28
28
28
21
24
24
21
21
21
21
22
28
21
29
22
29
29
29
29
21
21
21
SAROAD //
334520002
334520005
334520006
334620002
334680050
334680057
334880001
335200001
335360001
335520001
335780001
335800001
335910001
336340001
336560001
336580001
336580002
336580011
336580023
337320003
337320004
337320005
X-Coord .
103.5
99.5
104.0
81.5
68.0
68.0
76.0
75.5
71.5
89.5
67.0
90.5
88.5
160.0
54.5
129.5
139.0
119.0
118.0
78.5
82.0
85.0
Y-Coord .
65.5
85.0
79.0
90.5
75.5
75.5
107.0
114.5
127.0
99.0
107.0
65.0
97.5
89.0
110.5
91.0
79.0
73.0
83.5
108.5
125.5
129.0
Cm, 71
88.90
53.12
68.89
80.24
83.10
90.47
55.42
50.02
72.76
67.02
51.94
91.61
72.60
35.52
53.52
75.00
72.60
53.92
42.95
43.27
35.61
45.04
C,n. 73
57.83
41.99
48.79
64.88
72.97
77.48
46.29
44.26
65.37
51.03
50.65
73.33
58.41
43.16
57.69
52.46
40.65
51.42
48.30
40.04
32.30
36.51
-------
- 86 -
TABLE VII (continued)
TSP air quality data, 24 hr. Hi-Vol.
(geometric mean Cg, in
No.
122
124
125
Code
21
21
21
SAROAD 9
337320006
337480001
337620001
X-Coord.
77.5
82.5
73.0
Y-Coord.
103.5
101.0
93.0
Cm, 71
85.84
82.06
98.00
Cm, 73
59.44
56.68
90.92
-------
- 87 -
TABLE VIII TSP 24 hr. Hi-Vol. air quality data in 73/2 and 73/3
(geometric mean Cm in yg/m )
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Code
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
32
34
32
SAROAD #
070060001
070060001
070060002
070175001
070260002
070330001
070330002
070330003
070330004
070330007
070330008
070820001
070820005
071080001
071080003
071080004
071080010
071110001
071110005
310180001
311300002
312320001
X-Coord.
124.0
124.0
123.5
106.0
118.5
92.0
94.5
91.0
87.5
91.5
94.5
108.5
109.0
99.5
98.5
99.5
96.5
129.0
129.0
56.5
49.5
60.5
Y- Coord.
116.5
116.5
114.5
135.5
116.5
100.5
103.5
99.0
107.0
105.0
102.5
111.0
112.0
104.5
106.5
109.0
110.0
119.0
116.0
67.5
65.5
71.5
Cm, 73/2
—
42.16
53.29
73.30
40.94
46.28
63.09
51.20
41.40
33.03
62.45
51.13
60.53
—
—
129.90
42.76
—
53.99
—
—
__ ,
Cm, 73/3
—
50.00
65.13
66.08
48.43
49.95
63.80
53.69
52.53
44.53
70.40
72.37
63.94
—
—
83.98
82.89
-.-
47.00
—
—
—
-------
- 88 -
TABLE VIII (continued)
TSP 24 hr. Hl-Vol. air quality data in 73/2 and 73/3
(geometric mean Cm in
No.
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Code
33
35
38
39
32
33
37
39
38
36
36
33
33
31
31
36
31
31
33
32
33
32
SAROAD #
313480001
314140001
314220001
310060002
310180003
310400002
310500001
310560001
310820001
311100001
311100002
311160002
311380001
311440001
311460001
311540001
311560001
311560002
311820001
312180001
312280001
312320003
X-Coord.
52.0
53.0
43.5
64.0
57.5
51.5
24.5
61.0
48.5
24.0
23.5
43.0
45.0
57.0
65.0
36.5
67.0
67.5
62.0
68.5
47.5
60.5
Y-Coord.
70.0
90.5
51.0
28.0
66.5
83.0
58.5
13.5
58.5
88.0
89.0
75.0
75.0
93.5
81.0
77.5
84.0
85.0
89,5
74.0
72.0
73.0
CB, 73/2
—
—
—
53.52
83.10
—
—
34.10
70.52
—
37.08
132.39
39.07
42.80
79.48
33.26
—
46.14
--
116.93
55.49
108.96
Cn, 73/3
—
—
—
61.22
88.04
—
__
43.96
77.92
—
53.08
137.70
48.83
52.32
85.00
46.37
—
51.22
—
98.00
58.67
98.01
-------
- 89 -
TABLE VIII (continued)
TSP 2A hr. Hi-Vol. air quality data In 73/2 and 73/3
(geometric mean CB In jjg/m )
No.
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Code
32
34
38
38
38
38
39
39
36
33
33
38
33
35
35
38
34
39
39
34
38
38
SAROAD #
312320004
312580001
313020002
313060001
313060002
313060003
313180001
313180002
313300002
313480002
313480003
313500001
313980001
314100001
314140001
314220002
314440001
314500001
314500002
314760001
314920001
314920002
X-Coord.
58.5
47.0
38.0
39.0
44.5
23.5
41.5
45.0
29.5
54.5
53.5
31.0
47.5
55.5
54.0
44.0
44.0
59.5
58.5
45.5
37.5
40.5
Y-Coord.
71.5
61.5
54.5
52.5
46.5
39.0
21.0
17.5
79.5
72.5
71.5
49.5
76.5
134.0
139.5
52.0
61.0
35.5
34.5
65.5
47.5
48.5
Cm, 73/2
81.07
67.91
44.61
—
39.79
41.73
32.57
30.46
—
151.24
—
—
59.28
—
—
53.44
53.67
—
36.70
72.95
—
56.19
i
Ca, 73/3
110.30
88.59
58.87
—
60.42
57.54
42.00
49.45
—
118.48
—
—
62.11
—
—
66.26
66.33
—
50.52
76.94
—
78.26
-------
- 90 -
TABLE VIII (continued)
TSP 24 hr. Hi-Vol. air quality data in 73/2 and 73/3
(geometric mean €„, in
No.
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
Code
32
37
38
32
34
31
33
31
38
38
23
21
29
21
28
28
28
28
28
28
28
28
SAROAD f
314960001
315060002
315080001
315420001
315440001
315500001
315860001
315920001
316040001
316040002
334680001
337620001
330280001
331560001
332300001
332300002
332360001
332460001
332900001
332900003
332900004
332900005
X-Coord.
60.5
19.5
43.0
61.5
47.5
57.5
46.5
62.5
42.5
45.5
70.5
73.5
107.0
75.0
94.5
95.5
92.0
92.0
91.0
86.5
87.0
95.0
Y-Coord.
78.0
57.5
49.5
75.5
67.0
105.0
79.0
98.0
57.0
56.0
80.0
92.0
71.5
100.0
64.0
65.0
71.0
86.0
62.5
67.5
63.0
71.0
Cm, 73/2
70.40
48.54
52.21
80.24
46.97
35.15
57.03
46.03
68.89
50.65
—
—
52.94
50.52
—
50.85
52.63
—
53.98
59.04
50.06
71.31
Cm, 73/3
72.01
57.00
63.78
74.00
63.24
45.70
60.90
64.00
81.42
69.20
—
—
60.43
46.36
—
62.99
64.38
—
73.08
61.37
72.99
72.29
J
-------
- 91 -
TABLE VIII (continued)
TSP 24 hr. Hi-Vol. air quality data in 73/2 and 73/3
(geometric mean Cg, in
No.
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
Code
28
28
21
21
21
21
28
28
28
28
28
21
24
24
21
21
21
21
29
29
22
22
SAROAD t
332900007
333480001
334100001
334100002
334480001
334480003
334520001
334520002
334520004
334520005
334520006
334620002
334680050
334680057
334880001
335200001
335360001
335520001
335550001
335550002
335780001
335780002
X-Coord.
84.0
83.5
84.5
84.0
77.5
77.0
96.5
103.5
101.0
99.5
104.0
81.5
68.0
68.0
76.0
75.5
71.5
89.5
134.5
135.0
67.0
71.0
Y-Coord.
62.5
80.0
94.0
93.0
89.5
91.0
72.5
65.5
84.5
85.0
79.0
90.5
75.5
75.5
107.0
114.5
127.0
99.0
93.5
92.5
107.0
101.0
Cm, 73/2
74.57
40.10
54.46
73.64
—
76.08
53.17
60.09
—
42.58
46.30
58.19
75.97
83.92
40.74
48.04
73.29
50.46
—
—
53.08
50.36
C,,, 73/3
74.79
57.47
49.02
61.16
—
82.21
62.87
57.57
—
50.44
58.60
90.75
83.69
80.76
51.81
41.07
63.85
52.69
—
—
54.08
64.35
-------
- 92 -
TABLE VIII (continued)
TSP 24 hr. Hi-Vol. air quality data in 73/2 and 73/3
(geometric mean C^ in pg/ra3)
No.
Ill
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
Code
28
21
29
22
29
29
29
29
21
21
21
21
22
21
21
24
23
25
24
23
26
25
SAROAD 1
335800001
335910001
336340001
336560001
336580001
336580002
336580011
336580023
337320003
337320004
337320005
337320006
337400001
337480001
337620001
334680002
334680003
334680004
334680005
334680006
334680007
334680008
X- Coord .
90.5
88.5
160.0
54.5
129.5
139.0
119.0
118.0
78.5
77.5
85.0
77.5
66.0
82.5
73.0
68.0
72.0
78.0
65.5
72.5
69.0
77.0
Y-Coord .
65.0
97.5
89.0
110.5
91.0
79.0
73.0
83.5
108.5
75.5
79.0
103.5
119.0
101.0
93.0
81.5
84.0
73.0
75.0
87.0
58.5
77.5
CB, 73/2
67.60
52.96
43.54
65.23
45.43
38.28
48.18
48.84
35.81
30.78
32.45
52.04
53.02
52.72
—
84.99
84.20
55.25
80.44
71.22
63.35
73.62
Cm, 73/3
77.12
55.44
42.80
68.83
55.24
49.10
53.79
52.12
55.05
38.65
41.54
57.35
55.85
65.05
—
83.94
95.92
70.90
89.25
76.51
71.38
102.09
-------
- 93 -
TABLE VIII (continued)
TSP 24 hr. Hi-Vol. air quality data in 73/2 and 73/3
(geometric mean CB, in
No.
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
Code
23
26
26
24
25
25
24
26
26
25
26
23
26
25
25
27
27
27
27
27
27
24
SAROAD 8
334680009
334680010
334680011
334680014
334680015
334680016
334680017
334680018
334680019
334680020
334680021
334680022
334680025
334680029
334680030
334680031
334680032
334680033
334680034
334680035
334680036
334680037
X-Coord.
78.5
64.0
68.0
69.0
80.5
85.5
66.0
67.0
70.5
73.5
73.5
73.0
68.0
76.5
82.0
51.5
58.0
50.0
54.5
57.5
46.5
64.0
Y-Coord.
80.5
65.5
72.5
79.0
76.5
73.5
73.0
68.0
67.0
69.5
64.5
81.0
64.0
64.5
66.0
62.5
63.5
57.5
59.5
58.0
51.5
72.0
Cm, 73/2
53.89
85.52
87.03
78.53
57.08
60.13
85.45
50.94
69.49
74.62
—
84.76
—
56.81
53.12
75.46
75.52
79.58
60.50
68.10
53.65
62.63
C., 73/3
80.86
95.77
89.97
104.58
66.65
47.00
93.41
60.55
89.96
66.00
—
86.00
—
65.90
57.65
105.73
93.09
122.73
71.36
76.47
71.87
77.24
-------
- 94 -
TABLE VIII (continued)
TSP 24 hr. HI-Vol. air quality data In 73/2 and 73/3
(geometric mean CB in
No.
155
156
157
158
159
160
161
162
163
164
Code
23
23
25
25
25
25
26
26
25
26
SAROAD *
334680038
334680039
334680040
334680041
334680042
334680044
334680045
334680046
334680047
334680064
X-Coord.
77.0
77.0
71.5
74.0
70.5
83.0
62.5
72.5
77.0
66.0
Y-Coord.
83.0
87.0
77.0
75.5
75.5
69.5
63.0
62.5
57.0
62.0
€„, 73/2
61.87
71.61
86.41
65.43
83.46
73.72
67.70
63.90
103. 32
60.00
Cm, 73/3
70.43
75.67
111.00
62.00
93.39
62.00
77.00
104.68
115.08
82.97
-------
- 95 -
TABLE Cl Rank-order of monitoring stations according to the
first scheme (errors and PI in
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Station //
101
102
9
106
79
156
60
8
152
129
117
99
115
103
118
112
36
69
153
137
7
85
Code
24
24
11
21
29
23
38
11
27
24
29
28
29
21
29
21
31
38
27
25
11
28
2nd Q. Error
0.0
0.0
0.93
0.70
1.08
2.00
0.84
1.88
3.11
1.89
1.59
2.44
0.51
1.91
1.89
2.73
0.75
1.06
1.00
6.45
3.55
6.86
3rd Q. Error
0.0
0.0
0.56
1.11
1.45
0.70
1.97
1.60
0.72
2.18
2.48
1.70
3.88
3.17
3.25
2.77
5.34
5.59
5.66
0.46
3.68
0.93
PI
0.0
0.0
0.76
0.91
1.27
1.35
1.41
1.74
1.92
2.04
2.04
2.07
2.20
2.54
2.57
2.75
3.05
3.31
3.33
3.46
3.62
3.90
-------
- 96 -
TABLE Cl (continued) Rank order of monitoring stations according to the
first scheme (errors and PI in yg/m3)
Rank
23
24
25
26
27
28
29
30
31
32
33
3A
35
36
37
38
39
40
41
42
43
44
Station #
37
113
68
123
51
130
52
43
94
124
11
109
133
149
90
119
96
6
138
155
164
116
Code
31
29
37
22
39
23
39
33
21
21
11
22
23
27
28
21
28
11
25
23
26
29
2nd Q. Error
4.68
0.43
2.06
6.42
1.63
2.90
3.18
7.94
9.26
0.22
2.23
9.36
11.12
0.49
2.56
8.28
7.80
6.85
3.61
3.54
12.55
8.72
3rd Q. Error
3.22
7.82
6.30
2.30
7.22
6.17
6.42
1.89
1.08
10.45
8.70
1.91
0.20
10.92
8.98
3.38
4.25
6.63
10.02
10.14
1.13
4.98
PI
3.95
4.13
4.18
4.36
4.43
4.54
4.80
4.92
5.17
5.34
5.47
5.64
5.66
5.71
5.77
5.83
6.03
6.74
6.82
6.84
6.84
6.85
-------
- 97 -
TABLE Cl (continued) Rank order of monitoring stations according to the
first scheme (errors and PT in up/m^)
Rank
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Station #
74
86
27
80
126
110
159
139
50
57
30
122
19
13
136
144
135
5
73
12
131
87
Code
31
28
32
21
24
22
25
24
38
33
39
21
11
11
24
23
26
11
33
11
26
28
2nd Q. Error
3.46
5.47
7.56
0.92
14.09
0.98
7.97
8.01
8.31
9.43
6.81
9.20
7.66
8.54
4.72
1.35
13.88
8.40
8.70
10.21
0.70
17.17
3rd Q. Error
10.40
8.46
6.38
13.25
1.05
14.19
7.33
7.49
7.33
6.23
8.97
6.59
8.68
8.31
12.21
16.07
4.45
10.03
9.75
8.29
18.43
1.97
PI
6.93
6.97
6.97
7.09
7.57
7.59
7.65
7.75
7.82
7.83
7.89
7.90
8.17
8.44
8.47
8.71
9.17
9.22
9.23
9.98
9.57
9.57
-------
- 98 -
TABLE Cl (continued)
Rank order of monitoring stations according to the
first scheme (errors and PI in
Rank
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
Station #
70
111
46
83
104
98
2
66
95
161
76
160
49
142
67
82
141
128
3
147
63
158
Code
32
28
34
28
21
28
11
38
28
26
38
25
38
25
32
28
26
25
11
25
39
25
2nd Q. Error
7.75
13.67
4.10
13.71
7.27
9.25
9.89
6.48
13.96
7.84
14.88
16.87
13.85
11.58
15.13
12.84
7.10
10.99
10.06
17.29
14.72
3.99
3rd Q. Error
11.78
6.08
15.86
6.30
12.96
11.29
11.01
14.77
7.61
14.24
7.41
5.72
9.15
11.56
8.29
11.42
17.47
14.60
15.62
8.53
11.87
22.86
PI
9.77
9.88
9.98
10.01
10.12
10.27
10.45
10.63
]0.79
11.04
11.15
11.30
11.50
11.57
11.71
12.13
12.29
12.80
12.84
12.91
13.30
13.43
-------
- 99 -
TABLE Cl (continued)
Rank order of monitoring stations according to the
first scheme (errors and PI in pg/m-^)
Rank
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
Station #
121
92
88
4
64
26
157
89
91
146
127
31
61
75
47
162
71
151
148
105
100
44
Code
21
21
28
11
34
39
25
28
21
25
23
38
34
38
38
26
34
27
27
21
21
32
2nd Q. Error
11.08
18.89
18.48
22.72
20.41
18.05
7.87
22.15
17.53
6.94
12.30
2.71
15.58
18.56
16.45
1.13
22.27
12.35
9.70
24.11
10.99
31.92
3rd Q. Error
15.94
8.27
8.95
5.14
7.81
12.27
22.58
8.33
13.49
24.32
20.36
30.65
17.87
15.06
17.45
32,97
12.20
24.14
26.84
12.63
27.00
7.80
PI
13.51
13.58
13.72
13.93
14.11
15.16
15.23
15.24
15.51
15.63
16.33
16.68
16.73
16.81
16.95
17.05
17.24
18.25
18.27
18.37
19.00
19.86
-------
- 100 -
TABLE Cl (continued)
Rank order of monitoring stations according to the
first scheme (errors and PI in
Rank
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
Station #
72
10
154
45
114
134
42
33
150
140
17
40
163
54
120
132
16
38
35
34
Code
31
11
24
32
22
26
32
36
27
26
11
31
25
33
21
25
11
36
33
33
2nd Q. Error
22.25
23.52
26.47
33.06
24.74
24.25
34.62
27.76
12.36
28.53
62.19
37.19
42.06
62.67
33.68
31.07
82.17
54.49
61.83
88.67
3rd Q. Error
18.66
17.93
16.11
11.35
20.39
21.68
13.19
20.62
46.48
31.44
8.01
36.27
36.05
15.85
46.38
49.61
5.09
47.67
56.38
85.79
PI
20.46
20.73
21.29
22.21
22.57
22.97
23.91
24.19
29.42
29.99
35.1
36.73
39.06
39.26
40.03
40.34
43.63
51.08
59.11
87.23
-------
- 101 -
TABLE C2 Rank order of monitoring stations according to the
second scheme (mean error and PI in wg/m3)
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
1A
15
16
17
18
19
20
21
22
Station #
42
135
8
129
106
137
6
133
146
89
101
144
102
134
113
46
136
112
160
90
103
111
Code
32
26
11
24
21
25
11
23
25
28
24
23
24
26
29
34
24
21
25
28
21
28
# Receptors
0
0
1
2
2
2
2
2
2
2
1
1
1
2
3
2
1
2
3
3
6
3
Mean Error
0.0
0.0
0.155
0.332
0.503
0.573
0.746
0.948
0.972
0.974
2.026
2.181
2.673
1.394
0.953
1.458
3.252
1.731
1.251
1.315
0.665
1.438
PI
0.0
0.0
0.155
0.664
1.006
1.146
1.492
1.896
1.944
1.948
2.026
2.181
2.673
2.788
2.859
2.916
3.252
3.462
3.753
3.945
3.990
4.314
-------
- 102 -
TABLE C2 (continued)
Rank order of monitoring stations according to the
second scheme (mean error and PI in
Rank
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Station #
85
60
155
159
152
91
139
118
119
161
141
31
117
115
126
157
86
79
123
151
9
127
Code
28
38
23
25
27
21
24
29
21
26
26
38
29
29
24
25
28
29
22
27
11
23
9 Receptors
2
2
2
2
2
2
2
8
4
2
2
2
8
7
2
1
2
8
4
4
5
5
Mean Error
2.189
2.403
2.471
2.791
2.818
3.096
3.155
0.795
1.680
3.444
3.488
3.569
0.895
1.081
3.802
7.640
3.997
1.001
2.028
2.030
1.667
1.683
PI
4.378
4.806
4.942
5.582
5.636
6.192
6.310
6.360
6.720
6.888
6.976
7.138
7.160
7.567
7.604
7.640
7.994
8.008
8.112
8.120
8.335
8.415
-------
- 103 -
TABLE C2 (continued)
Rank order of monitoring stations according to the
second scheme (mean error and PI in
Rank
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
Station //
164
69
131
19
30
92
100
76
132
153
43
149
147
122
82
148
99
87
138
158
52
70
Code
26
38
26
11
39
21
21
38
25
27
33
27
25
21
28
27
28
28
25
25
39
32
# Receptors
3
9
4
3
5
2
3
3
1
5
2
3
1
7
6
1
11
2
3
3
6
4
Mean Error
2.857
0.976
2.461
3.282
1.980
5.020
3.567
3.592
10.965
2.231
5.698
3.935
11.844
1.708
2.020
12.888
1.175
6.534
4.389
4.568
2.342
3.584
PI
8.571
8.784
9.844
9.846
9.900
10.04
10.70
10.78
10.97
11.16
11.40
11.81
11.84
11.96
12.12
12.89
12.93
13.07
13.17
13.70
14.05
14.34
-------
- 104 -
TABLE C2 (continued)
Rank order of monitoring stations according to the
second scheme (mean error and PI In
Rank
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
Station #
156
130
37
10
11
128
110
27
7
142
109
83
94
124
96
71
2
74
162
154
120
61
Code
23
23
31
11
11
25
22
32
11
25
22
28
21
21
28
34
11
31
26
24
21
34
# Receptors
7
6
3
4
3
2
6
2
6
3
7
5
4
4
6
2
9
9
3
3
2
8
Mean Error
2.056
2.404
5.002
3.788
5.290
7.996
2.675
8.064
2.709
5.490
2.365
3.341
4.177
4.193
2.851
8.681
2.025
2.068
6.251
6.315
9.702
2.470
PI
14.39
14.42
15.01
15.15
15.87
15.99
16.05
16.13
16.25
16.47
16.56
16.71
16.71
16.77
17.11
17.36
18.23
18.61
18.75
18.95
19.40
19.76
-------
- 105 -
TABLE C2 (continued)
Rank order of monitoring stations according to the
second scheme (mean error and PI in
Rank
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
Station #
3
150
95
44
116
64
50
45
80
121
88
105
104
36
13
67
12
98
163
75
51
66
Code
11
27
28
32
29
34
38
32
21
21
28
21
21
31
11
32
11
28
25
38
39
38
# Receptors
7
3
9
3
8
6
10
3
5
3
6
3
7
10
7
5
9
7
1
5
8
9
Mean Error
2.840
6.643
2.243
7.303
2.764
3.750
2.334
7.973
4.855
8.278
4.157
8.555
3.803
2.747
4.072
5.919
3.428
4.433
31.464
6.335
4.227
3.790
PI
19.88
19.93
20.19
21.91
22.11
22.50
23.34
23.92
24.28
24.83
24.94
25.67
26.62
27.47
28.50
29.60
30.85
31.03
31.46
31.68
33.82
34.11
-------
- 106 -
TABLE C2 (continued)
Rank order of monitoring stations according to the
second scheme (mean error and PI in
Rank
111
112
113
114
115
116
117
118
119
120
121
122
123
12A
125
126
127
128
129
130
Station 0
57
49
140
26
35
72
5
73
68
114
47
40
4
63
17
38
54
16
33
34
Code
33
38
26
39
33
31
11
33
37
22
38
31
11
39
11
36
33
11
36
33
# Receptors
7
10
2
6
7
11
9
10
9
6
8
6
6
11
7
7
6
10
11
7
Mean Error
5.097
3.603
18.840
6.303
5.755
3.956
5.068
4.686
5.643
8.703
6.677
8.909
9.476
5.371
9.863
13.737
19.170
14.973
14.301
27.421
PI
35.679
36.03
37.68
37.82
40.29
43.52
45.61
46.86
50.79
52.22
53.42
53.45
56.86
59.08
69.04
96.16
115.02
149.73
157.31
191.95
1
-------
- 107 -
TABLE C3 Rank order of monitoring stations according to the
third scheme (mean error and PI in pg/m-*)
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Station #
42
135
8
106
129
137
133
146
89
6
144
159
102
136
134
113
46
85
103
164
161
160
Code
32
26
11
21
24
25
23
25
28
11
23
25
24
24
26
29
34
28
21
26
26
25
# Receptors
0
0
1
2
2
2
2
2
2
3
2
2
1
1
2
3
2
2
6
3
3
3
Mean Error
0.0
0.0
0.155
0.461
0.502
0.573
0.709
0.972
0.974
0.687
1.224
1.292
2.612
2.708
1.394
0.953
1.458
1.733
0.665
1.348
1.274
1.520
PI
0.0
0.0
0.155
0.922
1.004
1.146
1.418
1.944
1.948
2.061
2.448
2.584
2.612
2.708
2.788
2.859
2.916
3.466
3.990
4.044
3.822
4.560
-------
- 108 -
TABLE C3 (continued)
Rank order of monitoring stations accordinp to the
third scheme (mean error and PI in
Rank
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Station #
60
90
101
158
118
91
152
31
9
115
123
79
86
30
139
149
147
43
126
153
11
112
Code
38
28
24
25
29
21
27
38
11
29
22
29
28
39
24
27
25
33
24
27
11
21
# Receptors
2
3
3
3
8
3
2
2
5
8
5
10
5
5
2
4
4
2
3
5
3
5
Mean Error
2.403
1.852
1.994
1.964
0.795
2.162
3.289
3.354
1.402
0.952
1.626
0.877
1.969
1.980
5.219
2.720
2.852
5.726
3.857
2.359
4.229
2.187
PI
4.806
5.556
5.982
5.892
6.360
6.486
6.578
6.708
7.010
7.616
8.130
8.770
9.835
9.900
10.44
10.88
11.41
11.45
11.57
11.80
12.69
10.94
-------
- 109 -
TABLE C3 (continued)
Rank order of monitoring stations according to the
third scheme (mean error and PI in
Rank
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
Station #
19
122
156
52
130
117
141
76
109
132
70
7
83
111
82
69
50
142
100
155
3
154
2
119
Code
11
21
23
39
23
29
26
38
22
25
32
11
28
28
28
38
38
25
21
23
11
24
11
21
# Receptors
5
7
7
8
6
10
5
4
10
3
6
7
7
4
11
11
11
6
6
7
8
4
9
9
Mean Error
2.552
1.880
2.051
1.842
2.509
1.507
3.078
3.950
1.626
5.721
3.022
2.901
2.955
5.249
1.989
2.093
2.095
3.846
3.847
3.309
2.909
5.867
2.661
2.767
PI
12.76
13.16
14.36
14.74
15.05
15.07
15.39
15.80
16.26
17.16
18.13
20.31
20.69
21.00
21.88
23.02
23.04
23.08
23.08
23.16
23.27
23.47
23.95
24.90
-------
TRI-STATE REGION
COUNTIES AND PLANNING REGIONS
\
FIGURE 1. Tri-State Regional, SMSA*, and Study Areas
The boundary of the Tri-State Region is indicated by the solid line,
whereas that of the study area by the dotted line. The NE New Jersey-New
York SMSA is the same as the study area but excluding the Connecticut
portion.
*SMSA - Standard Metropolitan Statistical Area
-------
- Ill -
Figure 2 Standard network for environmental management.
-------
- 112 -
i
o VALID STATIONS
• INVALID STATIONS
Figure 3: Monitoring stations for the periods 71/2 and 73/2.
-------
- 113 -
Figure 4 Air monitoring stations reported valid data during 73/2
and 73/3.
-------
- 114 -
Figure 5 Concentration isopleths in 71/2 (/jg/m ).
-------
- 115 -
FigureG. Concentration isopleths in 73/2 (>ug/m3).
-------
- 116 -
1 >
10
E
^
•
0.
O
•*
<
T
*
~
e*
ji-i-i_
X
IUU
90
80
70
60
50
40
30
20
10
0
.9
.8
.6
.5
.4
.3
.2
.1
n
-
0 •
ft d 2
-
A A A
-
-
•
O
A
A
-
m m
•
M
^ a
a ^
•
• o
° •
o
•
a
i i i
%
o
A
A
71/2 AQp
73/2 AQp
71/2 AQ$
73/2 AQ8
•
a
o
71/2 HI
73/2 HI
71/2 Wl
73/2 Wl
i
Total School-Age Elderly Non-White
POPULATION
Figure 7: Changes in space average air quality, population
average air quality, health index, and welfare index
during 71/2 and 73/2.
-------
- 117 -
O
0.01 -
°°OI20 30 40 50 60 70 80 90 100 110 120 130
D*[>ug/m3]
Figure 8: Dosage spectrum distribution in the Tri-State Region
-------
- 118 -
O.OI
0.001
20 30 40 50 60 70 80 90 100 110 120 130
D* [/jg/m3]
Figure 9: Population dosage spectrum distribution for total
population.
-------
- 119 -
1.0
O.I
•
o
OOI
0.001
o Total population
A School-age population
a Elderly population
• Non-white population
I
I
20 30 40 50 60 70 80 90 100 110 120 130
D*[/jg/m3]
Figure I0: Population dosage spectra for four different
populations in 71/2.
-------
- 120 -
•
o
0_
o Total population
School-age population
a Elderly population
• Non-white population
0.01 -
0.001
20 30 40 50 60 70 80 90 100 110 120 130
D* [>jg/m3]
Figure II Population dosage spectra for four different
populations in 73/2.
-------
- 121 -
Figure 12= Risk probability of daily concentrations exceeding the primary
24 hour average air quality standard in 71 /2.
-------
- 122 -
Figure I3: Risk probability of daily concentrations exceeding the level of
the primary 24hour average air quality standard in 73/2
-------
- 123 -
Rgure I4: Risk probability of daily concentrations exceeding the
secondary 24 hour average air quality standard in 71/2.
-------
- 124 -
Figure 15= Risk probability of daily concentrations exceeding the
secondary 24 hour average air quality standard in 73/2.
-------
- 125 -
Figure I6: Risk probability of daily concentrations exceeding 75/jg/m3
in 71/2.
-------
- 126 -
Figure 17= Risk probability of daily concentrations exceeding 75/jg/m3
in 73/2.
-------
- 127 -
Figure 18= Risk probability of daily concentrations exceeding 60>ug/m3
in 71/2.
-------
- 128 -
Figure 19= Risk probability of daily concentrations exceeding 60/ig/m3
in 73/2.
-------
- 129 -
= 60>jg/m
annual
secondary
Cs=75
annual
primary
CS=I50
24 hour
secondary
= 250
24 hour
primary
i i i i
0.01 -
0 10 20 30 40 50 60 70 80 90 100
f
* 10,
Figure 20 Risk spectrum distribution in 71/2.
-------
- 130 -
CO
O
Cs=60/jg/m
annual
secondary
annual
primary
CS=I50
24 hour
secondary
Cs=250
24 hour
primary
0 10 20 30 40 50 60 70 80 90 100
f
* o
Figure 2h Risk spectrum distribution in 73/2.
-------
- 131 -
•o-o
I O.I
or
o_
0.01
Cs=60/jg/mT
annual
secondary
Cs=75
annual
primary
CS=I50
24 hour
secondary
Cs=250
24 hour
primary
n.
I I
0 10 20 30 40 50 60 70 80 90 100
f [%]
Figure 22: Population-at-risk spectrum distribution in 71/2.
-------
- 132 -
cr
o.
Cs=60jug/m
annual
secondary
cs=75
annual
24 hour
secondary
24 hour
primary
0 10 20 30 40 50 60 70 80 90 100
f*
Figure 23; Population-at-risk spectrum distribution in
73/2.
-------
- 133 -
•-p
E
Q_
T
"^i*
or
I.U
.9
.8
.7
.6
.5
.4
.3
.2
.1
0
.9
.8
.7
.6
.5
.4
.3
.2
.1
n
-
-
o
-
o
8
•— __
• 71/2
o 73/2
•
o
*
o
f 4
Cs=60 Cs=75 CS=I50 Cs=260
Figure 24 Regional risk index and population-at-risk index
in 71/2 and 73/2.
-------
- 134 -
Figure 25= Valid monitoring stations for 1971 and 1973.
-------
- 135 -
Figure 26 Concentration isoplethsin 1971 (jug/m3).
-------
- 136 -
Figure 27 Concentration isopleths in 1973 (jug/m3).
-------
- 137 -
100
90
to
80
Q.
O
70
O
60
50
i.o"
0,5
0
1971
— 1973
AQ,
AQS
-o— -o o -o
Total School-age Elderly Non-white
POPULATION
Figure 28= Values of air quality indices in 1971 and 1973.
-------
- 138 -
Q
CO
0.01 h
0.002
" ------ 50 60 70 80 90 100 110 120130
D*[>jg/m3]
Figure 23 Dosage spectrum distribution in 1971 and 1973.
-------
- 139 -
Q
Q-
0.002
0.01 -
0 10 20 30 40 50 60 70 80 90 100 IIO 120 130
D*
Figure 30: Population dosage spectrum distribution for total
population.
-------
- 140 -
1.0
•
O
001
0.002
-0-0-
o Total population
A School-age population
o Elderly population
• Non-white population
0 10 20 30 40 50 60 70 80 90 100 110 120 130
D* [jug/m3]
Figure 31= Population dosage spectra for four different populations
in 1971.
-------
- 141 -
1.0
O.I
o.oi
0002
• •
o Total population
A School-age population
a Elderly population
• Non-white population
0 10 20 30 40 50 60 70 8O 90 100 110 120 130
D*
Figure 32 Population dosage spectra for four different populations
in 1973.
-------
- 142 -
Figure 33= Isopleth map of geometric mean concentrations in 73/2
with 130 valid monitoring stations.
-------
- 143 -
Figure 34: Isopleth map of geometric mean concentrations in 73/3
with 130 valid monitoring stations.
-------
- 144 -
50 100
RANK[-]
Figure 35= Growth of the error induced at each station under
Scheme I
130
-------
- 145 -
oMOST IMPORTANT STATION
• LEAST IMPORTANT STATION
Figure 36: Locations of the 10 most and the 10 least important
stations by Scheme I.
-------
- 146 -
150 -
100 -
o:
o
a:
cc
Ld
u_
o
50 100
RANK [-]
Figure 37: Growth of the error in receptor concentrations under
Scheme IT.
130
-------
- 147 -
°MOST IMPORTANT STATION
• LEAST IMPORTANT STATION
Figure 38: Locations of the 10 most and the 10 least important
stations by Scheme TX.
-------
- 148 -
Rgure 39: Locations of the 10 least important stations by Scheme HI.
-------
- 149 -
150
io-
100
cr
o
cr
cr
u
50
100
50
RANK [-]
Figure 40: Growth of the error in receptor concentrations under
Scheme IE.
130
-------
- 150 -
Figure 41- tsopleth map of geometric mean concentrations in 73/2
from 68 odd numbered stations.
-------
- 151 -
Figure 42: Isopleth map of geometric mean concentrations in 73/2
from 62 even numbered stations.
-------
- 152 -
FIGURE 43 Key to the synbols used in Figures 44 through 47
Network subset
Total network
Even t network
Odd # network
Scheme I network
Scheme II network
Scheme III network
No. of stations
130
68
62
62
62
62
Symbol
—
•V
A
O
o
-------
- 153 -
100
90
80
7°
O 60
50
40
30
4
1
Tri-State N.Y.
N.J. New York Union
STATE
COUNTY
Rgure 44: Space average air qualities estimated from the
total network and from each of the five half size
sub-networks.
-------
- 154 -
100
90
80
10
E 70
v.
o»
0°" 60
50
40
30
I
Tri-State N.Y. N.J. New York Union
STATE COUNTY
Figure 45: Population average air qualities estimated from
the total network and from each of the five half
size sub-networks.
-------
- 155 -
IUU
90
80
70
60
50
40
30
20
10
n
-
-
-
-
-
-
5
- c
t
<
i
:
3
»
>
i
9
i
I
X
°
; ;
k
f
3
1
1
i
(
-i
f
i
9
t
. ,
t ~" f
> t
1
r-
H
v. 1 ,
H
V-(
Tri-State N.Y. N.J. New York Union
STATE
COUNTY
Figure 46= Health indices estimated from the total network
and from each of the five half size sub-networks.
-------
- 156 -
\JU
90
80
70
60
50
40
30
20
10
n
- /
-(
i
A
\
\
i
D
-.
t
t
J
f
- c
b
a
-
-
1
\
>_
t
3
t
«
1
I
/
*•
1
k
r
M •~"~^~~ ~
u li
1
k
»
r
-<
i i
"T
«4*
>-
D
Tri-State N.Y.
N.J. New York Union
STATE
COUNTY
Rgure 47: Welfare indices estimated from the total network
and from each of the five half size sub-networks.
-------
- 157 -
CO
X
X AXIS
Figure Ah Pictorial representation of variables appearing in
interpolation formula.
-------
- 158 -
120
100
10
o
LU
O
80
O 60
h»
40
O
<-> 20
0 L
(true) linear
2 4 68
DISTANCE X [—]
10
Figure A2: Performances of the linear and the pseudo
linear interpolation formula.
-------
- 159 -
LU
O
H
O
DISTANCE X
C =
1-1
Figure A3: Performance of the linear interpolation formula.
-------
- 160 -
UJ
O
z
IS
CO
O
DISTANCE X
= (Z Ci/n)/(Zl/n)
i«i i-i
Figure A4: Performance of the pseudo-linear
interpolation formula.
-------
- 161 -
120
100
10
80
O
O 60
tr
h-
2
LU
O
40
20
(true) parabolic
(true) parabolic
pseudo-parabolic
pseudo- parabolic
B
246
DISTANCE X [-]
8
10
Figure A5: Performances of the (true) parabolic and the
pseudo-parabolic interpolation formula.
-------
- 162 -
LU
O
CO
O
DISTANCE X
C=(IsjCj/rh/(Isj/r?)
Figure A6: Performance of the (true) parabolic
interpolation formula
-------
- 163 -
UJ
O
CO
o
DISTANCE X
c=(ZCi/n)/(H/n)
i-l i«l
Figure A7 Performance of the pseudo-parabolic
interpolation formula.
-------
- 164 -
120
100
10
E
^ 80
o
O 60
cr
h-
z:
UJ
o
40
8 20
0
w; *Cj
InC.
2468
DISTANCE X[—]
10
Figure A8: Performances of the three different weighted
pseudo-parabolic interpolation formulae.
-------
- 165 -
Ld
U
00
o
DISTANCE X
C=(f;Cj/Wir?)/(I I/win)
Figure A9: Performance of the weighted pseudo-parabolic
interpolation formula (wj=lnCj).
-------
- 166 -
DISTANCE X
3
Figure AIO: Performance of the weighted pseudo-
parabolic interpolation formula (wj = C5).
-------
- 167 -
UJ
O
o
Figure A
DISTANCE X
i'l
i'l
Performance of the weighted pseudo- 2
parabolic interpolation formula (Wj=Cj)
-------
- 168 -
eoo
180
o
UJ
t 140
0 160
-------
|