\ <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

-------