Document 997Nz9jGJxqm5mRa1ZemqBmaR
The relationship between EPA air lead levels and population density
Jayne's T . Pirkle M . D. Ph.P. Clinical Chemistry Division Center for Environmental Health Centers for Disease Control
Ethyl corporation used population density (linearly interpolated from 1970 and 1980 estimates) to represent air lead levels in their regression model. The ..question of whether air lead levels measured by EPA and population density are related can be analyzed from EPA air lead data and Bureau of Census population data for cities across the U.S.
Bureau of Census population data was collected in 1980., _ EPA also has air lead measurements summarized as quarterly averages during 1980 for many U.S, cities (National Aerometric Data Bank, EPA). Some cities have multiple sites of air lead measurement while others have only one. The 1981 Statistical Abstract (Bureau of Census) gives population data on 63 cities for which EPA also has air lead measurements for at least two quarters of 1980. The 1980 population density versus the 1980 quarterly average EPA air lead level for each city is plotted in Figure 1, Regression of air lead on population density (Table 1) shows that population density can account for less than 5 percent of the variation in the air lead levels. Thus, population density is an inadequate variable to represent EPA air lead levels.
TEH 0533077
AIR
LEAD
LEVELS
VS.
POPULATION
DENS: . i
r i
1.5
1 -U
1.3
1,2-3
1.1
2Bms -
1.0^
o.s-^
p,e^
rrn<nr" o io m r
0.7-
O.S
0.5-3
(ug/m3)
0,4 0-3-
0.2-
+ ++
0.J-:
+ *+
+ ++ *
+* +
+ {+ +.
*
f +*
+
+ +f
* *+
+
+
+ +
+' + 4*
o.o-
i ui|
2
tnri^ mtn irrfn'rrTrTTTirTi n'imjTiim m JT* rm mp m I nnjurvMTtm iumi ip. L-CT\-T\i~t J 7L.-v-1
3 4 5 6 7 8 ? 10 II 12 13 |<l_
POPULftTIGN DENSITT
15
(1000 persons/square mile)
Fiou<?e L
TEH 0533078
DUP050034348
Table 1
Regression of the 198.0 quarterly average EPA air lead levels vs. 1980 population density for 63 O.S. cities
DEP VARIABLE: MEAN
MEAN AIR LEAD LEVEL
SUM OF
MEAN
SOURCE MODEL
DF 1
SQUARES 0.190054
.SQUARE 0.190054
ERROR 61 3.973732 0.065143
C TOTAL 62 ROOT MSE
4.163785 0.255232
R-SQUARE
DEP MEAN
0.489400
ADI R-SQ
C.V. -
52.15197
F VALUE 2,917
0.0456 0,0300
VARIABLE DF
PARAMETER ESTIMATE
STANDARD T FOR HO: ERROR PARAMETERS)
INTERCEP DENSITY
1 1
0.401095
0.060883
0.016694 0.009773755
6.588 1.708
PROB>F 0.0927
PROS > |T| 0,0001 0.0927
/ \
TEH 0533079
DUP050034349