EPA-AA-TEB-EF-8 2-4
Ambient Versus Predicted Carbon Monoxide Levels
by
Mark Wolcott
September 1982
Test and Evaluation Branch
Emission Control Technology Division
Office of Mobile Sources
U.S. Environmental Protection Agency
-------
Ambient Versus Predicted Carbon Monoxide Levels
Introduction
MOBILE2 is a computer program that calculates emission factors for
hydrocarbons, carbon monoxide, and oxides of nitrogen for highway motor
vehicles. The program uses the calculation procedures described in the
Compilation of Air Pollution Emission Factors; Highway Mobile Sources(l),
to calculate emission factors for eight individual vehicle types in three
regions of the country. These emission estimates depend upon various
conditions, such as ambient temperature and vehicle usage. MOBILE2 can
estimate emission factors for any calendar year between 1970 and 2020.
Various groups have tried to characterize the accuracy of the MOBILE2
emission factors model by comparing its emissions results to air quality
measurements during the 1970-80 period. This report shows that such
comparisons are sensitive to the input assumptions used to generate the
emissions results.* It also compares the reduction in carbon monoxide
(CO) tailpipe.emissions predicted by the MOBILE2 model to the reduction
in ambient levels as they were measured by the nation's network of carbon
monoxide monitors. The comparison is made only for ambient CO since that
pollutant, more so than the others, is the result of emissions from motor
vehicles.
The ambient CO concentrations used in this investigation were taken from
SAROAD. SAROAD is the acronym for the Storage and Retrieval of
Aerometric Data system maintained by EPA. This system contains
information collected by federal, state, and local agencies on various
pollutants and hazardous chemicals, including carbon monoxide.
Discussion
Figure 1 shows the second highest eight hour moving average CO level from
SAROAD and demonstrates how significant the improvement in U.S. cities'
air quality has been during the decade of the 70s. The second highest
eight hour moving average CO level was used because that is the present
form of the National Ambient Air Quality Standard (NAAQS) for CO. On
each of the figures in this report the vertical axis represents the
summation of all the CO changes from the previous and current years.
Since very few monitors measure data continuously from 1970 to 1980, the
change is expressed as the difference in the base 10 logarithms of the
recorded concentrations from one year to the previous year. For example,
the 1970-1971 level is calculated from the logarithm of the 1971 second
highest eight hour moving average CO level measured by each monitor
subtracted from the logarithm of the corresponding 1970 monitor level.
These changes .are then accumulated over all years. The logarithms are
*While the assumptions used in this report may be appropriate for certain
local projections, national comparisons of alternate control strategies
have, heretofore, required the use of standard Federal Test Procedure
conditions.
-------
used since they are symmetric, while calculations based upon percentage
differences are not symmetric. For example, if during a three-year
period ambient concentrations are recorded as 20 ppm, 10 ppm, and 20 ppm,
the net difference based on logarithms is zero. Based on a percentage
calculation, the cumulative difference would be plus 50 percent
(10-20/20=-50%; 20-10/10=+!00%; -50%+100%=+50%).
Figure 1 also shows the reduction in fleet emissions calculated from
MOBILE2 under the standard conditions associated with the Federal Test
Procedure (FTP). These reductions are calculated by the same procedure
that was used to ascertain the changes in ambient carbon monoxide so that
a comparison between the two would be consistent. Such a comparison
suggests that, over the past decade, tailpipe emissions as predicted by
MOBILE2 have declined more than ambient concentration levels.
There are, however, several aspects of the underlying MOBILE2 model and
SAROAD data used to construct Figure 1 that should be noted to properly
interpret this figure.
1. The driving patterns associated with the elevated CO concentra-
tions represented in Figure 1 may not be reflected by the FTP
cycle.
2. Most violations of the CO National Ambient Air Quality Standard
occur during the months of November, December, and January when
temperatures are substantially colder than the 68°F - 86°F range
associated with the standard Federal Test Procedure*.
3. The rush hour traffic often associated with high ambient CO
concentrations may consist mostly of passenger vehicles rather
than the standard mix of vehicles assumed by MOBILE2.
4. The number of monitors fluctuates over time. For instance, only
43 monitors measured CO in both 1970 and 1971. This contrasts
with 178 monitors in 1973 and 1974 and with 247 monitors in 1979
and 1980. (See Table 1.)
5. MOBILE2 does not account for growth in the number of vehicle
miles traveled. It strictly calculates the grams of pollutant
emitted per mile traveled.
*Seventy-five degrees Fahrenheit has been used throughout this report to
characterize the FTP temperature range.
-------
The effects of each of these factors are addressed in the remaining parts
of this paper. It is important to realize that the discussion is
presented in terms of a sensitivity analysis. The operative factors at
any individual monitoring site may be different. The purpose is really
to point out that a straight FTP MOBILE2-ambient CO comparison may be
insufficient.
1. Driving Cycle
The Federal Test Procedure was developed in 1970 to model a
typical urban driving cycle (2,3). As defined, the driving
cycle is composed of two parts: the start-up phase and the
stabilized phase. The start up phase can either be a cold start
or a warmed-up start. The morning rush hour traffic on major
roads and in the central business district is probably best
characterized by the ' stabilized phase since vehicles will have
warmed up by the time they reach these heavy traffic areas. The
evening rush hour may be best characterized either by the cold
start plus stabilized phases (if most traffic consists of people
returning to their homes after working a full eight hour day) or
by the hot start plus stabilized phases (if most traffic
consists of people returning from short trips into the city).
Figure 2 displays the reduction in tailpipe emissions associated
with each of four driving patterns.
Of the four patterns, the stabilized phase of the Federal Test
Procedure may best characterize the driving patterns associated
with elevated CO concentrations. By the time the morning
suburban traffic reaches the central business district, most of
the vehicles will have warmed up. And in the afternoon, when
vehicles that have sat in parking lots all day get started and
reach congested roadways they, too, will have warmed up.
Afternoon cold start emissions away from the ambient monitor may
be important from the standpoint that such emissions tend to
increase background CO levels. See references 4 and 5 for a
discussion of this effect.
2. Ambient Temperature
Another concern that one might have with the comparison depicted
in Figure 1 is that most elevated CO levels of the carbon
monoxide standard are recorded during November, December, and
January. To assess the potential effect of cold temperature, it
was first necessary to determine the temperatures that actually
prevailed during periods of high ambient CO concentrations.
Table 2 lists the 30-year average temperatures during November,
December, and January at all of the Standard Metropolitan
Statistical Areas (SMSAs) in which there were CO monitors. The
overall average temperature in low-altitude, 49-state areas was
39.7°F(6), a value substantially below the standard Federal Test
Procedure's range of 68°F - 86°F. Figure 3 shows that slightly
less improvement in stabilized fleet CO emissions occurred at
-------
colder temperatures than occurred in this FTP temperature range.*
3. Vehicle Type
Another concern that might be expressed in the comparison of
MOBILE2 predictions to "real world" ambient carbon monoxide
levels is that the rush hour traffic often associated with high
concentration levels may consist mostly of passenger vehicles.
The default mix of vehicle types contained in MOBILE2 may not
really be representative of rush hour traffic. Figure 4 shows
the improvement in CO emissions averaged (1) over all vehicle
types and (2) over only personal passenger vehicles. The two
curves in the figure are both for stabilized operation at 40°F.
As Figure 4 shows, personal passenger vehicles are estimated to
have experienced a somewhat greater reduction in tailpipe CO
emissions than has the combined vehicle fleet.
Notice, too, that by 1980 the lower curve in Figure 4 is
essentially at the same position in that figure (-0.22) as is
the lower curve in Figure 1. That is, the net effect of the
adjustments considered heretofore is zero. While the net effect
of adjusting for specific local conditions may not be zero, this
analysis suggests that standard FTP conditions may, indeed, be
appropriate for national analyses.
4. Screening of Monitor Data
Monitoring data, like all other data, is subject to a variety of
errors. Due to limitations on the resources available for this
project, it was not possible to screen extensively all of the
monitoring data that went into this report. The total data set
consists of over ten million, eight-hour moving average
observations. Since the level of the CO NAAQS is not to be
exceeded more than once per year, the most important eight-hour
average each year is the second highest. Thus, the 10 million
observations from the original set were first reduced to
approximately 1000 second high values. Of these, four values of
greater than 70 mg/m-' were then deleted on the basis that the
highest confirmed CO value since 1977 has been 38.5 mg/m^.
While a few values higher than 38.5 mg/m^ were recorded prior
to 1977, those values greater than 70 mg/m^ definitely
appeared to be anomalous. The resulting data set is labeled
"edited" in Figure 5. From this edited set a subset was
constructed that excludes all observations for which there was
more than a 10 ing/m^ change during two consecutive years.
That data set is labeled "edited and screened" in Figure 5.
*Since MOBILE2 rounds input temperatures to whole degrees Fahrenheit, the
cold temperature curve of Figure 3 has been labeled "40°F".
-------
Three possibilities are offered to explain changes greater than
10 mg/m3 during two consecutive years. The first possibility
is data error. Examining the time series for suspect data
reveals that, in many instances, a large change in one direction
in one year is offset by an equally large change in the opposite
direction in the following year. This is the predominate reason
why the two curves in Figure 5 are so similar. The second
explanation offered is that traffic patterns may have shifted
from one year to the next. Such a shift may have occurred
because of intentional traffic control measures by the local
government. It may also have been the result of independent
decisions rendered by the area's drivers. For example, drivers
may want to avoid an intersection that has become more congested
over time or an area undergoing road or building construction.
The third possibility offered to explain large one—year
differences in ambient concentration levels is a change in
monitor calibration procedures. If any of these three
possibilities occurred, they would confound the comparison of
reduction in ambient concentrations to reduction in emissions.
5. Monitor Selection
Since the lower two curves in Figure 6 are similar, one can
infer that monitors within SMSAs have recorded about the same
reduction in ambient CO concentrations as have the monitors
within non-SMSA areas. However, if the monitors within SMSAs
that measure the highest second high CO concentrations are
compared without regard to whether the monitors are the same for
each comparison, then ambient CO concentration levels may not
have improved quite as much as would first appear. The curve
labeled "mixed SMSA monitors" shows the reduction recorded under
this circumstance.
As mentioned previously, one explanation for the difference in
estimated reductions is that traffic may have shifted away from
one monitor toward another. Ideally, a set of identical
monitors each surrounded by a constant volume of traffic could
be found. Such a set would allow one to isolate the effect of
reduced emissions on ambient CO concentrations. Unfortunately,
too few monitors come close to satisfying this criterion.
6. Growth in Vehicle Miles Traveled
Traffic in the vicinity of the ambient monitors may or may not
have appreciably shifted. However, it has increased.
Nationwide, personal passenger vehicle miles traveled (VMT) grew
at an overall rate of 2.3 percent per year during the past
decade. (7) Figure 7 shows the increase in VMT from 1970 to
1980. Traffic designated as "urban" by the Federal Highway
Administration (FHwA) grew at a greater rate than traffic
designated as "rural".
-------
Growth in vehicle miles traveled is an important factor to
consider in any comparison of MOBILE2 and the "real world",
since the emission trends predicted by MOBILE2 are expressed
simply in grams of CO emitted per mile traveled. MOBILE2
itself, without VMT growth, overpredicts the reduction in
ambient carbon monoxide levels.
According to the FHwA (8), there are two components of urban
growth. One component is geographic area, the other is traffic
density. During the past decade those locations designated
urban have come to encompass larger geographic areas. In
addition, other areas once designated rural, are now designated
urban. Meanwhile, traffic density within the core urban areas
has also increased.
It is important to consider both components when examining their
effect on ambient air quality. On one hand, it can be argued
that expansion of an urban area may increase background carbon
monoxide levels. That, in turn, may increase peak carbon
monoxide concentrations during CO episodes.(5) Accepting this
argument would lead one to adjust MOBILE2 emission factors by
the full 2.3 percent annual growth rate. On the other hand, it
could also be argued that only an increase in traffic density
will tend to increase ambient CO levels. Accepting this
alternate argument would lead one to adjust the emission factors
by less than 2.3 percent.
Summary
Figure 8 shows the effect of combining the assumptions made in the
previous sections. The second highest eight-hour moving average CO
levels measured by SAROAD monitors in the nation's low altitude, 49-state
regions were edited for potentially bad data and screened for outliers.
The resulting data set is plotted in Figure 8. Also plotted are the data
from the second highest averages within SMSAs. These two curves are then
compared with the predictions from MOBILE2, adjusted for nationwide urban
growth shown in Figure 7, and for personal passenger vehicles operating
in a stabilized mode at wintertime temperatures in the same low altitude,
49-state regions. As can be seen from the figure, the cumulative
reduction in ambient CO predicted from MOB1LE2 emission factors and
growth in urban VMT is greater than that recorded by the "mixed" SMSA
monitors but less than that recorded by the "matched" monitors.
It is also possible to compare the two ambient concentration curves with
two additional adjusted MOBILE2 curves. These latter two curves adjust
the MOBILE2 emission factors by the growth rate assumptions currently
used by EPA in air quality analyses performed in support of regulations
and in responses to congressional requests for information. The lower of
the two curves assumes that personal passenger VMT increases at a rate of
0.4 percent, while the higher curve assumes that VMT increases at a 2.4
percent per year rate. While the 2.4 percent per year rate is close to
the historical average, the 0.4 percent rate reflects the possibility
-------
that only traffic in the immediate vicinity of the monitors influences
the measurements recorded by them. Since many monitors are located in
central business districts and since many of the traffic corridors in
those districts operate near capacity, traffic may not be able to
increase there at the same rate it can in the rest of the city. Indeed,
the most recent EPA responses to congressional requests for analysis of
the low altitude CO problem have indicated a preference for the
assumption that personal passenger VMT will grow at a 0.4 percent annual
rate. Adjusting for this fairly modest growth rate, it appears that
MOBILE2 characterizes the ambient monitoring data fairly well.
-------
References
Compilation of Air Pollutant Emission Factors; Highway Mobile
Sources, EPA-460/3-81-005, U.S^Environmental Protection Agency, Ann
Arbor, Michigan 48105, 1981.
Kruse, R. and Huls, Development of the Federal Urban Driving
Schedule, U.S. Environmental Protection Agency, SAE Paper 730553,
May, 1973.
Huls, Evolution of Federal Light Duty Mass Emission Regulations,
U.S. Environmental Protection Agency,SAE Paper 730554, May, 1973.
Delman, A. , CO Hot Spot Analysis - The Uses and Limitations of
Emission Inventories, Department of Environmental Programs,
Metropolitan Washington Council of Governments, Washington, D.C. ,
April 1982.
Wolcott, M. , Carbon Monoxide Episodes, U.S. Environmental Protection
Agency, EPA-AA-TEB-EF-82-3, November, 1981.
Ruffner, J., Frank Bair, editors, The Weather Almanac, Second
Edition, Avon Book, 1979.
Highway Statistics, Federal Highway Administration, U.S. Department
of Transportation, 1970-1980.
Personal communication with J. Thwing, Federal Highway
Administration, U.S. Department of Transportation, February 12, 1982.
-------
FIgur e 1
Lou Rltllude. 49-Stotes
o>0
o
z!0
• 0
o
o>0
c
L
O-
o
>— _
I
I
L -
O
O _
E
3.
01
0_
12
10
08
06
OU
02
0
,02
,04
,06
. 08
. 10
. 12
. 14
. 16
. 18
. 20
. 22
. 24
. 26
i i
i i i I
Second high 8 -
hour overoge CO,
oil monitors,
edited
Federol test
proceedure
fleet emissions
70_71 72.73 74.75 76.77' 78.79
71.72 73.74 75.76 77.78 79.80
T e o P s
Figure 2
Driving Potterns, 75 F . Fleet Emissions
o>0
o
O
L
O-
O
>— _
t
I
L -
O
O _
£
3 -
tn
o _
3
6 ~
3
12
10
08
06
04
, 02
0
02
04
06
08
10
. 12
, 1 4
, 16
..18
. 20
. 22
. 24
26
Cold start plus
stabilized
Federol test
proceedure
• Stabilized onl
Worm stor.t plus
stabilized
70.71 72.73 74.75 76_77 78.79
71_72 73,74 75.76 77.78 79.80
Tears
-------
Figure 3 '.
Stoblllzed Driving, Fleet Emissions
0. 12
o>0 ,
o
=10
f.0
e>
o>0
c
°0
I
L
O
o
E
3.
l/l
O.
_>
3
E ~
3
10
08
06.
04
, 02
. 0
. 02
. 04
. 06
. 08
. 10
. 12
. 16
. 1 8
. 20
. 22
. 24
. 26
40 F
75 F
70.71 72_73 74_75 76_77 78_79
71_72 73_74 75_76 77_78 79_80
T e or s
Figure 4
Stoblllzed Driving ot 40 F
o>0
o
00
o
o>0
c
JO
°0
i.
o -
c
.<— _
I
I
t. -
o
o _
e
3.
tn
o _
12
1 0
08
06
04
02
0
02
04
06
08
1 0
1 2
1 4
16
1 8
20
. 22
24
26
J L
I I I I
Fleet emissions
Perconol
possenger
vehicles
70.71 72_73 74.75 76.77 78.79
71.72 73.74 75.76 77_78 79.80
-------
Figure 5
Screening of Monitor Doto
o>Q
o
do
«0
o
o>0
c
c
o-
I
o
o
o _
o
E
(ft
O _
_>
O
e ~
3
u -
12
10
08
06
01
,02
,0
.02
, 04
,06
. 08
. 10
. 12
. 14
. 16
. 18
.20
.22
. 24
. 26
J I I
Second high 8~
hour overoge CO,
oil monitors,
edited
Second high 8~
hour overoge CO,
oil monitors,
edited ond
screened
70_71 72_73 74_75 76.77 78.79
71.72 73_74 75.76 77.78 79.80
Y e o r s
Figure 6
tor Selection, Lou flllitude. 49-Stotes
o>0 ,
o
do.
0,0,
0
o>0 ,
c
L
O ~ ,
0
V- _
I
I
L - ,
0
O _
Second high 8~
hour overoge CO,
mixed SMSR
monitors, edltec
Second high 8~
hour overoge CO,
identlcol SHSfl
monitors,
edited on d
screened
Second high 8~
hour overoge CO,
oil monitors,
edited ond
screened
.71 72.73 74_7S 76.77 78.79
71.72 73.74 75.76 77_78 79.80
T e o r s
-------
Figure 7
Notlonwide Personol Passenger VMT Growth
o>
o .
a
c ,
o
jC
LJ
L
O
O •
E
3
ID
_J
O
150
mo
130
120
,110
, 100
.090
.080
. 070
. 060
. 050
. OUO
. 030
. 020
—• Urbon oreoe
Rural oreae
70.71 72.73 7i4_75 76.77 78_79
71_72 73.74 75.76 77.78 79.80
Stobt
0. 12
"o>0. 1 0
o
do. 08
0,0. 06
«0 . 0 4
c
j?0. 02
°0.0
o-. 02
06
08
10
12
14
16
18
20
22
24
Figure 8
llzed Driving. 40 F. Low Rltltude. 49-Stotes
o
£
3 _
in
o ~
>
-i -
_>
o _
3
6 '
D
<->-. 26
j_
J_
Second high 8~
hour overage CO,
mixed SM5R
monltors
Second high 8~
hour overoge CO,
oil monitors
LDV emissions
adjusted for
nationwide
urbon growth-
IDV emissions
adjusted Tor
2,4X growth
LDV emissions
adjusted for
0.4X growth
70.71 72_73 74.75 76-77 78.79
71.72 73.74 75.76 77.78 79_80
Tears
-------
Table 1
Number of Monitors
Years
Description 7O-7 I 71-72 72-73 73-74 74-75 75-76 76-77 77-78 78-79 79-8O
All monitors, <13 1O1 132 178 236 269 270 246 252 247
od i ted
All monitors, 39 9O 128 165 227 267 261 241 249 243
edited and screened
Identical SMSA monitors. 13 35 47 60 83 1O2 97 92 99 95
ed i ted
Identical SMSA monitors. 12 35 47 57 8O 1O2 95 91 99 95
edited and screened
Mixed SMSA monitors. 18 51 69 09 111 128 133 133 135 136
ed i ted
-------
Table 2
Monthly Winter Temperature
30-Year Average
Tempera lure (F)
Urban Area
State
Nov
Dec
Jan
Anchorage
B i rm i ngham
l-lun t sv i 1 t e
Mob i 1 e
Phoen i z
Tuscon
Anaheim-Santa Ana-Garden Gro
Bakersf f e 1 d
Fresno
Los Angeles-Long Beach
Modes to
Oxriard-Slml Va 1 ) ey-Ven tura
Riverside-San Bernard 1 no-On t
Sacramen to
Sal i nas -Seas i de-Mon terey
San Diego
San Francisco-Oakland
San Jose
Santa Barbara-Santa Maria-Lo
Santa Rosa
S lock ton
Val lejo-Falrfield-Napa
Colorado Springs
Denver-Bou 1 der
For t Col 1 ins
Fort Wayne
Gree 1 ey
Pueblo
Br idgepor t
Mar t f ord
Mew Britain
New Haven-West Haven
New London-Norwich
Norwa 1 k
Stamford
Wa terbury
Wash i ng ton , - DC
Fort Lnuderda 1 e-Ho 1 1 ywood
Jacksonv i 1 1 e
Miami
Or 1 ando
Pensaco 1 a
Ta 1 1 ahassee
Tampa-St. Petersburg
West Palm Beach-Boca Raton F
A 1 1 an ta
Honol u 1 u
Cedar Rapids
Des Molnes
Oubuque
Boise C 1 ty
AK
AL
A I.
AL
AZ
AZ
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CO
CO
CO
CO
CO
CO
CT
CT
CT
CT
CT
CT
CT
CT
DC
FL
FL
FL
FL
FL
FL
FL
FL
GA
HI
IA
IA
IA
ID
21.1
52 . 1
M
58.5
59.5
50.5
M
M
53 . 5
62 .7
M
M
M
53 .O
M
60.8
57 . 4
M
M
M
M
M
37 .5
39 . 4
M
M
M
M
M
41.3
M
M
M
M
M
M
48 .O
M
M
72.2
66 .6
M
M
66 .8
M
51.4
76.5
M
37 . 8
M
39 .8
13.0
45.2
M
52 .9
52 . 5
52 .O
M
M
45 .8
58. 1
M
M
M
15 . 8
M
56 . 7
52 .0
M
M
M
M
M
3 1 .O
32 .6
M
M
M
M
M
28 . 2
M
M
M
M
M
M
37 .4
M
M
68 . 3
61.5
M
M
61.6
M
43.5
73.7
M
25 .O
M
32. 1
11.8
44 . 2
M
51.2
51.2
5O.9
M
M
45 . 3
56.7
M
M
M
45. 1
M
55 . 2
50.9
M
M
M
M
M
28 .6
29 .9
M
M
M
M
M
24 .8
M
M
M
M
M
M
35.6
M
M
67.2
60.3
M
M
60.4
M
42 .4
72 . 3
M
19.4
M
29. O
3O-Year Average
Temperature (F)
Urban Area
State
Nov
Dec
Jan
Chicago
Davenport-Rock Is 1 and-Mol 1 ne
Peor la
Spr 1 ngf leld
Gary-Hammond-East Chicago
Ind 1 anapol 1 s
Terre Haute
Lawrence
Topeka
Wlchl ta
Evansv 1 1 1 e
Hunt 1 n ton-Ashl and
Lex Ing ton-Fa ye tte
Lou 1 sv 1 1 1 e
Owensboro
Baton Rouge
New Orleans
Bos ton
Lowe 1 1
Pi ttsf ield
Spr 1 ngf ield-Ch Icopee-Hol yoke
Worcester
Bal t tmore
Lew 1 s ton- Auburn
Detrol t
Grand Rapids
Muskegon-Norton Shores-Muske
Saglnaw
Du 1 u th-Super lor
M 1 nneapol 1 s-St . Paul
Roches ter
St. Cloud
Kansas City
Spr i ngf leld
St. Louis
Jackson
Bill 1ngs
Grea t Falls
L Incol n
Omaha
Sioux C1 ty
Ashevt 1 le
Charlotte-Gastonla
Greensboro-Wlnston-Salem-HIg
Ral e 1gh-0urham
Mariches ter
Nashua
Al 1 entown-Bethlehem-Easton
At lant 1c C1 ty
Jersey City
Long Branch-Asbury Park.
IL
IL
IL
IL
IN
IN
IN
KC
KC
KC
KY
KY
KY
KY
KY
LA
LA
MA
MA
MA
MA
MA
MD
ME
MI
MI
MI
MI
MN
MN
MN
MN
MO
MO
MO
MS
MT
MT
NB
NB
NB
NC
NC
NC
NC
NH
NH
NJ
NJ
NJ
NJ
4O.4
M
39.9
4 t .9
M
41.7
M
M
42.9
44.8
M
M
44 .6
45. 0
M
M
60. 1
45.2
M
M
M
M
46. 1
M
41.1
38.7
M
M
28.4
32.4
M
M
43.6
M
45. O
55.3
M
M
M
40. 0
M
46.3
M
M
50.0
M
M
M
46. O
M
M
28.5
M
28. 0
3O.5
M
30.9
M
M
31.8
34.5
M
M
35.5
35.6
M
M
54.8
33. O
M
M
M
M
35.3
M
29.6
27 .4
M
M
14 .4
18.6
M
M
32.3
M
34.6
48.9
M
M
M
28. O
M
38 .7
M
M
41.2
M
M
M
35. 1
M
M
24.3
M
23.8
26.7
M
27.9
M
M
28. 0
31 .3
M
M
32.9
33. 3
M
M
52.9
29.2
M
M
M
M
33.4
M
25.5
23.2
M
M
8.5
12.2
M
M
27.8
M
31.3
47. 1
M
M
M
22.6
M
37.9
M
M
40.5
M
M
M
32.7
M
M
-------
Table 2 (Continued)
Monthly Winter Temperature
3O-Year Average
Temperature (F)
30-Year Average
Temperature (F)
Urban Area
New Brunswick-Perth Amboy-Sa
Newark
Pa terson-C 1 1 F ton-Passa i c
Phi 1 adelphf a
Trenton
Vineland-Mi 1 Ivl le-Bridgeton
W 1 1 m 1 rig ton
A 1 buquerque
Las Cruces
Las Vegas
Reno
Albany-Schenectady-Troy
B i nghamton
Buf f a 1 o
E ) m i ra
Nassau-Suf f o 1 k
New York
Poughkeeps i e
Roches t er
Syracuse
U t i ca-Rome
Akron
Canton
C 1 nc 1 nnat 1
C 1 eve 1 and
Col umbus
Dayton
Mam 1 1 ton-M i dd 1 e town
Spr i ngf i e 1 d
S teubcriv i 1 1 e-We i r ton
To 1 "do
Youngs town -Warren
Ok 1 ahoma C 1 ty
Tu 1 sa
Eugene-Spr 1 ngf 1 e 1 d
Por t 1 and
Sa 1 em
Erie
Harr i sburg
Johnstown
Lancaster
Northeast Pennsylvania
P 1 t tsburg
Read i ng
York
Providence-Warwick-Pawtucket
Charleston-North Charleston
Columb ia
Greenv 1 1 1 e-Spar tanburg
Cha t tanooga
Johnson C 1 ty-K i ngspor t-Br 1 s t
State
NJ
NJ
NJ
NJ
NJ
NJ
NJ
NM
NM
NV
NV
NY
NY
NY
NY
NY
NY
NY
NY
NY
NY
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
OK
OK
OR
OR
OR
PA
PA
PA
PA
PA
PA
PA
PA
RI
SC
SC
SC
TN
TN
Nov
M
46.2
M
M
M
M
M
44 . 5
M
53 . 3
M
39 .6
M
39.8
M
M
47 . 4
M
40. 5
41.0
M
M
M
4-1 . 6
41.6
41.7-
M
M
M
M
39 .6
M
49. 2
49 . 4
M
45.3
M
M
43.8
M
M
M
44.1
M
M
43.3
56.3
M
M
M
M
Dec
M
34 .5
M
M
M
M
M
36.2
M
45.2
M
25 .9
M
27.9
M
M
35.5
M
20.3
28. 1
M
M
M
34 . 4
3O.3
30.7
M
M
M
M
28 .O
M
4O.O
39.8
M
4O. 7
M
M
32 .6
M
M
M
33. 3
M
M
31.5
49.3
M
M
M
M
Jan
M
31.4
M
M
M
M
M
35.2
M
44 .2
M
21.5
M
23.7
M
M
32 . 2
M
24.0
23.6
M
M
M
32 . 1
26.9
28.4
M
M
M
M
24 .8
M
36.8
36.6
M
38 . 1
M
M
3O. 1
M
M
M
30.6
M
M
28.4
48.6
M
M
M
M
Urban Area
Knoxv (lie
Memphi s
Nashvl 1 le-Davldson
Aus t in
Beaumont-Port Arthur-Orange
Corpus Chr tst 1
Dal las-Fort Worth
El Paso
Gal veston-Texas City
Houston
Odessa
San Antonio
Provo-Orem
Salt Lake Clty-Ogden
Newport News-Hampton
Norfolk-Virginia Beach-Ports
R 1 chmond
Roanoke
Seattle-Everett
Spokane
Tacoma
Yak Ima
Appleton-Oshkosh
Green Bay
damesv 1 1 1 e-Beo 1 t
Mad1 son
M 1 1 waukee
Rac 1 ne
Char 1 eston
Whee 1 Ing
State
TN
TN
TN
TX
TX
TX
TX
TX
TX
TX
TX
TX
UT
UT
VA
VA
VA
VA
WA
WA
WA
WA
WI
WI
WI
WI
WI
WI
WV
wv
Nov
49.2
50.9
48.4
M
M
M
55.8
51 .6
M
61.1
M
59.7
M
39. 1
M
51 .6
49. 0
46.7
44 .6
35.5
M
M
M
M
M
34.7
36.5
M
45.4
M
Dec
41.5
42.7
4O.4
M
M
M
47.9
44 .4
M
54 .6
M
53.2
M
30.3
M
42.3
39. 0
37 .4
40.5
29.0
M
M
M
M
M
21.9
24.2
M
36.2
M
dan
4O.6
40.5
38 .3
M
M
M
44 .8
43.6
M
52. 1
M
50.7
M
28.0
M
40.5
37.5
36.4
38.2
25.4
M
M
M
M
M
16.8
19.4
M
34.5
M
------- |