United States
Environmental Protection
Agency
Environmental Research
Laboratory
Corvallis OR 97333
Research and Development
EPA-600/S3-84-096 Dec. 1984
&EPA Project Summary
Meteorological Factors
Responsible for High CO
Levels in Alaskan Cities
Sue Ann Bowling
High latitude communities frequently
have severe air pollution problems. The
usual cause is the release of moderate
amounts of pollutants into the
atmosphere with extremely poor
dispersion which, in turn, is a direct
result of the high latitude radiation
balance. Winter is characterized by
short days and low solar elevation. At
locations north of 60°N, midwinter day
light may vary from 0 to just under 6
hours, and at noon the sun, if it rises at
all, is lower in the sky than it would be
45 minutes after sunrise in Los Angeles.
The result is a ground-based nighttime
inversion which continues through
peak traffic hours (throughout the day
in some places), coupled with a
complete lack of photochemical
reactions. Downtown mixing heights as
low as 10 m, combined with speeds less
than .5 m sec'1, have been measured in
Fairbanks. If development in high
latitudes is to proceed rationally, these
meteorological conditions must be
understood, and models developed
which take account of them.
This Project Summary was developed
by EPA's Environmental Research Lab-
oratory, Corvallis, OR, to announce key
findings of the research project that is
fully documented in a separate report of
the same title (see Project Report order-
ing information at back).
Introduction
The region north of 60° latitude in the
western hemisphere is not heavily
developed, but severe air pollution prob-
lems exist even m relatively small
settlements. Fairbanks. Alaska (64°50'N)
is the best studied example of high-
latitude air pollution, but Anchorage,
(61°10'N) also has a severe carbon
monoxide problem. As development
increases in northern regions, it is
important that the meteorological condi-
tions leading to these high pollution
levels be understood. In particular, the
physical and chemical consequences of
the high-latitude regime of solar radiation
must be recognized.
High-Latitude Meteorology
The 24-Hour Night Regime
The most critical factors behind the
poor winter dispersion conditions found
in many high-latitude cities are the very
low values of incoming solar radiation
and the tendency for cities to be in
locations sheltered from winds. With this
radiation regime, nocturnal inversion
conditions can occur 24 hours a day. If
warm clouds are present, or if winds are
high enough to force turbulent mixing,
inversion strengths may be low and
normal lapse rates may even occur.
However, surface inversions are present
in over 80% of all soundings taken in
Fairbanks during December and January.
Nominal solar times on these soundings
are 2 am and 2 pm; actual release times
may be as much as an hour earlier.
The effect of nocturnal inversions on
pollutant levels has been studied at lower
latitudes, where they also result in
significant pollution episodes even
though they occur at times when
emissions are relatively low. At latitudes
poleward of 61 °N, for at least a few days
each year, the potential for polar-night
inversions is continuous. The long
periods of possible inversion
development produce lapse rates which
-------
may be continuously inverted to as high
as 2 km. At Fairbanks, near-ground
inversion strengths away from town
commonly exceed 10°C/100 m when
winds are light and skies are clear, and at
times exceed 30°C/100 m in the lowest
30 m of the atmosphere, and 200°C/100
m in the first two meters. Complex
stepped temperature structures are
common. Furthermore, these intense
ground inversions continue through the
hours of maximum CO emissions. Our
studies show that Anchorage inversions
(measured near the shore of Cook Inlet)
may also persist throughout the day in the
latter half of December, with inversion
strengths of as much as 10°C/100 m.
These inversions are even steeper farther
inland. The result is that even relatively
small, non-industrial settlements such as
Fairbanks and Anchorage are subject to
high CO levels.
The Short Daily
Warming Reg/me
As the noon solar elevation exceeds
31/2°, some warming begins to occur at
midday on clear days, and as the season
progresses, dispersion conditions near
noon improve steadily. Based on the
observed situation at Anchorage, a 7°
noon solar elevation angle could result in
a 5°C rise in surface temperature at noon
on a clear, calm day. This, however, does
not guarantee that a ground inversion
fails to persist through the day. As an
example, Figure 1 shows a series of 1 am
and 1 pm temperature soundings for
Anchorage starting about a week befor
the winter solstice. The two number
after each date are the 1-hour mean C(
level in ppm and the temperature ii
degrees Celsius at the most polluted c
the Anchorage monitoring sites, Benso:
and Spenard. Note that the Benson am
Spenard site temperature is generall
lower than the airport temperature, ii
spite of heat island effects, so the inver
sion persistence is probably more pro
nounced inland.
By January, the 100-meter inversion a
the Anchorage airport (which is ver
close to the coastline near the tip of th<
peninsula) usually, though not always
vanishes or becomes very weak at 1 pn
compared with its 1 am value, which car
still approach 10°C/100 m. There is
Dec 14, 1 am, 7, -5.O Dec 15, 1 pm, 7, -2.8
Dec 14. 1 pm, 7, -3.3 Dec 16, 1 am. 2. -5.0
Dec 15, 1 am. 2,-3.3 Dec 16. 1 pm, 8. -2.8
500
o
-Q
400
300
200
;oo
0
-4
-2
T, °C
Dec 17, 1 am, 1. -3.3 —
Dec 17. 1 pm. 8. -3.3 —
Dec 18. 1 am, 5. -7.2 •••
1 1 1 — JT —
J
••''"' / 1
/ j
" '" '" s" /
/'•' •'/ / /
/.•'/' / x /
8 -6 -4 -2 C
T, °C
— Dec 18,
- Dec 19.
... Dec 19,
\i
^ I
VI
' J
' *s
'/
-
2 4
1 pm, 9, -7.2
1 am. 3. -7.8
1 pm. 7. -7.2
I
— Dec 20, / am. 3, -7.2
•— Dec 20, 1pm, 13, -6.7
— Dec 21, 1 am. 3,-11.7
500
400 •
Dec 21. 1 pm. 11. -8.9
Dec 23, 1 am, 5, -12.2 Dec 24. 1 pm, 12, -12.8
• ~ Dec 22, 1 am. 5, -12.2 Dec 23, 7 pm, 12. -14.4 Dec 25. 1 am. 7, -12.8
• ••• Dec 22. 1 pm, 10. -12.2 Dec 24. 1 am, 3. -13.9 Dec 25. 1 pm. 5. -11.7
o
-Q
Figure 1. Anchorage airport soundings for 12 days around the winter solstice, showing persistence of the nocturnal inversion through the dayligh
hours. The two numbers following each date and time give the 1 -hour mean CO (ppm) and the temperature (°C) at Benson and Spenarc
(an Anchorage shopping area) at the time of the sounding. Note that there is little systematic difference in inversion strength between
1 am and 1 pm.
-------
however, good reason to believe that
inversions persist farther iriland. Daily
maximum ; temperatures at the CO
monitoring sites are often lower than
daily minimum temperatures at the
airport, and mean hourly CO levels for
December at urban monitoring sites in
Anchorage show no midday minimum
(Figure 2, Benson and Spenard). Rather,
the pattern consists of a sharp rise during
the commuter peak from 7-8 am, fairly
steady CO values through the day, and a
second sharp rise around 5 pm. A
residential site (Garden Site, Figure 2)
does show a drop in CO near noon, but
this may represent no more than the local
traffic pattern.
Wee* - - -Sat
Fri Sun
Garden Site
1979-1980 H
Benson &
Spenard
1978-1980
4 8 12 16 20 24
HourofDay
Figure 2. Variation of mean December CO
levels with time of day and day
of week at three Anchorage
locations. 7th and C is an area of
downtown office buildings not
too far from the coast, Benson
and Spenard is on a heavily
travelled artery in a shopping
district, and Garden Site is in a
church parking lot in a quiet
residential area.
By February, when the noon solar
elevation angle at Anchorage is about
15°, a distinct and well-developed
midday minimum in CO levels is present
and the improved mixing is beginning to
overlap the evening traffic peak (Figure
3), delaying the corresponding CO peak.
An intermediate stage, which has been
observed in February in Fairbanks,
involves relatively good dispersion near
noon, with the nighttime inversion
remaining strong through the morning
commuter peak and becoming reestab-
lished before the evening rush hours.
This situation has led to alert levels of CO
(15 ppm or more) in Fairbanks.
Urban Modification of
Nocturnal Inversions
Substantial heat islands are known to
be associated with high latitude cities of
Week
Fri
- - Sat
Sun
7th &C
1977-1981
0 4 8 12 16 20 24
Hour of Day
Figure 3. Variation of mean February CO
levels with time of day and day
of week at three Anchorage
locations. 7th and C is an area of
downtown office buildings not
too far from the coast, Benson
and Spenard is on a heavily
travelled artery in a shopping
district, and Garden Site is in a
church parking lot in a quiet
residential area.
moderate size. These heat islands are the
surface expression of weakening of the
nocturnal inversion by anthropogenic
heating. Tethered balloon measurements
of the temperatures in the lowest 100 m
of the atmosphere just north of Fairbanks
and within 200 m of the CO monitoring
site were carried out in December 1981.
Hourly CO values during three pairs of
ascents made under polluted conditions
are shown in Figure 4; the data obtained
during the ascents are shown in Figures
5-7.
Our preferred interpretation of this
information is that vertical mixing was
probably complete through the
isothermal layer (30 m on December 15,
10 m on December 22 and 6 m on
December 23). Although the highest CO
level was not measured during the time of
the shallowest isothermal layer, the most
rapid increase in CO was. Wind minima
were observed above the isothermal
layers, another indication that these
represent mixing layers. Relatively warm
city temperatures above the isothermal
layer to the heights of 30-40 meters are
ascribed to partial mixing due to updrafts
Background
O
o
Dec. 15
Background
I ,City
I
O'
O
10
5
n
_l
1 1 1 1 1 1
_T
Dec. 22
i i
I
s
20 r-
15
10
Background
Dec. 23
10
12
14
Hours
16
18
Figure 4. CO levels around the tethered
balloon ascents. Width of arrows
shows the time over which the
ascents took place.
-------
generated along warm building walls.
(The majority of buildings in Fairbanks do
not exceed 10 to 15 meters in height, and
the tallest is only 35 m.) Temperature
differences between urban and rural
sites above 40 m can be readily explained
by gravity waves of some 15 m amplitude,
which are known to be common in the
Fairbanks area.
Neither CO levels nor rural inversion
strength during the ascents represent
worst-case conditions. Previous
measurements have shown background
lapse rates as great as 30°C/100 m for
the lowest -30 m, and 3 recent years of
Fairbanks data (1979-80, 1980-81 and
1982-83) give an average of 4 days a year
with more than 15 ppm CO on an 8-hour
average. (The ascent days had maximum
8-hour averages of 12-13 ppm.) A mixing
height of 10 m and a windspeed of 0.5 m
sec^1 should therefore be considered
generous estimates for worst-case
modeling of pollutants from surface
sources in Fairbanks. Repeated visual
observation of well-layered smoke below
street lights 12 m high provides
independent evidence for a mixing height
120
110
100 -
90
80
70
60
50
40
30
20
10
December 15, 1981
Surface Winds < .6 M s,~
210° to 240° E of N
A City
• Rural
A
A
-20
-15
-10
Temperature °C
Figure 5. Comparison of rural and downtown lapse rates at Fairbanks, Alaska, during an
episode of high CO on December 15, 1981. "Rural" sounding was taken at a wildlife
refuge just north of Fairbanks; "city" sounding was taken in a parking lot within 200
meters of the CO monitor.
of no more than 10 meters. Until betti
measurements become available, thes
values should be considered to app
generally to continental sites north i
60°N when low windspeeds prevail. Us
of models unable to handle such lo
mixing heights and windspeeds canni
be expected to give meaningful results i
high latitudes.
Modification of Nocturnal
Inversions by Topography
Local factors, such as the degree c
shelter from regional winds, can signifi
cantly affect both the strength and per
sistence of surface inversions, li
Fairbanks, several episodes of CO level
above 15 ppm have been associated will
winds adequate for good dispersion soutl
of town (including the airport) while stag
nation conditions prevailed in the cif
itself. In Anchorage, the weathe
observations are taken at the airpori
which is located on a point jutting out int<
Cook Inlet, with substantial oceanii
exposure. Rawinsonde data from this sit<
do not represent the near-surface ai
structure over the city. As an example, w«
used a tethered balloon to measure i
surface inversion strength of 10°C/10C
m at the Bus Barn, 10 km inland from the
airport, while the airport had norma
lapse rates (Fig. 8). Temperature
distributions across Anchorage suppon
the idea that this is a normal situation; the
CO monitoring sites (which should be
within the Anchorage heat island) can be
as much as 10°C colder than the airport
Therefore, it seems clear that careful
observation of local topography is
essential before assuming that a nearby
weather station can be used to give valid
estimates of winds or vertical stabilityat a
site even a very short distance away. I
Capping Inversions
Anchorage can develop strong
inversions overlying superadiabatic
layers from 50 m to more than 200 m in
depth when cold continental air crosses
Cook Inlet before arriving at the airport.
However, CO levels during these
episodes are high only when there is
evidence that ground inversions are
present at the CO monitoring sites or if
traffic concentrations are extraordinary
(e.g., 2 pm Christmas Eve, 1982, Fig. 1).
If temperatures at the CO monitoring
sites are not depressed by several
degrees relative to those at the airport,
CO levels remain at or below normal
when the airport lapse rate is normal for
4
-------
the first 50 m. This is an additional
argument that actual mixing layers
during high CO episodes in Anchorage,
as in Fairbanks, are very shallow.
Although capping inversions do not
presently lead to high pollutant levels in
Anchorage, their presence is a warning
against excessive release of pollutants
from tall stacks, either in Anchorage or on
the other side of Cook Inlet.
Winds: Variability in Time
and Space
In many high latitude regions, wind
speeds are normally high enough to
prevent development of steep surface
inversions, although the weaker overall
inversions may still persist. Such areas
include many island and exposed coastal
areas as well as upland areas further
inland. These sites, however, are subject
to the obvious problems of drifting snow
and high wind chill. Furthermore, river
travel was common when many
settlements were founded so that a large
fraction of the population is concentrated
in sheltered valley bottoms, such as in
Fairbanks and Whitehorse. Even
relatively exposed sites, such as
Anchorage, are subject to occasional
periods when wind speeds are low
enough to permit the formation of ground
inversions.
The few measurements available in
Anchorage show the existence of
persistent shears. Figure 9 documents a
period of three days with light winds
(mostly 2-3 m sec"1 or less) and a
substantial horizontal shear; the three
sites are in a triangle less than 10 km on a
side. Previous studies have documented
similar shears in Fairbanks
Synoptic Situations
The clear skies and low wind speeds
conducive to the formation of nocturnal
inversions are commonly seen under
anticyclonic situations. In Fairbanks,
anticyclones do in fact appear to
dominate periods of poor air quality.
Coastal cities such as Anchorage and
Juneau, however, are rarely under the
influence of anticyclones in winter, and
high CO episodes at Anchorage are
normally associated with easterly geo-
strophic winds. The city is located just
ENE of the abrupt mountain front of the
Chugach Range, and relatively calm,
clear conditions leading to nocturnal
inversions appear to be due to the
shielding effect of the mountains. Eight-
hour CO levels exceeding 9 ppm also
have been observed when the core of a
dissipating low-pressure system was
located directly over Anchorage. In
these cases, very light and variable winds
were associated with thin, high or broken
cloud cover.
Air Pollution Forecasting
Evaluation of the forecasting of CO
levels by the Fairbanks North Star
Borough indicated that alert situations
(1 5 ppm or more for an 8-hour average)
were being persistently underforecast.
Part of the problem was traced to poor
communications between the NOAA
Weather Service and the Borough
forecasters. As a result of our study, the
Borough now provides current CO levels
to the Weather Service as one input for
the dispersion forecasts that are then
provided to the Borough by the Weather
Service
Our study also identified several
meteorological situations likely to be
associated with very high CO levels. Two
have already been mentioned: winds
forecast (correctly) for the airport which
did not extend to the more sheltered
downtown area, and coincidence of the
evening rush hour with rapid reestab-
0)
QJ
5
120
110
100
90
80
70
60
50
40
30
20
10
0
1
December22, 1981
Surface Winds < .5 M s,~
65° to 100° £ of N
* City
• Rural
•
•
.«•'
J&&IN:?
-30
-i^A.
-25
-20
-15
Temperature °C
Figure 6. Comparison of rural and downtown lapse rates at Fairbanks, Alaska, during an
episode of high CO on December 22, 1981. "Rural" sounding was taken at a wildlife
refuge just north of Fairbanks; "city" sounding was taken in a parking lot within 200
meters of the CO monitor.
-------
lishment of a ground inversion after
sunset in February. Additional conditions
responsible for CO levels over 15 ppm
were episodes of warm air advection and
calm periods with a thin or high cloud
cover.
Conclusions
Winter air pollution episodes at high
altitudes are due to the persistence of
intense nocturnal inversions through the
hours of maximum release of pollutants
near the surface. Coldstart CO emissions
and high energy demands during cold,
dark winters contribute to the problem,
and any solution will depend on
controlling these emissions. However,
the fundamental problem remains that
the high latitude winter atmosphere is so
stable that only minute quantities of
pollutants can be dispersed.
Relocation of cities to windier sites
would reduce air pollution but cause
severe problems with snow drifting (a
major difficulty now on the windswept
North Slope of Alaska and parts of
northern Canada) and wind chill.
Furthermore, even generally windy sites
have calm days. Anchorage has problems
with wind storms as well as air pollution
(not simultaneously!) in winter. However,
any industry with air pollution potential
should only be considered for upland
sites with high frequencies of winds over
6 msec"1, and then be approved only after
tracer studies.
120
110
100
90
SO
70
2
a eo
SO
40
30
20
10
December 23. 1981
Surface Winds < .3 M s,~1
155° to 200° E of N
30 M Winds up to 1.7 Ms,''
310° to 330° E of N
A City
• Rural
*;'
^;
;.l'j*
'£
-25
-20
-15
-10
Temperature °C
Figure 7. Comparison of rural and downtown lapse rates at Fairbanks, Alaska, during an
episode of high CO on December 23, 1981. "Rural" sounding was taken at a wildlife
refuge just north of Fairbanks;" city" sounding was taken in a parking lot within 200
meters of the CO monitor.
-------
200
100
-17
Jan. 142:00 a.m.
Airport
Jan. 14 2:00 p.m.
Airport *
-76 -75 -14 -13 -12 -11 -10
Figure 8. Comparison of lapse rates at the A nchorage airport and the Bus Barn, 10 km inland,
January 14, 1983.
360
300
| e 200
|»
700
.\
Day, Dec. 77
Figure 9. Hourly wind directions from three sites in Anchorage over a 3-day period with low
wind speeds. Solid line airport (tip of peninsula), dashed line 7th and C (northeast
of airport), dotted line Tudor and Lake Otisfeast of airport and southeast of 7th and
C). The airport anemometer had a higher starting speed than the other two, hence
more missing data.
US GOVERNMENT PRINTING OFFICE 1985- 559-111/10749
-------
S. A. Bowling is with the University of Alaska, Fairbanks, AK 99701.
James C. McCarty is the EPA Project Officer (see below).
The complete report, entitled "Meteorological Factors Responsible for High CO
Levels in Alaskan Cities,"(Order No. PB 85-115 137; Cost: $11.50, subject to
change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Research Laboratory
U.S. Environmental Protection Agency
Corva/lis, OR 97333
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES F
EPA
PERMIT No. G-3!
Official Business
Penalty for Private Use $300
------- |