THE DEVELOPMENT AND MOVEMENT OF HAZE OVER THE CENTRAL AND
EASTERN UNITED STATES
NOAA/NWS Storm Prediction Center
ABSTRACT
The origin and nature of warm season haze over the central and eastern United
States is discussed. Summarizing the results of various studies dating back to
the mid 1970s, the microphysical basis of haze is described, with emphasis on
the fact that haze appears to be largely anthropogenic (man-made). The synoptic
meteorology of haze in the United States is then presented. It is shown that
haze forms primarily in modified polar air, not tropical air, as is
commonly believed, and that its movement may be estimated by trajectories at the
850 mb level. A brief discussion of haze-free "tropical surges" along the East
Coast follows, using synoptic charts to illustrate the evolution of typical
surge events.
1. INTRODUCTION
During summer, the central and eastern United States is often blanketed by a
veil of murky haze that can last for days. The haze shrouds the sky and dims the
sun --- sometimes making it disappear altogether before the end of the day
(Figure 1). Haze is so common in some areas that it is assumed to be natural and
is accepted as a fact of life.
But most haze is, in fact, not natural, and its presence warrants our
attention for many reasons. For example, it has been shown that haze directly
affects human health. Researchers at the New York University Medical Center
recently concluded that the acid droplets found in haze pose a serious hazard to
exposed tissues of the lungs and breathing passages (Pendick 1993). The same
study also concluded that when haze occurs in association with smog --- the
brownish, photochemically-enhanced form of air pollution derived from automobile
exhausts --- smog's destructive nature is significantly increased.
Haze also affects aviation. While not posing a serious threat to jet
aircraft, haze does shroud visual cues important to small plane pilots and can
significantly reduce visual range over thousands of square miles. The phenomenon
was, in fact, likely a factor in the fatal crash which took the life of John F.
Kennedy, Jr., his wife and his sister-in-law near Martha's Vineyard in July
1999.
Figure 1. The setting sun dimmed by dense haze over State College, Pennsylvania on 16 September 1992
Haze, however, does more than limit visibility and impair health. Because the same airborne compounds resposible for haze also produce acid rain (Urone and Schroeder 1978, Husar 1990), haze is necessarily linked to the destruction of vast stands of Douglas fir by acid precipitation over the East in recent years (Mitchell 1994), as well as to the accelerated crumbling of concrete, mortar and statuary in many cities.
Numerous studies have shown that clouds which form in hazy air tend to produce less rainfall than those which form in less contaminated environments (Monastersky 1992). In addition, when oriented in patches or bands, haze can alter patterns of differential heating. This, in turn, can affect convective development, the location of thunderstorms and regional precipitation (Lyons 1980).
Evidence is also mounting that haze plays a significant role in determining the global energy balance, since it not only diminishes the amount of solar energy reaching the surface but also reduces the rate at which the atmosphere radiates longwave energy to space (Ball and Robinson 1982, Penner et al. 1994). One study even concluded that because of haze, some parts of the eastern United States experience an annual solar radiation deficit of about 7.5% (Ball and Robinson 1982). A more recent estimate by William Chameides of the Georgia Institute of Technology (Raloff 1999) suggests that the reduction of photosysnthesis due to haze in China could be robbing farmers there of more grain than the entire nation imports.
Another reason why haze is of significance is more aesthetic: Haze profoundly
affects the clarity with which we view the environment. Haze not only obscures
the daytime sky and clouds (Figure 2), but also the stars, Milky Way and other
features of the nightime heavens. Rainbows, haloes, and vivid sunset colors all
fade when haze shrouds the view. Haze also robs the landscape of contrast and
hue. A New England hillside alive with color on a bright fall afternoon will
appear much less vibrant on a day with haze. And, while it might be argued that
haze adds depth to distant vistas and mountain scenery, ordinary scattering by
air molecules and dust will yield the same result without the loss of coloration
and visual range associated with haze.
Figure 2. Appearance of supercell thunderstorm clouds seen at similar distances looking south southeast at about 6 p.m. local standard time on (a) a haze-free day (12 July 1986 near Sedalia, Missouri), and (b) on a day with moderate haze (7 June 1990 near Topeka, Kansas). Haze can hamper storm spotting: A distant wall cloud (mesocyclonic lowering) is visible at the lower left in (a), whereas cloud bases are shrouded by haze in (b). Tornadoes occurred shortly after both photos were made.
The reduction in visual quality caused by haze is, however, more than cosmetic. Campbell (1983), Evans and Cohen (1987) and Zeidner and Shechter (1988) all indicate that poor visual air quality causes heightened levels of anxiety, tension and depression in humans. In a similar study, Jones and Bogat (1978) linked increased interpersonal aggression and hostility to reductions in visual air quality.
In summary, the impact of haze, though sometimes subtle, is seen to be quite significant. Yet for many in the operational meteorological community, haze remains poorly understood despite its importance and frequent occurence. This paper addresses development and movement of haze over the central and eastern United States, with emphasis on its largely anthropogenic (man-made) origin. Much of the material has been presented elsewhere, but generally not from the perspective of an operational meteorologist.
2. WHAT HAZE IS
A. Haze as pollution
As was already noted, haze is often assumed to be natural. Evapotranspirative
products from trees are probably the most often-cited source. Others attribute
haze to the minute particles of salt which are released during the breaking of
ocean waves. Popular meteorological works (e.g., Ludlum 1991) which use the term
"haze" to refer to strictly natural obstructions to vision perpetuate this view.
Certainly natural forms of haze do exist. But, as the following pages show,
the type of haze commonly seen over the eastern half of the United States during
summer is not predominantly natural. It is in fact a vast blanket of man-made
pollution. More specifically, this haze is a form of wet air pollution --- a
veil of tiny droplets (aerosols) of condensed pollutants.1 The
adjective "wet" is used to distinguish eastern haze from the dry forms of haze
which consist of fine dust particles and are more commonly observed elsewhere in
the world (e.g., the Harmattan haze of west Africa, or the shamal wind haze of
the Persian Gulf).
Numerous studies since the mid 1970s (see, for example, Ferman et al. 1981,
Wolff et al. 1982, Malm 1992, Husar and Wilson 1993) have consistently shown
that warm-season haze over the eastern United States is composed primarily of
sulfates ---- i.e., of sulfate aerosols. To understand the formation of sulfate
aerosols, then, is to understand the formation of warm-season haze.
----------
----------
B. The chemistry of haze
(1) SO2 and sulfates
Sulfates are sulfur-based compounds which contain sulfate ions (ions of the
form SO42-). The sulfates which comprise haze aerosols are
derived mainly from the oxidation of sulfur dioxide (SO2) and may
exist in solid or liquid form. Airborne sulfates also originate from bursting
air bubbles in ocean waves and, over arid regions, from wind abrasion of
sulfate-bearing soils (Wallace and Hobbs 1977). But these two sources produce
comparatively large particles (over 2.0 um in diameter) which, like salt
particles from ocean waves, settle out rather rapidly and do not represent a
major source of haze aerosols.
Since haze-producing sulfates originate from SO2, a few words are
in order regarding the origin and nature of atmospheric SO2. Sulfur
dioxide is a colorless gas that has the odor of burning sulfur. It may be both
natural or man-made (anthropogenic) in origin. Most natural SO2 is
produced from the oxidation of hydrogen sulfide gas (H2S), a
by-product of vegetative decay. Volcanoes are another significant natural source
of SO2 (de Pena 1982).
Anthropogenic SO2 is produced in great quantity during the
large-scale combustion of sulfur-bearing fossil fuels in power plants, oil
refineries and steel mills. Lesser amounts are produced during the manufacture
of pulp and paper. In this country, SO2 emissions are concentrated
along the "industrial crescent" extending from the mid Mississippi Valley east
into the lower Great Lakes and Mid Atlantic states (Figure 3). A separate
SO2 source region is associated with oil production along the upper
Texas and Louisiana Gulf Coasts.
Figure 3. Distribution of SO2 emissions during the summer over the eastern United States in metric tons per day (After Brueske 1990).
Nearly 2.7 x 107 metric tons of SO2 are produced in the United States each year, about one fourth of the world's total output (Brueske 1990). Coal-burning power plants in the Ohio Valley alone yearly contribute 3 million tons (Malm 1992). Worldwide, man-made SO2 emissions increased nearly 30-fold between the mid 1800s and the mid 1900s as a result of industrial development (Heicklen 1976).
Even in highly polluted environments, the concentration of SO2 in
the atmosphere rarely exceeds 1 part per million; over rural areas average
concentrations are on the order of one part per billion. More than half,
however, of the SO2 over continental areas is thought to be
anthropogenic, compared to about one third for the atmosphere as a whole (de
Pena 1982). This is especially true over the northern hemisphere, where more
than 90% of the total man-made output is produced (Heicklen 1976). A more
precise estimate of the total anthropogenic component of SO2 over a
given area is difficult to ascertain because of lingering unknowns in the many
sulfur budget equations involved.
(2) Formation of sulfate aerosols
The oxidation of SO2 in the atmosphere can occur in a number of
different ways. Some of these paths include only gas-phase reactions, while
others involve exclusively liquid-phase reactions or some combination of the
two. Those paths which result in the formation of sulfate aerosols are known as
"gas-to-particle conversions," since they involve the formation of liquid or
solid aerosols from a gas.
For example, one gas-phase SO2 oxidation path involves oxidation
of the SO2 to form gaseous sulfuric acid
(H2SO4). The latter gas is then converted to a liquid
sulfate aerosol either by contact with existing aerosols or by serving as a
nuclei for the condensation of water vapor (Rogers and Yau 1989). In contrast,
liquid-phase oxidation of SO2 occurs inside cloud droplets after the
gas has become dissolved in them.
Because gas-to particle conversions typically require the presence of
secondary pollutants such as ozone (O3), conversion reactions are
usually enhanced by sunlight. Reaction rates also vary with relative humidity.
For example, the oxidation rate of SO2 to sulfate increases by a
factor of eight when the relative humidity increases from 70 to 80% (Wallace and
Hobbs 1977).
The particular type of sulfate which forms during the oxidation of
SO2 is dependent not only whether the reaction occurs in the gaseous
or liquid phase, but also on the degree of neutralization which occurs during
the oxidation process. Neutralization, in turn, is governed by the number and
type of cations (positive ions) present during the reaction. Ammonium, calcium,
sodium and magnesium are the most common neutralization ions. For example,
complete neutralization by ammonium produces ammonium sulfate
((NH4)2SO4), whereas partial neutralization results in the
formation of ammonium bisulfate ((NH4)HSO4). The
production of sulfuric acid (H2SO4), meanwhile, represents
a complete absence of neutralization.
Even with some neutralization, the sulfates found over the eastern United
States tend to be very acidic. These particles generally exist as aqueous nitric
and sulfuric acid, the result of gas-phase reactions during which gaseous acids
are converted to liquid forms by the addition of water vapor. It is the
downstream transport of these aerosols and their eventual elimination from the
atmosphere in the form of acid rain and snow that results in the low pH
(hydrogen ion concentration) characteristic of precipitation over the
Appalachians and much of the remainder of the East.2
----------
----------
(3) Growth of sulfate aerosols and their appearance as
haze
All sulfate aerosols, depending upon their degree of neutralization, are at
least somewhat hygroscopic. This means that they adsorb water molecules from the
environment, even when the relative humidity is below 100 percent. As a result,
sulfates are excellent condensation nuclei. The aerosols grow (deliquesce) as
water vapor condenses on them, and they continue to grow until they reach a size
at which they are in equilibrium with their environment.3
If the environment is supersaturated with respect to water, sulfate droplets
eventually reach such a size that they become readily visible as a cloud. If,
however, the humidity in their vicinity is somewhat lower, say around 85
percent, the droplets might grow to only about .5 um in diameter. This is about
one tenth the size of most cloud droplets but may be twice the diameter of the
aerosol before it was "wetted." Aerosols of this size are in the range (.1 um to
1 um) over which scattering efficiency per unit particle mass reaches a maximum
(National Research Council 1990); in fact, particles of this size scatter
visible sunlight a million times more effectively than air molecules. Figure 4
illustrates graphically the relationship between scattering by aqueous sulfate
aerosols and increasing relative humidity.
Figure 4. Effect of relative humidity on the scattering of visible light by liquid sulfate aerosols, as measured by extinction coefficient, bext. Extinction coefficient is a measure of the attenuation of a beam of light as a result of scattering (as in this case) or absorption in a medium. (After Ferman et al. 1981).
The scattering of sunlight by molecules in the atmosphere is a form of Rayleigh scattering. Rayleigh scattering is a general term for the scattering produced by particles that are very small compared to the wavelength of the incident light. The degree of scattering which occurs is inversely proportional to the fourth power of the wavelength involved. Thus, since blue light is of a shorter wavelength than red (.39 vs. .76 um), a sky dominated by Rayleigh scattering by air molecules appears blue.4
In contrast to the scattering of sunlight by air molecules, the scattering produced by hygroscopic sulfate aerosols is not wavelength-dependent. Because such particles are comparable in size to the wavelength of light, the scattering they produce is a complex function of particle size and wavelength. If the aerosols are all the same size, the sky might take on a faint red or blue hue depending on the size of the aerosols and whether the part of the sky being viewed is seen by scattered or transmitted light. But if a range of aerosol sizes exists, as is most often the case, no particular color is favored and the sky appears white. In addition, since the aerosols represent a net increase in the number of particles normally present in the atmopshere, the tendency for whitening is further increased. These extra particles also enhance the atmospheric absorption of sunlight -- and therefore reduce the amount of light that reaches the ground.
----------
----------
(4) The natural component
Because haze is such a familiar sight in the eastern U.S. and has no obvious
source, casual observers find it difficult to believe that it is not natural in
origin. Some will note that the Blue Ridge and Smoky Mountains were named long
before the widespread industrial use of fossil fuels. Indeed, a small percentage
of the particles found in haze are derived from volatile hydrocarbons such as
terpenes that are released by certain trees (Chang 1990). In addition,
atmospheric sulfates are not entirely anthropogenic.
But studies using data dating back to 1940 show an incontrovertible link
between industrial sulfur emissions and the degree of haziness over the eastern
U.S., as shown, for example, in Figure 5 (Husar 1990, Malm 1992).5
And, research conducted in the Blue Ridge mountains has shown that
naturally-occurring aerosols are responsible for no more than 20% of the total
light extinction due to haze (Ferman et al. 1981). The primary reason for this
is that most organic aerosols (such as those from trees) are relatively
hydrophobic and therefore do not grow to such a size that they become visible as
haze.
Figure 5. Comparison of long-term sulfur emissions and degree of summertime haziness (as measured by extinction coefficient) over the southeastern United States. (From Malm 1992).
The Ferman study estimated that if all natural sources of pollution in the East were taken together, surface visibility would still exceed 25 miles --- well above the 4 to 6 mile visual range characteristic of summer afternoons over the region today. If any haze were visible at all, it would be of the thin, "blue" variety first noted by the Cherokees in the early 1700s (Powell 1968) --- not the murky pall that frequently obscures the Blue Ridge mountains today. Accounting for the mean warm-season airflow overthe East in summer, the study concluded that "observed airflow patterns are consistent with previous findings that Midwestern [sulfate] source areas are a major cause of widespread summertime haze in the eastern U.S."
More recent studies have continued to support these ideas, attributing more than three quarters of summertime light extinction in the central Appalachians to sulfates (Malm 1992, Sisler et al. 1993). This is perhaps not too surprising considering Malm's observation that the average sulfate concentration over the eastern states is some 5 to 10 times greater than the estimated natural backround value of approximately .8 ug/m3 (Ferman 1981, Trijonis et al. 1991).
----------
3. THE CLIMATOLOGY OF HAZE
A. A summer phenomenon
Sulfate aerosols are both small and chemically stable. As a result, they
settle out only very slowly and can remain airborne for days. Most are
eventually washed from the atmosphere in the form of acid rain. Although sulfate
aerosols are formed wherever the precursor gas-to-particle conversion reactions
occur, it is synoptic- and regional-scale weather conditions that govern their
overall concentration and, thus, the occurence and persistence of haze.
Because large scale atmospheric transport reaches a minimum during summer and
boundary layer relative humidity tends to be high, sulfate haze is primarily a
warm season phenomenon. As Figure 6 shows, haze in the United States is most
common east of the Mississippi River. Note that haze is less frequently observed
over south Florida and northern New England ---areas farther removed from major
SO2 source regions and areas more likely to experience air of a less
contaminated origin. The data also show an overall trend for increased haziness
in recent years, especially over the Southeast. This reflects the increased
summertime use of bituminous coal for electrical purposes such as air
conditioning (Husar and Patterson 1980).
Figure 6. Long-term spatial distribution of haze (expressed as average visibility in km) over the United States. Top row: spring (April-June); bottom row, summer (July-September). (From Husar 1990).
B. Why the East is hazy
Proximity to the semi-permanent Bermuda anticyclone makes the eastern third
of the United States especially prone to prolonged episodes of synoptic-scale
sulfate haze in summer. Large scale transport over the eastern states reaches a
minimum during July and August. Moist air parcels originating over the lower
Mississippi Valley may take 3 or 4 days to reach the eastern Great Lakes or mid
Atlantic states. Vertical mixing is also often quite limited, especially east of
the Appalachians where subsidence is enhanced closer to the mean position of the
anticyclone. This increases the atmosphere's loading of both SO2 and
its sulfate oxidation products. In addition, since summer days are long and
large-scale cloudiness is at a minimum, photo-enhanced gas-to-particle
conversions can proceed at a maximum rate.
The source region of the air commonly present over the Midwest and East in
summer is another factor which fosters the development of haze. While the Gulf
of Mexico is typically blamed for the regions's oppressive summertime
combination of heat, haze and humidity ("The Three Hs"), careful examination of
surface trajectories reveals that in most cases such air does not originate
over the Gulf, but is rather continental polar in origin. Typically this air
has undergone strong modification in the lowest layers and is returning poleward
after having overspread the eastern states two or three days before.
Since these air masses have had a recent history of subsidence, an inversion
is typically present around 850 or 900 mb. Because of the inversion, moisture
added by evapotranspiration --- and haze --- remains trapped in the shallow
boundary layer near the ground, while cleaner, drier conditions persist aloft.
The low "lid" placed on top of the haze layer enhances its negative impact on
visibility and encourages additional aerosol growth. True Gulf (or tropical)
air, in contrast, typically lacks a low-level subsidence inversion and is
characterized by thermodynamic instability through a comparatively deep layer.
Any haze that may be present mixes through that deeper layer and is therefore
less dense.
The top of a haze layer in a modified polar air mass is often quite distinct,
and the resulting "haze horizon" is a familiar sight to airline passengers
flying over the eastern United States in summer. The "horizon," of course, marks
the top of the well-mixed boundary layer. Because relative humidity reaches a
local maximum at the top of the boundary layer, haze is usually densest at an
altitude just below that of the "horizon."6
----------
----------
C. Haze in the West
The western United States is largely free of sulfate haze since
SO2 emissions are comparatively low and the climate is dry. In
addition, since the air is often of recent oceanic origin, it tends to be
"cleaner" (e.g., contains fewer background particles) than that found over the
eastern United States. As Figure 6 shows, however, other forms of haze do occur
in parts of the West. In particular, ozone smog and nitrate-based haze plague
much of southern California in and around the Los Angeles basin. These forms of
pollution are derived primarily from motor vehicle emissions (Sisler 1993).
4. HAZE FORECASTING
A. Introduction
The National Weather Service and WX-TALK (WeatherTalk) discussion group
postings from the 1990s shown in Figure 7 capture some of the frustration and
lack of understanding posed by haze from a forecasting perspective. Accurate
forecasting of haze remains problematic to this day.
Although studies dating back to
the mid 1970s have documented the growth and movement of sulfate haze "blobs"
over the eastern United States (see, for example, Hall et al. 1973, Gillani and
Husar 1976, Lyons et al. 1978, Husar et al. 1981, Wolff et al. 1981, or Wolff et
al. 1982), little of this material was written for or has found its way into the
operational environment. The following sections offer a distillation of some of the
more meteorologically-relevant material presented in those papers. Information is
also provided on the how different synoptic patterns affect the development and
movement of regional haze.
Figure 7 . Selected state forecast discussions mentioning haze from (a) WSFO Birmingham, Alabama (0900 UTC 21 July 1991), and (b) WSFO Taunton (Boston), Massachusetts (1930 UTC 21 July 1994 and 0140 UTC 22 July 1994); (c) excerpt from "WXTALK" list server chat board (14 August 1994). The Boston discussions reflect forecast thinking prior to and just after the passage of a typical "tropical surge" (See Section 5). Parts of discussions most relevant to haze in boldface.
From what has already been presented, it is apparent that haze forms wherever the sources of atmospheric sulfates overwhelm the sinks. This is why haze is usually associated with regions of large-scale stagnation. With both vertical and horizontal mixing at a minimum, whatever aerosols may be present will accumulate at first locally and then regionally over those areas experiencing minimal air flow and high relative humidity.
Favorable conditions for haze formation are realized on a regular basis each
summer over the SO2 source regions of the central and eastern United
States, a day or so after a polar air mass has settled over the area and the
surface winds have become light. Under sunny skies, evapotranspiration can
result in substantial moistening of the boundary layer in the previously dry air
mass. Dewpoints, which may have dropped into the 40s soon after cold frontal
passage, may rise into the upper 50s or low 60s by the end of the second day
following frontal passage. Because conditions are favorable for rapid aerosol
growth, and because the aerosol are confined to a shallow boundary layer, haze
typically becomes noticeable about 48 hours after a polar air mass has stagnated
over the eastern United States.
B. A typical example Because eastern haze is generally confined to the lowest 7 or 8 thousand feet
above the surface, its movement over a period of days is generally
well-correlated with the streamlines of flow at 850 mb (Wolff et al. 1981).
Figure 8 depicts the progress of a typical summertime haze event over the
east-central United States, with the areas of haze plotted as shaded areas on
the 850 mb charts. In this case, a mass of hazy air which formed over the Ohio
Valley was first shunted south to the Gulf Coast states ahead of a weakening
cold front.
Once over the Gulf Coast region, the air mass stalled and picked up an
additional load of sulfates from the refineries of southeast Texas and
Louisiana. The hazy air then made a return visit to the Ohio Valley as it was
drawn northward ahead of a new cold front over the central Plains and upper
Mississippi Valley. In this way, especially dense areas of haze may form as
sulfate aerosols accumulate in a given mass of air that makes one or more return
visits to an SO2 source region.
As is apparent in Figure 8, haze tends to concentrate in bands parallel to
fronts. This might seem perplexing considering that the depth of the boundary
layer tends to be greatest in the vicinity of fronts. But because large scale
ascent and low-level convergence are maximized there as well, relative humidity
is also at a maximum near fronts. Thus, sulfate production is enhanced and those
aerosols which do form grow to a comparatively large size. Weak fronts are
sometimes detectable in visible data satellite imagery when their associated
haze bands appear as a discontinuities in the brightness field.
Figure 8. A haze "blob" which twice affects the lower Ohio Valley in August 1979. Haze areas (visual range less than 10 km) shaded, with 850 mb streamlines and surface isobars/fronts. Map times: 1200 UTC (After Wolff et al. 1981).
The case shown in Figure 8 also illustrates the importance of trajectory analysis to forecasting haze movement. Trajectory analysis is used to determine air parcel motions forward or backward in time from a specific point or region (Yarnal 1991). Since the formation of haze is not an instantaneous process, and since haze can be advected from one area to another, knowledge of an air mass' recent history is essential to making an accurate forecast of haze movement. For example, in this event, a forecaster in Peoria, knowing that haze was reported earlier in the week at Little Rock and Memphis, should not be surprised by the appearance of haze over northern Illinois on 18 August. Even though the 850 mb flow over Illinois on the 18th was rather strong and from the west --- conditions not normally associated with haze in that part of the country --- back trajectories reveal that the air mass entering the state was indeed the same one which had brought haze to Arkansas and Tennessee two days earlier.
In contrast to the mid Mississippi Valley, it is in fact not uncommon for brisk westerly winds to accompany haze over the eastern Great Lakes and New England. Indeed, some of the worst haze episodes in the Northeast occur in conjunction with mid-summer heatwaves and moderate to strong westerly flow on the northern fringe of an elongated Bermuda High. The enhanced westerly flow is frequently associated with the approach of a shortwave trough over the Saint Lawrence Valley. The winds, which are typically strongest in advance of an east-west cold front associated with the trough, tap dense haze areas that may have been developing under nearly perfect sulfate-forming conditions over the Ohio Valley for several days. When the "blobs" arrive in New York or New England, it is often just in time to produce a fall of acid rain as thunderstorms erupt along the front.
Once an area of haze has developed, it remains in existence until it is
either advected to a new location or the aerosols are washed out by rain. Dry
deposition of haze (see Footnote 1) is minimal. Recent evidence also suggests
that boundary layer sulfates may be "vented" through the anvils of mesoscale
convective systems (McNaughton et al. 1994). Qualitative evidence certainly
indicates that surface visibilities increase in the wake of most mesoscale
convective complexes and larger mesoscale convective systems. In contrast,
although isolated thunderstorms may rid the boundary layer of local buildups of
pollutants, visibility often is not improved because the increase in boundary
layer humidity in the vicinity of such storms enhances the growth of remaining
aerosols.
C. Early season haze
Observations show that surface dewpoints in the 60s are most often associated
with the onset of haze in the summer, since lower dewpoints at that time of year
are indicative of fresh outbreaks of polar air. But haze can occur with lower
dewpoints, especially during spring and fall. A classic example of "low dewpoint
haze" is illustrated in Figure 9. This event occurred in early April with
surface dewpoints in the 40s and temperatures in the 60s. The haze was clearly
not derived from natural sources such as trees since very little plant growth
was in progress at the time of the event due to unusually cold weather during
the preceding weeks.
Figure 9. Surface fronts and streamlines associated with an early season haze event (12-15 April 1992) over the north central states. Haze areas (as reported in surface airway observations) shaded.
The main haze "blob" formed in a zone of persistent easterly flow north of an
east/west stationary front, where vertical mixing was limited by presence of a
strong frontal inversion (Figure 10). The sulfates originated in the industrial
area surrounding the lower Great Lakes and were carried west along the southern
fringe of a large polar anticyclone which had invaded the area earlier in the
week. Pollution from Chicago, Cleveland, Detroit and Milwaukee contributed to
the haze which spread as far west as the Nebraska high plains.
Figure 10. SkewT-logP diagram for Topeka, Kansas 0000 UTC 14 April 1992. Temperature and dewpoint lines solid and dashed, respectively; wind speeds in knots. Sounding shows presence of pronounced frontal inversion near 875 mb.
As Figure 9 shows, the haze formed in modified polar air, not tropical air;
the haze-producing air mass never moved south of the Ohio River. It is also
worth noting that some of the haze which affected the north central states
during this event did, however, arrive from the south --- in the form of a
returning "blob" that had been pushed into Arkansas and Oklahoma by a cold
frontal passage earlier in the week.
D. Haze and tropical cyclones
Some of the worst haze episodes in this country are associated with the
blocking of polluted air masses over the Ohio Valley or Midwest when tropical
cyclones are present along the East Coast. Figure 11 shows how a mass of hazy
air which formed over the Ohio Valley was forced to remain over that region for
two additional days as Hurricane David (1979) moved up the Eastern Seaboard.
Once David reached New England, northwest winds in the wake of the storm carried
the heavy pollution burden southward behind a cold front into the
Carolinas, Georgia and the Gulf Coast states. The GOES 7 depiction of a more
recent hurricane-related haze event is shown in Figure 12. In this case (Emily
1993), northeast winds on the fringe of the tropical system drove a plume of
dense haze that had been resident over the mid Atlantic states southwest into
the Carolinas.
Figure 11. Surface and 850 mb evolution during a haze event affected by Hurricane "David," September 1979. Map times: 1200 UTC (After Wolff et al. 1982).
Figure 12. GOES 7 visible data view of the eastern United States at 1331 UTC 30 August 1993 showing Hurricane Emily over the western Atlantic and an extensive area of clear but hazy skies (light grey areas) covering much of the eastern United States. Over the central Appalachians and along the mid Atlantic coast, a dense band of haze is being driven southwest into the Carolinas ahead of a "back-door" cold front on the northwest fringe of Hurricane Emily.
Even in the absence of tropical activity it is not uncommon for haze to
follow the passage of weak cold fronts over the Gulf Coast region and the
Southeast. This is because the post-frontal air typically originates over the
Ohio Valley or mid Atlantic states. Embryonic sulfate aerosols born in these
SO2 source regions move south with the subsident post-frontal flow,
eventually growing into haze as they encounter increased humidity and sunshine
over the Gulf and south Atlantic coastal states.
5. TROPICAL SURGES
One of the more pleasant though uncommon characteristics of the normally hazy
summer months over the mid Atlantic States and the Northeast is the arrival of
pure subtropical air from the western Atlantic in the form of a "tropical
surge." These events occur when the main axis of the Bermuda High rotates
clockwise from its usual east/west orientation along 30 or 35 degrees north
latitude. This may happen in response to the development of a strong trough over
the Mississippi Valley, or in response to the development of a closed
circulation along the south Atlantic Coast. Both scenarios result in deep south
to south-southwesterly flow over the mid Atlantic States and the Northeast
(Figure 13).
Figure 13. Standard NCEP 500 mb analyses associated with tropical surge events over the mid Atlantic States. In (a), anomalous southerly flow along the eastern seaboard is related to an unusually deep trough over the Mississippi Valley (1200 UTC 4 August 1974). In (b), a closed circulation east of Georgia produces a south to southeast flow along the mid Atlantic Coast (1200 UTC 4 August 1988). Ridge axes depicted by serrated lines.
Having had a long over-water trajectory, air streams of this type are
noticeably clean and unpolluted despite their high moisture content --- in sharp
contrast to the murky air masses they replace. Since the maritime boundary layer
is usually buoyant, cumulus buildups dot the sky, and some grow into small
thunderstorms. An example of a typical tropical surge sky is given in Figure 14
and in Corfidi (1993). There is often a stronger-than-normal southerly component
to the wind at the surface.
Figure 14. A classic "tropical surge" sky seen looking east from near Danville, Virginia around 1100 LST 14 July 1990. The azure blue sky and brightly-lit cumulus buildups are characteristic of surge events. The boundary layer in this case is especially transparent (note that cloud details are apparent even at the horizon) even though surface dewpoints were in the 70s.
A series of surface observations depicting the midday arrival of a tropical
surge at Wilmington, Delaware on 20 July 1994 is shown in Figure 15. This
particular set of observations was chosen because it clearly illustrates the
increase in visiblity that occurs during a surge in the absence of a major
change in dewpoint. The sequence is also largely free of diurnal effects. In
this case, visibility increased from 5 miles in haze at noon local time (1550
UTC) to 12 miles at 6 pm (2150 UTC). The dewpoint, meanwhile, remained in the
mid 70s.
Figure 15. Sequence of surface airway observations depicting the arrival of a tropical surge at Wilmington, Delaware around 1900 UTC (1 PM LST) 20 July 1994.
The doubling of visual range is significant since typically little or no afternoon visibility change is observed over this region during the summer in the absence of a frontal passage. In a sense, of course, there was a frontal passage --- strongly modified continental polar air with a high sulfate content was replaced by a cleaner air mass of Carribean or west Atlantic origin.
Surge "fronts," unfortunately, are not easily diagnosed by even the most sophisticated of today's objective analysis techniques. For example, the arrival of a surge generally produces no detectable change in surface equivalent potential temperature (theta-e). Automation of the surface observation network has exacerbated the detection problem since minimal information is now recorded regarding visibility changes above 6 miles. While the advance of a tropical surge can sometimes be tracked by satellite7, clouds often interfere with the view. Until observations of "airborne particulate mass" are routinely included in surface airway reports, it is probably safe to say that surge detection will remain problematic for some time to come.
The 500 mb analysis sequence for the Wilmington surge event is shown in
Figure 16. In this case, the combination of a seasonably strong trough over the
upper Great Lakes with a weaker trough along the south Atlantic Coast together
produced anomalously deep southerly flow along the mid Atlantic coast. Backing
of the upper flow as far north as Pennsylvania and New York on 21 July allowed
the surge to sweep northeast into New England the following day (Figure 16d).
Arrival of the surge in this area was rather abrupt, and elicited the comments
from the Boston area National Weather Service office shown in Figure 7b.
 
a.
 
b.
 
c.   d.
Figure 16. Standard NCEP 500 mb analyses for 1200 UTC (a) 19
July, (b) 20 July, (c) 21 July and (d) 22 July 1994 showing
evolution of mid-level flow during the tropical surge depicted in Figure 12. The
surge reached Wilmington, Delaware on the 20th and reached southern New England the
following day.
On average, tropical surges occur about once each summer along the mid Atlantic Coast. They are, however, notably absent in some years (e.g. 1992 and 1993), while several might occur in others (e.g., 1994 and 2004). Their frequency necessarily decreases from south to north along the eastern seaboard, and their effects usually do not extend west of the central and northern Appalachians. Tropical surges are also observed in the deep south, when the large scale circulation allows Caribbean or western Atlantic air to overspread the Gulf Coast states (Note comment regarding east-to-west flow in the forecast discussion shown in Figure 5a).
Surge events usually come to an end with the approach of a shortwave trough
in the westerlies. When these systems are weak, their passage is marked by only
a modest veering of the mean tropospheric flow. This shunts the maritime air
mass east into the Atlantic, allowing hazy continental air from the lower Great
Lakes or Ohio Valley to return. If the shortwave is seasonably strong, however,
a more decided veering of the large scale flow will occur. This will usher in an
air mass of a more northerly origin, and the haze-free maritime flow will be
replaced by a relatively clean air mass from northern Canada.
----------
6. CONCLUDING REMARKS
The origin and nature of sulfate haze over the central and eastern United States have been discussed. Persistent misconceptions about haze have been examined, and information has been provided so that forecasters may better anticipate its formation and movement. Although the impact of haze as a meteorological phenomenon is not immediate like that of a tornado, flash flood or hurricane, its long-term affects are no less significant.
It is worth noting that there has been a slight decrease in the incidence of sulfate haze over some parts of the Midwest and the Northeast in recent years due to a reduction in the use of high-sulfur coal (Husar 1990). And, in the Southeast, SO2 emissions have decreased as a result of stricter air pollution standards imposed by the 1970 Clean Air Act and its subsequent amendments.
But continued population growth and the increased demand for electrical power suggest that large-scale haze will remain a problem for some time to come. In addition, regional haze which is not predominantly sulfate-based but is still traceable to anthropogenic sources has begun to affect visibility in other parts of the country, such as the scenic national parks of Arizona and Utah (National Research Council 1990, Sisler et al. 1993). Haze is also increasing in rapidly developing parts of the world such as China, India and southeast Asia.
Haze, therefore, will continue to present operational meteorologists with a
challenge --- a challenge to not only better understand it, but to also heighten
public awareness that most haze is not natural and that more should be done to
eliminate it.
7. ACKNOWLEDGEMENTS
The author would like to thank the three reviewers who substantially enhanced
the clarity and technical content of this paper. Thanks also to Steve Weiss of
the Storm Prediction Center for helpful comments, and to Roger Edwards, Ryan
Jewell and Sarah Taylor for HTML assistance.
8. REFERENCES
Ball, R. J., and G. D. Robinson, 1982: The origin of haze in the central United States and its effect on solar radiation. J. Appl. Meteor., 21, 171-188.
Brueske, S. L., 1990: Forecasting Atmospheric Particulate Sulfur Concentrations Using National Weather Service Synoptic Charts. Master's thesis, The Pennsylvania State University, 132 pp.
Campbell, F. W., 1983: Ambient Stressors. Environ. Behavior, 15, 355-380.
Chang, J. S., 1990: Acid Deposition: State of the Science and Technology. Report 4. National Acid Precipitation Assessment Program (especially pp 27-34).
Corfidi, S. F., 1993: Lost horizons. Weatherwise, 46, 12-17.
de Pena, R. 1982: Sulfur in the atmosphere and its role in acid rain. Earth and Mineral Sciences, 51, 62-66 (Available from College of Earth and Mineral Sciences, The Pennsylvania State University, University Park)
Evans, G. W., and S. V. Jacobs, 1981: Air pollution and human behavior. J. Social Issues, 37, 95-125.
Ferman, M. A., G. T. Wolff, and N. A. Kelly, 1981: The nature and sources of haze in the Shenandoah/Blue Ridge Mountains area. J. Air Pollut. Control Assoc., 31, 1074-1082.
Gillani, N. V., and R. B. Husar, 1976: Synoptic scale haziness over the eastern U.S. and its long range transport. Proceedings of the 4th National Conference on Fire and Forest Meteorology, St Louis, MO, 232-239.
Hall, F. P., C. E. Duchon, L. G. Lee, and R. R. Hagan, 1973: Long- range transport of air pollution: A case study, August 1970. Mon. Wea. Rev., 101, 404- 411.
Heicklen, J., 1976: Atmospheric Chemistry. Academic Press, 406 pp.
Husar, R. B., 1990: Historical visibility trends. Acid Deposition: State of Science and Technology. Report 24. National Acid Precipitation Assessment Program, 68-76.
Husar, R. B., and D. E. Patterson, 1980: Regional scale air pollution: Sources and effects. Ann. NY Acad. Sci., 338, 399- 417.
Husar, R. B., and W. E. Wilson, 1993: Haze and sulfur emission trends in the eastern United States. Environ. Sci. Tech., 27, 12-16.
Husar, R. B., J. M. Halloway, D. E. Patterson , and W. E. Wilson, 1981: Spatial and temporal pattern of eastern U.S. haziness: A summary. Atmos. Environ., 15, 1919-1928.
Ludlum, D. M., 1991: The Audubon Field Guide to the Atmosphere. Alfred A. Knopf. 656 pp.
Lyons, W. A., 1980: Evidence of transport of hazy air masses from satellite imagery. Ann. NY Acad. Sci., 338, 418-433.
Lyons, W. A., J. C. Dooley Jr., and K. T. Whitby, 1978: Satellite detection of long-range pollution transport and sulfate aerosol hazes. Atmos. Environ., 12, 621-631.
Malm, W. C., 1992: Characteristics and origins of haze in the continental United States. Earth-Sci. Rev., 33, 1-36.
McNaughton, D.J., N. E. Bowne, R. L. Dennis, R. R. Draxler, S. R. Hanna, T. Palma, S. L. Marsh, W. T. Pennell, R. L. Peterson, J. V. Ramsdell, S. T. Rao, and R. J. Yamartino, 1994: Summary of the 8th Joint Conference on Applications of Air Pollution Meteorology. Bull. Amer. Meteor. Soc., 75, pp 2303-2311.
Menzel, W. P., and J. F. W. Purdom, 1994: Introducing GOES-I: The first of a new generation of geostationary operational environmental satellites. Bull. Amer. Meteor. Soc., 75, pp 757-781.
Mitchell, J. G., 1994: Legacy at risk. National Geographic, 186, 20-55 (October).
Monastersky, R., 1992: Haze clouds the greenhouse. Sci. News, 141, 232-233.
National Research Council, 1990: Haze in the Grand Canyon: An Evaluation of the Winter Haze Intensive Tracer Experiment. National Academy Press, 97 pp.
Pendick, D., 1993: Hazy summer days boost respiratory ailments. Sci. News, 143, pg 52.
Penner, J.E., R. J. Charlson, J. M. Hales, N. S. Laulainen, R. Lefier, T. Novakov, J. Ogren, L. F. Radke, S. E. Schwartz, and L. Travis, 1994: Quantifying and minimizing uncertainty of climate forcing by anthropogenic aerosols. Bull. Amer. Meteor. Soc., 75, 375-400.
Powell, W., 1968: North Carolina Gazeteer. University of North Carolina Press, Chapel Hill, NC.
Raloff, J., 1999: Sooty air cuts China's crop yields. Sci. News, 156, pg 356.
Rogers, R. R., and M. K. Yau, 1989: A Short Course in Cloud Physics. Pergamon Press. 290 pp.
Sisler, J. F., D. Huffman, and D. A. Latimer, 1993: Spatial and Temporal Patterns and the Chemical Composition of the Haze in the United States: An Analysis of Data from the IMPROVE Network, 1988-91. (Available from CIRA-Cooperative Institute for Research in the Atmosphere, Colorado State University, Ft. Collins, CO 80523)
Trijonis, J., R. Charlson, R. Husar, W. C. Malm, M. Pitchford, W. White, 1991: Visibility: Existing and Historical Conditions - Causes and Effects. In: Acid Deposition : State of Science and Technology. Report 24. National Acid Precipitation Assessment Program.
Urone, P. and W. H. Schroeder, 1978: "Atmospheric chemistry of sulfur-containing pollutants," Chapter 6 in Sulfur in the Environment, Part 1, J. O. Nriagu, ed. New York: Wiley Interscience.
Wallace, J. M. and P. V. Hobbs, 1977: Atmospheric Science: An Introductory Survey. Academic Press, 467 pp.
Wolff, G. T., N. A. Kelly and M. A. Ferman, 1981: On the sources of summertime haze in the eastern United States. Science, 211, 703-705.
Wolff, G. T., N. A. Kelly and M. A. Ferman, 1982: Source regions of summertime ozone and haze episodes in the eastern United States. Water Air Soil Pollut., 18, 65-81.
Yarnal, B., 1991: The climatology of acid rain. In: Air Pollution: Environmental Issues and Health Effects. S. K. Majumdar, E. W. Miller and J. Cahir, Eds. The Pennsylvania Academy of Science.
Zeidner, M., and M. Shechter, 1988: Psychological responses to air pollution: Some personality and demographic correlates. J. Environ. Psych., 8, 191-208.
Autobiographical data
Stephen Corfidi received the B.S. and M.S. degrees in meteorology from The Pennsylvania State University. Prior to becoming a Lead Forecaster at the NWS/NCEP Storm Prediction Center, he held various positions within the NCEP Hydrometeorological Prediction Center, the Baltimore-Washington NWS Forecast Office and the NWS Meteorological Development Laboratory. He has published previously on weather analysis and forecasting, mesoscale convective systems and severe local storms. Steve's interest in haze dates to sky watching as a youth in Baltimore.