Potter T.D., Colman B.R. (co-chief editors). The handbook of weather, climate, and water: dynamics, climate physical meteorology, weather systems, and measurements
Подождите немного. Документ загружается.
Convection was also understood to be an important driver of the general circulation,
or at least the tropical and subtropical circulation known as the Hadley cells.
By the turn of the century, the work of Hann and others showed that convection
was not the energy source for extratropical cyclones, the midlatitude migratory low-
pressure systems. Attention shifted from the role of latent heat release to the role of
the large horizontal temperature gradients that were systematically observed within
midlatitude low-pressure systems. In 1903, Max Margules, born in the Ukraine and
working in Austria, calculated the amount of kinetic energy that could be obtained
from the rising of hot air and the sinking of cold air in a low-pressure system. He
found that the amount of energy that could be converted into kinetic energy was
comparable to the actual amount of kinetic energy in a mature storm system.
Margules had identified the cor rect energy source for extratropical cyclones. The
process by which cyclones form and move was described by researchers working
under Vilhelm Bjerknes in Bergen, Norway, shortly after World War I. Finally, a
comprehensive theory by Jule Charney in 1947 explained that the structure and
intensity of low-pressure systems was a consequence of the growth of unstable
eddies on a large-scale horizontal temperature gradient.
Hurricanes did not have the prominent horizontal temperature variations found in
midlatitude weather systems; indeed, many hurricanes seemed almost perfectly
symmetrical. The prominence of convection throughout the hurricane, particularly
within the eyewall, suggested that convection was fundamentally important in the
development and maintenance of hurricanes. But how did the hurricane organize
itself or maintain itself against large-scale subsidence within its environment?
The currently accepted theory, published by Kerry Emanuel and Richard Rotunno
in 1986 and 1987, relies on radiative cooling of the subsiding air at large distances
from the hurricane. Emanuel also noted that as air spirals inward toward the eye it
would become more unstable because it would be gaining heat and moisture from
the sea surface at a progressively lower pressure.
Theories satisfact orily describing organized convection such as mesoscale
convective systems and supercells also have been slow to develop. One reason
for this is that no observing systems could accurately describe the structure of
organized convection until the development of radar. Indeed, the term ‘‘mesoscale’’
was coined specifically to refer to in-between sizes (10 to 500 km) that were too
large to be adequately observed at single locations and too small to be resolved by
existing observing networks. The widespread use of weather radar in the 1950s
helped fill the observation gap. The development of Doppler radar (for measuring
winds within precipitating systems) for resear ch purposes in the 1970s and as part
of a national network in the 19 90s helped even more. With comprehensive radar
observations, it became clear that the long lifetime of organized convection was
due to a storm keeping its updraft close to the leading edge of its cold, low-level
outflow. The storm could then take advantage of the ascent caused by the cold air
undercutting warm air without having its supply of warm air cut off completely.
But even radar was not sufficient. The development of the first dynamical descrip-
tions of supercells and squall lines in the 1980s, by Joseph Klemp, Richard
Rotunno, Robert Wilhelmson, and Morris Weisman, relied upon numerical
506 OVERVIEW FOR WEATHER SYSTEMS
computer-generated simulations of the phenomena to provide an artificial data set
unavailable from observations.
While the above discussion has focused on self-contained weather systems,
certain ‘‘weather producers’’ may be thought of as byproducts of these weather
systems. Many forms of severe weather, such as hail, lightning, and tornadoes,
are essentially side effects of convection (organized or otherwise) but have little
or no influence on the convection itself. Other mesoscale phenomena, such as sea
breezes, upslope precipitation, and downslope windstorms, do not represent instabil-
ities at all. Instead they are called ‘‘forced’’ weather phenomena because they are
driven by such external features as solar heating gradients and topographic obstacles
to the large-scale flow.
5 FORECASTING
The massive strides in weather forecasting during the past 50 years are due in large
part to our growing understanding of the nature and dynamics of important weather
systems. Forecasting of many phenomena has evolved from an exercise in extra-
polation to specific predictio ns of the evolution of weather systems to computer
simulations that accurately forecast the evolution of weather systems. Currently,
the most skillful weather forecasts involve the prediction of large-scale extratropical
weather systems.
One very important advance in our ability to forecast large-scale weather systems
was the development of the omega equation, first presented in simplified form by
Richard Sutcliffe in 1947. The omega equation is named after the Greek symbol that
represents the change in pressure of an air parcel. In its original form, the omega
equation is based upon a simplification of the equations of motion that assumes that
all motions are large-scale and evolve slowly. When this approximation is made, it is
possible to diagnose vertical motion entirely from large-scale pressure and tempera-
ture variations. Vertical motion is important for weather prediction because it is
fundamentally related to clouds and precipitation, but large-scale vertical motion
also can be used to infer the evolution of low-level and upper-level wind patterns.
Vertical motion became the key to understanding the evolution of weather. In fore-
cast offices, maps were designed and widely distributed that made it easy to look at
the large-scale fields and diagnose vertical motion.
A second important advance was the development of numerical weather predic-
tion, or NWP. NWP, discussed more extensively in the chapter by Kalnay in the
dynamics part of this Handbook, has completely transformed the forecasting of
large-scale weather systems. In the past, forecasters relied on their diagnosis o f
the current weather patterns to make forecasts. Nowadays, computers can make
much more accurate forecasts than humans alone, and the best possible forecast is
obtained by humans working with computer forecast output. The forecaster’s task
has become one of identifying likely errors in the model forecast, based on the
forecaster’s knowledge of systematic errors within the model, errors in the initial
analysis, and compu ter forecast scenarios that run counter to experience.
5 FORECASTING 507
In contrast to large-scale weather systems, individual convective storms are not
directly simulated, nor accurately forecasted, by the numerical weather prediction
models currently in use. Forecasting convective storms still relies heavily on extra-
polation and a sound knowledge of what sorts of storms are likely to be produced by
certain larger-scale weather conditions. The modern era of forecasting severe
weather began in 1948, when two Air Force weather forecasters (Maj. Ernest
Fawbush and Capt. Robert Miller) had responsibility for forecasting at Tinker Air
Force Base near Oklahoma City. After a damaging tornado struck the base in 1948,
the two meteorologists were given the task of identifying the days when tornadoes
were likely to strike the base. We now know that forecasting the specific path of a
tornado is essentially impossible, but when a similar low-pressure system evolved a
few days later, the forecasters were compelled by the base commander to use a newly
minted severe weather warning system to issue a forecast of a possible tornado at the
base. Amazingly, the forecast came true, and preventive measures taken before the
second tornado struck the base prevented considerable casualties to aircraft and
personnel. This weather forecast, possibly the most serendipitous in the history of
humankind, led directly to modern tools for diagnosing the likelihood of severe
weather from large-scale conditions and laid the foundation for modern tornado
forecasting.
6 CONCLUSION
This historical review has emphasized that progress in our understanding of the
weather has required four ingredients: the need to understand the science of
the atmosphere before one can hope to make accurate forecasts; a complete kn ow-
ledge of the basic physical underpinnings of atmospheric behavior; observations
adequate to test the theor ies and point the way toward new ones; and numerical
models to provide simulations of the atmospher e more complete than any set of
observations. The following chapters of this part of the Handbook discuss the
current understanding of the weather phenomena discussed above, improvements
in which have lead to vastly improved weather forecasts.
BIBLIOGRAPHY
Frisinger, H. H. (1983). The History of Meteorology: To 1800, Boston, American
Meteorological Society.
Kutzbach, G. (1979). The Thermal Theory of Cyclones: A History of Meteorological Thought
in the Nineteenth Century, Boston, American Meteorological Society.
Miller, R. C., and C. A. Crisp (1999). The First Operational Tornado Forecast—Twenty Million
to One, Weather and Forecasting, 14(4), 479–483.
508
OVERVIEW FOR WEATHER SYSTEMS
CHAPTER 26
LARGE-SCALE ATMOSPHERIC
SYSTEMS
JOHN W. NIELSEN-GA MMON
1 INTRODUCTION
The structure and evolution of large-scale extratropical weather systems is dominated
by a fundamental contradiction: The airflow within such systems represents an almost
exact balance among the forces affecting each air parcel, but the slight departures from
balance are essential for vertical motion and the resulting clouds and precipitation, as
well as changes in intensity of the systems. Furthermore, balanced flow could not be
maintained without slight departures from balance. This section will explore the
morphology and dynamics of large-scale weather systems and will never stray far
from the concepts of balance and adjustment to balance.
The figures in this Chapter of the Handbook will typically use pressure as a
vertical coordinate, in keeping with the standard practice of displaying all meteor-
ological information above Earth’s surface on levels of constant pressure. Constant-
pressure surfaces are so nearly flat that they can be treated as horizontal for most
purposes. Vertical motion in this coordinate system, represented as o, is defined as
dp=dt rather than dz=dt, and since pressure increases downward rather than upward,
negative o corresponds to upward motion.
2 HORIZONTAL BALANCE
In its simplest form, the balance of forces in the horizontal plane is geostrophic
balance (see chapter by Salby in this Handbook) between the Coriolis force and the
Handbook of Weather, Climate, and Water: Dynamics, Climate, Physical Meteorology, Weather Systems,
and Measurements, Edited by Thomas D. Potter and Bradley R. Colman.
ISBN 0-471-21490-6 # 2003 John Wiley & Sons, Inc.
509
horizontal press ure gradient force. Since the Coriolis force is proportional to the
wind speed and directed to the right of the wind direction (to the left in the Southern
Hemisphere), a balance of forces can only be attained if the horizontal pressure
gradient force is directed to the left of the wind direction (right in the Southern
Hemisphere), with sufficient speed that the magnitude of the Coriolis force equals
the magnitude of the horizontal pressure gradient force. Thus, balanced flow is
parallel to the isobars (in Cartesian coordinates) or height contours (in pressure
coordinates) with a strength proportional to the horizontal pressure gradient. This
balance (and the others described in this chapter) is generally stable, in the sense that
air parcels initially out of balance will tend to approach balance, on a time scale of a
few hours.
The full horizontal wind may be divided into geostrophic and ageostrophic
components. The geostrophic wind is, by definition, that wind that would represent
an exact balance between the horizontal pressure gradient force and the Coriolis
force. For large-scale extratropical weather systems, the geostrophic wind very
nearly equals the total wind. The ageostrophic wind is associated with an imbalance
of forces and, therefore, acceleration. It may be thought of as that portion of the total
wind whose associated Coriolis force is not balanced by a horizontal pressure
gradient force. Thus, acceleration is proportional to the magnitude of the ageos-
trophic wind and is directed in the same direction as the Coriolis force, to the right of
the ageostrophic wind (to the left in the Southern Hem isphere).
Other forms of balance may be defined that include nonzero ageostrophic winds.
For example, circular flow represents a balance of three forces: the horizontal pres-
sure gradient force, the Coriolis force, and the centrifugal force. In this circumstance,
the Coriolis force associated with the geostrophic wind balances the horizontal
pressure gradient force, and the Coriolis force associated with the ageostrophic
wind balances the centrifugal force. The wind remains parallel to the isobars or
height contours but is weaker (subgeostrophic) in a cyclonic vortex and stronger
(supergeostrophic) in an anticyclonic vortex. Since ageostrophic winds can often be
associated with balanced flow, it is often more useful to subdivide the wind a s
divergent and nondivergent rather than geostrophic and ageostrophic. The geos-
trophic wind is always nondivergent (except for effects due to the variation of the
Coriolis parameter with latitude), and the divergent wind is directly related to verti-
cal motion through the continuity equation.
3 VERTICAL BALANCE
The vertical balance of forces, known as hydrostratic balance, is between the vertical
pressure gradient force and gravity. This balanc e may be assumed to be maintained
exactly for extratropical weather systems. Indeed, the balance is so close that vertical
accelerations must be deduced by indirect means, through diagnosis of the divergent
wind.
As horizontal balance introduced a close connection between the instantaneous
pressure and wind fields, vertical balance introduces a close connection between the
510 LARGE-SCALE ATMOSPHERIC SYSTEMS
instantaneous pressure and density=temperature fields. To balance gravity, the verti-
cal pressure gradient force must be strong where air density is large and weak
where density is small. Equivalently, the vertical separation between two given
pressure surfaces (known as thickness) is small where the air is relatively cool and
large where the air is relatively warm. Horizontal variations in temperature imply
vertical differences in the horizontal variation of pressure; that is, the horizontal
pressure gradient at one level will differ from the pressure gradient at another
level if air of differing densities (temperatures) lies between the two levels. If one
simultaneously assum es geostrophic (horizontal) and hydrostatic (vertical) balance,
one obtains a relationship between temperature and wind known as the thermal wind
law: If one represents the vertical derivative of the horizontal wind (vertical wind
shear) by a vector, it will lie parallel to the isotherms with relatively warmer tempera-
ture to the right (to the left in the Southern Hemisphere). Thus, temperature, pres-
sure, and wind are all constrained by each other in the limit of exact balance.
Vertical motion carries a strong influence on atmospheric temperature. Following
an air parcel within which phase changes are not taking place and which is not
exchanging heat with its surroundings, upward motion causes the parcel to cool at
the dry adiabatic lapse rate, 9.8
C=km, and downward motion causes an identical
amount of warming. In most circumstances away from the daytime boundary layer,
the atmosphere is stably stratified, meaning that the instantaneous vertical derivative
of temperature is greater than 9.8 K=km, so that, for example, an ascending air
parcel replaces an air parcel that is warmer (and less dense) than itself. Conse-
quently, at a given level, upward motion causes cooling at a rate proportional to
the magnitude of upward motion and the degree of stratification. Similarly, down-
ward motion causes warming. This stratification is called stable because an ascend-
ing air parcel, finding itself cooler and more dense than its surroundings, would tend
to sink back down toward its original level, and a subsiding air parcel would be
warmer than its surroundings and would tend to rise.
Extratropical large-scale weather systems can be directly affected by condensa-
tion of water vapor within ascending, cooling air parcels. This condensation releases
latent heat, causing the air parcel to be warmer than it would have been without the
release of latent heat. As a result, in areas of saturated ascent, the temperature at a
given level does not fall as rapidly as it would in ascending dry air and may in fact
rise if the atmosphere is unstable to moist convection. Condensation, and the genera-
tion of clouds and precipitation, represents the primary internal heat source for large-
scale weather systems.
4 POTENTIAL VORTICITY
When the atmosphere is in approximate horizontal and vertical balance, the wind
and mass fields are tightly interconnected. The distribution of a single mass or
momentum variable may be used as a starting point to infer the distribution of all
other such variables. We choose to focus on the variable known as potential vorticity
(approximately equal to the vorticity times the stratification) for three reasons: (1) it
4 POTENTIAL VORTICITY 511
tends to have a simple three-dimensional distribution; (2) it is conserved in the
absence of diabatic and frictional processes, so it tends to have a simple evolution;
(3) it is of direct relevance to the fundamental dynamics of atmospheric behavior.
Potential vorticity is related to the three-dimensional mass and wind fields
through equations that involve three-dimensional inverse second-order operators
whose exact form depends on the level of approximation. One consequence of the
inverse Laplacian-like operator is that potential vorticity variations in one location
are diagnostically related to variations in the height and wind fields at a distance.
Indeed, the relationship between perturbation (deviation from some standard mean
state) potent ial vorticity and perturbation height (or pressure) is mathematically and
conceptually analogous to the relationship between electric charge and electric
potential. Areas of locally high potential vorticity correspond to locally (and region-
ally) low heights, and therefore cyclones, while areas of locally low potential vorti-
city correspond to anticyclones.
While it is possible for isolated regions of high potential vorticity to form
anywhere in the atmosphere, many potential vorticity anomalies are found at the
tropopause. There, midlatitude potential vorticity changes suddenly from around
0.5 10
6
m
2
/s K kg (henceforth, 0.5 PVU) to around 5 PVU. Horizontal and verti-
cal displacements in the position of the tropopause lead to sizable potential vorticity
variations.
Figure 1 shows the wind and mass fields associated with a prototypical axially
symmetric potential vorticity anomaly at the tropopause. This cyclonic anomaly
(positive in the Northern Hemisphere) is associated with a cyclonic circulation
that is strongest at the level of the anomaly and decreases downward and away
from the anomaly. Temperatures are anomalously warm above the vortex (isentropic
surfaces are deflected downward) and anomalously cold below the vortex. While
consistent with the given simple potential vorticity distribution, the wind and mass
fields are also in vertical and horizontal balance. The general characteristics of the
Figure 1 Wind (V ) and potential temperature (y
0
) structure associated with a balanced
axisymmetric vortex at the tropopause. Wind speeds are in m=s and are out of the page on the
left and into the page on the right. Potential temperature (in K) is the departure from a uniform
surface potential temperature. (From Thorpe, 1986).
512
LARGE-SCALE ATMOSPHERIC SYSTEMS
wind and temperature distribution are common to cyclonic potential vorticity ano-
malies, with analogous atmospheric states for anticyclonic potential vorticity anoma-
lies. Winds a nd temperatures associated with complex potential vorticity
distributions may, to a first approximation, be recovered by treating the potential
vorticity distribution as an assemblage of discrete anomalies, each with its own
impact on the nearby wind and temperature fields.
5 FRONTS
Fronts are elongated zones of strong temperature gradient separating regions of
relatively weak temperature gradient. Fronts can form as a consequence of deforma-
tion, convergence, and differential heating or cooling. Fronts produced by differen-
tial heating or cooling tend to be relatively small in scale, such as a gust front
(produced by evaporative cooling) or a sea breeze front (produced by daytime heat-
ing over land). Such mesoscale fronts are discussed in Chapter 30 (Brooks et al).
Here, we focus on synopt ic-scale fronts, general ly formed by a combination of
deformation and convergence. Depending on their motion and structure, synoptic-
scale fronts are called warm fronts, cold fronts, stationary fronts, or occluded fronts.
Characteristics of Fronts
While fronts are defined in terms of temperature, other atmospheric variables are
also affected by the presence of a front. Often, the air on either side of a front
originated from widely different locations, and a shar p humidity gradient will be
present as well. Pressure and wind are also affected by the density contrast across a
front. A pressure trough, wind shift, and vorticity maximum tend to be present along
the warm side of the front. The temperature gradient also tends to be strongest at the
warm edge of the front, and it is at this edge of the strong temperature gradient that
the front itself is deemed to be located. The temperature gradient can be so strong
that most of the temperature variation across the frontal zone takes place over a
distance as small as 100 m. Under many circumstance s, the leading edge of the
frontal zone is marked by a discontinuity of low cloud or even a long,
narrow band of low cloud apparen t in visible satellite imagery and known as a
rope cloud.
Cyclones (here used in the common meteorological sense to refer to synoptic-
scale low-pressure systems) have a symbiotic relationship to fronts. Cyclones
frequently form along preexisting surface frontal zones. If a preexisting frontal
zone is not present, however, a cyclone is quite capable of generating its own surface
fronts. A typical life cycle is shown in Figure 2. As the cyclone intensifies, both a
warm and cold front are present, which may or may not intersect in the vicinity of the
surface low. The warm front is marked by geostrophic wind from warm to cold air,
and the cold front is marked by the opposite. As the cyclone matures, the cold front
moves ahead of the low, and the former warm front between the cyclone and the
5 FRONTS 513
intersection of warm and cold fronts becomes the occluded front, marked with
strong temperature gradients on both sides and the warmest air along the front.
Fronts are typically associated with a variety of clouds and precipitation. The
precipitation may be ahead of the front, behind it, or both. Some distinguish between
anafronts, with an updraf t slopi ng upward toward the cold air and clouds and
precipitation behind the front, and katafronts, with an updraft sloping upward
toward the warm air and the bulk of the precipitation within the warm air. The
distribution of clouds and precipitation associated with the surface fronts of a devel-
oping cyclone is shown in Figure 3.
Surface fronts tend to be strongest (i.e., possess the largest tem perature gradient)
near the ground and decrease in intensity upward. Typically, a surface front will lose
its frontal characteristics by an altitude of 4 km or 625 mb (Fig. 4), although deeper
fronts have been observed. Fronts are also favored at heights of 6 to 9 km. Such
fronts are called upper-level fronts. An especially strong upper-level front is shown
in Figure 5. On rare occasions, a surface and upper-level front may merge, yielding a
single frontal zone stretching from the ground to the top of the troposphere (Fig. 6).
Figure 2 Typical life cycle of an intensifying offshore extratropical cyclone: (I) incipient
frontal cyclone, (II) frontal fracture, (III) T-bone structure, and (IV) warm-core seclusion.
Top panel: sea level pressure (solid), fronts (bold), and clouds (shaded). Bottom panel:
temperature (solid), cold air currents (solid arrows), and warm air currents (dashed arrows).
(From Shapiro and Keyser, 1990).
514
LARGE-SCALE ATMOSPHERIC SYSTEMS
Figure 3 Distribution of clouds and fronts associated with a prototypical developing
extratropical cyclone. Surface weather is represented by standard meteorological symbols.
Figure 4 Cross section through strong front observed over the central Pacific on March 9,
1987. Data taken from dropwinsondes (potential temperature in K, solid lines; wind barbs, 1
long barb ¼10 knots) and airborne Doppler radar (open circles with section-normal wind
component in m=s beneath). Wind analyzed with bold contours in m=s; negative directed out
of the page. (From Shapiro and Keyser, 1990).
5 FRONTS 515