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.
Figure 2 Schematic cross section through a typical squall line system. All flow is relative to the squall line, which is moving from right to left.
Circled numbers are typical values of wet-bulb potential temperature, directly related to equivalent potential temperature (in
C). The convective region
occupies the first 30 km behind the leading edge, and stratiform precipitation the next 100 km. The air descending in the mesoscale downdraft remains
above the cooler air, which descended in the convective downdrafts and spread out in the lowest layers (after Zipser, 1977).
626
Figure 3 Soundings shown are taken behind squall lines, mostly within or toward the rear of
the stratiform precipitation region. The temperature and dew-point curves are far apart in the
mesoscale downdraft air below the melting level and near the surface, signifying low relative
humidity even in the rain area (after Zipser, 1977).
3 SQUALL LINE AND MCS STRUCTURE 627
4 SQUALL LINES, MC Cs, OR DISORGANIZED MCSs
What are the environmental conditions that favor MCSs organizing in the form of
squall lines? Empirically and theoretically (and confirmed in simulations) the deter-
mining factor is the low-level wind shear (LeMone et al., 1998). Substantial low-
level wind shear usually results in convection rapidly organizing into lines perpen-
dicular to the shear, moving downshear while leaning upshear and transporting line-
perpendicular momentum opposite to the direction of motion (upshear).
What are the environmental conditions that favor MCCs (mesoscale convective
complexes)? These are the largest and longest-lived of the MCS family. The question
is what favors extremely large and persistent relative inflow of high y
e
air into the
system at low levels? Local or purely orographic effects are unable to provide such
inflow, which helps to explain why MCCs are often associated with a low-level jet
stream. Also, there is often a region of forced ascent for the low-level jet, eith er a
synoptic-scale front, orography, or large cold pool left behind by previous
MCSs=MCCs (Laing and Fritsch, 2000). Squall lines and MCCs are not mutually
exclusive. Only the criterion of approximately circular shape prevents many large
squall lines from satisfying the other MCC threshold criteria.
5 EVOLUTION OF MCS s
All MCSs that have been documented in case studies evolve in much the same
manner (Houze, 1993). A group of cumulonimbus clouds forms fairly close to
one another for a variety of reasons. The convective rain volume often peaks rapidly
but occasionally can persist at a high level for many hours (the typical MCC
precursor). Stratiform precipitation usually rises in phase with the convective preci-
pitation but displaced a few hours later in time. If the system is weak or short-lived,
there may be little overlap in time between convective and stratiform regions—one
may appear to evolve into the other. For strong, long-lived systems, incl uding squall
lines and MCCs, stratiform and convective regions may coexist in relative steady
state for many hours. The end stages of MCSs inevitably consist of decaying strati-
form-only regions.
6 CANDIDATE SYSTEMS THAT ORGANIZE RAINFALL AND MAIN
CONTRIBUTORS TO RAINFALL IN SELECTED REGIONS
Organization of precipitation systems is governed by disturbances with a variety of
horizontal length scales. The main requirement is that they include regions where
ascending motion is organized. The frequency and amount of precipitation in a
region is often determined by the nature of the disturbances, while the character
and intensity of the rainfall system (e.g., strength of convection, MCS or not) is often
determined by the local wind shear and thermodynamic profiles. There is too much
variety to cover the entire range of tropi cal environments, so selected illustrative
examples are given.
628 TROPICAL PRECIPITATING SYSTEMS
As a general rule, there tends to be an inverse cor relation between total monthly
or annual rainfall, and the intensity of the rainfall systems. For example, the heaviest
monsoon rains over land, and the heavy rains of the oceanic intertropical conver-
gence zones (ITCZ), often come without lightning and wi th poorly organized MCSs.
Where are the most violent storms in the tropics? This may seem counterintuitive,
but they are often on the fringes of deserts or during the premonsoon seasons or
‘‘break monsoon’’ periods. Some examples are given below.
Orographic Forcing (1) Strong flows of moist low-level air capped by warm
dry air. The classic example is the Hawaiian Islands, where persistent, steady trade
winds are forced to ascend the windward slopes. Annual rainfall in the adjacent
oceans is probably <300 mm, yet parts of the windward slopes of most islands
experience >300 mm per month. In the extreme case of Mt. Waileale on Kauai,
annual rainfall exceeds 10 m. These orographic rains are not at all steady but consist
of shallow convective clouds that produce periodic heavy showers. Other regions
with similar orographic enhancement include Jamaica and Puerto Rico,
January–May.
Orographic Forcing (2) Weak flow, with the moist layer weakly capped. This
distinctly different kind of orographic enhancement is triggered by the heating of
elevated terrain, which, in turn, forces a diur nal upslope flow. The organized
mesoscale flow often triggers thunderstorms, which may be closely spaced and in
turn organize into MCSs or MCCs, occasionally propagating off the mountains as
organized squall lines. The Sierra Madre Occidental of western Mexico is the site of
a classic example. These mesoscale rain events extend northward into southeaste rn
Arizona during July and August, sometimes referred to as a monsoon. Rainfall can
exceed 500 mm monthly in Mexico and 200 mm monthly in southeastern Arizona
during this regime. Many other parts of the world experience similar regimes,
including the tropical Andes, parts of east Africa, mountainous islands in the
Indonesian region, and Puerto Rico in summer.
Land–Sea Breeze Circulation Systems Conditions for effective rain produc-
tion from sea breeze circulations are similar to that for orographic forcing (2) above.
Florida in summer is a classic example. Adjacent oceans have little rain except for
tropical disturbances, while peninsular Florida generates some 200 mm rain and 20
to 25 thunderstorms per month, some of which can be locally severe, and=or
organize into MCSs, including squall lines. It is obvious from satellite and surface
data that these are generated by the ascent along sea-breeze convergence zones from
either coastline, and numer ous case studies of these storms have been undertaken.
Synoptic-Scale Waves The best known is the easterly wave. These may be the
most regular synoptic waves on Earth, generated in Africa, coasting across the
Atlantic into the eastern Pacific. They are less common and regular in the Pacific.
They have a period of 4 to 5 days, a wavelength abou t 2500 km, propagating
westward at about 6 to 8 m =s. Some 10% of these waves encounter favorable
6 CANDIDATE SYSTEMS THAT ORGANIZE RAINFALL 629
conditions for intensification into tropical cyclones. They modulate convection and
MCSs strongly, with the wave troughs wetter than the wave ridges (Reed et al.,
1977); in this way they are analogs of traveling waves in the higher-latitude
westerlies.
The northern half of an easterly wave usually has a strong midlevel easterly jet
that dictates squall line formation, while the southern half of the same waves may
generate heavily raining but poorly organized MCSs in the light wind shear regime.
Also of importance is the relative motion of the MCS and the system that may have
generated it. The squall line moves at roughly twice the wave speed; therefore, it
forms preferentially near the wave trough but tends to dissipate as it approaches the
wave ridge. This resembles the case of MCSs in the United States, where it is
common for the system to move eastward more rapidly than its larger-scale forcing
mechanism. If it moves into an unfavorable area, it may dissipate rapidly. If it can
regenerate convection on its right rear flank, it may not only live for a long time but
also move slowly enough to be a serious flash flood threat (Chappell, 1986). Exces-
sive rain events in the tropics are rarely attributable to a simple wave passage but to
unusual interactions among systems of different scales.
Many other regions of the tropics are affected by westward traveling disturbances.
Most of these do not fit the description of easterly waves but can be Rossby waves or
Rossby-gravity waves. Their existence is well-documented by satellite data analyses,
but case studi es are rarely under taken due to almost complete lack of appropriate in
situ data.
Cyclones Other than tropical cyclones (Marks, F. D., Jr., Chapter 32), there are a
variety of other cyclones that can organize tropical rainstorms. Once again, their
existence is well known but definitive case studies are quite rare. They are often
categorized by whether the maximum circulation (vorticity) is found in the upper,
mid, or low troposphere.
The tropical upper tropospheric trough (TUTT) is a semipermanent feature of the
summer hemisphere in the north Atlantic and north Pacific. It is a subtropical feature
and generally exists over dry midtropospheric air mass es. Upon occasion, some of
the common low-pressure centers along this trough become strong en ough to orga-
nize deep convection and MCSs, usually in the subtropical west Atlantic and west
Pacific. It is possible that such activity requires the low to extend downward from its
usual upper tropospheric location.
Midtropospheric cyclones may exist in the vicinity of Hawaii, where they are
known as Kona lows, and they can generate heavy rainfall events, especially when
they move slowly (Barnes, 2001). Similar cyclones have been documented in the
Indian Ocean. It is not clear how often these cyclones propagate downward into
the low levels to form monsoon depressions, as they are called, or whether the latter
are generated independently in low levels. In any event, some of the heaviest rain-
falls of the Indian subcontinent are produced in association with the passage of these
depressions. They are most common in the Bay of Bengal and frequently move over
northern India from there. They are not to be confused with tropical cyclones, which
are unable to survive in the strong wind shear regime that dominates south Asia
630 TROPICAL PRECIPITATING SYSTEMS
during the heart of the summer monsoon (strong low-level westerlies and strong
upper-level easterlies).
Intertropical Convergence Zone: Oceans It has been known, literally for
centuries, that the trade-wind flows of the Northern and Southern Hemispheres
converge along rather narrow zones in the oceans, known as the ITCZ, along which
synoptic-scale ascent must take place on average (not at all times). Rain systems are
frequent along the oceanic ITCZ, although without great convective intensity.
Typical rainfall totals are about 300 mm monthly; where the ITCZ stays over a
given location for many months of the year, annual totals may exceed 3 m. There is a
large-scale convergence zone extending southeastward from New Guinea toward
Tahiti known as the South Pacific convergence zone, within which the rainfall may
have similar properties to the ITCZ.
Intertropical Convergence Zone: Land, Monsoons The off-equatorial
heating of the continents forces much stronger cross-equatorial flows well into the
summer hemisphere. At very low levels, the ITCZ as defined by wind convergence
often moves far from the equator, and the low-level flow of moist air attempts to
follow suit. However, the deep convection and heavy rainfall does not usually reach
the location of this ITCZ but is distributed over a large region.
A prime example is north Africa in August. The wind convergence marking the
ITCZ at the surface appears to reach the central Sahara Desert, while the heaviest
rainfall remains well to the south, along about 10
N. This is a region of strong
temperature gradie nt, connected with the midlevel easterly jet and the generation
of easterly waves. Very strong thunderstorms and squal l lines form in the desert
margins near 20
N, modulated by the waves, but rare in any one location. The Sahel
zone near 15
N is in the strong rainfall gradient, with heavy rainfall during the July–
September monsoon season when the large-scale flow brings in moist air from the
southwest at low levels and the waves and squall lines are strongest (Zipser, 1994).
Short monsoon seasons are also experienced in western North America, noted
above, and northern Australia in December to March. The latter has been the subject
of several field programs near Darwin and the Tiwi Islands just to the north. The
thunderstorms over these islands are so common in season that Darwinites call them
by name (Hector). As in west Africa, the surface convergence zone extends well
beyond the area of heaviest rain (near 10
S), into interior Australia. Near Darwin,
research has established that the main modulation of these storms and rain systems is
not by synoptic-scale waves but by longer period oscillations in the flow (also known
as the MJO, see below). During monsoonal low-level westerly wind periods, rainfall
is heavy, oceanic in character, and tropical cyclones may form. During the ‘‘break ’’
periods, westerlies weaken or become easterlies, and rainfall decreases, but is
concentrated into strong thunderstorms and squall lines moving from the continent
(Rutledge et al., 1992).
The classic Asian monsoon affects billions of people and requires far more
research before its rain systems have been properly described and understood,
since few field programs have been undertaken with observations of appropriate
6 CANDIDATE SYSTEMS THAT ORGANIZE RAINFALL 631
scale. It is well known that the southwest monsoon flow is from ocean to continent
during northern summer over India, China, and all of Southeast Asia. There is great
variety of local and regional details, well beyond our scope here to describe . Over
India, there is a characteristic alternation between rainy and break periods on a scale
of weeks. Where persistent moist flow is forced to ascend rapidly, as over the hills of
Assam in northeast India, world record rainfalls have been recorded.
Madden–Julian Oscillation (MJO) First discovered by Madden and Julian
(1972, 1994), there is often a strong variability in tropical winds and rainfall on this
intraseasonal time scale of 30 to 60 days. While many wave modes may come into
play, the most significant one appears to be an equatorially trapped Kelvin wave
(Meehl, G., Chapter 5). It is strongest within 10
of the equator (but its effects extend
beyond), and from the western Indian Ocean eastward through the central Pacific
Ocean (but its effects extend beyond). In these regions, it is considered normal for
rainfall to be excessive for a few weeks and deficient for the next few weeks. The
break period has anomalous easterlies at low levels and westerlies at high levels,
while the rainy period has westerly wind anomalies at low levels and easterly
anomalies at upper levels. During the rainy west wind periods near the equator,
tropical cyclones may form on either side of the equator, sometimes on both sides
(double vortex). Periods with especially strong west winds are known as westerly
wind bursts. These phenomena have been extensively studied during a major field
program in 1992–1993 (Godfrey et al., 1998).
7 DISTRIBUTION OF TROPICAL=SUBTROPICAL RAINFALL AND
MESOSCALE RAIN SYSTEMS
Maps of annual rainfall totals in the tropics are often accuracy challenged but show
the same major features. The ‘‘doldrums’’ or ITCZ regions near the meteorological
equator are very rainy; the subtropical high-pressure regions and trade wind regions
are generally dry except where orographic lifting occurs. Interestingly, the regions of
heaviest rainfall, and the regions with numerous large, strong MCSs, are often quite
different.
Total Rainfall Distribution
Figure 4 can be considered a ‘‘best estimate’’ of total rainfall in the tropics, using a
combination of satellite and rain gauge data. The rainfall maps include the region
covered by the Tropical Rain Measuring Mission (TRMM) satellite (36
N–36
S) and
include one month from each season in 1998–1999. During the first 4 months of
1998 one of the strongest warm episodes of El Nin
˜
o–Southern Oscillation (ENSO)
occurred, and for the next 12 months one of the strongest cold episodes. The rainfall
anomalies associated with these events are clear ly seen in the expected areas: equa-
torial central and east Pacific (rainy in January and April 1998, dry in January and
632 TROPICAL PRECIPITATING SYSTEMS
April 1999); Indonesia (rainier in January and April 1999 than in the same months
of 1998).
Despite these unusually strong anomalies, these rainfall maps clearly show the
‘‘normal’’ seasonal cycle of rainfall over the tropics and subtropics. Asia is dry in
winter, wet in the summer monsoon, with similar extent in both years (despite
different amounts in some regions). African rainfall migrates north and south of
the equator with season. The oceanic ITCZ migrates very little in the tropical
Atlantic and Pacific with the seasons.
MCC Distributions Using Satellite Infrared (IR) Measurements
Laing and Fritsch (1997) have systematically mappe d out the regions subject to
MCCs throughout the globe (Fig. 5). There are concentrations of MCCs in the
central United States, Panama–Columbia, South America east of the Andes from
15
to 35
S, Africa between 0
and 15
N, northeast India–Bangladesh, and lesser
concentrations in China and northern Australia. Laing and Fritsch (2000) show how
the environments of MCCs in these regions vary, but the sim ilarities are striking: a
low-level jet bringing high y
e
air, and an approaching disturbance (which can be
weak) providing ascent, triggering inte nse convection downwind of elevated terrain.
Some tropical regions are notable for high rainfall, but a near absence of MCCs.
These include the entire oceanic ITCZ belt, equatorial South America, Southeast
Asia, and the entire Maritime continent (Indonesian Archipelago). These regions are
characterized by persistent deep and high cloudiness, and by low values of average
outgoing longwave radiation (Laing and Fritsch, 1997). The contrast between the
Amazon and Congo basins is especi ally striking. McCollum et al., (2000) have
recently noted that the satellite estimates of rainfall are biased low in equatorial
Africa but are approximately correct in the Amazon, for reasons yet to be deter-
mined, but which they speculate are related to differing properties of MCSs and their
microphysics.
MCS Distributions Usi ng Passive Microwave Measurements
Passive microwave radiances at 37 and 85 GHz can be used to identify, categorize,
and map MCSs. The principle used to diagnose deep precipitating convection is
that ice particles are efficient scatterers but poor absorbers and emitters of radiation
at frequencies in this range. Therefore, regions with large depth of precipitation-
sized ice particles, especially graupel, have low brightness temperatures compared
with their surroundi ngs. This approach has the advantage that precipitation parti-
cles are sensed directly, compared with IR techniques that sense only the cloud
tops, even if they consist mainly of nonprecipitating cirrus. Data are available at
microwave frequencies from a series of SSM=I Special Sensor Microwave Imager
(SSM=I) instruments on DMSP satelli tes, since 1987, and on TRMM since 1997.
The main disadvantage compared to IR techniques is that they are not yet available
at high temporal frequencies from geosynchronous orbits.
7 DISTRIBUTION OF TROPICAL=SUBTROPICAL RAINFALL 633
Figure 4 Average precipitation (mm=day) for selected months of 1998–1999 for the TRMM-
merged analysis (Adler et al., 2000).
634
TROPICAL PRECIPITATING SYSTEMS
Figure 5 Relationship among mesoscale convective complex (MCC) population centers, elevated terrain, and prevailing midlevel flow (Laing and
Fritsch, 1997).
635