Forecasters Guide to Tropical IViefeoroiogy

Chapter 1

INTRODUCTION

1.1 GENERAL

The urge to understand tropical meteorology has been [Link]). For most generating regions, about 100 years
driven by a variety of forces that include the of daily tropical cyclone positions provide the basis
devastation caused by tropical cyclones, demands by for detailed, authoritative climatologies. Apart from
military operations (for example, World War II, the monsoon depressions, not even 1 year of tracks has
Marshall Islands weapons tests, and the conflict in been plotted for other tropical vortices. This
southeast Asia), and the belief of influential mid- disproportion, although slightly reduced by squall-
latitude meteorologists that the tropics may hold the line studies in GATE, Venezuela, and the Amazon,
key to successful forecasts of global weather patterns and a 1987 field study of early-summer rain systems
and climate change. Efforts by World Weather Watch off southeast China (Taiwan Area Mesoscale
(WWW) and the Global Atmospheric Research Experiment-TAMEX), means that a meteorologist
Program (GARP), including the GARP Atlantic trying to forecast tropical cyclones is served by a
Tropical Experiment (GATE) of 1974 and the Winter variety of subjective, numerical, and statistical
and Summer Monsoon Experiments (WMONEX and techniques, special services from global prediction
SMONEX) of December 1978 and June-July 1979, centers, and at least hourly satellite updates. But for
gave many meteorologists their first experience of the the other 99.9 percent of the time, forecasters are
tropics. Research papers proliferated, concentrating pretty much on their own. This manual tries to redress
on the small time and space scales described by the the balance. It devotes one chapter to tropical
special observing systems (aerial reconnaissance, cyclones, but emphasizes other weather Systems,
rawinsondes, and radars) set up for the experiments. background climatologies, and forecasting hints and
pitfalls.
But where does all this leave the field forecaster? A
case study dependent on data not operationally Observational information on the tropics increased
available and supported by brief, doubtful statistics, significantly through the 1970s, but progress has been
doesn't help much in the day-to-day grind of tropical patchy over Africa and the Americas. Weather radars
analysis and forecasting. operate at many tropical stations, and there is great
potential for more aircraft weather reports. Compared
Predicting tropical cyclones taps almost all the to 20 years ago, there are fewer human weather
resources available to the tropical forecaster. Riehl observations, while automated observations,
(1979a) devotes a chapter to numerical hurricane especially those from satellites, are vastly increased
prediction but nothing to other tropical forecasting. and not yet fully exploited.
On an average day in the tropics, less than one tenth
of one percent of the area is affected by a tropical There is an urgent need to synthesize from these
cyclone. Only about 80 cyclones develop each year, diverse observational systems a coherent basis for the
but since 1950 more than 5,000 operational and analysis and prediction of tropical circulations and
research reconnaissance flights have been made into associated weather. Our ability to digest and use this
cyclone centers. In contrast, only one Arabian Sea information in tropical numerical models has not kept
mid-tropospheric cyclone, one monsoon depression, pace with numerical weather prediction in the middle
and two upper-tropospheric cold lows have been latitudes. This is partially due to the greater
reconnoitered. Fifteen single-aircraft research complexity of weather systems and interactions
traverses have been made across the near-equatorial between various scales of motion in the tropics. Thus,
trade-wind convergence, but none into the mysterious although worldwide predictions are being made,
weak circulations that cause disastrous floods (see tropical weather forecasting (at least in the near
Chapter 1

future) will continue to depend heavily on subjective of tropical meteorology in much greater detail than
judgment, aided by selected numerical products such is possible here.
as objective wind analyses, global long-wave
predictions, and computer-processed satellite data. Weather forecasters are viewed as applied researchers
who can evaluate hypotheses and come up with new
West Pacific typhoons and Atlantic hurricanes are the ideas. They should realize that weather mechanisms
same phenomenon; they need not be described in the tropics are not often understood. "Authoritative"
separately. Thunderstorms also possess the same opinions may differ, for example, on the prevalence
characteristics worldwide. What is not so well-known of easterly waves, on the cause of the quasi-biennial
is that some other phenomena occur throughout the oscillation, on predicting El Nino, and on the
tropics. These include surges, cold lows, linear generation, movement, and dissipation of surges.
convergence lines and shear lines, squall lines, low- Even when there is general agreement, such as on
level jets, heat lows, upwelling and dust storms. Once tropical cyclone dissipation, diurnal variations along
each of these phenomena is explained, all forecasters mountainous coasts, and the role of surface cooling
need to know are distributions and frequencies. in causing fog, apparent exceptions may occur.

Although this manual discusses various theoretical Chapters 2 through 6 focus on annual cycles and
hypotheses, it uses little mathematical formulation. monthly averages while making some attempt to
It assumes that readers are familiar with the pressure/ interpret them in terms of weather. Chapter 3
wind and thermal wind relationships; with divergence, discusses data. Chapter 7 covers diurnal effects. The
convergence, and vertical motion; and with vorticity remaining chapters concern weather analysis and
and lapse rate. It also assumes that they generally forecasting where astronomical control by the annual
understand Mid-latitude synoptic models. The and diurnal cycles is muted by complex feedbacks
comprehensive list of references in the bibliography among atmosphere, ocean, and terrain.
serves to facilitate further study on selected aspects
 Introduction

1.2 TERMINOLOGY

Definitions in the Glossary of Meteorology (Huschke, takes on the character of a trough if the curvature is
1959) are generally used here, but this section expands cyclonic, or a ridge if the curvature is anticyclonic.
some of these definitions, adds new terms introduced The term has not been widely accepted, and has been
to the tropics since 1958, and discusses terms that criticized by Leroux (1983) "because there is
have been made misleading by a variety of definitions. generally no interruption of the stream."
Terms used only once in the text (e.g., "hodograph")
are defined where they appear. Cold Low (upper troposphere) In comparison
to its environment, this low is most intense and coldest
Buffer Zone was defined by Conover and Sadler in the upper troposphere. It is therefore weaker at
(1960). In Figure 1-1, a buffer zone straddles the lower levels and can be hard to find below 500 mb.
equator and separates two oppositely-directed wind Apart from a few cumulonimbus clouds that may form
streams: The easterly trades and the monsoon near the center, it is generally subsident and cloud-
westerlies. The sense of rotation in the buffer zone is free, except for rising motion and cloud on its eastern
clockwise in the Northern Hemisphere (NH) summer periphery. In summer over the tropical Atlantic and
and anticlockwise in the NH winter. The zone tends Pacific Oceans, as well as the South Indian Ocean, a
to persist; it may contain circulation cells that are line of cold lows may comprise the tropical upper
generally stationary and it may last 3 to 4 days. If it tropospheric trough (TUTT)-see Figures 8-17 to
moves more than a few degrees from the equator it 8-22.

MONSOON TROUGH TRADE WIND TROUGH

-20N

-^ r-y^ yf-^
NETWC
/ / /

NEAR EQUATORIAL BUFFER ZONES

NH SUMMER NH WINTER
ION

-lOS-
Figure 1-1. Schematic representations of monsoon and tradewind troughs and near-
equatorial tradewind convergence (top) and near-equatorial buffer zones (bottom).
Stippling denotes areas of unsettled weather.
 Chapter 1

Electromagnetic Radiation. All MAIN

ULTRAVIOLET VISIBLE INFRARED (IR)
electrooptical devices and all the CATEGORIES
SUB- NEAR MIDDLE FAR FAR FAR
MILLIMETER (mml

SUB- MILUMETER
MICROWAVE

instruments on meteorological CATEGCHilES IR IR IR IR MILLIMETER

satellites are triggered by

electromagnetic waves that are 0.001
(4000 A)
I ym 0,74 pm
WAVELENGTH'-
emitted from, or reflected by, objects
or gases in the atmosphere. These
waves are usually specified in terms
of wavelength (e.g., a 10-cm radar)
or frequency (e.g., 3 GHz) or, more
Figure 1-2. Electromagnetic energy categories, frequencies, and
descriptively, as visible or infrared.
wavelengths (based on Elrick and Meade, 1987).
Frequency (in cycles per second)
equals (speed of light (2.997 x 10'" cm s"') divided by tropospheric air sinks and weather is fine. Over the
the wavelength in cm. For example, a 10-cm radar deserts, as winter follows summer and as surface air
operates at a frequency of 2.977 GHz. Figure 1-2 cools and becomes denser, surface pressure increases
illustrates the range of electromagnetic waves of and anticyclones replace heat lows. On the eastern
interest to the meteorologist. and westem edges of heat troughs, moister air and
some clouds reduce insolation; the surface pressure
The following units are used: minimum, especially over the oceans, is less marked.
However, the troughs are still heat troughs, with the
Frequency: lowest pressure coinciding with the highest surface
1 Hertz = 1 Hz = 1 cycle sec' temperature. That, in turn, is ensured by a relative
1 Gigahertz = 1 GHz = 10« Hz cloud minimum and light variable winds (oceanic
doldrums) along the trough axis.
Wavelength:
1 micrometer = l|im = 10"^ m The term "monsoon trough" has been incorrectly
applied to all persistent near-equatorial troughs with
Heat Low/Heat Trough Surface heat lows, westerlies on the equatorward side and easterlies on
occasionally strung along a trough (as over the deserts the poleward side-refer to Figure 1-1. Although this
of North Africa and southwest Asia in summer, and usage is too well-established to be readily changed,
over southern South America, southern Africa and readers should remember that in this report,
along the Great Rift of Africa throughout the year- "monsoon" troughs may be year-long, (as over South
see Figures 4-1 to 4-4), form above surfaces made America and southern Africa), and not just summer
relatively hot by insolation through clear skies. features (as over North Africa).
Heating of the overlying air causes its pressure to fall
to a relative minimum and a depression to form. Air Figure 1-3 (opposite) is a schematic model of the
converging into the heat low rises, but a strong rainfall, pressure, and low-level circulation associated
subsidence inversion, reflecting convergence in the with the eastem North Pacific "monsoon" trough. The
upper troposphere, allows only shallow clouds to meridional section, based on satellite observations,
form; this is quickly evaporated through mixing with reveals that maximum cloudiness and rainfall do not
very dry air from above the inversion. Axes of heat occur along the trough line, but in the westerlies south
troughs have been termed "heat equators." of the trough (see Figure 6-16). A slight secondary
rainfall maximum is evident just north of the trough.
In the family of tropical vortexes (see Figure 6-32), Along the trough line, clouds are scattered or even
heat lows/troughs are linked to surface anticyclones absent. This confirms that the trough is really a heat
and upper tropospheric (cold) lows in which mid- trough where light, variable winds and fewer clouds
 Introduction

allow the sun to maintain a relatively warm sea surface Maritime Continent This region includes
and relatively low air density. Similar distributions Indonesia, Malaysia, Mindanao, the western Caroline
have been observed over central India (see Figure Islands, Papua New Guinea, and the Solomon Islands.
6-19), in the south Indian Ocean (Ramage, 1971, Along with equatorial Africa and equatorial South
1974) and across West Malaysia (Lim, 1979). America, it is the source of most tropical
thunderstorms (see Figures 10-1,10-2, and 10-3). The
Intertropical Convergence Zone (ITCZ) large, high islands generate continental-like
Although this term is still widely used, its meaning convection, while the warm surrounding ocean is an
has been hopelessly corrupted. Besides "ITCZ," the inexhaustible moisture source. The Maritime
extensive low-latitude surface pressure trough has Continent is outlined in the locator chart (Figure l-5a,
been called "equatorial trough," "intertropical front," Page 1-10).
"heat trough," and "monsoon trough." As shown in
Figure 1-4 (next page), the ITCZ has been variously l\1id-Tropospheric (Subtropical) Cyclone This
defined as coinciding with (a) the axis of the surface system originates in baroclinic environments and is
trough; (b) a zone determined by "wind observations" most intense in the middle troposphere where
(Stechnovsky and Krouskova,1970); (c) "the fluence horizontal temperature gradients are least. It weakens
asymptote associated with the equatorial trough, both upward and downward from this level because,
without regard to cloudiness" (Simpson, 1967); and relative to the environment, it is warmer in the upper
(d) the line of maximum cloudiness (Sadler, 1975a). troposphere and cooler in the lower troposphere.
Finally, the semantic impossibility of two ITCZs Although a vertical eye does not develop, mid-
along the same meridian has also been perpetrated. tropospheric cyclones may be mistaken for tropical
This manual, then, replaces "ITCZ" with more cyclones in a weather satellite picture. Extensive
specific terms: "heat trough," "monsoon trough" nimbostratus with embedded cumulonimbus
(already somewhat contaminated), and "near- sometimes gives heavy rain. These systems are most
equatorial tradewind convergence" (NETWC). common over the subtropical North Pacific and North

RAINFALL PRESSURE
4^ W ISO* I40» 130'
20*N ^^20 N" 20"N

I5*N

IO*N

5*N

5»S 5*S

Figure 1-3. Model of the low-level summer circulation of the tropical North Pacific and associated meridional
rainfall and pressure patterns. The dashed line marks the monsoon trough (after Sadler, 1967).
Chapter 1
Atlantic (where they are known as "subtropical Quasi-Biennial Oscillation (QBO) Winds in the
cyclones") and in summer over the northern Arabian tropical stratosphere undergo a downward-
Sea where the term "mid-tropospheric cyclone" is propagating oscillation between easterlies and
generally used. See Figures 8-7 through 8-16. westerlies on an average of once every 26 months
(hence the name). The oscillation maximum occurs
Monsoon Trough See Heat Low/Heat Trough. at 75,000 to 85,000 feet (23-26 km) above the equator
and decreases downward and poleward. The cause is
Near-Equatorial Tradewind Convergence still debated. See Figure 4-17.
(NETWC) Throughout the year, surface pressure
gradients are directed from east to west in near- Rains/Showers These terms, as used in this
equatorial parts of the central Pacific and eastern manual, are given the following meanings:
Atlantic/South America. Coupled with small
equatorward-directed meridional pressure gradients, • Rains fall from deep nimbostratus with embedded
this ensures that the northeast tradewinds of the NH cumulonimbus when wind shear in the vertical and
converge with the southeast tradewinds of the SH in lower tropospheric convergence are both large. They
a persistent zone of unsettled weather—the near- occur in the circulations of tropical cyclones, mid-
equatorial tradewind convergence. Over the oceans, tropospheric cyclones, and monsoon depressions, as
the NETWC shifts between about 5° N in April and well as in squall lines and vigorous cloud clusters.
10° N in September. Over northern South America, Rains vary in intensity, but since skies stay mainly
the annual range is from about 2° S to 10° N. Between overcast, heaviest falls are most likely during the night
February and May, a secondary NETWC sometimes and around dawn.
forms near 5° S over the eastern Pacific—see Figures
8-55, 8-56 and 9-10.

60» 50" 40" 30" 20" 10" W 0" E 10" 20'

Figure 1-4. Four definitions of the "intertropical convergence zone" (ITCZ) over the tropical North Atlantic
(Sadler, 1973).
 Introduction
• Showers fall from scattered cumulus or monsoons, greater speeds may suddenly appear
cumulonimbus when wind shear in the vertical and (surge) and last for a few days. Downwind of the speed
lower tropospheric convergence are both small. In the maximum of the surge, convergence worsens weather
moist tropics, showers prevail when the weather is and the associated vertical circulation accelerates.
generally fair. Diurnal local winds, which develop Above, in the upper troposphere, flow in the opposite
readily during a showers regime, in turn determine direction from the surface winds is divergent and
the diurnal variation of the showers. exports heat to colder latitudes. Surges have been
linked to the development of tropical cyclones as well
ResultantWind Forclimatological purposes, scalar as monsoon depressions, squall lines, equatorial
quantities can be simply averaged; their standard convection, and the onset of El Niiio. Their causes
deviations indicate variability. On the other hand, seem to be varied and not well-observed or
wind must be averaged vectorially by first resolving understood. See Figure 8-38.
each wind observation into components from north
and east, summing over the climatological period, Tradewind Trougli The first edition of AWS TR
obtaining the averages, and finally reconverting the 240 defined "tradewind trough" as the pressure trough
average north and east components into a single along which the tradewinds of each hemisphere
vector—the resultant wind. The steadiness or converge. Research aircraft traverses have since
persistence of the wind is defined as the ratio of the shown that the trough is very broad and weak, and
magnitude of the resultant wind to the average speed that the convergence, which is much narrower and
of the wind without regard to direction. It is usually more well-defined, may not coincide with the lowest
expressed as a percentage. A resultant of winds all pressure. This edition, then, replaces "tradewind
with the same direction would have a steadiness of trough" with the more specific "near-equatorial
100 percent. tradewind convergence" (NETWC).

South Pacific Convergence Zone (SPCZ) Tropical Upper Troposphieric Trough (TUTT)
Climatological charts show a zone of maximum See Cold Low.
cloudiness extending east-southeastward from New
Guinea throughout the year. During winter, cloudiness Wall<er Circulation Bjerknes (1969) identified a
is enhanced by troughs in the upper tropospheric weak cUmatological feature in the equatorial zonal
westerlies that tend to become stationary just west of circulation. In this feature, surface flow from the east
the zone. During summer, surface convergence in the was lifted and gave rain over Indonesia, returned from
zone is associated with a trough in which cyclogenesis the west in the upper troposphere and sank over the
occurs. Mechanical and thermal effects generated by central Pacific. Others have postulated similar
Australia and the maritime continent probably cause circulations associated with the equatorial African and
the SPCZ. See Figures 6-1 to 6-4, 6-22, 8-41, South American thunderstorms. The Walker
and 8-42. circulation is much weaker than the Hadley circulation
and is often not apparent on the synoptic scale. See
Surge Within persistent surface wind regimes that Figure 2-7.
flow toward the heat equator, such as tradewind and
Chapter 1
1.3 UNITS OF MEASUREMENT

The World Meteorological Organization (WMO) has There has been some progress; upper-air temperatures
established a set of units, based on the centimeter- and dew points in the U.S. are expressed in ° C,
gram-second system, in which to express potential temperatures in K, and mixing ratios in gm
meteorological measurements. Ideally, this manual Kg-'
should conform. In the United States, however, surface
air temperature is still regrettably measured and Whenever U.S. units appear in the text, they are
reported in degrees Fahrenheit (° F) instead of degrees followed by the approximate metric equivalents set
Celsius (° C) and rainfall is measured in inches instead in brackets, except for speed; the transfer between
of millimeters (mm) or centimeters (cm). In aviation knots and m s' can be readily approximated by
operations, heights are in feet instead of meters or applying a factor of one-half. When convenient,
kilometers (km), distances are in nautical miles figures include both sets of units.
instead of kilometers (km), and speeds are in knots
instead of meters per second (m s')-

1.4 LOCATOR CHARTS

The map foldouts opposite should help readers in described on page 1-6 is outlined by the dashed lines
locating the geographical points of reference on the map of the Eastern Hemisphere, Figure l-5a.
metnioned in the text. The "Maritime Continent"
 60' 80' 100" 120* 140" 160
T JAPAN
C H1N A

HaiFi m/wo Jima

• INDIA C<''"''«5JV-BURMA!.^ .TVIETV,^ ZJT^^ TAJ//AN^T-

20' V r»Chi3ngMTiir Cj SOUTH r^

ARABIAN \ / BENCA£^i^Ci'''^'*^°"°"^'''"^'^'SP"^
SEA Aodroi
•AncJcneii AB
Guan,
KERALAt /*^^j:«bu City (Macfon AB)
N SRI
i^ongltlo Pcko,
% LANKA V ^ C AR O n NE ISLANDS
NORTH INDIAN OCEAN
\
0' *G/>o

^ is^
Pis/ GUINEA>;%^
<^v
t^
BORDER OF ^-«
"MARITIME CONTINENT"
• m//,s
5 O UTH INDIA N OCEAN
20'

60* 80" 100' 140"

Figure l-5a. Locator chart. Eastern Hemisphere. The dashed line encloses the "Maritime Continent."
 160* 180' W 160'
I 140' 120"
'AN

^ Midway

NORTH
Jima
Kauai
[Link].« Honolulu
Maul '•'*''■ Moltatalo
HAWAII/<?«'»
20"
-9 *Wak»
Kaoiutla
O Johnston

•Andersen A 8
o /» ^ Cl F I C
MARSHALL
Guam
^Eniwttak ISLANDS
• Kwafo/e*n
^^m Mafuro
E ISLANDS o r* Washington
•3 ^Fanning

\
7r.
arawa
• Christmas

• Canton

-^'^GUINEA-\C^;> ^Funafuti

V"'
*Pago Pago
• Willis
^^^/v,
C^Vili Pop..t..'°C/^
Lfvu I TAHl
20'
Caledonia
5 o^r/y /P^ C I F I C

160° 180* w 160* 140« 120*

nt."
N

O'

120° 60' 40°

Figure l-5b. Locator chart, Western Hemisphere.