Professional Documents
Culture Documents
Tropical Meteorology All
Tropical Meteorology All
GEOGRAPHY
For meteorological purposes, the Tropics may be roughly defined as that region
lying between the boundaries separating the easterlies and the westerlies in the
middle troposphere.
This boundary varies
seasonally
longitudinally
A knowledge of geography is important for a meteorologist. Below is a list of
places and locations that I expect you to be able to find on a map.
India Marshall Islands Midway Island
Madagascar U.S. Virgin Islands Peru
Somalia The Hawaiian Islands Pacific Ocean
Japan Guam Atlantic Ocean
Vietnam Wake Island Indian Ocean
Cuba Lesser Antilles Caribbean Sea
Taiwan Tahiti Persian Gulf
Malay Peninsula Fiji Bay of Bengal
Baja California Indonesia North Arabian Sea
Australia The Philippines South China Sea
One-half of the Earth’s surface area is contained between 30S and 30N.
where U is the characteristic wind speed, f is the Coriolis parameter, and L is the
characteristic horizontal length scale.
The Rossby number can be thought of as showing the ratio of the inertial forces
(Term A) to the Coriolis force (Term C).
2
A small Rossby number (R0 << 1) implies that the inertial terms are
unimportant, and that the pressure gradient force balances the Coriolis force
(geostrophic balance).
A large Rossby number (R0 >> 1) implies that the Coriolis term is
unimportant, and that the inertial terms and the pressure gradient term
balance (cyclostrophic balance).
For Rossby numbers on the order of unity, no term can be ignored, and
gradient balance holds.
For synoptic scale motions we typically use values of U ~ 10 m s−1 and L ~ 106 m.
The graph below shows the Rossby number as a function of latitude for these
values of U and L.
Rossby Number
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
0 20 40 60 80
latitude
Poleward of 20°° the Rossby number is small enough that the geostrophic
wind and actual wind are fairly close.
Equatorward of 20°° the Rossby number is no longer small, and the
geostrophic and actual wind can differ greatly, especially as the Equator is
approached.
The comparatively large values of Rossby number in the Tropics vs. the Mid-
latitudes means that on the synoptic scale quasi-geostrophic theory isn’t very
useful in explaining the dynamics of synoptic-scale tropical circulations.
3
On the planetary scale, where L ~ 107 m, the Rossby number does remain small
through most of the Tropics. Therefore, quasi-geostrophic theory may be
carefully applied to planetary-scale circulations such as the monsoon, the
Walker circulation, etc.
WIND-PRESSURE RELATION
In the mid-latitudes, due to the smallness of the Rossby number, Term A in the
horizontal momentum equation can often be ignored, so that there is a direct
relationship between the pressure field and the wind,
1
∇p = − kˆ × f V .
ρ
This is why in mid-latitudes we analyze isobars on the surface chart, from
which we can just look at the map and readily infer wind direction and
speed.
In the Tropics, all three terms must be retained, and gradient wind balance is
the norm.
There is no longer a simple relation between the pressure field and the wind.
Pressure gradients are quite weak in the Tropics except in tropical cyclones.
Because of the weak pressure gradients and lack of a simple wind-pressure
relationship, meteorologists prefer to analyze streamlines and isotachs rather
than isobars.
4
So, the 850 and 200 mb charts, analyzed using streamlines, are the two most
common (but by no means the only) charts used for tropical analysis and
forecasting.
STREAMLINE ANALYSIS
Streamline analysis consists of drawing lines that are everywhere parallel to the
wind.
Some features that commonly show up on streamline analyses are discussed as
follows:
Asymptotes (diverging or converging)
These are streamlines from with other streamlines are either diffluent or
confluent.
Although in general diffluence and confluence don’t necessarily imply
divergence and convergence, it is common to refer to a diffluent asymptote as
a divergent asymptote, and a confluent asymptote as a convergent asymptote.
5
Cusps
6
Inflow and outflow
Pure inflow and outflow doesn’t usually appear on the synoptic scale.
Usually it is accompanied by rotation, and so appears in conjunction with
cyclones and anticyclones.
Cyclones and anticyclones
7
They are usually accompanied by either inflow or outflow, and so purely
circular or closed streamlines are not usually analyzed.
Although cyclones usually have inflow and anticyclones usually have outflow,
it is possible to have outflow cyclones and inflow anticyclones.
Some general rules and reminders:
Streamlines are everywhere parallel to the wind
Neutral points should be kept relatively small, and in areas of light wind.
Streamlines are not necessarily the same as trajectories (they are equivalent
only if the wind field is steady-state.)
The spacing of the streamlines is rather arbitrary. You should draw enough
to be useful. (Some analysts draw more streamlines in areas of strong wind,
and fewer in areas of weak wind, but this is not universal.)
Here are two rather arbitrary rules that I was taught to follow by veteran
analysts while in the Navy. I’ve never found these written down in any reliable
source, and have never been able to find a dynamical argument for these two
rules. I’ve often seen them violated, so there is nothing sacred about them. I
just present them here, because it’s the way I was taught. They are:
8
Neutral points should not be connected to each other.
ABSOLUTE VORTICITY
The synoptic scale vorticity equation is
D
(ζ + f ) = − ( ζ + f ) ∇ • V .
Dt
Because the absolute vorticity is comparatively weak in the Tropics, a given
amount of convergence or divergence will have a much smaller effect on absolute
vorticity than it would in the midlatitudes.
Areas of inflow and outflow with little associated rotation appear on
streamline charts in deep Tropics.
Strong anticyclones are not common in deep Tropics, because they would have
negative absolute vorticity.
ATMOSPHERIC TIDES
Another interesting difference in the tropics compared to the higher latitudes is
the relatively large amplitude, diurnal and semi-diurnal variation of sea-level
pressure due to atmospheric tides.
Both the diurnal and semidiurnal tides are due to solar heating and cooling, and
are not gravitationally induced.
The solar heating excites three main pressure oscillations in the atmosphere,
having periods of 24, 12, and 8 hours.
The horizontal and vertical structure of each of these modes of oscillation are
different.
The 12-hour (semidiurnal) wave has the largest amplitude at the surface, and is
the most meteorologically significant of the modes, since it has a range of 1 to
2 mb.
The amplitude of the semidiurnal variation decreases with height, and is less
than 0.5 mb at the tropopause.
10
The pressure maxima occur around 1000 and 2200 local time, with the
minima at about 0400 and 1600 local.
Due to the superposition of the 24 and 8-hr period waves with the 12-hr wave,
the morning pressure maximum is slightly larger than nighttime maximum,
while the afternoon minimum is slightly lower than the early morning minimum.
The diagram below shows a 5 day pressure trace Lime Tree Bay, VI (from the
National Data Buoy Center website). The semidiurnal variation is readily
apparent, with the maxima and minima occurring at the same time every day.
Times are GMT, and not local.
Since solar heating and cooling is the primary factor for producing the
atmospheric tides, the daily range of the pressure variation can vary with
cloudiness, water vapor content, and surface moisture.
In general:
The tidal range is greater near the Equator, and decreases as you go
poleward.
The tidal range is greater over land than over water.
The asymmetry of the maximums and minimums is also more pronounced
over land.
Though the signal from atmospheric tides can be deduced in the midlatitudes, it
is much smaller than in the tropics and isn’t really considered important in
midlatitude meteorology.
11
The tides must be accounted for when comparing the pressure tendency at a
single station in the Tropics.
Failure to account for the tides may lead a forecaster to mistakenly conclude
that a tropical cyclone is developing or approaching, or may even lead to
missing such development.
12
ESCI 344 – Tropical Meteorology
Lesson 2 – Tropical Cyclones: Definitions and Advisory Responsibilities
SOME ACRONYMS
CPHC – Central Pacific Hurricane Center
JMA – Japan Meteorological Agency
JTWC – Joint Typhoon Warning Center
NHC – National Hurricane Center
RSMC – Regional Specialized Meteorological Center
TCWC – Tropical Cyclone Warning Center
WMO – World Meteorological Organization (part of the United Nations)
2
SUBTROPICAL CYCLONES
Cyclones occasionally form over the tropical ocean that have a mixture of
tropical and extratropical characteristics. These are referred to as subtropical
cyclones.
The energetics are a mix of baroclinic instability as well as latent heating
from the sea.
Characteristics of subtropical cyclones are
Maximum winds not concentrated near center of storm.
Asymmetric wind distribution.
Asymmetric distribution of convection, not coincident with low center.
Two main types of subtropical cyclones are:
Those that are surface manifestations of cold-core upper-level lows.
Those that are mesoscale circulations (cold or warm core) initiated near a
frontal zone or shear line.
Subtropical cyclones can transition into tropical cyclones.
3
from World Meteorological Organization web site: http://www.wmo.ch/index-en.html
The RSMC’s and TCWC’s are operationally run by the respective weather
service of the host country. The associations of these centers with the host
country’s weather service are shown in the table below.
Center Organization
RSCM Miami U.S. National Hurricane Center (NHC)
RSMC Honolulu U.S. Central Pacific Hurricane Center (CPHC)
RSMC Tokyo Japan Meteorological Agency (JMA)
RSMC New Delhi India Meteorological Department
RSMC La Réunion Météo-France
RSMC Nadi Fiji Meteorological Service
TCWC Perth Bureau of Meteorology, Australia
TCWC Darwin Bureau of Meteorology, Australia
TCWC Brisbane Bureau of Meteorology, Australia
TCWC Port Moresby National Weather Service, Papua New Guinea
TCWC Wellington Meteorological Service of New Zealand, Ltd.
In addition to the WMO centers, the U.S. Navy and Air Force maintain the Joint
Typhoon Warning Center (JTWC) in Pearl Harbor, Hawaii, which provides
advisories and forecasts to U.S. government assets in the South Pacific, the
Northwest Pacific and Indian Oceans (both hemispheres).
JTWC is maintained so that the U.S. government, and U.S. military in
particular, do not have to rely on foreign advisories and guidance, which
may not be available in times of crisis or conflict.
Other than in the North West Pacific, the JTWC only provides warning for
tropical cyclones having sustained winds of 34 knots or greater, and refers to
these systems generically as “tropical cyclones”.
Many national weather services also do their own tropical cyclone forecasts and
issue their own warnings.
There is inconsistency in what averaging times are used for reporting sustained
winds in tropical cyclones among the various agencies.
The U.S. agencies (National Hurricane Center, Pacific Hurricane Center,
and the Joint Typhoon Warning Center) use 1-minute average winds
4
Note that standard U.S. surface observations use a 2-minute average wind
when reporting wind speed (FMH-1).
Most other agencies use 10-minute average winds, while some even use 3-
minute average winds.
For the same wind conditions, a 1-minute averaged wind will be reported
as stronger than a 10-minute average wind.
For the same reported wind speed, a 10-minute averaged wind will be
about 15% stronger than a 1-minute averaged wind (e.g., a 55 kt 10-
minute averaged wind is roughly equivalent to a 63 kt 1-minute averaged
wind).
The difference in averaging explains why sometimes JTWC will sometimes
report a stronger intensity than will RSMC Tokyo.
There is also inconsistency as to the minimum sustained winds required for
issuance of tropical cyclone advisories. This varies from region-to-region and
agency-to-agency, with some providing advisories at the depression level, while
others wait until the winds are 34 knots or greater.
NHC, CPHC, JMA, and JTWC (Northwest Pacific only) provide advisories
on tropical depressions once the sustained winds are 25 to 30 knots.
In the Indian Ocean and the Southwest Pacific, JTWC provides advisories
only for those cyclones with winds 34 knots or greater.
New Delhi provides advisories only for winds 34 knots or greater.
In the recent past, names were assigned by regional agencies, and were not
necessarily internationally recognized.
For example, up through the 1990’s, JTWC assigned names to Western
Pacific cyclones, while JMA simply referred to them by number.
However, pretty much all regions now assign names to those cyclones whose
winds are 34 knots or greater, and the name lists are coordinated by the WMO
member countries in the respective regions.
5
SAFFIR-SIMPSON SCALE
The Saffir-Simpson scale is a classification scheme for tropical cyclones based on
the maximum sustained wind (1-minute mean) and also type and extent of
damage.
It is mainly used by RSMC Miami, RSMC Honolulu, and JTWC. The wind
categories are
Saffir-Simpson Max Sustained Wind
Category knots mph
1 64 - 82 74 - 95
2 83 - 95 96 - 110
3 96 - 113 111 - 130
4 114 - 134 131 - 155
5 135 + 156 +
6
ESCI 344 – Tropical Meteorology
Lesson 3 – General Circulation of the Tropics
Reading: McGregor and Nieuwolt, Chapters 5 and 6 (thru page 110 only)
Vincent, “The SPCZ: A Review” (e-reserve)
Krishnamurti, “Walker Circulation” Figure 1 (e-reserve)
Lau and Yang, “Walker Circulation” (e-reserve)
James, “Hadley Circulation” (e-reserve)
Waliser, “Intertropical Convergence Zones” (e-reserve)
Hastenrath, “Tropical Climates” (e-reserve)
Madden, “Intraseasonal Oscillation (MJO)” (e-reserve)
TERMINOLOGY
Boreal refers to the Northern Hemisphere
Austral refers to the Southern Hemisphere
The Bowen ratio can actually be negative, which would occur when the air is
warmer than the surface, but evaporation is occurring.
ROLE OF CONVECTION
The overwhelming majority of heat transferred to the atmosphere from the
surface is in the form of latent heat stored in water vapor.
2
The heat is released to the atmosphere when the water vapor condenses in
convective clouds.
Convection is therefore an extremely important process for the general
circulation of the atmosphere, especially in the tropics.
In two papers, Malkus and Riehl estimated that the heat balance of the tropics
can be maintained by around 30 synoptic-scale disturbances consisting of a total
of several thousand giant cumulonimbus clouds.
3
ZONAL CIRCULATIONS
In addition to the meridional circulation of the Hadley cell, numerous attempts
have been made to establish the existence of large-scale zonal circulations in the
Tropics.
There have been several, often contradictory theoretical and observational
studies which have attempted to describe and document the structure of these
circulations.
These studies differ mainly in the number and locations of the cells. However,
the existence of three large cells centered over the Pacific, Atlantic, and Indian
Oceans is generally agreed upon.
Krishanmurti’s 1971 map of velocity potential and divergence field shows:
Large upper-level divergence maximum in summer over Bay of Bengal.
Large upper-level convergence maximum in summer off tip of Baja
California.
Pacific Ocean – The largest cell if located over the Pacific, and involves rising
motion in the Western Pacific in the vicinity of Indonesia (Indonesia is often
referred to as the Maritime Continent), with sinking motion in the Eastern
Pacific off of South and Central America.
This circulation is known as the Walker Circulation.
Atlantic Ocean – Over the Atlantic there is rising motion in the West, with
sinking motion in the East.
Indian Ocean – In the Indian Ocean there is rising motion over Indonesia, with
sinking motion off of the east coast of Africa.
4
Image provided by the NOAA-CIRES Climate Diagnostics Center, Boulder, Colorado,
USA, from their Web site at http://www.cdc.noaa.gov/.
January
July
The upper-level high is usually located to the west of the surface high.
Because of the asymmetry in the upper-level versus surface high, as well as the
fact that ocean temperatures are generally warmer in the western ocean basins,
the weather characteristics associated with the subtropical highs are distributed
asymmetrically as follows:
5
Characteristic Western portion Eastern portion
Vertical motion Weak subsidence or rising Strong subsidence
motion
Surface wind From Equator From Pole
Advection Warm Cold
Temperature Warmer Colder
Dew point Higher Lower
Stability Weakly stable Strongly stable
Clouds Open-cell Clear or closed-cell
The longitudinal axis of the high can often be inferred from satellite imagery
based on the transition from open-cell to closed-cell convection.
Formation and maintenance of the subtropical highs is complex, and is likely the
result of a combination of dynamic and thermal processes. Possible important
processes are:
Poleward moving air in the upper-branch of the Hadley cell experiences
upper-level convergence in the subtropics, resulting in higher surface
pressure.
This process by itself could not produce the subtropical highs because the
resulting subsidence would cause adiabatic warming, which would push
the lower-level pressure surfaces downward and work against creating of
a ridge (recall the hypsometric equation).
If diabatic cooling of the air were added to the above process, it could
offset the adiabatic warming during descent. Two sources of diabatic
cooling are:
Long-wave radiational cooling of the upper atmosphere.
Cooling of lower atmosphere from cold ocean currents or wintertime
continents.
Another possible dynamic process is the non-linear Beta effect. This effect
results in Polar anticyclones moving equatorward, providing mass to
maintain the subtropical highs.
6
THE TRADE WINDS
The trade winds result from the flow on the equatorial side of the subtropical
highs.
The trades blow generally from ENE in the NH and from ESE in the SH.
The trades extend over about 20°° of latitude in the summer hemisphere, and
about 30°° of latitude in the winter hemisphere.
They blow more toward the equator in winter than in summer.
Mean velocities are 3.6 – 7.2 m/s, and are stronger in winter than in summer.
The trades are very steady, though there is large inter-annual variability.
The trades have a 3-layer vertical structure:
Sub-cloud layer – The layer below the cloud bases.
Cloud layer
Inversion layer – Characterized by negative lapse rate, and therefore, the
tops of the convective clouds.
The trade-wind inversion is a subsidence inversion.
The inversion height and strength vary spatially as follows:
Zonally Meridionally
Height Increases toward the west Increases toward the Equator
Strength Decreases toward the west Decreases toward the Equator
As the inversion height and strength vary, so do the type and height of the
clouds, with small cumulus or stratocumulus prevailing in the east, while taller
cumulus clouds become more prevalent toward the west, or toward the Equator.
The spatial variation of the inversion is explained by the fact that subsidence is
strongest, and convection weakest, to the east of the subtropical high, resulting
in a low, strong inversion in that region.
As the air embedded in the trades moves westward and equatorward it
encounters less subsidence, as well as enhanced convection (since it is picking up
latent and sensible heat from the ocean surface.) This results in a weaker, higher
inversion toward the Equator, or toward the west.
7
The warming of the air as it moves westward helps to maintain the trade winds
by resulting in a lowering of surface pressure to the west (recall again the
hypsometric equation).
The maximum wind in the trades is usually found near the bottom of the cloud
layer.
Friction causes the wind to increase with height.
Thermal wind is westerly, so causes trades to decrease with height.
Result of these two effects is a wind max near the bottom of the cloud layer.
This vertical shear causes the trade-wind cumulus clouds to have their
characteristic appearance of leaning upstream.
8
The deceleration of the flow downstream from the region of maximum
anticyclonic curvature (but before reaching the equatorial trough) results in
convergence.
The separation of the regions of maximum cloudiness and minimum pressure is
necessary to maintain the thermal trough.
If the max cloudiness were directly over the thermal trough, the solar energy
at the surface would be decreased, which would be a negative feedback for
maintaining the thermal trough and maximum in surface temperature.
The role of the sea-surface temperature (SST) maximum is not completely clear.
Some argue that SST maximum directly contributes to the formation of the
equatorial trough.
Others (e.g., Ramage) argue that the SST maximum is caused by the
convergence of the ocean surface waters from the converging trades, and is
therefore an artifact, rather than a cause, of the trough.
There is a positive feedback mechanism between the upper ocean and the
atmosphere that also aids in the formation of the equatorial trough.
Where the surface winds are strong there is more mixing of the upper ocean,
resulting in cooler surface temperatures.
In the region of the trough the surface winds are lighter, resulting in less
ocean mixing and warmer ocean temperatures, which result in warmer
atmospheric temperatures and lower surface pressure.
If the earth’s surface were uniform the position of the ITCZ would be oriented
along the lines of latitude, and its annual migration would be symmetric with
respect to the Equator.
Because of the land-water contrasts, the ITCZ is not oriented exactly zonally,
but meanders north and south.
The ITCZ takes its largest poleward excursions in the summer hemisphere over
large land masses.
Because the NH has much more land than the SH, the ITCZ is closer to the poles
in boreal summer than in austral summer.
Mean position of ITCZ
15 N in boreal summer
9
5 S in austral summer
Maps of highly reflective clouds gives an idea of the location of the ITCZ. Below
are monthly means (Image provided by the NOAA-CIRES Climate Diagnostics
Center, Boulder, Colorado, USA, from their Web site at http://www.cdc.noaa.gov/.)
10
There are two general types of circulations associated with the equatorial
trough.
A trade-wind trough occurs when there is a confluence of the trade winds
from each hemisphere, with no directional shear across the trough.
The trade-wind trough occurs over the open ocean areas of the North
Atlantic, and over the Northeast and North Central Pacific.
A monsoon trough occurs when the trades from the winter hemisphere
recurve as they cross the Equator, and so there is a westerly flow on the
equatorial side of the trough.
The monsoon trough occurs near large continental areas, and is prevalent
in the Western Pacific and Indian Oceans.
A region where the cross-equatorial winds recurve from easterly to westerly is
referred to as a near-equatorial buffer zone.
Since the buffer zone is so close to the Equator, and Coriolis is negligible in
this region, terms such as cyclonic or anticyclonic have rather nebulous
meaning when applied to circulations in the buffer zone.
11
THE SOUTH PACIFIC CONVERGENCE ZONE
Region of persistent convection in South Pacific.
Generally oriented east-southeastward.
More zonally oriented in the western portion, and diagonally oriented in the
eastern portion.
Present year round, but more active in austral summer.
Several theories for origin and maintenance.
SST gradients – Gradients in SST lead to moisture convergence in the low-
levels, with enhanced convection. Most applicable to zonal portion.
Land-sea distribution.
Equatorial/midlatitude wave dynamics, interaction, and energy convergence.
Tropical/mid-latitude interaction – Most applicable to diagonal portion.
The strength and location of the diagonal portion appears related to the
subtropical jet.
12
ENSO really isn’t an abnormal phenomenon. It can be thought of as just one of
several stable global climate modes. For reasons not completely understood, the
general circulation of the ocean/atmosphere system switches periodically
between modes.
The question then becomes, “Why aren’t the shifts regular and predictable”?
The reason is because “weather” is superimposed upon these climate modes.
The entire system is non-linear (perhaps chaotic?), so the response to
“weather” cannot be predicted.
13
ESCI 344 – Tropical Meteorology
Lesson 4 – Seasonal-mean Wind Fields
2
Upper-level troughs, termed tropical upper-tropospheric troughs (TUTT) have
formed in North Pacific and North Atlantic.
Anticyclone established over Mexico.
October
Anticyclone in Asia-Africa region moved back toward Equator.
NH subtropical jet reestablished.
Anticyclone that was over Mexico has moved south.
Variable westerlies reestablish over Atlantic and North Eastern Pacific.
SH subtropical jet weakens somewhat.
3
The gradients across the upper-tropospheric front are due in part to the
upper air convergence in this region as the upper branch of the Hadley cell
converges with the mid-latitude westerlies.
Though there is no surface weather associated with the subtropical jet, it does
interact with the midlatitudes and does play a role in midlatitude dynamics and
influences midlatitude weather.
4
ESCI 344 – Tropical Meteorology
Lesson 5 – Tropical Cyclones: Climatology
GLOBAL CLIMATOLOGY
Note: Most of the statistics given in this section are from Gray, W.M., 1985:
Tropical Cyclone Global Climatology, WMO Technical Document WMO/TD-72, Vol.
I, 1985.
About 80 tropical cyclones per year world-wide reach tropical storm strength
(≥
≥ 34 kts).
About 50 – 55 each year world-wide reach hurricane/typhoon strength
(≥
≥ 64 kts).
The rate of occurrence globally is very steady.
Global average annual variation is small (about 7%).
Extreme variations are in the range of 16 to 22%.
Variability within a particular region is much larger than global variability.
Most (87%) form within 20°° of the Equator.
Those that form farther than 20°° from the Equator are usually in the
Northern Hemisphere.
Vast majority form in or near monsoon troughs.
Most of the remainder form in tropical waves
A few form along old frontal zones or shear lines.
Cyclogenesis tends to cluster in 2-3 week active periods that occur after 2-3 week
inactive periods.
NORTH ATLANTIC
11.6% of global total tropical storm strength and higher.
12% of global total hurricane strength and higher.
97% occur between June 1 and November 30
There is no month without at least one tropical cyclone
Peak of season is September 10
Position and strength of subtropical jet is key factor in seasonality, along with
variations in SST.
Genesis regions migrate throughout season in response to upper-level winds
and SST.
Per year there are on average
10.1 named cyclones
5.9 hurricanes
2.5 category III or greater
Variability
Have been as few as 1, to over 20 cyclones
Modes of genesis different from global averages
2
More than half form in tropical waves coming off of Saharan Africa.
Less than half form along ITCZ, with a few forming in baroclinic zones
associated with old fronts or shear lines.
Possible 25 to 40 year cycle in numbers (not enough data to conclusively
determine).
SOUTH ATLANTIC
Tropical cyclones are very rare in the South Atlantic.
No ITCZ
Strong vertical shear
In satellite era there have been two tropical cyclones in the South Atlantic.
One, in March, 2004, becoming a Category I hurricane.
Formed from upper-level low.
In April, 1991 there was a nominal tropical storm off of Africa.
3
WESTERN NORTH PACIFIC
Granddaddy of all basins
30.7% of global total tropical storm strength and higher.
35.7% of global total hurricane strength and higher.
Noted for high occurrence of very large and very intense storms.
Lowest SLP on record was 870 mb in Super Typhoon Tip (October, 1979).
Only basin that is active throughout the year.
Highly seasonal, with peak in late-July/early-August
Can occur in any month
Per year the average is
26 named storms
16 typhoons
Genesis regions migrate seasonally with position of monsoon trough.
Vast majority of cyclones form in monsoon trough, with a small minority
forming in tropical waves.
Tropical upper-tropospheric trough (TUTT) can play significant role in genesis
and development.
4
During height of summer monsoon, the monsoon trough is well inland, and
there is also strong vertical shear due to the Tropical Easterly Jet, which
suppresses cyclogenesis.
November peak is more pronounced than May peak.
Majority occur in the Bay of Bengal, with far fewer in the Arabian Sea.
Per year there are on average
5 – 6 tropical storm strength cyclones per year
Range is 1 – 10
2 to 3 hurricane strength cyclones per year
5
SOUTHWEST PACIFIC OCEAN (East of 142°°E)
10.8% of global total tropical storm strength and higher.
9.5% of global total hurricane strength and higher.
Seasonal (October through April)
Peak in February/March
Per year the average is
9.0 tropical storm strength cyclones
4.3 hurricane strength cyclones
Monsoon trough and SPCZ are important genesis features.
Twin cyclones (one on each side of the Equator) sometimes form in early and
late season, often in conjunction with westerly wind burst along the Equator.
6
ESCI 344 – Tropical Meteorology
Lesson 6 – Local and Diurnal Circulations
Reading: “The interaction of trade wind and sea breeze, Hawaii” (e-reserve)
“The effects of a large island upon the trade-wind air stream” (e-reserve)
GENERAL
Diurnal variation of solar radiation and temperature in Tropics is much larger
than annual variation.
In midlatitudes, annual variation is larger than diurnal variation.
Diurnal and mesoscale process are very dominant in the Tropics.
LAND-SEA BREEZE
Land-sea breezes caused by the differential heating across land-water
boundaries occur both in the midlatitudes and the tropics. However, the
influence of land-sea breezes is generally more pronounced in the tropics.
Inland or offshore extent of land-sea breeze is greater in the tropics compared
with the midlatitudes (~100 km, vs. ~10 km).
The difference in extent is explained by the weaker Coriolis acceleration in the
tropics.
Rossby radius of deformation is the ratio of the gravity wave speed divided by
the Coriolis parameter,
λR = c f .
The radius of deformation is essentially a measure of how large a circulation
must be in order for the effects of the earth’s rotation to be important.
Circulation smaller than the radius of deformation are not affected by the
earths rotation.
Coriolis helps to limit the horizontal effects of the land-sea breeze by turning
the wind and therefore limiting its inland-offshore influence.
In the tropics the radius of deformation is nearly an order of magnitude
larger than in the midlatitudes, so a circulation such as the sea breeze can
have a much larger horizontal extent without being turned by the Coriolis
acceleration.
The type of cloudiness and/or precipitation associated with the land-sea breeze
depends on the orientation of the coast with respect to the prevailing wind, as
well as the inland topography.
In a 1949 study, Leopold categorized four main types of land-sea breeze
interactions with the trade winds over islands. His four categories are:
Lanai type – Appropriate for small, low island in which the trade winds blow
up and over the peaks, and do not split.
Sea-breeze front is narrow and perpendicular to trades.
Maui type – Appropriate when peaks are tall enough to split trade wind flow.
Flanks of sea-breeze front exhibit shear lines trailing downwind.
Mauna Kea type – Appropriate for windward side of island with very tall
peaks.
Cloudiness forming on windward side during daytime as trades are
reinforced by sea breeze.
At night, land breeze front causes convergence offshore, with offshore
cloudiness and showers.
Kona type – Appropriate for leeward side of islands with very tall, broad
peaks.
During daytime, sea breeze moves in unimpeded by trade flow, which has
been blocked by tall, broad peaks.
Convective showers form inland as sea breeze brings in moist air that is
heated from below.
At night, down-slope land breeze suppresses clouds.
Leopold’s categories likely apply to other islands besides Hawaii.
2
Land-breeze interaction with trades has been documented as producing
enhanced rainfall at night and early morning on the east coasts of Brazil and the
Ivory Coast.
Large, mountainous islands (such as Puerto Rico) often have a clear ring of
suppressed cloudiness surrounding them in the afternoon.
This is due to the compensating subsidence from the vigorous convection
over the interior.
Land-lake breeze circulations occur over and around large lakes, such as Lake
Victoria in Africa.
MOUNTAIN CIRCULATIONS
Mountain-valley breezes are common in the tropics.
Valley breezes occur in afternoon, leading to cloudiness over mountains, with
clear skies in subsiding air over valleys.
Mountain breezes occur at night, leading to clearing over the mountains, and
cloudiness in valleys.
Mountain rainfall patterns in tropics often show a band of maximum rainfall
somewhere at around 1000 meters above the valley floor, with decreasing
rainfall at the foot or toward the top of the mountain.
This is mainly associated with areas where convective clouds dominate the
rainfall.
Lower slopes don’t experience as much orographic lift, and also experience
evaporation of rain below cloud base.
3
Upper slopes suffer from lower moisture availability.
Where stratiform clouds dominate, the rainfall maximum tends to be near the
top of the mountain.
In order to balance the friction and Coriolis terms, the following relation must
hold,
tan ϕ = Fr f V
4
Convergence and moister conditions over the coastline if higher pressures
are inland.
This may partially explain the dry conditions over certain coastal regions such as
the North coast of Venezuela, and other locations in the Tropics.
5
ESCI 344 – Tropical Meteorology
Lesson 8 – Tropical Weather Systems
GENERAL
Deep convection in Tropics tied generally to where there is upper-level
divergence and outflow, not so much on static stability.
Tropical weather systems are typically divided into three categories:
o Waves
o Vortexes
o Linear disturbances
Observational studies of tropical weather systems find that often the lowest
observed relative humidities are actually located in the region of highest
precipitation!
o This is due to the fact that even in the areas of greatest precipitation and
convection, compensating subsidence occupies most of the area, and so
statistically the observations will be taken in subsiding (dry) regions.
2
Waves in the Easterlies are primarily a Northern Hemisphere phenomena
o This is because they need deep, easterly current to form an propagate,
and these are more prevalent in the Northern Hemisphere versus the
Southern Hemisphere.
o If the wave is moving slower than the mean wind speed (U – C > 0), then
the air parcels approach the wave axis from the East, and the
convergence will lie behind (to the East) of the wave axis.
o If the wave is moving faster than the mean wind speed (U – C < 0), then
the air parcels approach the wave axis from the West, and the
convergence will lie ahead of the wave axis.
In most cases the winds in the lower troposphere are faster than the wave speed,
while winds in the mid and upper levels are slower than the wave speed. The
simplified divergence/convergence pattern would look like the figure below.
3
o This pattern of convergence/divergence is associated with upward motion
and cloudiness behind the wave axis, with subsidence and clearing ahead.
The wave needs anticyclonic outflow aloft for maintenance of the convection.
Convection often amplifies as the waves approach the eastern Caribbean Sea.
o This is due to the presence of the TUTT, which can enhance the upper-
level outflow from the convection.
Waves in the easterlies are usually thought to be a North Atlantic phenomenon.
There is considerable differences of opinions among researchers and operational
meteorologists as to whether or not there are true waves in the easterlies in the
Western Pacific, or to whether or not Atlantic waves actually traverse Central
and South America and maintain their form into the Eastern Pacific.
BAROTROPIC INSTABILITY
One form of hydrodynamic instability that can occur in the atmosphere is
barotropic instability. The derivation of the condition for barotropic instability is beyond
the scope of this course. But, the condition for barotropic instability involves the
horizontal shear of the mean wind. The necessary condition for barotropic instability to
occur is that, somewhere within the flow, the following condition must be true:
d 2u
0. (1)
dy 2
This means that for barotropic instability to occur that the second derivative of the mean
zonal wind must be equal to somewhere in the flow. Condition (1) can also be written
as
d du
f 0 (2)
dydy
or
4
d
0. (3)
dy
We can interpret this to mean that the absolute vorticity must have a minimum or
maximum value somewhere in the flow in order for barotropic instability to occur.
There are absolute vorticity minima and maxima on both flanks of the jet, near the
locations of the inflection points in the velocity profile. Thus, the condition for
barotropic instability is met in these two regions.
However, the presence of an inflection point does not automatically mean that there is a
minimum or maximum in the absolute vorticity. If beta is large compared to the second
derivative of the velocity, such as for a broad, weak jet at low latitudes, as shown below,
then there will not be any maxima or minima in vorticity, even though there are inflection
points in the velocity profile. Thus, beta acts as stabilizing influence against barotropic
instability.
5
ENERGETICS OF BAROTROPIC DISTURBANCES
Barotropic disturbances derive their energy from the mean flow. Energy
considerations show that for a barotropic disturbance to grow it must tilt opposite to
du dy .1 Since midlatitude disturbances tend to tilt in the same direction as du dy , they
actually lose energy back to the mean flow due to barotropic instability. Thus, barotropic
instability is not a viable way for midlatitude disturbances to form and grow. However,
interestingly enough, since midlatitude disturbance decay due to barotropic instability,
they give up energy to the mean flow and help maintain the mean flow against friction.
Thus, barotropic instability is somewhat important for the maintenance of the mean flow
in the midlatitudes.
AFRICAN WAVES
Western Africa is major source of waves in the easterlies.
Their dynamics appears to be a combination of both barotropic and baroclinic
instability associated with the West African mid-level (easterly) jet.
o Convection may play a role acting as a forcing mechanism to create a
perturbation in the flow from which the disturbance can grow into a
wave.
Wave axis generally tilted from southwest to northeast, and is opposite to
horizontal shear south of the jet.
1
See Haltiner and Williams, pp.74-75.
6
o When tilt and shear are opposite, barotropic instability removes energy
from the jet and puts it into the perturbation (wave).
o Thus, growth due to barotropic instability will mainly be south of the jet
core.
Often have multiple vorticity centers, with southern one at around 600 mb and
south of the jet axis, and the northern one nearer to 850 mb and north of the jet
axis.
o Thorncroft and Hodges found that those 850 mb vorticity centers that are
found over the ocean are not continuations of the 850 mb over the land.
Instead, it appears that as the 600 mb vorticity center passes the coast
that development occurs downward, possibly influenced by latent
heating, so that a new 850 mb vorticity center is formed while the
previous, northerly 850 mb vorticity center dissipates.
These waves are preferentially formed in the half-year associated with the
Boreal summer.
Waves increase in amplitude as they approach the west coast of Africa, and
weaken as they head out to sea.
Divergence/convergence pattern opposite of waves over the North Atlantic.
o Low-level convergence, clouds and rain ahead of wave axis, rather than
behind.
Waves in the Easterlies account for the majority of all North Atlantic tropical
cyclones.
An increased number of waves means an increased number of tropical cyclones.
Meteorological factors in the wave formation region over Africa impact the
Atlantic tropical cyclone season.
UPPER-LEVEL CYCLONES
Most intense in the 200 – 300 mb level.
Many don’t even show up on the 700 mb chart.
Associated with upper-level convergence and suppressed convection.
Often surrounded by rings of thin cirrus.
7
Occasionally convection fires in center of low since cool upper-levels are
associated with static instability.
o Convection usually short-lived due to:
upper-level convergence
entrainment of dry air
Upper-level lows rarely transform directly into warm-core surface lows, though
they can interact with surface features to form new surface lows to the east of
the upper-level cyclone.
SQUALL LINES
Non-frontal lines of active thunderstorms.
Can be hundreds of miles in length
Persists much longer than lifetime of individual cells that make up line.
o Usually last for 3 – 15 hours
o May last for several days in West Africa
Occur in northern Australia, northwestern India, Bangladesh, malysia,
Indonesia, South America/Caribbean, and West Africa
Most frequently occur in mid to late afternoon.
Require moist low levels with relatively dry mid levels.
Tropical squall lines have trailing anvil, while mid-latitude squall lines have
preceding anvil.
8
Leading edge of gust front sparks new development and propagates line.
SUMATRA
Form during SW monsoon over Straits of Malacca
Usually form at night
Last 1 to 2 hours
Three main factors
o Daytime heating over Sumatra followed by nighttime radiative cooling
over the Straits of Malacca
o Orographic lifting over Malaysia
o Converging land breezes from Malaysia and Sumatra
SHEAR LINES
Equatorward extension of a midlatitude cold front
No large temperature or humidity shift
Zone of convergence
Characterized by stratiform clouds with embedded convection
Can stall out and cause low ceilings and rain for extended periods.
COLD SURGES
Surge of cold air from midlatitudes into Tropics.
Surge from winter hemisphere can enhance Equatorial westerlies in summer
hemisphere, and increase cyclonic vorticity.
o Can result in development of tropical cyclones in monsoon trough.
9
ESCI 344 – Tropical Meteorology
Lesson 9 – The Tropical Oceans
January
July
2
WIND STRESS AND THE EKMAN SPIRAL
The horizontal momentum equations for the ocean are
Du 1 ∂p 1 ∂τ x
=− + fv +
Dt ρ ∂x ρ ∂z
Dv 1 ∂p 1 ∂τ y
=− − fu +
Dt ρ ∂y ρ ∂z
where the stress terms (those involving τ ) are due to vertical turbulent
momentum fluxes (Reynolds stress).
For steady flow the equations are
1 ∂ p 1 ∂τ x
− fv = − +
ρ ∂x ρ ∂z
1 ∂ p 1 ∂τ y
fu =− +
ρ ∂ y ρ ∂z
The flow can be broken up into a part driven by the pressure gradient and that
driven by the stresses. The part driven by the stresses is governed by the
equations
1 ∂τ x
v=−
f ρ ∂z
1 ∂τ y
u=
f ρ ∂z
In the early 1900’s an oceanography students named Ekman solved the above
two equations to find the vertical structure of flow driven by the wind stress. His
result is
u ( z ) = exp(γ z )[U 0 cos(γ z ) − V0 sin (γ z )]
v( z ) = exp(γ z )[V0 cos(γ z ) + U 0 sin (γ z )] .
where U0 and V0 are the zonal and meridional components of the surface
current, and
γ ≡ f 2K
where K is the eddy viscosity.
If the solution above is plotted on a hodograph it traces a decaying clockwise
spiral with depth (in the Northern Hemisphere). This is known as the Ekman
spiral.
3
The depth of the Ekman layer is taken to be that point at which the spiral has
made one complete rotation (or the e-folding scale of the decay). Therefore, the
depth of the Ekman layer is
1 2K
DE = = .
γ f
Two interesting and important facts about the Ekman spiral:
ο #1 – The surface current flows at exactly 45° to the right of the surface wind (in
the NH).
Objects such as icebergs, life rafts, etc. will actually drift to the right of
the prevailing wind.
ο #2 - The vertically-integrated mass transport in the Ekman layer is directed at
90° to the right of the surface wind.
5
An eastward flowing equatorial counter current between the equatorial
currents.
Strong, western boundary currents in the subtropical and subpolar gyres,
contrasted with weaker return flows east of the gyre centers.
One striking feature of this circulation pattern is the strong, well-defined western
boundary current, as opposed to the weaker, broader flow to the east of the gyre.
The western boundary currents are due to the fact that Coriolis changes with
latitude. If Coriolis were constant, the western boundary currents wouldn’t
exist.
EQUATORIAL CURRENTS
The surface currents in the Pacific and Atlantic have a similar structure, and
can be explained at least in part by convergence and divergence associated with
Ekman transport.
6
The diagram above shows the air flow (open arrows) and the resultant Ekman
transport (dark arrows).
The “DIV” and “CONV” denote regions of divergence and convergence in the
Ekman transport.
Regions of divergence will result in a lowering of the sea-surface, while
regions of convergence will raise the sea surface.
Other than within a degree or so of the Equator, the ocean flow will be parallel
to the sea-surface contours with low heights to the left in the Northern
Hemisphere, and to the right in the Southern Hemisphere.
The ocean currents and their directions are indicated by the “W” and “E”
annotations, with “W” indicating a westward current, while “E” indicates an
eastward current.
The resultant surface currents are the
North Equatorial Current – Westward flowing
Equatorial Counter Current – Eastward flowing, more-or-less aligned with
the ITCZ.
South Equatorial Current – Westward flowing, and in both hemispheres.
There is also an Eastward flowing Equatorial Undercurrent that more-or-less
flows along the Equator at depth.
The equatorial currents in the tropical Indian Ocean differ in that since the
atmospheric flow switches directions seasonally due to the monsoon, so do the
currents.
7
ESCI 344 – Tropical Meteorology
Lesson 10 – Tropical Cyclones: Structure
GENERAL
Tropical cyclones are primarily driven by latent heating.
As the air spirals in toward the center, it picks up latent heat and sensible heat
through evaporation from the ocean.
Once the air is saturated, it can still pick up some sensible heat, but latent
heating is the dominant mechanism.
As the air approaches the center of the vortex, it rises in convection either in the
eye wall, or in the spiral convective bands. As it rises, it cools and the water
vapor condenses, giving off its latent heat.
The warming of the air due to the latent heat release results in low-level height
falls, and upper-level height rises, which helps maintain the low-level
convergence of the warm-moist air.
The air is exhausted out and away at the upper levels.
DISTRIBUTION OF WIND
The diagram below shows the typical tangential wind structure in a tropical
cyclone.
2
r is positive outward from the vortex
v is positive for a cyclonic vortex, and negative for an anticyclonic vortex.
If the vortex is steady, axisymmetric, and friction is ignored, then the tangential
wind is
v2 1 ∂p
+ fv − = 0, (4)
r ρ ∂r
and the vortex is in gradient balance.
Dividing (4) by fv, we get
v 1 ∂p
+1− = 0. (5)
fr f vρ ∂r
The first term is the Rossby number, RO.
In the core of the vortex, the Rossby number is large, so the Coriolis effects
can be ignored. Therefore, the core of the storm is in cyclostrophic balance.
Outside the core, the Rossby number is of the order of unity, so gradient
balance holds.
In either case, the wind speed depends on the pressure gradient.
Broadly speaking, the lower the central pressure, the faster the maximum wind
will be.
The relation between maximum winds and the central pressure is closely
approximated by
vmax = A p∞ − pc (6)
If the parcel started out at rest at a distance of 500 km, and spiraled in to a
radius of 15 km, it would have attained a tangential velocity of over 600 m/s
if its angular momentum were conserved.
Obviously, parcels don’t conserve angular momentum as they spiral into the
center of a tropical cyclone.
In fact, the absolute angular momentum decreases toward the center of tropical
cyclones, which means that air parcels must be losing angular momentum as
they spiral inward.
The parcels lose angular momentum through turbulent dissipation.
The absolute vorticity of an axisymmetric vortex is
∂v v
η = f0 + + , (10)
∂r r
where the second and third terms are just the shear and curvature terms for
relative vorticity. The absolute vorticity and the specific absolute angular
momentum are related via
1 ∂M a
η= . (11)
r ∂r
4
In pressure coordinates the momentum equations for an axisymmetric vortex
without friction are
Du ∂Φ v 2
=− + + f 0v (12)
Dt ∂r r
Dv uv
= − − f 0u . (13)
Dt r
If an air parcel starts out in gradient balance, but is impelled outward at a
velocity u, the change in radial acceleration with time is given by taking the time
derivative of (12),
D Du D ∂Φ D v 2 D
=− + + ( f 0v ) . (14)
Dt Dt Dt ∂r Dt r Dt
D v 2 1 Dv 2 v 2 Dr 2v Dv uv 2 2v uv uv
2
3uv 2 2 f uv
= − 2 = − 2 = − − f 0u − 2 = − 2 − 0 (16)
Dt r r Dt r Dt r Dt r r r r r r
D Dv uv
( f 0v ) = f 0 = − f 0 − f 02u . (17)
Dt Dt r
Substituting (15) thru (17) into (14) results in
D Du ∂ 2 Φ 3uv 2 3 f 0uv
= − u 2
− 2 − − f 02u
Dt Dt ∂r r r
or
D 2u ∂ 2Φ 3v 2 3 f0v
2
+ 2 + 2 + + f 02 u = 0 . (18)
Dt ∂r r r
Defining
2∂ 2Φ 3v 2 3 f 0v
ω = 2 + 2 + + f 02 (19)
∂r r r
we can write (18) as
D 2u
+ ω 2u = 0 . (20)
Dt 2
which expands to
6
2v ∂ v v 2 ∂ v 3v 2 3 f 0 v
− 2 + f0 + 2 + + f 02 > 0
r ∂r r ∂r r r
and through the following steps
2v ∂ v v 2 ∂ v 2v 2 v 2 2 f 0 v f 0 v
− 2 + f0 + 2 + 2+ + + f 02 > 0
r ∂r r ∂r r r r r
2v ∂ v ∂ v 2v 2 2 f 0 v f 0 v
+ f0 + 2 + + + f 02 > 0
r ∂r ∂r r r r
2v ∂ v 2v v 2v
+ f 0 + + f0 + + f 0 f 0 > 0
r ∂r r r r
results in
2v ∂v v
+ f0 + + f0 > 0 . (26)
r ∂r r
The second term in brackets is the absolute vorticity, and so we finally have the
criteria for inertial stability as
2v
+ f 0 η > 0 . (27)
r
Equation (27) shows that negative absolute vorticity is not a sufficient condition
to have inertial instability. A vortex with negative absolute vorticity can still be
inertially stable as long as
2v
+ f0 < 0 . (28)
r
In straight-line flow, or for weak vortexes where the curvature term in (27) can
be ignored, the stability condition can be written as
η f0 > 0 . (29)
This is why you often hear it said that anytime absolute vorticity is negative
that the flow is inertially unstable. However, keep in mind that for stronger
vortexes, particularly at low latitudes, negative absolute vorticity doesn’t
automatically imply inertial instability.
7
And finally, a reminder that the oscillation frequency of a stable vortex is given
by the stability condition [recall Error! Reference source not found.], so that for a
stable vortex
2v
ω 2 = + f 0 η . (30)
r
Sometimes it is helpful to write the stability condition or oscillation frequency in
terms of the specific absolute angular momentum,
1 ∂M a2
ω2 = . (31)
r 3 ∂r
ROSSBY RADIUS OF DEFORMATION
A fundamental horizontal length scale for a disturbance in a rotating fluid is the
Rossby radius of deformation.
The Rossby radius of deformation is the distance that a gravity wave (which is
the means by which the fluid adjusts to equilibrium) will travel in one inertial
period (ω−1),
λR = cgτ f = cg ω . (32)
which is the form of the Rossby radius of deformation most commonly used
in dynamic meteorology textbooks. However, (34) is just a less general form
of (33).
The gravity wave group velocity, cg, depends on the stratification of the fluid.
For a barotropic fluid, it is given by c g = gH where H is the depth of the
fluid.
8
For a baroclinic fluid, there are multiple baroclinic modes of oscillation, each
with its own group velocity. For waves that travel primarily horizontally
(long wavelengths) the group velocity can be approximated as
cg ≅ NH nπ . (35)
10
Radius of maximum winds is generally in the eye wall.
There is a sharp increase in speed near the radius of maximum winds.
Winds decrease aloft.
Maximum winds tilt outward with height.
Maximum winds are found in the front-right quadrant of the cyclone.
This is true, even if the forward speed of the storm is factored out.
The reason isn’t completely understood.
Strongest horizontal temperature gradients are near the radius of maximum
winds.
Warm core is narrow in the lower troposphere, and bulges outward aloft.
Some features of the core region are the eye and rainbands.
Some definitions:
Rainband – Also called spiral bands, or feeder bands, they are areas of cloud
and of precipitation spiraling in toward the center of the circulation.
Convective ring – Rainbands that near completely circle the center.
Moat – Clear region between two convective rings.
Eyewall – Innermost convective ring.
Eye – Clear region at center of circulation, surrounded by the eyewall.
11
The dynamics and fluctuations in the core are dominated by primarily by
convection and heating.
THE EYE
An eye generally doesn’t form until the cyclone reaches hurricane or typhoon
intensity.
The mechanisms by which an eye is formed and maintained are not completely
known. One conceptual model is presented below.
As air spirals inward toward the center of the cyclone, its speed will increase
due to conservation of angular momentum.
However, if angular momentum were truly conserved, wind speeds of 600
m/s or more would be reached.
Therefore, angular momentum must be dissipated via convection,
turbulence, and other processes that retard the flow.
The angular momentum balance at a point is given by
∂M a ∂M a ∂M a
= −u −w − Fr . (36)
∂t ∂r ∂z
The three terms on the right-hand-side are radial advection, vertical
advection, and dissipation of angular momentum.
At some distance near the center of the vortex, the dissipation of angular
momentum can no longer keep up with the horizontal advection. This leads
to a horizontal convergence of angular momentum.
This horizontal convergence of angular momentum leads to an increase in
the winds above the value that the pressure gradient can support (i.e., they
are super-gradient).
The super-gradient winds develop a radially outward component, since the
pressure gradient cannot supply the required centripetal acceleration. The
consequences of this are (see figure):
There is radial convergence at distances outside of the radius of
maximum winds.
There is radial divergence at distances inside the radius of maximum
winds.
12
The convergence leads to upward motion, and results in the convection
within the eye wall.
The resultant convection within the eye wall serves to balance the
horizontal convergence of angular momentum through vertical advection.
The upward motion in the eye wall results in outflow aloft, and high
perturbation pressure over the center of the storm.
The upper-level pressure perturbation, combined with the low-level
divergence within the eye itself, results in compensating subsidence in the
eye.
Due to the strong inertial stability in the eye wall, the strongest compensating
subsidence is found on the inside of the radius of maximum winds.
This results in an abrupt clearing just inside the eye wall.
As the air subsides, the resultant compressional warming actually works against
subsidence (through buoyancy).
All that is needed in order to keep the eye relatively clear is enough subsidence
to balance the buoyancy.
Thus, in the steady state, there doesn’t have to be strong subsidence in the
eye in order to maintain the eye.
The maintenance of the eyewall is a balance between the horizontal and vertical
advection of angular momentum. The radius of the eye can expand or contract
depending on this balance.
13
Eye wall replacement usually results in a lowering of the intensity, since the new
eye wall is at a larger radius, and from angular momentum arguments, would
have a lower wind speed.
Intensity may build again one the new eye wall contracts.
The causes and dynamics of eye wall replacement cycles are not well known.
Eye wall replacement tends to occur in very intense cyclones.
RAINBANDS
Rain bands are bands of convection that occur in the core region, and also can
extend into the outer region of the cyclone.
Rain bands can be tens of kilometers wide, and hundreds of kilometers in length.
Convection often forms on the inside of the rain bands and moves up and
forward through them.
The region between the rain bands is characterized by subsidence.
Rain bands may be of three types:
Moving spiral bands – These bands appear to rotate with the circulation.
Convective rings – A rain band that completely encloses the cyclone.
A convective ring may transition into an eye wall
Principle spiral bands – A non-moving rain band that wraps into the cyclone
on the east side of the cyclone, and feeds lower latitude air into the vortex.
In a rough sense, principle spiral bands separate the inner region of the
cyclone from the outer regions.
In the inner region, air makes several circuits as it spirals into the core,
where in the outer region the air doesn’t make it to the core.
14
In the outer region the mass field adjusts to the wind field.
The tangential circulation can extend 1000 – 2000 km from the center of the
storm.
The radial circulation is much smaller in scale (~600 km).
There is usually subsidence in the outer region, with areas of weak upward
motion.
A typical ring-like area of subsidence is found around 5° latitude or so from the
center, and is sometimes referred to as the moat, since it results in clearing.
OUTFLOW REGION
The intense updrafts in the eye wall carry cyclonic vorticity upwards. This
results in the upper-level outflow near the cyclone center having cyclonic
rotation.
As it spreads outward, it loses positive vorticity and gains negative vorticity, so
that it becomes anticyclonic as it moves away.
The outflow region has weak inertial stability, and so doesn’t resist horizontal
flow.
The outflow region interacts readily with the environmental flow.
The outflow mainly takes place in one or two outflow jets or channels.
These outflow channels are shallow, and are in the upper troposphere.
Cyclones with two outflow channels tend to be more intense than those with a
single outflow channel.
DIURNAL CYCLES
Tropical cyclone convection shows a pronounced diurnal cycle.
Convection is enhanced in the early morning hours (0300 to 0600 local), likely
due to the same reasons that tropical convection over the open oceans has an
early morning maximum.
Areal coverage in cirrus clouds is maximum in the late afternoon (around 1800
local time).
15
ESCI 344 – Tropical Meteorology
Lesson 11 – Tropical Cyclones: Formation, Maintenance, and Intensification
GENERAL CONSIDERATIONS
Tropical convection acts as a heat engine, taking warm moist air from the
surface and converting the latent heat into kinetic energy in the updraft, which is
then exhausted into the upper troposphere.
If the circulation can overcome the dissipating effects of friction it can become
self-sustaining.
In order for a convective cloud cluster to result in pressure falls at the surface,
there must be a net removal of mass from the air column (net vertically
integrated divergence).
Since there is compensating subsidence nearby, outside of a typical
convective cloud, there really isn’t much integrated mass divergence.
Pressure really won’t fall unless there is a mechanism to remove the mass
that is exhausted well away from the convection.
Compensating subsidence near the convection also serves to decrease the
buoyancy within the clouds, because the subsiding air will also warm. This
reduces the temperature difference between the in-cloud and outside air.
This illustrates the importance of upper-level outflow in tropical cyclones.
Even if there are pressure falls at the surface, the atmosphere could just
generate gravity waves and readjust back to the original pressure, unless the
Rossby radius of deformation is small enough to allow a residual circulation
to form.
Frictional effects will work against formation as well.
The mechanisms by which a tropical cloud cluster can overcome the above
effects and develop a sustained circulation are complex, and much is still
unknown.
Some literature refers to these mechanisms as providing a ‘crank-start’ for
the tropical cyclone, in analogy to starting internal combustion engine.
2
Most studies of CISK have involved a linear approximation in which the
different wavelengths, or modes, do not interact and exchange energy.
It is also possible that the non-linear interaction of the wave modes is
important, which would allow energy that is input into one mode to be
transferred into other modes that do have a horizontal scale closer to their
Rossby radii of deformation.
Although CISK has trouble explaining how cloud clusters form into a tropical
cyclone, it can be used to explain how a tropical cyclone, once formed,
intensifies, since as the vortex becomes stronger the Rossby radius of
deformation in the core can become small enough to allow a mass-field
disturbance to force a velocity field adjustment.
3
In the
he diagram the maximum in vorticity is in the middle, with decreasing
intensity above and below.
Since the MCV is in thermal wind balance there is a cold anomaly below, and
a warm anomaly
aly above the vorticity maximum.
Mesoscale convective vortices often form in and are associated with mesoscale
mes
convective systems (MCS)
MCS).
An idealized mesoscale convective system contains two regions:
A convective region where new convective cells are formed and progress to
maturity
In the convective region there is generally low to mid-level
level convergence,
with divergence aloft.
A stratiform region formed from the remnants of the old cells from the
convective region.
In the stratiform region there is generally mid
mid-level
level convergence, with
low-level
level and upper
upper-level divergence.
The three recent theories all attempt to explain how an MCV associated with an
MCS can result in the
he formation and/or amplification of a vortex at the surface,
which can then further develop into a tropical cyclone.
The three theories are broadly classified as: Top-down Merger; Top-down
Top
Showerhead; and Bottom
ottom-up Development
4
TOP-DOWN MERGER THEORY FOR DEVELOPMENT
In the Top-down Merger theory, two or more smaller mid-level vortices merge
into a larger mid-level vortex. The influence of this larger vortex is then felt
through a deeper depth of the atmosphere, influencing development at the
surface.
The depth or vertical thickness to which a vortex penetrates is given by the
Rossby penetration depth,
DR = ω L N (1)
where ω is the inertial frequency, L is the horizontal scale of the vortex, and N is
the Brunt-Vaisala frequency.
The larger in horizontal extent, or the more intense a vortex is (more intense
vortices have higher inertial frequencies), the deeper the Rossby penetration
depth.
If two or more smaller, mid-level vortices merge then they will form a larger
vortex that has a deeper Rossby penetration depth, and can therefore build
down to lower levels.
5
BOTTOM-UP DEVELOPMENT THEORY FOR DEVELOPMENT
In the theory of Bottom-up Development a low-level potential vorticity (PV)
anomaly produced from a separate convective updraft moves underneath the
MCV.
The interaction between the PV anomaly and the MCV shifts the profile of
divergence and convergence such that there is enhanced low-level convergence
and spin-up of a low-level vortex, at the expense of the original MCV.
6
Once the vortex is strong enough to maintain itself on its own it breaks away
from the parent wave.
The marsupial hypothesis gets its name from an analogy to how a marsupial
such as a kangaroo rears its young in a protective pouch until the young are
ready to fend for themselves.
MAINTAINING INTENSITY
The coupling between the strong winds in the boundary layer and the transfer of
latent and sensible heat into the air as it spirals into the cyclone.
Frictional convergence of the tangential winds plays a key role in
transporting latent heat into the core of the vortex.
The stronger the winds, the greater the frictional convergence into the core.
Also, the stronger the winds, the greater the transfer of latent and sensible
heat from the sea surface into the inflowing air.
The formation of an anticyclone aloft by the heating aids in the upper-level mass
divergence which is necessary to sustain and intensify the cyclone.
Tropical cyclones can maintain themselves as long as there is sufficient inflow of
warm, moist air into the cyclone, and there is adequate outflow aloft. Factors
that can cause a tropical cyclone to fluctuate in intensity (either up or down) are:
Variations in SST.
Interaction with land, which can result in less evaporation and latent heat
inflow.
Tropical cyclones can momentarily increase in intensity as they make
landfall due to enhanced low-level convergence due to the increased
friction over land.
Enhanced or suppressed outflow.
Increased vertical shear.
Vertical shear can cause the upper-level anticyclone and outflow to
decouple from the low-level inflow. If the outflow is weakened, then the
mass that is ejected into the upper troposphere by the convection can
subside in the vicinity of the tropical cyclone and weaken the convection,
as well as limit the surface pressure falls.
7
SYNOPTIC SCALE INFLUENCES
Since tropical cyclone formation requires a pre-existing cyclonic disturbance,
they normally form in
Monsoon troughs
Tropical waves
Old frontal zones or shear lines
Any enhancement of vorticity is favorable for formation.
A cold surge in the wintertime hemisphere often will enhance the Equatorial
westerlies, and is favorable for formation.
Existing storms can influence the formation and development of new storms.
For storms that have not yet recurved, the path ahead of the storm is
unfavorable for formation of another tropical cyclone, while the wake area
behind it is favorable.
This is due to the large scale vertical motion pattern forced by the
tropical cyclone.
The outflow from an existing storm can also sometimes provide too much
shear over a region and suppress formation and development.
Formation of upper-tropospheric outflow jets is key for development, since the
upper-level mass divergence must be larger than the lower-level mass
convergence.
A north-east quadrant outflow jet is enhanced by linking with
the subtropical jet
a tropical upper-tropospheric trough (TUTT)
a deep mid-latitude trough
an upper-level cold low
The position of the cyclone in relation to the above features is crucial for an
enhanced north-east outflow jet. Although they enhance outflow, which aids
deepening of the surface low, if they are too close then the increased westerly
shear can inhibit growth.
A south-west quadrant outflow jet is enhanced by an intense upper-level
anticyclone in the opposite hemisphere.
8
ESCI 344 – Tropical Meteorology
Lesson 12 – Tropical Cyclones: Motion and Analysis
GENERAL
Tropical cyclone motion can be thought of as the cumulative effect of the
following three influences:
Environmental steering
The beta effect
Asymmetrical convection
Of the three, environmental steering is usually of primary importance, followed
by the beta effect.
Asymmetrical convection is mainly responsible for short term “eye wobble”,
but is of little importance for long term motion.
ENVIRONMENTAL STEERING
To a first approximation, environmental steering can be thought of as “cork-in-
a-stream”, with the tropical cyclone merely advected by the mean environmental
flow.
The environmental flow can be represented by various means, either through a
single level, or through a mass-weighted mean flow.
A deep-layer mean is best for intense, mature cyclones.
A medium-layer mean or even shallow-layer mean is more suited to weak
systems, especially if they are highly sheared.
In any case, the cyclone circulation needs to be removed from the wind field, in
order to determine the environmental flow.
Synoptic-scale influences are very important for the environmental steering
currents.
An approaching trough can alter the steering current such that the cyclone
recurves, or at least tracks more northerly.
A strong subtropical ridge will keep a cyclone entrenched in the trade winds,
and tracking westward.
An adjacent tropical cyclone can influence the steering flow.
This is called the Fujiwhara effect, and can result in some unusual tracks as
the cyclones spiral around each other, occasionally even merging.
where u is the speed of mean zonal flow, and k and l are the zonal and
meridional wave numbers.
The dispersion relation (equation 2) shows two important characteristics of
Rossby waves:
They will move westward in the absence of a mean flow.
They are dispersive, with the shorter waves moving more slowly than the
longer waves.
If a tropical cyclone is thought of as the superposition of several different linear
Rossby waves, then in the absence of a mean flow the cyclone would be expected
to move westward.
This tendency for the cyclone to move westward in the absence of a mean flow is
known as the linear beta effect.
2
NONLINEAR BETA EFFECT
If the Rossby waves comprising the cyclone were truly linear, there would be no
interaction between the different wavelengths, and the waves would just pass
through each other, with the longer waves moving faster.
However, if nonlinear interactions are allowed, the resultant motion of the
cyclone is not only westward, but also poleward.
This is known as the nonlinear beta effect.
The nonlinear beta effect can be explained somewhat qualitatively by imagining
that the cyclone is comprised of two Rossby waves.
The outer cyclone is represented by a long wavelength (small wave number)
wave while the inner cyclone is represented by a short wavelength (large
wave number) wave.
We will imagine that initially the cyclone is symmetric, and the vorticity
isopleths will be concentric with the streamfunction contours (see left side of
diagram below).
There will initially be no advection of relative vorticity.
The cyclone will move westward due to the advection of planetary vorticity.
Since Rossby waves are dispersive, at some time later the outer Rossby wave will
have traveled farther westward than the inner Rossby wave.
The cyclone is no longer symmetric, and the vorticity maximum will no
longer be in the center of the cyclone (see right side of diagram).
The result of the dispersion is that there is now positive advection of relative
vorticity poleward of the cyclone, which will cause the cyclone to move poleward.
3
The net result of the nonlinear beta effect is a westward and poleward cyclone
track in the absence of a mean flow.
Some qualities of the nonlinear beta effect are:
It is the size and strength of the outer region that are most important for
determining the beta effect.
Larger cyclones will have a stronger beta effect.
The intensity of the inner core has little influence on the beta effect.
The beta effect acts in addition to the mean flow steering.
The interaction between the beta effect and the mean flow steering may not
be linear.
ASYMMETRICAL CONVECTION
Though the inner core of a tropical cyclone has large inertial stability, there are
asymmetries in convection.
The asymmetrical convection can be due to several factors, including
SST gradients
Differential stress (different surface roughness, especially near land).
The asymmetrical convection can lead to local pressure falls in the eyewall,
which essentially can displace the eye in the direction of the pressure falls.
This leads to short-term “eye wobble” often seen in tropical cyclone tracks.
This eye-wobble is one reason why the longer term tropical cyclone motion
should not be estimated from two eye positions, but should be based on a longer
term average of several positions or fixes.
4
ANALYSIS
One important key for a successful prediction of the cyclone track and intensity
is an accurate analysis of the current position and intensity of the cyclone.
Location is determines primarily by satellite imagery, except near land, when
radar and aircraft can be brought to bear.
Synoptic fixes (positioning based on adjacent synoptic observations) was the
staple of location techniques in the pre-satellite era, but is now a “quaint”
pastime for bored analysts.
Satellite imagery can give a very accurate fix when there is an eye, or an easily
visible, exposed low-level circulation center.
Positions based on curved cloud features or when the eye is obscured by a
central dense overcast (CDO) are less accurate.
The best intensity estimate is a direct observation from an aircraft penetration.
The aircraft can not only measure flight-level winds, which can be
extrapolated to surface winds, but pressure measurements can also be made
using drop sondes.
If the cyclone is close enough to land, Doppler radar observations can be used to
estimate intensity.
In the absence of aircraft observations or Doppler radar observations, the
Dvorak technique is the staple for determining intensity, and is described below.
DVORAK ANALYSIS
The Dvorak technique is based on appearance of the storm from IR and visible
satellite imagery.
The Dvorak technique assigns a T-number based on the appearance from the
satellite imagery.
The T-number is itself a combination of a CF (central feature) number and a
BF (banding feature) number.
The “T-number”, along with some rules which are formulated to avoid wildly
differing intensity estimates from one observation to the next, are used to assign
the current intensity, or CI.
5
The CI is usually close to the T-number for developing storms, but is higher for
weakening storms.
The assumption is the cloud features dissipate faster than the momentum of
the circulation.
The relation between CI and intensity is shown in the table below.
CI 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
Intensity 25 25 30 35 45 55 65 77 90 102 115 127 140 155 170
(knots)
A CI of 2.5 implies a tropical storm, while a CI of 4.0 implies a
hurricane/typhoon.
TRACK PREDICTION
There are several different methods for predicting tropical cyclone tracks.
The historical progression of forecasting techniques began with persistence,
which is useful in the first few hours of the forecast.
Later, methods based on climatology were developed, such as finding analogs, or
past storms that had similar characteristics and synoptic environments.
Statistical methods use regression techniques to correlate information about the
storm and its environment with the likely track.
The earliest successful prediction technique was a statistical blend of
persistence and climatology, called CLIPER.
CLIPER had no dynamical input, and yet was very successful. It is often
used as the reference by which new methods are judged. If they can’t do
better than CLIPER, then they aren’t worth the effort.
Those statistical techniques that use dynamical properties as independent
variables are known as statistical-dynamical methods.
As understanding of the physical processes affecting motion was gained, pure
dynamical methods were developed.
Examples of the dynamical methods are
Numerical models
BAM (Beta plus Advection Models) which used output from numerical
models to define the steering flow, to which calculations of the beta effect
were added.
6
FBAM, MBAM, and SBAM stood for full BAM, medium BAM and
shallow BAM, and denoted whether a deep-layer, medium-layer, or
shallow-layer mean was used for the steering flow.
Output tracks from numerical models are also ensembled, in a sense creating a
type of statistical dynamical method.
7
ESCI 344 – Tropical Meteorology
Lesson 13 – Monsoons
GENERAL
• Monsoon comes from “mausim,” which is Arabic for season.
• There are a number of different definitions.
• Ramage’s definition
o Prevailing wind directions shifts by at least 120°° between January and July.
o Prevailing wind direction persists at least 40% of the time in January and
July.
o Mean wind speed exceeds 3 m/s in either January or July.
o Fewer than one cyclone-anticyclone alternation every 2 years in either
January or July.
• According to Ramage’s definition, only Asia, Australia, and Africa have distinct
monsoons.
• A more relaxed definition commonly used:
o Wind must reverse in direction between summer and winter.
o Summer season must be very wet, and winter season very dry.
• Using this definition, North America also experiences a monsoon (in the
Southwest U. S. and Northern Mexico.)
• Using the relaxed definition, over ½ of the tropics and ¼ of entire globe
experience monsoon-type climates.
• Three main driving mechanisms for the monsoon:
o Differential heating of land and ocean
o Moist processes
o Rotation of the Earth
2
o Moist land acts more like ocean.
o If land is dry, rising motion will occur closer to the coast, since the land will
be very warm.
o As land becomes wet from precipitation, the rising motion will move inland
over drier land.
o Precipitation will progress inland, allowing coastal area to dry out. Cycle
will then repeat itself.
o This is one factor in monsoon variability and monsoon “breaks.”
LAND/OCEAN GEOMETRY
• Latitude and orientation of land masses effect monsoon circulation, as shown in
figure below, in which grey areas are land.
• Areas on or near Equator do not experience much rotation.
• Uniform surface (either all land or all water) result more in trade-wind trough
ITCZ, and not monsoon (e.g., Central Pacific Ocean).
• Western boundary results in atmospheric jet (e.g., East African Low-level Jet).
3
After “Physics of Monsoons: The Current View,” J.L. Young, in Monsoons (Fein, ed)