NOAA's%2033rd%20Climate%20Diagnostics%20and%20Prediction%20Workshop/CLIVAR%20Drought%20Workshop%20(CDPW)%20October%2020-24,%202008%20in%20Lincoln,%20Nebraska

1
Hydrodynamics of the Caribbean Low-level Jet and
its Relationship to Drought K.H. Cook and
E.K. Vizy Jackson School of Geosciences University
of Texas at Austin
NOAA's 33rd Climate Diagnostics and Prediction
Workshop/CLIVAR Drought Workshop (CDPW)October
20-24, 2008 in Lincoln, Nebraska
2
Seen in the NARR domain, the CLLJ is a prominent
feature, with the highest low-level zonal wind
speeds in the domain July, 925 hPa, m/s
Caribbean low-level jet (CLLJ)
3
Zonal Wind Speed July, 925 hPa, m/s
Cuba
Hispaniola
Puerto Rico
Venezuela
Columbia
4
Seasonality of the CLLJ
Zonal wind speed (m/s)
Month
U wind averaged 80-60W and 12-18N, 925 hPa
5
Seasonality of the CLLJ
0 m/s to -10 m/s (easterly)
Zonal wind speed (m/s)
Month
U wind averaged 80-60W and 12-18N, 925 hPa
6
Seasonality of the CLLJ
Zonal wind speed (m/s)
Strong in DJF
Strongest in June/July
Month
U wind averaged 80-60W and 12-18N, 925 hPa
7
Seasonality of the CLLJ
Weakest in October
Zonal wind speed (m/s)
Weak in May
Month
U wind averaged 80-60W and 12-18N, 925 hPa
8
October through April CLLJ largely disconnected
from the flow (and moisture) entering the US and
northern Mexico
May through September Caribbean and Great
Plains low-level jet system
9
There are also significant seasonal differences
in the elevation and vertical structure of the
CLLJ.
10
Zonal wind cross section in latitude and height
at 70W
January
February
April
March
May
June
11
Zonal wind cross section in latitude and height
at 70W
July
August
September
October
December
November
Note that the minimum in Sept/Oct at 925 hPa
shown previously is partly due to this elevation
change in boreal fall
12
Diurnal Cycle July 925 hPa
Daytime CLLJ 2m/s weaker than the nighttime
jet. Secondary minimum at 4 AM LT
Local time at CLLJ region GMT 5 hr
13
Purpose
What controls the CLLJ on seasonal and
diurnal time scales? What is the relationship
of the CLLJ to drought
. over the central U.S.?
. over central
America and southern Mexico?
14
Dynamics of the CLLJ

• Examine the horizontal momentum balance to
  understand
• Diurnal cycle of the jet
• Differences between seasons

Start with July .
15
(No Transcript)
16
First-order v-momentum balance
geostrophy
17
Note the diurnal cycle of the height gradient,
weakening through the night and building up
during the day.
18
Afternoon 4 PM LT
Late Night 4 AM LT
The diurnal cycle of the meridional height
gradient is caused by the diurnal cycle of
heating over far northern South America.
19
This diurnal cycle in height gradient does not
directly generate the CLLJs diurnal cycle. The
diurnal cycle of the jet has a different shape
and the jet weakens during the day. Instead, the
diurnal heating and cooling over northern South
America generates meridional acceleration.
20
Daytime Land warms and geopotential heights
fall. This strengthens the meridional
geopotential height gradients and leads to
onshore flow (southward, negative local
acceleration). Night Offshore flow and
northward (positive) acceleration. This
solenoid circulation sets up because of the space
scales involved and because of the low latitude
Coriolis accelerations are relatively weak.
21
July, u-momentum equation
22
The July u-momentum balance is more complicated
than the v-momentum balance
The only small terms are meridional and
vertical advection.
23
Strong zonal geopotential height gradients
associated with low heights over Central
America/ southern Mexico to the west and the
NASH to the east.
24
Strong westward frictional accelerations
associated with the easterly jet.
25
jet decelerates
Should have the weakest jet speeds at 3 PM, and
a secondary minima at 3 AM. (We saw this
above.)
26
jet decelerates
The minimum at 3 PM occurs mainly because of
the Coriolis force, but also because the
zonal pressure gradient force weakens.
The northerly off-shore flow anomaly is strongest
at 10 AM, after which it begins to weaken.
27
jet decelerates
The minimum at 3 AM is related to a weakening
of the zonal pressure gradient force.
28
How is the momentum balance different in other
months, and what does this tell us about how the
seasonal cycle works?
29
Seasonality of the CLLJ
Magnitude CLLJ
Magnitude GPLLJ
Jet has its maximum in July (which we just
examined), and has minima in May and Oct, with a
secondary max in December/January
30
V-momentum balance is about the same as in July,
i.e., to first order geostrophic. So we need to
understand the seasonality of meridional height
gradients.
31
For example, why are the meridional geopotential
height gradients similar in magnitude and,
especially, sign in January and July, and weak in
May and October?
32
(No Transcript)
33
Even though the CLLJ is strong in both January
and July, it has very different downstream
connections in those months.
34
October through April CLLJ largely disconnected
from the flow (and moisture) entering the US and
northern Mexico
May through September Caribbean and Great
Plains low-level jet system
35
The u-momentum equation is different because the
Coriolis force in January is negative, while it
was (mostly) positive in July.
36
925 hPa geopotential heights and winds
MJJAS
ONDJFMA
V gt 0
V lt 0
37
925 hPa geopotential heights and winds
Difference
38
(No Transcript)
39

• What is the relationship between the CLLJ and
  drought?

40
1. Is the CLLJ anomalous during dry years in the
central US and/or Central America?

• Is precipitation anomalous during
• strong or weak CLLJ years?

41
Examine the NARR for 1978-1996 to identify dry
years in these regions
Central U.S, 100-90W and 30-50N, for JJA 1980,
1988, 2006 are anomalously dry
42
In June 1988, the CLLJ is quite weak at ½ its
climatological value. But not In 1980 and 2006.
43
In July 1980 the CLLJ is weak by ¼ of its
climatological value. But not in 1988 or 2006
its even strong in 2006.
44
In August, the CLLJ is weak in 1980 and 1988, but
strong in 2006. We dont find a consistent
relationship between the CLLJ and the 3 years of
central US drought.
45
For S. Mexico/Central America?
Only one year (2003) is identifiable as a drought
year in the NARR.
46
1. Is the CLLJ anomalous during dry years in the
central US and/or Central America?

• Is precipitation anomalous during
• strong or weak CLLJ years?

47
Define a CLLJ Index 925 hPa u JJA 75 - 70W and
12 - 15N
48
Standard Deviations from the 1979-2006 JJA mean
zonal wind
Weaker CLLJ
Stronger CLLJ
49
Drought Years in the U.S.
Weaker CLLJ
Stronger CLLJ
50
Drought Year in Central America
Weaker CLLJ
Stronger CLLJ
51
Correlation between the CLLJ strength and JJA
rainfall
52
Correlation between the CLLJ strength and JJA
rainfall
Positive correlation at the 99 CI
Negative correlation at the 99 CI
53
Correlation between the CLLJ strength and JJA
rainfall
Strongest, most coherent responses over land are
along the Pacific coast of southern Latin
America, over the Caribbean Islands, and the
eastern Yucatan.
54
Correlation between the CLLJ strength and JJA
rainfall
There is not a significant correlation between
the summer CLLJ strength and summer rainfall
over the Central US
55
Maybe we need to look at smaller timescales to
better understand .
56
Define Strong and Weak CLLJ events on monthly
timescale
Monthly CLLJ Standardized Anomaly
57
Correlation between the monthly CLLJ index and
monthly rainfall anomalies
Strong positive correlation between springtime
rainfall in the southeastern U.S. and the
strength of the CLLJ
58
May precipitation (shaded mm/day) Vertically
integrated water vapor transport (x 105 kg/m)
40 increase
29 decrease
59
May 925 hPa Heights (m) Wind (m/s)
60
Correlation of monthly CLLJ strength to monthly
rainfall anomalies
Strength of CLLJ and Central American
rainfall are negatively correlated during the
months when the CLLJ and the GPLLJ are
connected (MJJAS) Hypothesis Dry conditions
in Central America lead to warming and an
enhancement of the zonal geopotential height
gradient, adding to the positive meridional flow
over the Caribbean and further depriving the
region of moisture.
61
Conclusions
A solenoid circulation (land/sea breeze) is
generated by diurnal heating and cooling over
northern South America, and Coriolis forces
acting on this meridional flow contribute to the
diurnal cycle of the CLLJ.
Meridional height gradients and the CLLJ are
strong in January because of low heights over
northern South America, and in July because of
the position of the NASH.
When Central America/southern Mexico warm in the
spring, the resulting zonal height gradients
introduce northerly flow across the western
Caribbean and direct moisture flux into the Gulf
of Mexico.
62
Relationships between the CLLJ and Regional
Rainfall
There is no significant, systematic correlation
between the summer CLLJ strength and summer
rainfall over the Central US (although the
CLLJ was quite weak in the summer of 1988)
There is a strong positive correlation between
springtime rainfall in the southeastern U.S. and
the strength of the CLLJ
Central American rainfall is negatively
correlated with the strength of the CLLJ during
the months when the CLLJ and the GPLLJ are
connected (MJJAS).
63
 The End
64
(No Transcript)
65
(No Transcript)
66
On diurnal time scales - when the height gradient
and Coriolis forces do not balance,
meridional acceleration occurs. Non-linear terms
(advection) and friction approximately balance,
are smaller than the local acceleration term,
and dont explain diurnal variations.
67
(No Transcript)
68

• Close up of the region showing the
• jet is related to the strong meridional
• geopotential height gradients between
• the northern coast of South America and
• the Caribbean Islands (Cuba, Hispaniola,
• Puerto Rico)
• Low over northern South America
• North Atlantic subtropical high, and its
• eastward extension (Bermuda high)
• Geopotential height lines pinch together,
• And the flow associated with these high
• Meridional gradients is the CLLJ

69
Difference between the drought climatology (3
central US drought years) and the 29 year NARR
climatology 900 hPa U JJA CLLJ is a lot more
narrow and weaker in June, not much change in
July in all three months it is anomalously weak
over the Yucatan/eastern Caribbean
70
So how consistent is this response in each of the
3 years identified as drought in the NARR time
period?
71
Diurnal Cycle July 925 hPa
Meridional wind has a very different diurnal
signal. (Hint geostrophic processes are not
controlling diurnal variations.)
72
Coriolis force is playing a role in the diurnal
cycle.
73
Strong westward frictional accelerations
associated with the easterly jet.
Coriolis force is playing a role in Contributing
to
74
The diurnal cycle of the Coriolis force is
related to the land/sea breeze. Note that the
Coriolis force changes sign, consistent with the
land-sea breeze super- imposed over a weak
northward flow. Coriolis and zonal
advection terms are comparable during much of the
day.
75
Positive zonal acceleration, which decelerates
the zonal jet, occurs from 10 PM until 3 AM, and
resumes at about 5 AM until 2 PM. The jet begins
to strengthen after 2 PM until 10 PM
(du/dtlt0), consistent with the max at 10 PM, Due
to decreases in Coriolis (land/sea) Which are
amplified by a decrease in Zonal advection and
then supported after 1 PM by a strengthening of
the zonal height Gradient due to the warming of
the land To the west (Central America).
76
Jet strengthens
77
(No Transcript)
78
Related to sensible heating over central
America In the early spring maybe orographic
effect? Heights fall over central America,
setting up a zonal Geopotential height gradient
and, by geostrophy, Positive (northward)
meridional accelerations
Draw boxes 3D showing balance of forces in
each season In temperature u-momentum equation
maybe it will relate to Why the jet is stronger
in July than in Jan or whatever
79
(No Transcript)
80
(No Transcript)
81
(No Transcript)
82
(No Transcript)
83
(No Transcript)
84
(No Transcript)
85
(No Transcript)
86
(No Transcript)
87
(No Transcript)
88
(No Transcript)
89
(No Transcript)
90
(No Transcript)
91
(No Transcript)
92
(No Transcript)
93
(No Transcript)
94
(No Transcript)
95
(No Transcript)
96
(No Transcript)
97
(No Transcript)
98
(No Transcript)
99
(No Transcript)