Diapositiva 1

1
The Intra-Americas Seas Low Level JetAn
Analysis of Atmospheric Sounding
Observations Jorge A. Amador1, Erick Rivera1,
Marcela Ulate1 andChidong Zhang2 1 Center for
Geophysical Research, University of Costa Rica2
RSMAS, University of Miami30th CDPW October 2005
2
The IAS (Intra-Americas Seas) is defined here as
consisting of the Caribbean Sea (CS), the Gulf of
Mexico (GM) and the eastern tropical Pacific
adjacent to Central America (etPac). The IALLJ
(Intra-Americas Low level Jet) is an outstanding
climate feature of the IAS The trade wind
enters the Caribbean intensifying and forming the
IALLJ, whose core is near 15 N at the 925 hPa
level (Amador, 1998)
3
Stensrud 1996 (LLJsdefinitions/ review
4
(No Transcript)
5
Maul G.A, 1993. Climate Change in the
Intra-Americas Sea Implications.., Eadward
Arnold, London, 389 pp.
6
Kalnay et al 1996
7
Austral summer IALLJ bifurcates (mainly) into
two branches. - central branch (CALLJ) crosses
CA and penetrates into the etPac - southern
branch goes into SA directly from the Atlantic
Ocean veering southwards and connecting with the
SALLJ along the eastern slopes of the
Andes Late boreal spring and summer - central
branch is located near the same place as in
austral summer but extends upwards to 850
mb. - northern branch separates from the central
branch and veers toward the GM where it
connects with the Great Planes Low Level Jet
(GPLLJ)
8
(No Transcript)
9
Amador and Mo, J. Climate, submitted Oct. 2005
(P4.9)
10
The strength of the southern branch peaks in
February and the northern branch peaks in
July The central branch exhibits a semi-annual
cycle with maxima in February and July. The
strength of this branch varies with ENSO phases
(Amador et al 2003, Amador and Mo P4.9) such
as Weaker (stronger) than normal wind
velocities at 925 hPa tend to occur at a warm
(cold) ENSO phase during winter, and the opposite
occurs during summer
11
Amador and Mo, J. Climate, submitted Oct. 2005
(P4.9)
12
(No Transcript)
13
Influence of the IALLJ Strong surface winds
associated with this low-level jet offshore the
Gulf of Papagayo influence SST over the etPac
(Xie et al 2005) and the Costa Rica dome
dynamics. Cooling due to coastal upwelling
induced by the CALLJ might contribute to the
semi-arid regions of northern Venezuela and the
Netherlands Antilles (Granger 1985). The IALLJ
plays an important role in modulating summer
precipitation over the Caribbean region (Amador
1998, Amador and Magaña 1999). Classical
dynamical concept of a low level easterly jet
holds.
14
Sea surface temperature image showing the cold
upwelling areas associated with the Papagayo
winds. Wind vectors are from the ERS-1
scatterometer. Image courtesy of D. Ballestero
and E. Coen, Universidad Nacional, Costa
Rica. daacuso_at_daac.gsfc.nasa.gov
Effect on the coastal circulation estimated
long-term mean kinetic energy (courtesy of A.
Trasvina) trasvi_at_cicese.mx)
15
(No Transcript)
16
entrance
exit
conv
div
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
CALLJ
22
CALLJ
23
CALLJ
24
CALLJ
25
CALLJ
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
New Central American topographic map derived from
Shuttle Radar Topography Mission (SRTM) data. The
Papagayo wind blows across the low elevations
near Lake Nicaragua from the Caribbean Sea. Map
courtesy of NASA Jet Propulsion Laboratory
32
Convective activity over the Caribbean might be
reduced due to an increased low level vertical
wind shear during summer (Amador et al., 2000,
Amador and Mo P4.9) with less favorable
conditions during warm ENSO years. The location
and strength of the easterly winds influence the
rainfall pattern over southern Mexico, Central
America and the etPac. The peak of the CALLJ in
July might be related to the mid-summer drought
(MSD) in part of Central America (Magaña et al.
1999 Magaña and Caetano 2005).
33
Dr. A. S. Oersted, Danish naturalist came
to Costa Rica in 1846 and provided the
first quantitative evidence of the veranillo or
MSD in Central Valley (1300 m.a.s.l.)
Veranillo (MSD)
?T()
1 15 May-5 June 2 16 June-15
July 3 16 July-5 August 4 16
August-15 September 5 16 September-15
October 6 16 October-15 November 8
16 November-15 December
34
(Pacific slope)
MSD
I
years
months
35
Water vapor transported by the IALLJ from the
IAS into South and North Americas is also vital
to precipitation there. In austral summer, the
southern branch of the IALLJ supplies moisture to
rainfall associated with the South American
Monsoon. In boreal summer, the northern branch
of the IALLJ, together with the GPLLJ,
contributes to the moisture transport that
supplies moisture to rainfall in the central
United States (Ropelewski and Yarosh 1998). The
strength of the CALLJ and GPLLJ and rainfall over
the central United States reaches their maxima in
June-July (Mo and Berbery 2004). Almost one-third
of all the moisture that enters the continental
US is transported by the GPLLJ (Helfand and
Schubert 1995).
36
The role of the IAS as a water vapor source for
summer rainfall in North America has been studied
by many (e.g., Hastenrath 1966 Rasmusson 1967,
1968, 1971 Brubaker et al 2001 Mestas-Nuñez et
al. 2005). The low-level jet serves not only as a
conveyer belt for moisture, but also as a
moisture collector that modulates surface
evaporation. The signature of the IALLJ at the
surface is well captured by satellite
scatterometer wind data (Quikscat). Evaporation
can be substantially enhanced by the IALLJ, with
its strong wind speed extending from its core at
the 925 hPa level to the surface. Moisture
transport from the IAS into the central US in
boreal summer varies on inter-annual timescales
independently of the ENSO (Hu and Feng 2001).
Stronger CALLJ/GPLLJ and their associated
moisture transport lead to positive rainfall
anomalies over the central United States (Mo et
al. 1997), which is associated with the negative
rainfall anomalies along the Gulf Coast and East
Coast (Higgins et al. 1997). This inter-annual
variability of moisture transport is often
related to extreme rainfall events in the central
US (Paegle et al. 1996 Trenberth and Guillemot
1996 Mo et al. 1997). 38
37
(No Transcript)
38
The importance of the moisture transport by the
low-level jet is further underlined by the
finding that the largest discrepancies in
rainfall estimates over US during spring and
summer between the NCEP and NASA global
reanalyses are directly related to the
differences in their low-level jets (Mo and
Higgins 1996). A composite of precipitation
based on the NCEP Eta regional reanalysis
(Mesinger et. al. 2004) shows the enhanced
rainfall pattern over the United States
associated with strengthened IALLJ. It also shows
negative rainfall anomalies covering the Gulf of
Mexico and the Caribbean. The composite of 925
hPa wind anomalies indicates that the branch of
moisture transport from the Caribbean through the
Gulf of Mexico to the Great Plains strengthens.
Over the Caribbean, strong zonal transport
implies a strong CALLJ.
39
The association between the IALLJ and
precipitation over the Great Plains can also be
established by the composite of precipitation
anomalies based on the magnitude of the CALLJ,
which can be measured by a CALLJ index defined as
the mean zonal wind at 925 hPa averaged over the
area (12.5-17N, 70-80W). The signal over North
America is strongest about 4 days after the CALLJ
index is over 1.2 standard deviations (Amador and
Mo, P4.9).
40
(No Transcript)
41
Most of the above results on the IALLJ are based
on either reanalyses or simulations of global
models. There has not been a systematic
document on the structure and variability solely
based on in situ observations of atmospheric
soundings or PACS data in the IAS region, which
are available for several decades. Despite the
sparse coverage of the sounding sites, data gaps,
and coarse vertical and temporal resolutions and
the fact that most of these soundings are
available to data assimilation for global
reanalyses, a document of the IALLJ serves as
ground truth for quantifying systematic errors
and biases in global model reanalyses and
simulations. It might be reasonable to believe
that global model reanalyses capture the gross
features of the IASLLJ, however, to raise the
credibility of research results using them it is
vitally important to quantify their systematic
biases and errors.
42
Objective to provide a ground truth benchmark
that can be used to quantify systematic biases
and errors of global model reanalyses in their
descriptions of the IALLJ. The general strategy
is, within the limits of available sounding
data To examine the structure and variability
of the IALLJ for its three branches respectively.
Special attention will be paid on the structure
of the central branch along its axis from its
entrance into the Caribbean Sea to its exit over
Central America, the structure of the northern
branch along its axis from the entrance into the
Caribbean Sea to its exit across the Gulf coast
of North America, and the structure of the
southern branch when it enters South America.
The analysis of the mean structure of the IALLJ
using the sounding data is therefore performed in
the context of its seasonal cycle. The
inter-annual variability of the IALLJ is examined
mainly through its composite during recent 10
ENSO warm (El Nino) and 6 cold (La Nina) events.
The intra-seasonal variability of the IALLJ is
examined only during the peak seasons of each
branch.
43
2. Data and Procedures The sounding data used in
this study are from 26 sites in the IAS region
covering from January 1973 through December
2004. They were retrieved from, National Weather
Services, CARDS data set and the archives at
University of Wyoming. At most sites, soundings
are available once a day at 12 or 00 Z and only
at the mandatory pressure levels. During a
limited period, namely, 16 June 6 September
2004, 4 times a day soundings were launched at
San Jose, Costa Rica as part of the North
American Monsoon Experiment (NAME, Higgins et al
2005). This special set of soundings, quality
controlled for a vertical resolution of 25 hPa,
will be referred to as NAME-CR soundings.
Possible biases of the sounding analysis due to
limited temporal and vertical resolutions of the
data will be assessed using the NAME-CR
soundings.
44
(No Transcript)
45
(No Transcript)
46
3. Results Mean seasonal cycles of winds are
shown at (8) selected sites representing
different parts of the three branches of the
IALLJ. Winds at four sites near the central
branch (or the Caribbean low-level jet, CALLJ)
from its entrance into the Caribbean Sea
(Barbados and Point a Pitre) across the Sea
(Curazao) to its exit (San Andres) are shown.
Based on global reanalysis data, these sites are
on the southern side of the CALLJ core in boreal
summer. On the poleward side of the central
branch Grand Cayman was selected. It is known
from reanalysis that the CALLJ undergoes a strong
seasonal cycle with two peaks one in boreal
winter and the other in boreal summer. The
sounding data confirm this. In addition to the
two peak in the strength of the CALLJ, another
feature in its seasonal cycle is the change in
its vertical extent. The CALLJ exists all the
time at the 800 hPa level but expands upward
during boreal summer. In consequence, while the
core of the CALLJ remains at 800 hPa all year
around, the CALLJ becomes thicker during boreal
summer.
47
(No Transcript)
48
(No Transcript)
49
R2 There is an interesting east-west
inhomogeneity in the seasonal cycle in the CALLJ.
Its primary peak in amplitude occurs in July at
Barbados but in January-February at San Jose,
where the low-level jet becomes a near surface
jet because of the topography (shown later). In
between (San Andres), the two peaks of the CALLJ
are comparable. The CALLJ is dominated by the
zonal wind at both ends (Barbados, San Andres and
San Jose) but has strong northerly components in
between in boreal winter. It seems that the
vertical momentum mixing from the jet core to the
surface is more efficient inside the Caribbean
Sea (e.g., at San Andres) than outside (at
Barbados), leading to a stronger surface easterly
at San Andres than at Barbados. Reasons for these
east-west discrepancies in the seasonal cycle and
structure of the CALLJ need further
investigations. 53
50
Aloft in the upper troposphere, the seasonal
cycle is dominated by strong westerlies
(Barbados) and southwesterlies (San Andres)
within a layer of 300 100 hPa during boreal
winter and spring. The amplitude of the
upper-tropospheric westerlies increases toward
the east in this longitudinal sector. The lower
and upper tropospheric winds at these sites are
out of phase in their seasonal cycles. In this
sense, there appears to be no classic zonal or
meridional vertical overturning circulation at
these sites that tends to vary seasonally in
coherence vertically with an identifiable heating
source. In other words, the CALLJ and the upper
tropospheric winds are likely to be driven by
different mechanisms.
51
(No Transcript)
52
(No Transcript)
53
(No Transcript)
54

• R3
• Interannual variability (ENSO)
• 30 years sufficient?
• Intraseasonal and synoptic scale variability
• PACS
• Possible biases
• Shorter time scales
• ECAC data (twice daily observations 10-23 July
  2001)
• NAME-CR soundings

55
CALLJ
B
boreal summer ENSO (W-C) lt O
C
SA
56
Northern Branch Yucatan
?
Southern Branch Cayenne
57
PACS Managua May-August1998
58
ECAC 2001 (Warm Pools Climatic Experiment),
UNAM, UCR, (IAI-CRN 073)
59
ECAC 2001 (Warm Pools Climatic Experiment),
UNAM, UCR, (IAI-CRN 073)
SE
SSE
E
SSE
E
NE
SE
SE
E
NE
NNE
E
60
Diurnal cycle of the zonal wind over Costa Rica
at 00 Z (red), 06 Z (black), 12 Z (blue), 18 Z
(green) and the mean zonal wind (purple) during
August 2004 (NAME04).
61
Future work!!! Inter-annual variability --- 30
years sufficient? 50 years? - monthly
correlation with Nino SST (3?) for the peak
season at all pressure levels at selected sites
(or at all sites plotted on the site map) with
significant levels - composite of the three
branches during their peak seasons for
warm-normal and cold-normal. Intraseasonal
variability - peak season time series -
correlation with local precipitation? Possible
biases ECAC data (twice daily
observations) NAME-CR soundings plus one month
2005 4xdaily soundings - daily mean using all
data vs. 12Z and 00Z - full vertical resolutions
vs mandatory pressure levels only
62
Acknowledgements UCR Kingtse Mo and Chidong
Zhang for their interest and encouragement
during the course of this work NAME
63
Gracias!