Tracking and Mean Structure of Easterly Waves over the Intra-Americas Sea

Yolande L. Serra Institute of Atmospheric Physics, The University of Arizona, Tucson, Arizona

Search for other papers by Yolande L. Serra in
Current site
Google Scholar
PubMed
Close
,
George N. Kiladis NOAA/Earth System Research Laboratory/Physical Sciences Division, Boulder, Colorado

Search for other papers by George N. Kiladis in
Current site
Google Scholar
PubMed
Close
, and
Kevin I. Hodges Environmental Systems Science Centre, University of Reading, Reading, United Kingdom

Search for other papers by Kevin I. Hodges in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

Easterly waves (EWs) are prominent features of the intertropical convergence zone (ITCZ), found in both the Atlantic and Pacific during the Northern Hemisphere summer and fall, where they commonly serve as precursors to hurricanes over both basins. A large proportion of Atlantic EWs are known to form over Africa, but the origin of EWs over the Caribbean and east Pacific in particular has not been established in detail. In this study reanalyses are used to examine the coherence of the large-scale wave signatures and to obtain track statistics and energy conversion terms for EWs across this region. Regression analysis demonstrates that some EW kinematic structures readily propagate between the Atlantic and east Pacific, with the highest correlations observed across Costa Rica and Panama. Track statistics are consistent with this analysis and suggest that some individual waves are maintained as they pass from the Atlantic into the east Pacific, whereas others are generated locally in the Caribbean and east Pacific. Vortex anomalies associated with the waves are observed on the leeward side of the Sierra Madre, propagating northwestward along the coast, consistent with previous modeling studies of the interactions between zonal flow and EWs with model topography similar to the Sierra Madre. An energetics analysis additionally indicates that the Caribbean low-level jet and its extension into the east Pacific—known as the Papagayo jet—are a source of energy for EWs in the region. Two case studies support these statistics, as well as demonstrate the modulation of EW track and storm development location by the MJO.

Corresponding author address: Yolande L. Serra, Department of Atmospheric Sciences, The University of Arizona, 1118 E. 4th St., P.O. Box 210081, Tucson, AZ 85721-0081. Email: serra@atmo.arizona.edu

Abstract

Easterly waves (EWs) are prominent features of the intertropical convergence zone (ITCZ), found in both the Atlantic and Pacific during the Northern Hemisphere summer and fall, where they commonly serve as precursors to hurricanes over both basins. A large proportion of Atlantic EWs are known to form over Africa, but the origin of EWs over the Caribbean and east Pacific in particular has not been established in detail. In this study reanalyses are used to examine the coherence of the large-scale wave signatures and to obtain track statistics and energy conversion terms for EWs across this region. Regression analysis demonstrates that some EW kinematic structures readily propagate between the Atlantic and east Pacific, with the highest correlations observed across Costa Rica and Panama. Track statistics are consistent with this analysis and suggest that some individual waves are maintained as they pass from the Atlantic into the east Pacific, whereas others are generated locally in the Caribbean and east Pacific. Vortex anomalies associated with the waves are observed on the leeward side of the Sierra Madre, propagating northwestward along the coast, consistent with previous modeling studies of the interactions between zonal flow and EWs with model topography similar to the Sierra Madre. An energetics analysis additionally indicates that the Caribbean low-level jet and its extension into the east Pacific—known as the Papagayo jet—are a source of energy for EWs in the region. Two case studies support these statistics, as well as demonstrate the modulation of EW track and storm development location by the MJO.

Corresponding author address: Yolande L. Serra, Department of Atmospheric Sciences, The University of Arizona, 1118 E. 4th St., P.O. Box 210081, Tucson, AZ 85721-0081. Email: serra@atmo.arizona.edu

1. Introduction

Easterly waves (EWs) are prominent features of the intertropical convergence zone (ITCZ), found in both the Atlantic and Pacific during the Northern Hemisphere summer and fall, where they commonly serve as precursors to hurricanes over both basins. In addition to hurricanes, convection coupled to these waves accounts for a significant fraction of the boreal summer precipitation events within the Intra-Americas Sea (IAS) region and influences precipitation patterns throughout the North American monsoon region, including northern Mexico and the southwestern United States.

The fate of easterly waves once they leave the west coast of Africa has been the subject of numerous studies (e.g., Simpson et al. 1968; Carlson 1969; Chang 1970; Frank and Clark 1980; Nitta and Takayabu 1985; Shapiro 1986; Tai and Ogura 1987; Reed et al. 1988b; Lau and Lau 1990; Avila 1991; Thorncroft and Hodges 2001; Hopsch et al. 2007; Berry et al. 2007; Chen et al. 2008; Kerns et al. 2008, 2009). However, there remain several unanswered questions as to the genesis and lysis regions of these waves away from the African continent. Easterly waves tend to diminish in strength over the relatively cool central Atlantic water to the west of a tropical upper-tropospheric trough (TUTT; Carlson 1969); however, they pick up strength in the western Atlantic, possibly through interactions with the subtropical ridge at upper levels in this region. Nevertheless, the wave structure is highly variable across the Atlantic and IAS, where EWs are thought to serve as seeds for more than half of the Atlantic hurricanes and are also believed to be the primary source of east Pacific hurricanes (Avila and Guiney 2000; Avila et al. 2003; Pasch et al. 2009). A lack of both land and ocean observations in the tropics, as well as the inability of global reanalyses to resolve potentially important orographic forcing and other subgrid-scale processes, such as low-level jets, continues to hinder progress in studying the physical processes on synoptic and smaller time and space scales that affect EWs. Numerical weather prediction (NWP) analyses are typically at higher resolution than global reanalyses, making them useful for case studies; however, they suffer from frequent changes in model configuration and data assimilation systems, making climate studies difficult. At the same time, the IAS region is highly affected by the weather events associated with these synoptic-scale waves and would greatly benefit from a better understanding of the processes that maintain and intensify them locally (Amador et al. 2006).

Shapiro (1986) demonstrates that vorticity at both low levels and 200 hPa propagates from the Caribbean into the Pacific for some EW disturbances observed in July 1975. Several other studies also find waves crossing from the Caribbean into the east Pacific (Farfán and Zehnder 1997; Molinari et al. 1997, 2000; Molinari and Vollaro 2000; Berry et al. 2007; Kerns et al. 2008). In other studies, easterly wave activity is identified crossing the IAS region north of 20°N, where they have an influence on the North American monsoon (NAM; e.g., Stensrud et al. 1997; Higgins et al. 2004; Adams and Stensrud 2007; Douglas and Englehart 2007; Johnson et al. 2007). These studies are in contrast to Tai and Ogura (1987), who find little correspondence between the disturbances in the east Pacific and Caribbean using data from May to September 1979. Similarly, a study by Lau and Lau (1990) using reanalyses for the Northern Hemisphere summer 1980–87 period also finds little correspondence between east Pacific and Caribbean disturbances.

Tracking of easterly waves in the IAS region is complicated by their interactions with the Sierra Madre along the west coasts of Mexico and Central America (Zehnder et al. 1999). In addition, the North American monsoon midtropospheric ridge (an extension of the subtropical ridge) and associated circulations and mesoscale convective systems also complicate wave structures in this region (Nitta and Takayabu 1985). The lack of correspondence between the two basins found by Tai and Ogura (1987) and Lau and Lau (1990) may have resulted from their use of 850-hPa winds in their analyses. We will show that while the flow within waves at 850 hPa is not coherent across Central America, the flow above this level does indicate continuity of some disturbances between the two basins.

Over West Africa and the east Atlantic, barotropic and baroclinic instabilities within the African easterly jet (AEJ) contribute to the genesis and intensification of easterly waves in this region (e.g., Norquist et al. 1977; Thorncroft and Hoskins 1994a,b; Hsieh and Cook 2005, 2007; Hall et al. 2006). Some studies suggest that easterly waves intensify in the Caribbean and east Pacific because of instabilities in the mean easterly flow brought on by reversals in the background meridional potential vorticity (PV) gradient in these regions (Molinari et al. 1997, 2000). Maloney and Hartmann (2001) and Aiyyer and Molinari (2008) find that low-level westerly periods related to the MJO enhance barotropic growth of eddies on submonthly time scales in the east Pacific, Gulf of Mexico, and over Central America, providing a more favorable environment for the growth of synoptic disturbances and hurricanes, similar to the results of Lau et al. (1989) for the western Pacific. Zehnder (1991) and Zehnder et al. (1999) show that easterly flow incident on idealized topography representative of the Sierra Madre can excite westward-propagating Rossby waves. These waves may then interact with barotropically unstable regions within the east Pacific ITCZ, causing the flow to break down into a series of eddies through a process like that described in Ferreira and Schubert (1997) and Wang and Magnusdottir (2005, 2006). Easterly waves incident on idealized Sierra Madre topography may also result in lee cyclogenesis, potentially giving rise to a source of EWs along with tropical depressions (Zehnder 1991; Zehnder et al. 1999). Observational evidence for such breakdowns related to EW activity is abundant (Wang and Magnusdottir 2005, 2006).

The earlier-mentioned studies provide evidence for several forcing mechanisms of EWs and hurricanes in the IAS region. In this study, we assess the relative importance of local intensification and genesis processes versus wave propagation through the region using easterly wave track statistics, regression analyses, and eddy energy conversion statistics for the June–November season, when the waves are most active. We also present two case studies to demonstrate the important physical mechanisms borne out in the statistical analyses. We take advantage of the European Centre for Medium-Range Weather Forecasts (ECMWF) interim reanalysis (ERA-I; Simmons et al. 2007a,b; Uppala et al. 2008), produced at TL255 (∼70 km at the equator) horizontal resolution and with greater resolution in the vertical, concentrated within the boundary layer and near the tropopause. Representation of the tropics is significantly more realistic than other reanalysis products generally produced at T159 or lower horizontal resolution and with fewer levels in the vertical (Simmons et al. 2007a,b; Uppala et al. 2008).

2. Data and methods

a. Vorticity tracking

Easterly wave track statistics across the tropical Atlantic and northeast Pacific are estimated using objective feature tracking following the method of Hodges (1995, 1999). Relative vorticity from the ERA-I is used in this study. ERA-I products (produced at TL255) are provided on a 1.5° × 1.5° grid with 37 levels in the vertical and are currently available from 1989 to the present at 4 times daily resolution. The analyses in this study spans the 1989–2007 period. In addition, we focus on June–November, when easterly waves are most active (Roundy and Frank 2004). The 850-, 825-, 800-, 775-, 750-, 700-, 650-, and 600-hPa raw relative vorticity from ERA-I is vertically averaged and then smoothed to T42 resolution (∼280-km grid spacing at the equator) using a Gaussian spatial filter. The spatial smoothing is necessary to prevent tracking of vorticity features much smaller than the scale of easterly waves. Smoothed positive vorticity features are tracked on a sphere, where we have used a threshold of +0.5 × 10−5 s−1. We also require systems to travel at least 1000 km and persist for at least 2 days. These criteria are the same as those used by Thorncroft and Hodges (2001). Genesis and lysis density are also derived from the tracking method, where the first (last) point in each track is considered a genesis (lysis) location. The density is then computed as the number of track, genesis, or lysis points per month per unit area (∼106 km2) using the spherical kernel method (Hodges 1996).

The objective feature tracking technique has been used previously with 15-yr ECMWF Re-Analyses (ERA-15) to study midlatitude stormtrack statistics (Hoskins and Hodges 2002), as well as to study easterly wave track statistics over West Africa and the tropical Atlantic (Thorncroft and Hodges 2001; Hopsch et al. 2007). Hodges et al. (2003) extended the Hoskins and Hodges (2002) analyses to several additional reanalyses. The latter study suggests that differences in stormtrack density and storm intensity are due to both differences in model resolution as well as model formulation. ERA-I was selected for use in this study for its improved representation of the tropics over the National Center for Environmental Predication–National Center for Atmospheric Research (NCEP–NCAR) reanalyses and ERA-40 products.

Objective tracking of vorticity features does not necessarily provide a direct correspondence to the tracking of easterly waves. First, waves in the tropics have smaller intensities than those of midlatitudes, and Hoskins and Hodges (2002) point out that the tracks obtained in the tropics are likely biased toward the strongest waves. We offset this bias in part by selecting a low threshold vorticity of +0.5 × 10−5 s−1. In addition to this issue, more than one positive vorticity center may be associated with a given open trough, and these features generally have different lifetimes than the wave itself (Reed et al. 1988a; Pytharoulis and Thorncroft 1999; Kerns et al. 2008). Reed et al. (1988a) also find that skill in locating vorticity centers in the reanalyses can vary by region, with the data-sparse regions over the Atlantic coincident with erratic behavior of these centers. Vertically averaging the vorticity from 850 to 600 hPa will not help with multiple vorticity maxima within one wave trough, but it will improve the temporal coherency when a vorticity maximum shifts between levels. Despite these potential difficulties with the objective feature tracking technique, the tracks obtained have proven useful for establishing general characteristics of wave activity and intensity in the tropics consistent with more subjective tracking studies (e.g., Chang 1970; Reed et al. 1988b; Avila 1991). In this study the wave activity and intensity will be compared with independently derived lagged regression analyses of wave structure and energy conversions to assess the preferred tracks and likely genesis regions for easterly waves in the IAS region.

b. Regression analyses

EW cloudiness locations in the IAS region are isolated through space–time filtering of the National Oceanic and Atmospheric Administration’s Advanced Very High Resolution Radiometer (AVHRR) outgoing longwave radiation (OLR) following Wheeler and Kiladis (1999), who found spectral peaks in OLR corresponding to easterly waves referred to as the tropical disturbance band (“TD band”) after Takayabu and Nitta (1993). The space–time region we use as a filter includes 2–6-day fluctuations in westward-moving zonal wavenumbers more than 6. Magnusdottir and Wang (2008) have shown that a similar peak is observed in 850-hPa relative vorticity. Kiladis et al. (2006) and Serra et al. (2008) used TD-filtered OLR for the analysis of African easterly waves and Pacific easterly waves, respectively. The OLR data are on a 2.5° × 2.5° grid and are available from 1979 to the present at 2 times daily resolution. Winds, air temperature, streamfunction, and humidity from ERA-I are linearly regressed onto the TD-filtered OLR time series at select base points to obtain statistical representations of wave structure. A −20 W m−2 TD-filtered OLR anomaly is used for scaling in all regression analyses in this study.

3. Results

a. Climatology of cloudiness and winds in the IAS region

The 1989–2007 seasonal mean OLR and 700-hPa zonal winds are shown in Figs. 1a,b for June–August (JJA) and September–November (SON), respectively. In JJA the deepest land-based convection is observed to extend from the Mexican Sierra Madre down to the northern part of South America. This convection extends into the central and east Pacific as the ITCZ, roughly centered on 10°N. The convection is particularly strong over Panama and in the east Pacific near 10°N, 100°W. The pattern is similar in SON but the convection over South America is further developed because of the onset of the South American monsoon, while the convection over North America and the east Pacific is diminished.

The zonal winds are seen to have a strong longitudinal dependence throughout the IAS region. The upper portion of the Caribbean low-level jet (CLLJ) near 15°N, 75°W is apparent and is strongest during JJA, with a maximum of greater than −10 m s−1. The CLLJ vertical profile has a maximum at 925 hPa (Fig. 2), but it extends up to approximately 600 hPa climatologically (e.g., Wang 2007; Cook and Vizy 2010). The −6 m s−1 contour during JJA and the −4 m s−1 contour during SON both indicate that the easterly flow through the Caribbean passes into the east Pacific via the Papagayo jet across Costa Rica (see also Fig. 2). Zonal contours also suggest flow around the Sierra Madre through the Isthmus of Tehuantepec near 17°N, 95°W. Meridional flow is accelerated in this region (e.g., Fig. 2). Both the Caribbean and the region of strong easterlies extending into the east Pacific are places where the flow is potentially barotropically unstable (Molinari et al. 1997, 2000).

Seasonal standard deviations in twice-daily TD-filtered OLR for the IAS region are shown in Figs. 1c,d, respectively, along with the seasonal standard deviations in twice-daily OLR for comparison, calculated after removing the seasonal cycle. Large standard deviations in TD-filtered OLR are seen in the IAS during JJA, and they correspond well with high variability in subseasonal OLR during this season (Fig. 1c) as well as high values of mean OLR (Fig. 1a). As the ITCZ as a whole becomes less active during SON (Fig. 1b), the TD-filtered OLR becomes a smaller fraction of the subseasonal OLR variability (Fig. 1d).

Figure 2 shows 1989–2007 JJA 925-hPa wind vectors and relative vorticity overlaid on ERA-I topography for the IAS region. The figure highlights the complex mean flow at low levels in this region. For instance, the CLLJ centered at 13°N, 75°W has a seasonal maximum at 925 hPa of −12.5 m s−1. This jet produces a positive vorticity maximum near 10°N, 75°W, between the jet maximum flow and the mountainous terrain over Colombia and Venezuela to the south. Comparisons with observations such as those reported in Amador et al. (2006) show that the ERA-I CLLJ is realistic both in magnitude and location. The extension of the CLLJ across Costa Rica and into the east Pacific in the ERA-I is also in agreement with Quick Scatterometer (QuikSCAT) winds for this same period (Amador et al. 2006, their Fig. 19). A region of positive low-level vorticity is also seen to extend from the Caribbean into the east Pacific equatorward of the strong easterly flow. In the east Pacific, the positive vorticity is maintained by both the enhanced easterlies to the north (primarily near the coast) and the weak westerly flow to the south in this region. The CLLJ creates a region where the low-level flow meets the necessary conditions for dynamic instability (Molinari et al. 1997, 2000). Thus, the CLLJ is a possible genesis/intensification region for easterly waves, making its accurate representation in the reanalyses of particular importance to this study.

In addition to the CLLJ, gap winds through the Isthmus of Tehuantepec in southern Mexico and across the lowlands of Nicaragua and northern Coast Rica (Papagayo jet) contribute to the complex pattern of low-level relative vorticity shown in Fig. 2. These gap winds are the result of wind being forced through narrow mountain passages and have a significant seasonal cycle associated with the seasonal variations in surface pressure between the Atlantic and east Pacific basins (e.g., Chelton et al. 2004). Vorticity dipoles develop on either side of these jets and the Sierra Madre in relation to horizontal shear in the winds. Zehnder et al. (1999) and Molinari et al. (2000) have presented evidence that low-level vorticity associated with these winds can be important for cyclogenesis in the east Pacific.

b. Track statistics

The June–November 1989–2007 track densities for the 850–600-hPa vorticity are shown in Fig. 3a along with the TD-filtered 7 W m−2 OLR contour for the June–November standard deviation for reference. The 75th percentile track density contours are in bold, highlighting the regions of greatest easterly wave activity. The geographical distribution of the tracks is similar to that obtained for 850 hPa by Hodges et al. (2003), although they showed that the actual densities differed greatly among various reanalyses products. Hodges et al. (2003) also show that for all but one reanalysis product, 850-hPa tracks are discontinuous (<2 tracks per unit area per month) between the west Atlantic and Caribbean. We find that the location of the 75th percentile contour is not highly sensitive to the threshold vorticity chosen for the ERA-I product.

Track density is low entering the eastern Caribbean and increases to a local maximum south of the CLLJ. Tracks then continue across Central America into the east Pacific, where track densities are also at a local maximum and coincident with high convective activity in this region. Thorncroft and Hodges (2001) examined the 850- and 600-hPa-level track densities separately. In their study, only one density maximum occurs in the east Pacific, and their area of maximum density extends into the Caribbean only at the 600-hPa level. The localized maximum in the Caribbean in the ERA-I data may result from the better horizontal resolution over ERA-15 and ERA-40 used in the Thorncroft and Hodges (2001) and Hopsch et al. (2007) studies, respectively, as well as the improved data assimilation scheme. As will be discussed in a later section, the depiction of the CLLJ in the reanalysis product affects the assessment of the degree of instability in the flow over the region.

The tendency of the tracks to pass over the southern portion of Central America is not surprising considering the mountainous terrain along the west coast of Mexico and Central America north of Costa Rica (e.g., Fig. 2). It appears that these mountains divert the waves over lower elevations in this region.

Overall, track density is maximized within the band of high TD-filtered OLR variability, showing consistency between these independent measures of wave activity. However, the extension of the track density west of 110°W is somewhat to the north of the variance in TD-filtered OLR. This offset most likely results from the fact that the track density highlights the most coherent vorticity structures in the region, while the standard deviation in TD-filtered OLR captures the regions of greatest variability, including smaller scale, less organized convection within the TD band.

Figures 3b,c show genesis and lysis density statistics for the 850–600-hPa levels. As in Fig. 3a, the 75th percentile genesis and lysis density contours are highlighted. Genesis maxima occur south of the CLLJ and in the east Pacific, where track densities are also high. By comparison, Thorncroft and Hodges (2001) only observed a genesis maximum at the 600-hPa level in the Caribbean in ERA-40; however, they found, as we do, two additional maxima off the coast of Panama and in the east Pacific at both the 600- and 850-hPa levels.

Lysis is observed near 60°W at the entrance to the Caribbean, where track density falls to a local minimum in Fig. 3a. Lysis is also a maximum in the western Caribbean and in a small region of the east Pacific, between and to the west of the track density maximum in these areas. It appears that these lysis regions are in fact not due to analysis deficiencies contributing to discontinuities in vortices across the ocean basins, since satellite measures of cloudiness variations (e.g., Fig. 1) also show such discontinuities, and weakening of EWs in the central Atlantic (roughly between 32° and 56°W) in particular has been noted previously (e.g., Carlson 1969; Thorncroft and Hodges 2001).

c. Structure of easterly waves in the IAS region

The mean fields, standard deviations in TD-filtered OLR, and the tracking statistics indicate that the western Caribbean Sea, in the region south of the core of the CLLJ, is the primary location for observing easterly wave signals over the western Atlantic. On the basis of this result, Fig. 4 shows regressions of OLR and 700-hPa winds and streamfunction onto TD-filtered OLR for the base points at 12.5°N, 80°W for lags −4 to +4 days. This base point was chosen to highlight potential changes in EW morphology as they cross from the Caribbean region into the east Pacific. Every other wind vector with significance greater than or equal to 95% is shown.

At lag −4, a strong cyclonic center is seen near 57°W, with positive OLR anomalies to the east and negative OLR anomalies to the west of the center (Fig. 4a). At −2 lag, the structure begins to change as the cyclonic center encounters the upper portion of the CLLJ in the eastern Caribbean, indicating a northwest-to-southeast tilt north of the wind maximum and a northeast-to-southwest tilt south of this maximum. The tilt of the wave across the strong easterlies in the Caribbean is consistent with barotropic energy conversions of zonal mean to eddy kinetic energy. At lag 0, the convective anomalies have entered the east Pacific and form along the west coast of Central America and Mexico (Fig. 4c). Large-eddy circulations over the Pacific and well separated from the main wave near 120°W lie between the equator and 10°N. These drift westward, but they appear to have eastward energy dispersion, as evidenced by their successive centers forming farther east over time in Fig. 4. This behavior is indicative of mixed Rossby–gravity waves, which along with easterly waves form a large class of hybrid convectively coupled systems over the Pacific ITCZ (see Kiladis et al. 2009), the nature of which is currently a topic of investigation by the authors.

As the main wave energy crosses the American landmass at lags +2 to +4, a series of vortices are seen traveling up the west coast of Central America and Mexico, where the gap winds and associated mean vorticity maxima are observed in the seasonal mean low-level wind fields shown in Figs. 1, 2. These traveling vortices are smaller in scale than the original EW structure itself entering the Caribbean at lag −4. The region of maximum track density oriented northwest to southeast at all levels along the coast seen in Fig. 3 also corresponds well with the location of the traveling vortices.

The convection associated with the traveling vortices is much stronger in the regressions than the convection associated with the main wave train seen centered on 10°N, 120°W at lag 0. The most significant wind vectors cross the lower elevations of Central America and southern Mexico, consistent with the track density shown in Fig. 3a. Amador et al. (2006) note that the Papagayo jet is essentially an extension of the CLLJ during boreal summer and that it can pulse on time scales similar to those of easterly waves (see their Fig. 18). Our own analyses shows that both the gap winds across the Isthmus of Tehuantepec, as well as the Papagayo jet, have significant spectral energy in the 3–6-day band during May–November (not shown). These results support modeling studies of Zehnder et al. (1999) and observational studies of Wang and Magnusdottir (2006), and they strongly suggest that easterly waves interact with orographically forced circulations in the region. Comparison with Fig. 2 additionally indicates that the scales of the vortices along the west coast of Central America and Mexico associated with an easterly wave disturbance in the Caribbean are similar to those present because of the gap winds, implying that the low-level flow may influence the scales of these disturbances along the coast.

The structure of the wave in the vicinity of the CLLJ and across Central America is further investigated in Fig. 5, which shows a vertical cross section of the meridional wind, temperature, and humidity regressions at 12.5°N and lag 0, along with TD-filtered OLR also at 12.5°N in the top panel. The minimum ERA-I surface pressure between 9.5° and 15.5°N is also shown as a function of longitude to indicate the local orography. At 12.5°N, 75°W there is a near-surface maximum in the meridional wind in the vicinity of the CLLJ embedded within the wave signature that extends throughout the troposphere. The vertical structure has a slight eastward tilt with height through the lower troposphere, topped by a westward tilt above 300 hPa that increases further west. A secondary wind maximum near 200 hPa is also present. This structure is consistent with both the waves over Africa (Kiladis et al. 2006) and the east Pacific (Serra et al. 2008); however, a reversal in the meridional winds with height above 300 hPa is not observed in Fig. 5 over the Caribbean as it is for African easterly waves and Pacific easterly waves.

The temperature structure is typical of a classical easterly wave (Riehl 1979), with cold signals in the low to midtroposphere and warm temperatures aloft in phase with the convection and the southerlies. The humidity maximum is also in phase with the convective maximum and extends to at least 300 hPa, the apparent limit to the ERA-I humidity signal. The tilt of the humidity with height is also consistent with previous analyses (Serra and Houze 2002, 2008; Kiladis et al. 2006) and indicates a moistening of the near-surface layers prior to the arrival of the convection.

The progression of the wave into the east Pacific is further examined in Fig. 6, which shows the lag −2-, 0-, and +2-day vertical structure in the meridional wind only, along with the TD-filtered OLR for these same lags. At lag −2, the maximum wave signature is near 750 hPa and there are distinct lower- and upper-tropospheric maxima. At lag 0, as discussed earlier, the lower-tropospheric maximum appears to be split between approximately 950 and 550 hPa, with both maxima much smaller in scale than the maximum near 60°W. The wave tilt has also shifted from slightly westward with height near 60°W to slightly eastward with height at this location. At lag +2, the positive phase of the wave is over Central America and the far eastern Pacific and has a maximum near 500 hPa, while the secondary maximum near 950 hPa has disappeared. However, there is evidence that the wave signature redevelops near the surface west of 90°W and that the winds nearly reverse with height at 100°W, consistent with the Pacific easterly waves in Serra et al. (2008). Together Figs. 5 and 6 indicate that the EW maintains structure in the vertical across the IAS and that EW vertical structures survive intact while crossing Central America, although they are somewhat disrupted at the lowest levels. This is presumably due to interactions with the land surface and orography, although there is also some indication of an interaction between the wave and the CLLJ. If so, this would be expected to show up in energy conversion statistics, which we calculate next.

d. Energetics

Following the method of Serra et al. (2008), eddy energy terms are estimated from a statistical representation of easterly waves against a 19-yr climatology. Thus, variables in this system can be written as a = aclim + a′, where aclim is the 1989–2007 June–November climatology and a′ is the TD-wave-filtered perturbation from our regressions. To account for local evolution, we further divide the climatological potential temperature θclim into a regional mean (denoted by 〈·〉) and deviation from the regional mean Δ:θclim(x, y, p) = 〈θclim〉(p)|θclim(x, y, p). The mean eddy kinetic energy KE and potential energy AE are then defined as
i1520-0442-23-18-4823-e1
i1520-0442-23-18-4823-e2
where R is the universal gas constant for dry air; u and υ are the zonal and meridional wind components, respectively; and T is temperature. The overbar with the t indicates a time average over ±4-day lag, and the angle bracket denotes a spatial average taken over ±20° from the base point in longitude and over 20° in latitude, a region encompassing a typical EW. The eddy energy budget terms are then expressed as
i1520-0442-23-18-4823-e3
i1520-0442-23-18-4823-e4
where CAE→KE is the conversion from eddy available potential energy to eddy kinetic energy, CKM→KE is the conversion from mean to eddy kinetic energy, and CAM→AE is the conversion from mean available potential energy to eddy potential energy. Also present are the advection by the mean flow terms ADVKE and ADVAE, the fluxes of FKE, and eddy available potential energy FAE at the boundaries of a defined region, as well as the dissipation of eddy kinetic energy (D) and sources and sinks of eddy available potential energy due to heating S. As in Serra et al. (2008), we make no attempt to calculate these additional terms primarily because the reanalysis data and regression technique used to define the statistical waveform in this study do not permit reliable estimates of them. Here we are primarily interested in estimating the barotropic conversions associated with the CLLJ and gap winds in the region.
The conversion terms from the earlier-mentioned equations are
i1520-0442-23-18-4823-e5
i1520-0442-23-18-4823-e6
i1520-0442-23-18-4823-e7
where cp is the specific heat at constant pressure, ω = dp/dt, and is the three-dimensional divergence operator. We will focus on the largest two terms found for the locations in this study: CAE→KE and the second of the six terms in the expansion of (6), that is, uυt(∂uclim/∂y).

Figure 7 shows conversions from CAE→KE and conversions from meridional shear in the mean zonal wind to eddy kinetic energy (second term in the expansion of CKM→KE). Conversion terms that are not shown are of similar or smaller magnitude to those shown here. Easterly wave temperature perturbations at lag 0 are overlaid on the CAE→KE profiles, while 1989–2007 June–November mean zonal wind is overlaid on the CKM→KE: term 2 profiles. The figure shows calculations of latitude-height sections at the longitude of three base points: 12.5°N, 60°W; 12.5°N, 80°W; and 10°N, 95°W. Positive values in all panels indicate the conversion to eddy kinetic energy.

Upwind of the CLLJ near 60°W, eddy energy conversions are dominated by CAE→KE conversions. The baroclinic overturning suggested by these conversions results from the vertical structure of the classical EW, which is cold core in the lower troposphere and warm core aloft (e.g., Reed et al. 1977). This term is also found to be important for African EWs over land (Norquist et al. 1977). However, at 12.5°N, 80°W, CKM→KE: term 2 positive conversions are of equal magnitude to CAE→KE conversions but more limited in latitude. These latter conversions are observed at low levels just south of the maximum in the CLLJ. Further west at 10°N, 95°W, positive CKM→KE: term 2 conversions are also observed at low levels with a similar latitudinal extent but with a somewhat weaker maximum than at 12.5°N, 80°W. However, because positive CKM→KE: term 2 conversions extend deeper into the troposphere, the column-integrated energy conversions are comparable to those of CAE→KE at this location and exceed the integrated magnitude of the barotropic conversions in the Caribbean. As seen in Fig. 1 and noted by Amador et al. (2006), the CLLJ extends into the east Pacific via the Papagayo jet. This flow is strongest in the JJA season and is apparent even at 700 hPa (Fig. 1a). In addition a narrow region of westerly flow is observed between 3° and 10°N at this location, further enhancing the meridional gradient in the zonal wind. Thus, the CKM→KE: term 2 conversions remain an important source of low-level kinetic energy for easterly waves in the region.

In a previous study (Serra et al. 2008) based on NCEP–NCAR reanalyses at 2.5° × 2.5° horizontal resolution, the barotropic conversions were found to be weaker than the baroclinic conversions shown in Fig. 7 at 10°N, 95°W. Upon further investigation we found that the NCEP–NCAR reanalyses do not resolve the CLLJ or the flow off the coast of Central America as well as the ERA-I reanalyses, with the NCEP–NCAR reanalyses having a weaker CLLJ than observations from radiosondes and QuikSCAT surface winds and lacking the structure in the low-level vorticity seen in Fig. 2, extending across Central America and up the west coast of Mexico (not shown). The global models used for the ERA-I and NCEP–NCAR reanalyses differ in many aspects, including data assimilation techniques, model physics parameterizations, and model orography, as well as having different horizontal and vertical resolutions. It is likely that a combination of all of these things results in the differences in the wind field over the IAS. However, given that the strongest du/dy south of the CLLJ is over less than 5° in latitude and that the Sierra Madre are not as well represented in the NCEP–NCAR reanalyses compared to ERA-I (for instance, there is no land connecting Central America to South America in the NCEP–NCAR reanalyses), it is likely that model resolution plays a significant role in the results for the IAS region.

e. Relationship between the CLLJ, easterly waves, and tropical storm (TS) activity

The CLLJ has significant seasonal variability, with a maximum during boreal winter and summer and a minimum during boreal spring and fall, and is closely connected to the variability in sea level pressure over the region (Wang 2007; Muñoz et al. 2008; Cook and Vizy 2010). EWs also have a strong seasonal cycle, with a maximum in boreal summer and little or no activity in boreal winter. To better understand the relationship between the CLLJ and EW activity, we examine composites of track density at 700 hPa related to anomalous easterly (strong) and westerly (weak) phases of the CLLJ during boreal summer using a CLLJ index that removes the seasonal cycle in the jet activity. This index is defined by first calculating the monthly-mean ERA-I 925-hPa zonal wind over the area 12.5°–17.5°N, 80°–70°W, the same region used by Wang (2007) for the 1989–2007 period. We then subtract the respective 19-yr monthly climatology from this time series. This index I is centered and normalized and is then used to create weights w, where the weights are defined as w(mn) = tanh(±1.5I(mn)), if ±1.5I(mn) > 0 and 0 otherwise. Here, mn is the month and w is the value of the weight. These weights are introduced into the spherical kernel estimators to compute composite track density statistics for positive and negative phases of the CLLJ and similarly to compute weighted means of various fields. See the appendix of Bengtsson et al. (2006) for a more complete discussion of this compositing technique.

Figure 8 shows the May–November 1989–2007 925-hPa zonal wind composites for the strong and weak phases of the CLLJ index and their differences. The box in the Caribbean indicates the region over which our CLLJ index is defined. The strong phase composite clearly shows that when there is stronger easterly flow in the CLLJ region, these easterlies tend to extend further west into the eastern Pacific and further east into the Atlantic. These differences are up to 3 m s−1 in the core CLLJ region and in the eastern Pacific, while 1–2 m s−1 differences are observed in the western Atlantic.

The May–November 1989–2007 composite 700-hPa track densities for the strong and weak phases of the CLLJ and their differences are shown in Fig. 9. As in Fig. 3, the 75th percentile track density is highlighted in bold in the top two panels. A region of higher track density is seen within and downstream of positive du/dy anomalies in Fig. 8c, especially over the east Pacific. This result is consistent with the energetics analysis presented in the previous section, which shows that this region is a source of eddy kinetic energy for EWs. Thus, in the strong phase of the CLLJ, the flow supports more EW activity, which develops once the waves cross Central America.

Wang (2007) notes that the September–October period is coincident with maximum rainfall and hurricane formation in the Caribbean (Wang 2007; Shieh and Colucci 2010). The situation during this time is similar to that seen in the weak composite shown in Fig. 8, in that the CLLJ is typically approaching a minimum (Wang 2007; Cook and Vizy 2010). This result is somewhat surprising given the strong relationship between EWs and hurricanes. However, strong low-level easterlies in the Caribbean create unfavorable conditions for hurricane formation because of strong lower-tropospheric shear and associated moisture flux divergence (Wang 2007; Shieh and Colucci 2010).

To investigate the relationship between the CLLJ and TS activity further, we examine the number of named tropical storms by the National Hurricane Center (NHC) in the east Pacific and Atlantic for all years from 1989 to 2007 and compare these values to the June–September CLLJ index when the jet is at a maximum climatologically (Fig. 10). The least squares fit (LSF) lines for the east Pacific and Atlantic data are also shown along with the corresponding correlation coefficients. For the east Pacific, the correlation is positive with a value of 0.62, significant at the 99% confidence level, assuming each year is an independent measurement. In contrast, a significant negative correlation (−0.69) is observed between the number of Atlantic storms and the CLLJ index. No significant correlation is observed between hurricane intensity and the CLLJ index in either basin (not shown). Thus, while the Atlantic storms, including the Caribbean, are reduced during a strong CLLJ, east Pacific hurricanes increase in number. This may be one factor linking the inverse relationship between overall activity in the east Pacific and Atlantic (−0.36), which was also found by Wang and Lee (2010) to be statistically significant (−0.40) using an accumulated cyclone intensity index for each basin back to 1949. Barrett and Leslie (2009) show that such a negative correlation could arise from modulation by the MJO. This result, together with Fig. 10, suggests that the CLLJ may also be modulated by the MJO. A detailed analysis of these relationships is currently underway.

f. Case studies

Here we examine two case studies from 1993 of eastern Pacific hurricane development from EWs that crossed into the basin from the Caribbean during a near-neutral and weakly positive phase of the MJO. We chose 1993 because it has a strongly positive CLLJ index during boreal summer but is a neutral year with respect to ENSO (available online at http://www.cpc.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.shtml).

Our first case is the development of east Pacific Hurricane Fernando. According to the NHC, Fernando formed from an EW that traveled across the Atlantic. However, after 2 August when the wave entered the southeast Caribbean, the convection weakened and the wave could no longer be tracked until 4 August when convection again increased near Panama. Fernando was identified as a tropical depression at 0600 UTC 9 August.

Figure 11a shows streamlines at 1000 hPa for 20-day smoothed winds on 8 August. The 850–600-hPa vorticity track for the EW that eventually became Fernando is also shown, with dates indicated at 1200 UTC along the track. The broken lines and diamonds indicate the track prior to being designated a named storm by the NHC. The solid lines indicate when the track was designated as a tropical depression (open circles), tropical storm (solid circles), or hurricane (solid squares). We only show the 850–600-hPa vorticity tracks because they are nearly indistinguishable from the NHC best-track data and also provide the position of the EW prior to becoming a named storm.

The 1000-hPa streamlines indicate strong flow through the Isthmus of Tehuantepec on both dates, converging with the southerly and southwesterly monsoon flow in the east Pacific to form a confluence region along 10°–13°N, marking the ITCZ. The EW that became Fernando traveled across Central America and the east Pacific along this confluence region. Fernando’s track is also coincident with the region where the meridional gradient in PV at the 300-K level changes sign (Fig. 12a). The 300-K level is at approximately 900 hPa in this region. The east–west orientation along 10°N of the confluence region seen in Fig. 11a is similar to that shown in Aiyyer and Molinari (2008) for the nonconvective (easterly) phase of the MJO, when eddies tend to propagate more zonally. Analysis of the reconstructed 850-hPa zonal winds using the MJO multivariate index of Wheeler and Hendon (2004) indicates that 6–11 August 1993 was a period of transition for the MJO from an easterly to a westerly phase (not shown).

Figures 13a,b show the development of Fernando in the east Pacific using barotropic energy conversion estimates calculated from 700-hPa 2–10-day filtered winds for the eddy component and 6-day smoothed winds for the background shear. Only the conversions associated with the meridional shear in the zonal wind (CKM→KE:2) are shown here to illustrate conversions from the low-level jets to the eddy field, as discussed in section 3d. As stated previously, these terms account for the majority of the energy conversions in Eq. (6). The shading highlights regions of positive meridional gradients in the 6-day smoothed zonal wind ≤−5 × 10−6 s−1, indicating the region where barotropic conversions would be favored. This region extends from the western Caribbean into the east Pacific following the confluence at 1000 hPa shown in Fig. 11a. We additionally show 700-hPa streamfunction to indicate the location of the easterly wave troughs (bold black lines).

At 1800 UTC 8 August, 2 days after the NHC reidentified the EW trough associated with Fernando near Panama, eddy kinetic energy conversions are seen to the east of the EW trough near 13°N, 97°W. The EW trough associated with Fernando tilts southwest to northeast, consistent with barotropic conversions, and remains in this orientation through its development into a hurricane (not shown).

By 1800 UTC 9 August, the disturbance in the east Pacific has become Tropical Storm Fernando. The eddy kinetic energy generation associated with the disturbance remains high and to the east of the trough. At 1800 UTC 10 August, Fernando was declared a hurricane.

For our second case study, we examine Atlantic Hurricane Bret’s transition into east Pacific Hurricane Greg after dissipating at 1800 UTC 11 August (not shown). Figure 11b shows the 1000-hPa streamlines for 14 August as well as Greg’s 850–600-hPa vorticity track. In contrast to the background flow associated with Fernando, the development of Greg is associated with strong south-to-southwesterly flow across the east Pacific at 1000 hPa and a shift in the confluence region to along the coast of Mexico and Central America. The confluence region is not as well defined as on 8 August, as eddies dominate the 20-day streamlines at 108°, 95°, and 90°W at this time. Nevertheless, Greg’s track generally follows the region of confluence as seen with Fernando. Figure 12b indicates that Greg’s track also meanders through a region of potentially unstable flow prior to amplifying, starting on 14 August near 12°N, 105°W. On 14 August the MJO was in a westerly phase with westerly anomalies of 0.5–1 m s−1 observed between 100° and 120°W from 16–21 August (not shown). The location of the confluence region and resulting EW track occurs within the region of enhanced low-level cyclonic vorticity off the southwestern coast of Mexico, as is favored during the convective (westerly) phase of the MJO (Aiyyer and Molinari 2008).

The development of Hurricane Greg is further examined in Figs. 13c,d. At 1200 UTC 13 August, Atlantic Tropical Storm Bret can still be seen as a disorganized trough south of and through the Isthmus of Tehuantepec (Fig. 13c). The location of the dissipated Bret is marked by a diamond at 10°N, 100°W. Little meridional shear in the zonal winds is seen at this time, except north of Panama and within the Isthmus of Tehuantepec. A limited region of eddy energy conversions is seen west of the trough, as well as over Panama and Costa Rica.

By 1200 UTC 14 August, these centers of eddy energy have merged and now lie in a region of enhanced meridional shear of the zonal wind. The trough is better defined, and the vorticity track is now at 13°N, 104°W. Conversions are also developing on the east side of the trough, though a large center of conversions remains on to the west as well. At 0000 UTC 15 August, Greg was identified as a tropical depression. At this time high eddy kinetic energy is observed around the center of the storm, except on the open wave side of the trough to the southwest (not shown).

Fernando’s and Greg’s development are in line with the statistical analyses of the previous sections and suggest that, for these cases, EW and associated storm development in the east Pacific involves conversions of mean flow to eddy energy through the meridional shear in the zonal wind, which occur preferentially in regions of negative meridional gradients in PV. The statistical relationship between tropical storm genesis, EWs, and the states of the CLLJ and MJO need to be examined further.

4. Discussion and conclusions

Easterly waves are a dominant mode of variability in the IAS region during boreal summer when the ITCZ is north of 5°N (e.g., Amador et al. 2006). However, unlike the open ocean areas of the Atlantic and Pacific where winds are generally deep easterlies and only weakly dependent on longitude, the lower-level flow within the IAS region is strongly affected by topography and is highly variable in time and with longitude (e.g., Maloney and Hartmann 2001; Kessler 2006; Amador et al. 2006).

Many previous studies have suggested a strong link between EWs in the Atlantic and Pacific basins, but these have generally been in the context of individual case studies. Track, genesis, and lysis density statistics in this paper show that synoptic-scale positive vorticity centers frequently enter the IAS region from the Atlantic. However, regions of lysis across the central Atlantic indicate that this is not necessarily the primary source of synoptic activity in the IAS region. On the other hand, genesis is observed in the Caribbean as well as off the coast of Central America and Mexico, extending out into the east Pacific, pointing to a local source for EWs in the IAS.

Horizontal and vertical wind structure from regression analyses also show continuity between Atlantic and Pacific easterly waves, in line with the tracking statistics. However, the vertical structure also implies the interaction between easterly waves and the CLLJ, consistent with regions of EW genesis. In addition, a significant number of vortex anomalies appear to initially develop along the west coast of Central America and travel northwestward toward the Baja Peninsula.

Eddy energy conversions from meridional shear in the zonal wind to eddy kinetic energy are significant in the Caribbean and east Pacific in association within the CLLJ and Papagayo jet, as are conversions from eddy potential to kinetic energy. The latter is the result of the eddy vertical structure, which is cold at low levels and warm aloft (e.g., Reed and Recker 1971; Reed et al. 1977; Serra and Houze 2002; Kiladis et al. 2006; Serra et al. 2008). The former strongly suggests that easterly waves intensify in this region because of the unstable flow associated with the CLLJ and Papagayo jet, as found by Molinari et al. (1997, 2000). This result is also supported by track density composites for the strong and weak phases of the CLLJ, which indicate a greater number of EWs in the west Caribbean and east Pacific during the strong phase of the jet. We also find that the frequency of east Pacific storms is positively correlated with the CLLJ index, while the Atlantic storms are negatively correlated with this index.

Two case studies from 1993 demonstrate the tendency for EWs and associated storms to develop within the region of confluence between the south-to-southwesterly flow associated with the monsoon and the easterly flow from the Atlantic. Within the confluence zone, the meridional gradient in PV changes sign associated with enhanced positive meridional shear in the zonal wind, resulting in strong barotropic conversions of mean to eddy energy for both case studies. The location of the EW tracks and storm intensification for Fernando and Greg are consistent with those observed by Aiyyer and Molinari (2008) for the nonconvective and convective phases of the MJO, respectively, suggesting that even a weak MJO event (<1 m s−1 anomalies) can modulate the location of storms and storm precursors.

The behavior of the CLLJ and EWs over the IAS region is strongly modulated by the large-scale basic state, which in turn affects the frequency and location of tropical storm genesis over the region. Several studies have found that periods of enhanced low-level westerly flow over the east Pacific related to the MJO enhance convective activity and eddy activity on synoptic time scales in the region (Maloney and Esbensen 2007), as well as favor cyclogenesis (Maloney and Hartmann 2001; Knabb et al. 2008; Aiyyer and Molinari 2008). The present study shows that periods of strong easterly flow are associated with a greater frequency of easterly waves and storms in the Pacific, with the opposite effect in the Atlantic. Further investigation is required to understand the time scales and relative importance of these two mechanisms for enhancing synoptic activity and cyclogenesis in the IAS region. The smaller vorticity structures on the lee side of the Sierra Madre seen in our regression analyses also require further investigation to understand the possible role of the orography in setting the scale of EWs in this region and/or initiating such disturbances due to interactions of Atlantic zonal flow and EWs with the terrain, as suggested by Zehnder et al. (1999). We are currently pursuing these topics further.

Acknowledgments

YLS was supported by the NOAA Climate Program Office’s Climate Predication Program for the Americas (CPPA) program. GNK was supported by NOAA’s Office of Global Programs under Grant GC05-156. ERA-I data used in this study were provided by the ECMWF data server.

REFERENCES

  • Adams, J. L., and D. J. Stensrud, 2007: Impact of tropical easterly waves on the North American monsoon. J. Climate, 20 , 12191238.

  • Aiyyer, A., and J. Molinari, 2008: MJO and tropical cyclogenesis in the Gulf of Mexico and eastern Pacific: Case study and idealized numerical modeling. J. Atmos. Sci., 65 , 26912704.

    • Search Google Scholar
    • Export Citation
  • Amador, J. A., E. J. Alfaro, O. G. Lizano, and V. O. Magaña, 2006: Atmospheric forcing of the eastern tropical Pacific: A review. Prog. Oceanogr., 69 , 101142.

    • Search Google Scholar
    • Export Citation
  • Avila, L. A., 1991: Atlantic tropical systems of 1990. Mon. Wea. Rev., 119 , 20272033.

  • Avila, L. A., and J. L. Guiney, 2000: Eastern North Pacific hurricane season of 1998. Mon. Wea. Rev., 128 , 29903000.

  • Avila, L. A., R. J. Pasch, J. L. Beven, J. L. Franklin, M. B. Lawrence, S. R. Stewart, and J-G. Jiing, 2003: Eastern North Pacific hurricane season of 2001. Mon. Wea. Rev., 131 , 249262.

    • Search Google Scholar
    • Export Citation
  • Barrett, B. S., and L. M. Leslie, 2009: Links between tropical cyclone activity and Madden–Julian oscillation phase in the North Atlantic and northeast Pacific basins. Mon. Wea. Rev., 137 , 727744.

    • Search Google Scholar
    • Export Citation
  • Bengtsson, L., K. I. Hodges, and E. Roeckner, 2006: Storm tracks and climate change. J. Climate, 19 , 35183543.

  • Berry, G. J., C. Thorncroft, and T. Hewson, 2007: African easterly waves during 2004—Analysis using objective techniques. Mon. Wea. Rev., 135 , 12511267.

    • Search Google Scholar
    • Export Citation
  • Carlson, T. N., 1969: Some remarks on African disturbances and their progress over the tropical Atlantic. Mon. Wea. Rev., 97 , 716726.

    • Search Google Scholar
    • Export Citation
  • Chang, C-P., 1970: Westward propagating cloud patterns in the tropical Pacific as seen from time-composite satellite photographs. J. Atmos. Sci., 27 , 133138.

    • Search Google Scholar
    • Export Citation
  • Chelton, D. B., M. G. Schlax, M. H. Freilich, and R. F. Milliff, 2004: Satellite Measurements Reveal Persistent Small-Scale Features in Ocean Winds. Science, 303 , 978983.

    • Search Google Scholar
    • Export Citation
  • Chen, T-C., S-Y. Wang, and A. J. Clark, 2008: North Atlantic hurricanes contributed by African easterly waves north and south of the African easterly jet. J. Climate, 21 , 67676776.

    • Search Google Scholar
    • Export Citation
  • Cook, K. H., and E. K. Vizy, 2010: Hydrodynamics of the Caribbean low-level jet and its relationship to precipitation. J. Climate, 23 , 14771494.

    • Search Google Scholar
    • Export Citation
  • Douglas, A. V., and P. J. Englehart, 2007: A climatological perspective of transient synoptic features during NAME 2004. J. Climate, 20 , 19471954.

    • Search Google Scholar
    • Export Citation
  • Farfán, L. M., and J. A. Zehnder, 1997: Orographic influence on the synoptic-scale circulations associated with the genesis of Hurricane Guillermo (1991). Mon. Wea. Rev., 125 , 26832698.

    • Search Google Scholar
    • Export Citation
  • Ferreira, R. N., and W. Schubert, 1997: Barotropic aspects of ITCZ breakdown. J. Atmos. Sci., 54 , 261285.

  • Frank, N. L., and G. Clark, 1980: Atlantic tropical systems of 1979. Mon. Wea. Rev., 108 , 966972.

  • Hall, N. M. J., G. N. Kiladis, and C. D. Thorncroft, 2006: Three dimensional structure and dynamics of African easterly waves. Part II: Dynamical modes. J. Atmos. Sci., 63 , 22312245.

    • Search Google Scholar
    • Export Citation
  • Higgins, R. W., W. Shi, and C. Hain, 2004: Relationships between Gulf of California moisture surges and precipitation in the southwestern United States. J. Climate, 17 , 29832997.

    • Search Google Scholar
    • Export Citation
  • Hodges, K. I., 1995: Feature tracking on the unit sphere. Mon. Wea. Rev., 123 , 34583465.

  • Hodges, K. I., 1996: Spherical nonparametric estimators applied to the UGAMP model integration for AMIP. Mon. Wea. Rev., 124 , 29142932.

    • Search Google Scholar
    • Export Citation
  • Hodges, K. I., 1999: Adaptive constraints for feature tracking. Mon. Wea. Rev., 127 , 13621373.

  • Hodges, K. I., B. J. Hoskins, J. Boyle, and C. Thorncroft, 2003: A comparison of recent reanalysis datasets using objective feature tracking: Storm tracks and tropical easterly waves. Mon. Wea. Rev., 131 , 20122037.

    • Search Google Scholar
    • Export Citation
  • Hopsch, S. B., C. D. Thorncroft, K. Hodges, and A. Aiyyer, 2007: West African storm tracks and their relationship to Atlantic tropical cyclones. J. Climate, 20 , 24682483.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B. J., and K. I. Hodges, 2002: New perspectives on the Northern Hemisphere winter storm tracks. J. Atmos. Sci., 59 , 10411061.

    • Search Google Scholar
    • Export Citation
  • Hsieh, J-S., and K. H. Cook, 2005: Generation of African easterly wave disturbances: Relationship to the African easterly jet. Mon. Wea. Rev., 133 , 13111327.

    • Search Google Scholar
    • Export Citation
  • Hsieh, J-S., and K. H. Cook, 2007: A study of the energetics of African easterly waves using a regional climate model. J. Atmos. Sci., 64 , 421440.

    • Search Google Scholar
    • Export Citation
  • Johnson, R. H., P. E. Ciesielski, B. D. McNoldy, P. J. Rogers, and R. K. Taft, 2007: Multiscale variability of the flow during the North American Monsoon Experiment. J. Climate, 20 , 16281648.

    • Search Google Scholar
    • Export Citation
  • Kerns, B., K. Greene, and E. Zipser, 2008: Four years of tropical ERA-40 vorticity maxima tracks. Part I: Climatology and vertical vorticity structure. Mon. Wea. Rev., 136 , 43014319.

    • Search Google Scholar
    • Export Citation
  • Kerns, B., K. Greene, and E. Zipser, 2009: Four years of Tropical ERA-40 vorticity maxima tracks. Part II: Differences between developing and nondeveloping disturbances. Mon. Wea. Rev., 137 , 25762591.

    • Search Google Scholar
    • Export Citation
  • Kessler, W. S., 2006: The circulation of the eastern tropical Pacific: A review. Prog. Oceanogr., 69 , 181217.

  • Kiladis, G. N., C. D. Thorncroft, and N. M. J. Hall, 2006: Three-dimensional structure and dynamics of African easterly waves. Part I: Observations. J. Atmos. Sci., 63 , 22122230.

    • Search Google Scholar
    • Export Citation
  • Kiladis, G. N., M. C. Wheeler, P. T. Haertel, K. H. Straub, and P. E. Roundy, 2009: Convectively coupled equatorial waves. Rev. Geophys., 47 , RG2003. doi:10.1029/2008RG000266.

    • Search Google Scholar
    • Export Citation
  • Knabb, R. D., L. A. Avila, J. L. Beven, J. L. Franklin, R. J. Pasch, and S. R. Stewart, 2008: Eastern North Pacific hurricane season of 2005. Mon. Wea. Rev., 136 , 12011216.

    • Search Google Scholar
    • Export Citation
  • Lau, K-M., and N-C. Lau, 1990: Observed structure and propagation characteristics of tropical summertime synoptic scale disturbances. Mon. Wea. Rev., 118 , 18881913.

    • Search Google Scholar
    • Export Citation
  • Lau, K-M., L. Peng, C. H. Sui, and T. Nakazawa, 1989: Dynamics of Super Cloud Clusters, Westerly Wind Bursts, 30-60 Day Oscillations and ENSO: A Unified View. J. Meteor. Soc. Japan, 67 , 205219.

    • Search Google Scholar
    • Export Citation
  • Magnusdottir, G., and C-C. Wang, 2008: Intertropical convergence zones during the active season in daily data. J. Atmos. Sci., 65 , 24252436.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., and D. L. Hartmann, 2001: The Madden–Julian Oscillation, barotropic dynamics, and North Pacific tropical cyclone formation. Part I: Observations. J. Atmos. Sci., 58 , 25452558.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., and S. K. Esbensen, 2007: Satellite and buoy observations of boreal summer intraseasonal variability in the tropical northeast Pacific. Mon. Wea. Rev., 135 , 319.

    • Search Google Scholar
    • Export Citation
  • Molinari, J., and D. Vollaro, 2000: Planetary- and synoptic-scale influences on eastern Pacific tropical cyclogenesis. Mon. Wea. Rev., 128 , 32963307.

    • Search Google Scholar
    • Export Citation
  • Molinari, J., D. Knight, M. Dickinson, D. Vollaro, and S. Skubis, 1997: Potential vorticity, easterly waves, and eastern Pacific tropical cyclogenesis. Mon. Wea. Rev., 125 , 26992708.

    • Search Google Scholar
    • Export Citation
  • Molinari, J., D. Vollaro, S. Skubis, and M. Dickinson, 2000: Origins and mechanisms of eastern Pacific tropical cyclogenesis: A case study. Mon. Wea. Rev., 128 , 125139.

    • Search Google Scholar
    • Export Citation
  • Muñoz, E., A. J. Busalacchi, S. Nigam, and A. Ruiz-Barradas, 2008: Winter and summer structure of the Caribbean low-level jet. J. Climate, 21 , 12601276.

    • Search Google Scholar
    • Export Citation
  • Nitta, T., and Y. Takayabu, 1985: Global analysis of the lower-tropospheric disturbances in the tropics during the northern summer of the FGGE year. Part II: Regional characteristics of the disturbances. Pure Appl. Geophys., 123 , 272292.

    • Search Google Scholar
    • Export Citation
  • Norquist, D. C., E. E. Recker, and R. J. Reed, 1977: The energetics of African wave disturbances as observed during phase III of GATE. Mon. Wea. Rev., 105 , 334342.

    • Search Google Scholar
    • Export Citation
  • Pasch, R. J., and Coauthors, 2009: Eastern North Pacific hurricane season of 2006. Mon. Wea. Rev., 137 , 320.

  • Pytharoulis, I., and C. Thorncroft, 1999: The low-level structure of African easterly waves in 1995. Mon. Wea. Rev., 127 , 22662280.

  • Reed, R. J., and E. E. Recker, 1971: Structure and properties of synoptic-scale wave disturbances in the equatorial western Pacific. J. Atmos. Sci., 28 , 11171133.

    • Search Google Scholar
    • Export Citation
  • Reed, R. J., D. C. Norquist, and E. E. Recker, 1977: The structure and properties of African wave disturbances as observed during phase III of GATE. Mon. Wea. Rev., 105 , 317333.

    • Search Google Scholar
    • Export Citation
  • Reed, R. J., A. Hollingsworth, W. A. Heckley, and R. Delsol, 1988a: An evaluation of the performance of the ECMWF operational system in analyzing and forecasting easterly wave disturbances over Africa and the tropical Atlantic. Mon. Wea. Rev., 116 , 824865.

    • Search Google Scholar
    • Export Citation
  • Reed, R. J., E. Klinker, and A. Hollingsworth, 1988b: The structure and characteristics of African easterly wave disturbances as determined from the ECMWF operational analysis/forecast system. Meteor. Atmos. Phys., 38 , 2231.

    • Search Google Scholar
    • Export Citation
  • Riehl, H. H., 1979: Climate and Weather in the Tropics. Academic Press, 623 pp.

  • Roundy, P., and W. Frank, 2004: A climatology of waves in the equatorial region. J. Atmos. Sci., 61 , 21052132.

  • Serra, Y. L., and R. A. Houze Jr., 2002: Observations of variability on synoptic timescales in the east Pacific ITCZ. J. Atmos. Sci., 59 , 17231743.

    • Search Google Scholar
    • Export Citation
  • Serra, Y. L., G. N. Kiladis, and M. G. Cronin, 2008: Horizontal and vertical structure of easterly waves in the Pacific ITCZ. J. Atmos. Sci., 65 , 12661284.

    • Search Google Scholar
    • Export Citation
  • Shapiro, L. J., 1986: The three-dimensional structure of synoptic-scale disturbances over the tropical Atlantic. Mon. Wea. Rev., 114 , 18761891.

    • Search Google Scholar
    • Export Citation
  • Shieh, O. H., and S. J. Colucci, 2010: Local minimum of tropical cyclogenesis in the eastern Caribbean. Bull. Amer. Meteor. Soc., 91 , 185196.

    • Search Google Scholar
    • Export Citation
  • Simmons, A., S. Uppala, and D. Dee, 2007a: Update on ERA-Interim. ECMWF Newsletter, No. 111, ECMWF, Reading, United Kingdom, 5.

  • Simmons, A., S. Uppala, D. Dee, and S. Kobayashi, 2007b: ERA-Interim: New ECMWF reanalysis products from 1989 onwards. ECMWF Newsletter, No. 110, ECMWF, Reading, United Kingdom, 25–35.

    • Search Google Scholar
    • Export Citation
  • Simpson, R. H., N. Frank, D. Shideler, and H. M. Johnson, 1968: Atlantic tropical disturbances, 1967. Mon. Wea. Rev., 96 , 251261.

  • Stensrud, D. J., R. L. Gall, and M. K. Nordquist, 1997: Surges over the Gulf of California during the Mexican monsoon. Mon. Wea. Rev., 125 , 417437.

    • Search Google Scholar
    • Export Citation
  • Tai, K-S., and Y. Ogura, 1987: An observational study of easterly waves over the eastern Pacific in the northern summer using FGGE data. J. Atmos. Sci., 44 , 339361.

    • Search Google Scholar
    • Export Citation
  • Takayabu, Y. N., and T. Nitta, 1993: 3-5 Day-Period Disturbances Coupled with Convection over the Tropical Pacific Ocean. J. Meteor. Soc. Japan, 71 , 221246.

    • Search Google Scholar
    • Export Citation
  • Thorncroft, C., and B. J. Hoskins, 1994a: An idealized study of African easterly waves. I: A linear view. Quart. J. Roy. Meteor. Soc., 120 , 953982.

    • Search Google Scholar
    • Export Citation
  • Thorncroft, C., and B. J. Hoskins, 1994b: An idealized study of African easterly waves. II: A nonlinear view. Quart. J. Roy. Meteor. Soc., 120 , 9831015.

    • Search Google Scholar
    • Export Citation
  • Thorncroft, C., and K. Hodges, 2001: African easterly wave variability and its relationship to Atlantic tropical cyclone activity. J. Climate, 14 , 11661179.

    • Search Google Scholar
    • Export Citation
  • Uppala, S., D. Dee, S. Kobayashi, P. Berrisford, and A. Simmons, 2008: Towards a climate data assimilation system: Status update of ERA-Interim. ECMWF Newsletter, No. 115, ECMWF, Reading, United Kingdom, 12–18.

    • Search Google Scholar
    • Export Citation
  • Wang, C., 2007: Variability of the Caribbean Low-Level Jet and its relations to climate. Climate Dyn., 29 , 411422.

  • Wang, C., and S-K. Lee, 2010: Is hurricane activity in one basin tied to another? Eos, Trans. Amer. Geophys. Union, 91 .doi:10.1029/2010EO100002.

    • Search Google Scholar
    • Export Citation
  • Wang, C-C., and G. Magnusdottir, 2005: ITCZ breakdown in three-dimensional flows. J. Atmos. Sci., 62 , 14971512.

  • Wang, C-C., and G. Magnusdottir, 2006: The ITCZ in the central and eastern Pacific on synoptic time scales. Mon. Wea. Rev., 134 , 14051421.

    • Search Google Scholar
    • Export Citation
  • Wheeler, M., and G. Kiladis, 1999: Convectively coupled equatorial waves: Analysis of clouds and temperature in the wavenumber-frequency domain. J. Atmos. Sci., 56 , 374399.

    • Search Google Scholar
    • Export Citation
  • Wheeler, M., and H. H. Hendon, 2004: An all-season real-time multivariate MJO index: Development of an index for monitoring and prediction. Mon. Wea. Rev., 132 , 19171932.

    • Search Google Scholar
    • Export Citation
  • Zehnder, J. A., 1991: The interaction of planetary-scale tropical easterly waves with topography: A mechanism for the initiation of tropical cyclones. J. Atmos. Sci., 48 , 12171230.

    • Search Google Scholar
    • Export Citation
  • Zehnder, J. A., D. M. Powell, and D. L. Ropp, 1999: The interaction of easterly waves, orography, and the intertropical convergence zone in the genesis of eastern Pacific tropical cyclones. Mon. Wea. Rev., 127 , 15661584.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Mean OLR and 700-hPa zonal winds for (a) JJA and (b) SON. Standard deviation of twice-daily OLR anomalies (contours) and TD-filtered OLR (shading) for (c) JJA and (d) SON. All seasonal means and standard deviations shown are for the 1989–2007 period.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Fig. 2.
Fig. 2.

ERA-I orography and mean JJA 1989–2007 925-hPa wind vectors and relative vorticity (shading). Orography contours are every 200 m, starting at 200 m. Every other wind vector is shown for readability. The core of the CLLJ and the Gulf of Tehuantepec (Mexico) and Gulf of Papagayo (Costa Rica) gap winds are indicated with black arrows.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Fig. 3.
Fig. 3.

(a) Track, (b) genesis, and (c) lysis densities for the 850–600-hPa levels for the period June–November 1989–2007. Also shown in (a) is the 7 W m−2 contour of the standard deviation of TD-filtered OLR for the same period for reference. Density is in units of number of tracks per unit area per month, where the unit area is a 5° spherical cap equivalent to ∼106 km2. The 75th percentile density contour is in bold. Contours are every 2 density units in (a) and 0.25 density units in (b) and (c). A rectangle highlights the region used to calculate the CLLJ index.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Fig. 4.
Fig. 4.

Regressions of OLR (shading) and 700-hPa winds and streamfunction (contours) onto TD-filtered OLR at 12.5°N, 80°W for (a) lag −4, (b) lag −2, (c) lag 0, (d) lag +2, and (e) lag +4 days. Every other 95% significant wind vector is shown. Streamfunction contours are every 1 × 105 m2 s−1; negative contours are dashed.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Fig. 5.
Fig. 5.

Vertical regressions at lag 0 along 12.5°N of (a) OLR and (b) meridional wind, (c) temperature, and (d) specific humidity for the 12.5°N, 80°W base point. Contour intervals are 0.2 m s−1, 0.04°C, and 0.08 g kg−1 in (b)–(d), respectively. Minimum surface pressures between 9.5° and 15.5°N are also shown to represent the orography.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Fig. 6.
Fig. 6.

As in Fig. 5, but for (a) OLR and (b) meridional wind at lag −2, (c),(d) lag 0, and (e),(f) lag +2 days.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Fig. 7.
Fig. 7.

Conversions at the base-point longitude of (left) CAE→KE and (right) CKM→KE: term 2 for base points at (a),(b) 12.5°N, 60°W; (c),(d) 12.5°N, 80°W; and (e),(f) 10°N, 95°W. Top graph above each profile shows vertically integrated values. Also shown are contours of the (left) lag-0 temperature anomalies and (right) June–November 1989–2007 zonal wind. Contour intervals are every 0.03° and 1 m s−1, respectively. Negative contours are dashed. Color bars apply to all colored contours in the panels above them. Anomalies are based on TD-filtered OLR regressions.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Fig. 8.
Fig. 8.

May–November 1989–2007 composite 925-hPa zonal wind for the CLLJ (a) strong and (b) weak phases and (c) the strong-minus-weak-phase zonal wind differences. The box in the Caribbean indicates the region used to calculate the CLLJ index.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Fig. 9.
Fig. 9.

May–November 1989–2007 700-hPa track density composite for the CLLJ (a) strong and (b) weak phases and (c) the strong-minus-weak-phase track density differences. Track density is in units of number of tracks per unit area per month, where the unit area is a 5° spherical cap equivalent to ∼106 km2. The 75th percentile track density is in bold in (a) and (b). Contours are every 1 density unit in (a),(b) and every 0.5 density unit in (c). The box in the Caribbean indicates the region used to calculate the CLLJ index.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Fig. 10.
Fig. 10.

The number of named storms in the east Pacific and Atlantic for the 1989–2007 seasons plotted against the June–September CLLJ index. Also shown are the LSF lines to these data along with the associated correlation coefficients and their significance.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Fig. 11.
Fig. 11.

Streamlines of 20-day smoothed winds at 1000 hPa for (a) 8 and (b) 14 Aug 1993. The vorticity tracks that eventually became Fernando and Greg are also shown in (a) and (b), respectively. Time points are at 1200 UTC on the date indicated. The broken lines indicate the track of the easterly wave prior to intensification to the named storm by the NHC. The solid lines indicate the vorticity track of the named storm (NHC tracks not shown but are nearly indistinguishable). A diamond designates an unnamed trough, an open circle designates a tropical depression, a filled circle designates a tropical storm, and a filled square is a hurricane.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Fig. 12.
Fig. 12.

The 6-day-averaged meridional gradient of PV on the 300-K isotherm for values ≥2 × 10−13 m−1 s−1 (light shading) and values ≤−2 × 10−13 m−1 s−1 (dark shading) for (a) 8 and (b) 14 Aug 1993. The bold dashed line indicates the vorticity track for (a) Fernando and (b) Greg.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Fig. 13.
Fig. 13.

The 6-day-average meridional gradient of the 700-hPa zonal wind for values ≤−5 × 10−6 s−1 (shading) centered on (a) 1800 UTC 8 Aug, (b) 1800 UTC 9 Aug, (c) 1200 UTC 13 Aug, and (d) 1200 UTC 14 Aug. Solid contour lines are barotropic energy conversions associated with the meridional shear in the zonal wind. Contour lines are every 2 m s−1 day−1 until 10 and are then spaced every 10 m s−1 day−1. Dashed contour lines are 700-hPa streamfunction drawn every 2 × 106 s−1. The symbols are as in Fig. 11. Troughs are also shown as heavy black lines.

Citation: Journal of Climate 23, 18; 10.1175/2010JCLI3223.1

Save
  • Adams, J. L., and D. J. Stensrud, 2007: Impact of tropical easterly waves on the North American monsoon. J. Climate, 20 , 12191238.

  • Aiyyer, A., and J. Molinari, 2008: MJO and tropical cyclogenesis in the Gulf of Mexico and eastern Pacific: Case study and idealized numerical modeling. J. Atmos. Sci., 65 , 26912704.

    • Search Google Scholar
    • Export Citation
  • Amador, J. A., E. J. Alfaro, O. G. Lizano, and V. O. Magaña, 2006: Atmospheric forcing of the eastern tropical Pacific: A review. Prog. Oceanogr., 69 , 101142.

    • Search Google Scholar
    • Export Citation
  • Avila, L. A., 1991: Atlantic tropical systems of 1990. Mon. Wea. Rev., 119 , 20272033.

  • Avila, L. A., and J. L. Guiney, 2000: Eastern North Pacific hurricane season of 1998. Mon. Wea. Rev., 128 , 29903000.

  • Avila, L. A., R. J. Pasch, J. L. Beven, J. L. Franklin, M. B. Lawrence, S. R. Stewart, and J-G. Jiing, 2003: Eastern North Pacific hurricane season of 2001. Mon. Wea. Rev., 131 , 249262.

    • Search Google Scholar
    • Export Citation
  • Barrett, B. S., and L. M. Leslie, 2009: Links between tropical cyclone activity and Madden–Julian oscillation phase in the North Atlantic and northeast Pacific basins. Mon. Wea. Rev., 137 , 727744.

    • Search Google Scholar
    • Export Citation
  • Bengtsson, L., K. I. Hodges, and E. Roeckner, 2006: Storm tracks and climate change. J. Climate, 19 , 35183543.

  • Berry, G. J., C. Thorncroft, and T. Hewson, 2007: African easterly waves during 2004—Analysis using objective techniques. Mon. Wea. Rev., 135 , 12511267.

    • Search Google Scholar
    • Export Citation
  • Carlson, T. N., 1969: Some remarks on African disturbances and their progress over the tropical Atlantic. Mon. Wea. Rev., 97 , 716726.

    • Search Google Scholar
    • Export Citation
  • Chang, C-P., 1970: Westward propagating cloud patterns in the tropical Pacific as seen from time-composite satellite photographs. J. Atmos. Sci., 27 , 133138.

    • Search Google Scholar
    • Export Citation
  • Chelton, D. B., M. G. Schlax, M. H. Freilich, and R. F. Milliff, 2004: Satellite Measurements Reveal Persistent Small-Scale Features in Ocean Winds. Science, 303 , 978983.

    • Search Google Scholar
    • Export Citation
  • Chen, T-C., S-Y. Wang, and A. J. Clark, 2008: North Atlantic hurricanes contributed by African easterly waves north and south of the African easterly jet. J. Climate, 21 , 67676776.

    • Search Google Scholar
    • Export Citation
  • Cook, K. H., and E. K. Vizy, 2010: Hydrodynamics of the Caribbean low-level jet and its relationship to precipitation. J. Climate, 23 , 14771494.

    • Search Google Scholar
    • Export Citation
  • Douglas, A. V., and P. J. Englehart, 2007: A climatological perspective of transient synoptic features during NAME 2004. J. Climate, 20 , 19471954.

    • Search Google Scholar
    • Export Citation
  • Farfán, L. M., and J. A. Zehnder, 1997: Orographic influence on the synoptic-scale circulations associated with the genesis of Hurricane Guillermo (1991). Mon. Wea. Rev., 125 , 26832698.

    • Search Google Scholar
    • Export Citation
  • Ferreira, R. N., and W. Schubert, 1997: Barotropic aspects of ITCZ breakdown. J. Atmos. Sci., 54 , 261285.

    • Search Google Scholar