1. Introduction
Tropical moisture is exported to higher latitudes on the western flank of marine anticyclones, often forming atmospheric rivers (Boettcher and Wernli 2013). The convective outflows are modulated by El Niño–Southern Oscillation (ENSO) and equatorial Walker cell (Dai and Wigley 2000; Chen 2002). During an El Niño event, warm sea temperatures and atmospheric convection in the eastern Pacific generate anomalous divergence and vorticity (Hoskins and Karoly 1981) that sets off a northeastward Rossby wave train that spreads ENSO influence (Trenberth et al. 1998) via the Pacific–North America (PNA) pattern and intraseasonal Madden–Julian oscillation (MJO) (Lopez and Kirtman 2019). Rossby wave amplitudes tend to grow next to diabatic heat sources when the polar vortex weakens (Steinschneider and Lall 2016; Jiménez-Esteve and Domeisen 2018; Riboldi et al. 2022). Jet stream bifurcation accompanied by blocking highs can slow the eastward progression of Rossby waves < 5 m s−1, sustaining moist poleward airflow and repetitious storm tracks (Barnes 2001; Chang et al. 2002; Fröhlich et al. 2013; Martineau et al. 2017) that can reach into the subtropics during spring.
The larger western and central Caribbean Sea islands experience bimodal rainfall peaking in spring (April–May) and autumn (August–November) (Giannini et al. 2001; Chen and Taylor 2002; Jury et al. 2007). In other seasons trade wind subsidence limits the depth of convection (Amador 1998; Muñoz et al. 2008). The spring rainfall is triggered by westerly troughs in the subtropical jet with relative vorticity of >10−4 s−1 (Kosaka and Nakamura 2010). These tend to penetrate the Caribbean more frequently following Pacific El Niño (Laing 2004) via evaporation from sea surface temperatures (SST) followed by convective heating that alters the large-scale wind flow (Duran-Quesada 2012). Spring rains are welcome at the end of the dry season but narrow southwest (SW) cloud bands can overflow small river catchments.
Allen and Mapes (2017) find the initial peak of rain in the western Antilles occurs in May–June as local SST exceed 26°C. Martinez et al. (2019) studied how the regional moisture budget drives the seasonality of Caribbean rainfall using principal component analysis of daily data. They found that airflow from the tropical east Pacific and west Atlantic Oceans converged onto the western flank of the North Atlantic anticyclone during early summer wet spells over the Dominican Republic and Puerto Rico. Their analysis reveals an axis of moisture extending from Panama across the Antilles toward Bermuda, involving a poleward deflection of Caribbean trade winds and reduced moisture export to the Pacific.
Here, the meteorological behavior of springtime SW cloud bands is studied to address the following questions: What is their statistical context over central Caribbean islands? How do tropical–midlatitude interactions sustain SW cloud bands? Is the interannual forcing by ENSO PNA robust? Are intraseasonal oscillations, atmospheric rivers or Rossby wave trains implicated? What are the long-term trends in Caribbean springtime rainfall and its forcing? What new insights can be gained in comparison with earlier work (Jury and Sanchez 2009)? The research proceeds as follows: section 2 reviews how daily high-resolution fields were analyzed by empirical orthogonal functions (EOF) and top-10 composites; section 3 outcomes are subdivided into statistical context, processes that sustain cloud bands, case studies that highlight wave trains, and long-term trends. New insights are summarized in section 4.
2. Data and methods
Springtime SW cloud bands were analyzed via modern data-assimilated reanalysis fields (CFS2; Saha et al. 2014; ERA5; Hersbach et al. 2020): winds, vertical motion, geopotential height, temperature, specific humidity, and rainfall. Convection was quantified by NOAA satellite net outgoing longwave radiation (OLR; Lee 2014). A daily rainfall time series was constructed from the ERA5 overland maximum in the “local” area 17.5°–20°N, 70°–65°W, covering Dominican Republic and Puerto Rico (cf. dashed area in Fig. 1b). Gauge comparisons indicate optimal performance by ERA5 due to 25-km grid spacing over small mountainous islands. The annual cycle mean and upper quintile were calculated for comparison with 500-hPa zonal winds. Daily maximum river discharge over Dominican Republic and Puerto Rico (Harrigan et al. 2020) was statistically analyzed.
Caribbean island area: (a) mean annual cycle and upper 2.5% of ERA5 maximum daily rainfall (black and purple lines, respectively) and mean annual cycle of ERA5 monthly 500-hPa zonal wind (red symbols). (b) Standardized loading pattern of ERA5 daily EOF mode-2 500-hPa geopotential height (500Z) for April–May of 1980–2021, with “local” area (dash outlined). (c) Lag correlation of ERA5 daily April–May local rain and 500Z time series; the inset is a scatterplot of wet spells.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0126.1
EOF was applied to standardized daily ERA5 500-hPa geopotential height (500Z) fields April–May 1980–2021. This calculation objectively extracts the leading stationary dipole patterns and associated time series over the wider Caribbean area 0°–35°N, 90°–45°W. The second mode, accounting for 15% of variance, is significantly correlated with local rainfall. The daily local rainfall and 500Z mode-2 time series were analyzed for lag autocorrelation to determine springtime oscillations.
A statistical context was provided by lag correlation of April–May local rainfall and 500Z mode-2 time series onto the daily MJO index in the east Pacific, the daily PNA geopotential height index, and the monthly Pacific Niño-3.4 SST; all climate indices derive from the CPC (2021). 90% confidence is reached for monthly values with correlation coefficient r > 0.24 (sample N = 42) and for daily values r > 0.11 (N = 2562); considering background persistence. Lag correlations from −12 to +6 days indicate how short-lived wet spells relate to zonal propagating atmospheric waves, while lag correlations from −12 to +6 months describe the seasonal influence of ENSO. An alternative statistical approach uses ERA5 March–May rainfall EOF modes over the central Caribbean. The mode-1 time score was correlated with March–May fields of tropical SST, wind and net OLR, to distinguish ENSO influences on ocean–atmosphere coupling. This method includes March to link prior development with later impacts.
Daily time series of local rainfall and mode-2 500Z were ranked, and cases > 50 mm day−1 and > 0.20 standard deviation were identified in the period April–May 1980–2021. The top 10 days (Table 1) were used to form composite anomaly maps of satellite net OLR and 500-hPa winds that describe SW cloud bands. Similarly composite latitude–height sections averaged 65°–70°W of NCEP-2 (Kanamitsu et al. 2002) winds and moisture anomalies were calculated for the top 10 days. These datasets, with 100–200-km grid spacing, are suitable for analysis of large-scale circulation and convection. Atmospheric Rossby wavelength, wave amplitude, and zonal movement was studied by Hovmöller longitude–time plots averaged 10°–30°N for two wet spells in April 1983 and May 1986. Case studies of SW cloud bands on 17–21 April 1983 and 10–14 May 1986 used HYSPLIT back-trajectory simulations at lower and upper levels (Stein et al. 2015) similar to that of Duran-Quesada (2012). Radiosonde profiles and CloudSat radar reflectivity slices describe the vertical structure of SW cloud bands in selected cases.
Caribbean island ERA5 rainfall rank (mm day−1) and 500Z mode-2 value (standardized fraction); boldface type identifies case studies.
The time series of maximum daily local rainfall in April–May was analyzed for trend, slope, and variance and compared with river discharge. Regional trends in 700-hPa specific humidity and winds over central Caribbean islands were analyzed via NCEP-2 fields April–May 1980–2021. Long-term projections from the CNRM6 coupled model with SSP585 scenario (Voldoire et al. 2019; Gidden et al. 2019) simulate the trends in upper geopotential height and rainfall April–May 1980–2100. This model was used because of its optimal representation of localized rainfall over central Caribbean islands (Jury 2022). Dataset acronyms and attributes are listed in Table 2.
Datasets used in the analysis.
3. Results
a. Statistical context
SW cloud bands are somewhat infrequent but supply rains to central Caribbean islands at the end of the dry season. The annual cycle of maximum daily rainfall 1980–2021 (Fig. 1a) reveals many days in the upper quintile >50 mm. During springtime the mean 500-hPa winds are westerly ∼ 5 m s−1 and favor baroclinic weather systems, yet thermodynamic energy is minimal due to regional SST ∼ 27°C and convective available potential energy (CAPE) ∼ 700 J kg−1. The conditional instability requires a kinematic trigger in the form of a 500-hPa low–high dipole pattern represented by the second mode of April–May geopotential variability (Fig. 1b). The 500Z dipole centers of action are located at 28°N, 75°W and 15°N, 50°W, forming a trough-west–ridge-east pattern that draws southwesterly winds over central Caribbean islands. The lag correlation between the dipole mode and rainfall time series (Fig. 1c) has significant values from −2 to +4 days; while the scatterplot shows that mode-2 covers 27% of April–May rainfall variance.
Significant autocorrelation in the daily time series (Fig. 2a) occurs at ∼7-day intervals in the 500Z record and at ∼40-day intervals in the rainfall record, the former associated with transient midlatitude Rossby waves, and the latter with equatorial MJO, which modulates wind shear (Jury 2020). Monthly lag correlations with Niño-3.4 SST (Fig. 2b) suggest that the mode-2 500Z dipole pattern is anticipated by ENSO phase (r > 0.4 from −6 to −2 months). Rainfall is noisier and has a weak response to ENSO, but its sign and phase follow the 500Z dipole. Daily lag correlations between the MJO index and 500Z dipole (Fig. 2c) are significant from −4- to 0-day lag. Rainfall has a similar lag-correlation structure, but values are about one-half. For the PNA index and 500Z dipole, significant correlations extend from −4- to +2-day lags. Hence mode 2 may be considered an eastward extension of the PNA and its standing Rossby waves. Again, rainfall follows the lag-correlation structure of 500Z, with values barely significant. The outcome of statistical analysis is that slow subtropical and equatorial atmospheric waves are implicated in Caribbean springtime wet spells, and that warm-phase ENSO acts in conjunction with PNA to anchor Rossby wave trains that increase the chances for springtime floods.
(a) Lag autocorrelation of ERA5 daily April–May local rain (blue) and EOF mode-2 500Z (dashed red) time series, with arrows denoting significant periods. (b) Lag correlation of ERA5 April–May local rain and 500Z time series vs CPC monthly Niño-3.4 SST index. Also shown is the lag correlation of ERA5 daily April–May rain and 500Z time series vs (c) CPC east Pacific MJO index and (d) CPC daily PNA index. Monthly and daily correlations have N = 42 and N = 2562, respectively, and reach 90% confidence with r > 0.24 and r > 0.11, respectively (gray lines).
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0126.1
b. Composite structure
Composites are constructed with the top 10 days having April–May rainfall > 50 mm and positive values of 500Z dipole as listed in Table 1. Composite maps of net OLR and wind anomalies (Figs. 3a,b) reflect a SW cloud band extending from Colombia, which draws tropical moisture on the leading edge of a Rossby wave. A ridge east of the trough concentrates the SW cloud band, similar to earlier work (Jury and Sanchez 2009). There is a subtropical jet streak characterized by cyclonically curved 500-hPa wind anomalies > 10 m s−1. The North Atlantic anticyclone is displaced so trade winds weaken to ∼5 m s−1 north of Colombia and export less moisture from the Caribbean. Composite vertical sections averaged 65°–70°W (Figs. 3c,d) reveal deep rising motion from South America northward beyond the Antilles Islands, associated with thermal advection as outlined in Allen and Mapes (2017), creating moisture anomalies > 6 g kg−1 in the layer 850–500 hPa from 20° to 30°N. Westerlies > 20 m s−1 are present in the layer 400–150 hPa from 20° to 30°N, but these are near normal. In contrast, the equatorial zone shows low-level easterly–upper-level westerly wind anomalies > 20 m s−1 typical of MJO influence (Jury 2020).
Composite anomalies of (a) NOAA satellite net OLR (W m−2) with area of 5 m s−1 weaker trade winds (thick dashed outline), and (b) NCEP-2 500-hPa winds (m s−1) for top-10 cases (cf. Table 1). Composite height sections averaged 65°–70°W of (c) NCEP-2 meridional circulation over specific humidity anomalies (shaded >4 g kg−1) and (d) NCEP-2 zonal wind (contour) and anomalies (shaded <−10 and >+20). Case study analysis of vertical structure: (e) radiosonde profile at Santo Domingo, Dominican Republic, at 0000 UTC 13 May 1986, with CAPE = 1773 J kg−1. (f) CloudSat reflectivity sections for the case of 25 (top panel) and 26 (bottom panel) Apr 2016 (last row of Table 1).
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0126.1
c. Case studies
The Santo Domingo radiosonde profile for the May 1986 case (Fig. 3e) indicates 15 m s−1 southwesterly winds at 700 hPa. Dewpoint depression is < 5°C from 1000 to 600 hPa, and CAPE is 1773 J kg−1. CloudSat reflectivity slices just east of Puerto Rico in April 2016 (Fig. 3f) show thunderstorm clusters > 30 dBZ from 2 to 8 km, revealing mesoscale structure within the poleward atmospheric river. Maximum daily river discharge in the two case studies is listed in Table 3. Sudden increases are noted: from 84 to 500 m3 s−1 17–18 April 1983 and from 209 to 1073 m3 s−1 10–11 May 1986, indicative of flash floods.
Maximum river discharge (m3 s−1) in the Caribbean area for the two case studies.
Two of the top 10 cases (April 1983, May 1986) are analyzed for large-scale Rossby wave structure via 200-hPa meridional wind (Figs. 4a,b). Subtropical wave trains extend across the Pacific–Atlantic sector with alternating V winds at ∼60° longitude intervals. Hovmöller plots of 500-hPa V wind and 700-hPa specific humidity averaged 10°–30°N (Figs. 4c,d) demonstrate that eastward-moving Rossby waves slow and retrograde before the Caribbean wet spell (dashed line in Fig. 4c). Specific humidity builds as the Rossby waves decrease in length and increase in amplitude because of blocking. The beta effect is resolved according to W = V(ΔZ)(β/f). With V wind ∼ +10 m s−1, ΔZ layer thickness ∼ 104 m, βdf/dy ∼ 2 × 10−11 s−1, and Coriolis force f ∼ 5 × 10−5 s−1; the outcome is rising motion W ∼ 0.04 m s−1 (cf. Fig. 3c).
Rossby wave trains in ERA5 200-hPa V wind: (a) 19 Apr 1983 and (b) 12 May 1986. Hovmöller plots of (c) ERA5 500-hPa meridional wind (m s−1) and (d) ERA5 700-hPa specific humidity (g kg−1) (averaged over 10°–30°N) for April 1983 (left panels) and May 1986 (right panels); dashed lines in (c) suggest Rossby wave retrograding, in (d) the X marks the greatest rainfall and dashed lines suggest blocking.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0126.1
The troughs dip into the tropics entraining moisture that initiates deep convection during springtime (Fig. 5a). Wet spell rainfall over the central Caribbean is pulsed > 20 mm h−1 (Fig. 5b) leading to rapidly rising river flows (cf. Table 3). Low-level airflow trajectories in the April 1983 and May 1986 cases originate from the south (Fig. 5c), whereas upper-level trajectories undergo a zonal oscillation (Fig. 5d) associated with a quasi-stationary Rossby wave trough and ridge over the east Pacific and Central America, respectively.
For (left) 17–21 Apr 1983 and (right) 10–14 May 1986, (a) map of ERA5 accumulated rainfall (mm), (b) temporal graph of Caribbean island hourly rain rate, (c) HYSPLIT 6-hourly back-trajectories arriving at 18°N, 67.5°W, 1.5 km, and (d) same back-trajectories but arriving at 10-km elevation with wider perspective, based on NCEP-2.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0126.1
The 700-hPa circulation and thermodynamic conditions for April 1983 and May 1986 wet spells (Fig. 6a) emphasize confluence over the central Caribbean islands. The airflow is drawn poleward by a subtropical trough next to a tropical ridge. In the April 1983 case, the westerly airflow over Florida splits downstream. In both cases a tongue of moisture extends more than 1000 km north of Colombia (Fig. 6b). Over the southeastern United States the 700-hPa air temperatures are < 0°C, whereas the southern Caribbean is > 10°C (Fig. 6c). Thermal gradients accelerate the subtropical jet stream, which undergoes cyclonic curvature around the quasi-stationary trough.
Wet spell maps of (a) ERA5 700-hPa wind, (b) 700-hPa specific humidity with key area, and (c) ERA5 700-hPa air temperature for (left) 18 Apr 1983 and (right) 13 May 1986. Local area is dash outlined in (b).
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0126.1
d. Trends and projections
Historical trends in 700-hPa specific humidity reveal a wet-west–dry-east pattern across the Caribbean (Fig. 7a) that is related to an anticyclonic circulation trend with equatorward airflow to the northeast and poleward airflow to the northwest (Fig. 7b). The meridional circulation shows an accelerating trend in April–May (Fig. 7c) with rising motion over South America, northward upper-level airflow, and sinking motion over the Atlantic (20°–30°N) that appears as a Hadley cell. Peak daily rainfall per annum (Rx1day) over the central Caribbean (Fig. 7d) has declined since the 1980s. Weaker springtime wet spells have suppressed April–May river discharges by −1.86 m3 s−1 yr−1 over the period 1980–2021. The CNRM6 model simulation with SSP585 anticipates a drier climate over the twenty-first century. Processes that underpin the drying trend emerge in April–May projections of upper geopotential height and rainfall (Figs. 7e,f). A ridge is expected to build across the tropics that may inhibit the penetration of westerly troughs to low latitudes. The drying trend in April–May rainfall sweeps northward from South America across the Caribbean, along the axis of the Gulf Stream. Drier springtime conditions are also implicated for the east Pacific and southwestern United States. The key feature is subsidence in an accelerating Hadley cell that may inhibit tropical–midlatitude interactions necessary for springtime wet spells in the Caribbean.
Linear trends of April–May: (a) ERA5 700-hPa specific humidity (g kg−1 yr−1), (b) NCEP-2 700-hPa winds (vectors; m s−1 yr−1), and (c) NCEP-2 meridional circulation (vectors; m s−1 yr−1). (d) ERA5 historical and CNRM6 projected Caribbean island daily maximum April–May rainfall (mm day−1 yr−1) and linear trend. CNRM6 SSP585 projected trends in April–May: (e) 200-hPa geopotential height (m yr−1) and (f) rainfall (mm day−1 yr−1) for 1980–2100.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0126.1
e. Alternative statistical outcome
We have analyzed the Caribbean spring climate starting with an upper geopotential height EOF mode that emphasized midlatitude features. An alternative basis is March–May rainfall EOF modes. The mode-1 loading pattern (47% variance) is an all-wet Caribbean (Fig. 8a), whereas mode-2 (12%) is a dry-west–wet-east dipole. The mode-1 time score contains a 5–6-yr cycle consistent with ENSO. Simultaneous correlations of mode-1 time score with tropical SST, winds and netOLR (Figs. 8b–d) reveal a central Pacific El Niño that links convection to the Caribbean via a Walker circulation with equatorward–poleward upper airflow on subsiding 150°W–rising 70°W limbs (Fig. 8c). The <-shaped correlation pattern of Pacific SST and netOLR (Figs. 8b,d) reflects a coupled ocean Rossby wave whose westward phase speed is faster near the equator (∼0.3 m s−1) than in the subtropics ∼0.1 m s−1 (White et al. 2003; Abe et al. 2016). The downwelling wave is driven by anticyclonic wind stress curl that deepens the thermocline, inducing El Niño events at 5–6-yr intervals. Thus, statistical methods based on March–May rainfall offer complementary insights to those based on 500Z. The equatorial Walker cell induces a quasi-stationary trough in the northern subtropical jet accompanied by suppressed 150°W–enhanced 70°W convection. Warming in the central Pacific anchors springtime SW cloud bands over the Caribbean and coincides with opposing responses in the equatorial Atlantic. In contrast, the less frequent east Pacific El Niño (Kao and Yu 2009) shifts SW cloud bands farther east, creating a dry–wet dipole in the Caribbean.
(a) Standardized mode-1 loading pattern of ERA5 March–May rainfall 1980–2021 (blue is wet); the mode-2 dry area is enclosed by the red dashed line. Field correlations of ERA5 rain mode-1 time score with March–May: (b) SST, (c) section of NCEP-2 zonal circulation (vectors) and meridional wind (red contours), and (d) NOAA satellite netOLR (blue is wet, the dashed line is the convective axis); labels indicate association. In (c), the topographic profile is included, the vertical motion is exaggerated, the Walker cell is the gray rotor, and the section averaged is 5°S–10°N.
Citation: Journal of Applied Meteorology and Climatology 62, 2; 10.1175/JAMC-D-22-0126.1
4. Discussion and summary
The response of the large-scale atmospheric circulation to east Pacific warming is a low-north–high-south pattern (Gill 1980), yielding an anomalous southwesterly airflow that weakens northern subtropical anticyclones and trade wind subsidence. Atmospheric Kelvin and Rossby waves spread the circulation response away from the zone of anomalous equatorial SST via reduced heat fluxes, coastal downwelling and ocean Rossby wave coupled feedbacks (Matsuno 1966). Numerical model experiments often reproduce these features, which observational statistics cannot firmly establish. Yet certain linkages emerge: the PNA response to east Pacific warming forms a standing Rossby wave train that aligns springtime SW cloud bands east of Central America. 500Z mode-2 correlations with Niño-3.4 SST are robust, but rain impacts are weaker (cf. Figs. 2b,d) and suggest a random weather component due to transient localized growth in the amplitude of subtropical jet stream troughs.
The analysis demonstrated that springtime wet spells over the central Caribbean are related to Rossby wave patterns that redirect trade wind moisture poleward as noted in Martinez et al. (2019). The 500Z mode-2 accounts for 15% of total variance and its loading pattern (in wet phase) is composed of a trough near Florida and a ridge over the tropical Atlantic, associated with Pacific El Niño in the preceding winter. A new insight was gained: westerly wind shear from equatorial MJO was implicated in redirecting tropical airflow to the Northern Hemisphere. Rising motion (cf. Fig. 3c) via +df/dy and warm advection (Allen and Mapes 2017) linked tropical outflows with midlatitude forcing. A composite of top-10 cases revealed a subtropical trough penetrating toward Colombia, weakening the trade winds 5 m s−1 and converging tropical moisture into a SW cloud band. The April–May local rain and 500Z mode-2 time series exhibit spectral cycling at intervals of ∼7 and 40 days, reflecting eastward passage of midlatitude and equatorial atmospheric waves, respectively.
Another important new insight was gained in understanding the meteorology of flood events in April 1983 and May 1986. These cases exhibited subtropical Rossby wave trains of ∼60° wavelength × 30° amplitude across the entire western hemisphere. Hovmöller plots indicated a slowing of eastward movement that briefly anchored SW cloud bands over the Caribbean, bringing heavy rain and river flows exceeding 1000 m3 s−1 (cf. Table 3). The April 1983 case followed an intense El Niño, whereas the May 1986 case coincided with a lull in the polar vortex and retrograding Rossby wave trains (cf. Fig. 4c).
Martinez et al. (2019) show that spring rainfall in the Caribbean requires a decoupling of the continental and marine anticyclones either side of the Gulf Stream. Instead of exporting moisture westward, the trades turn poleward and draw moisture from the equatorial Atlantic and Pacific. The work here indicates a PNA pattern of quasi-stationary midlatitude Rossby waves (cf. Fig. 5d) play a role in dividing the continental and marine anticyclones, enabling the western flank of the North Atlantic high to sweep atmospheric rivers over the western Antilles (cf. Figs. 6a,b).
A second statistical method based on rainfall EOF modes, gave new insights on how the Pacific El Niño circulation sustains springtime SW cloud bands. A wet Caribbean was related to the Walker circulation (cf. Fig. 8c) and ocean–atmosphere coupling with application to seasonal forecasts. Historical trends and long-range projections by the CNRM6 model suggest that an accelerating Hadley cell and retreating jet stream may inhibit midlatitude–tropical interactions, leaving the western and central Antilles Islands with less springtime rainfall. Adaptation to a delayed wet season will be needed.
Acknowledgments.
Statistical analyses were enabled by the websites of the Climate Explorer KNMI, Colorado State University, IRI Climate Library, NOAA PSL, NOAA Ready ARL, University of Hawaii APDRC, and University of Wyoming.
Data availability statement.
A spreadsheet of data analysis is available from the author on request.
REFERENCES
Abe, H., Y. Tanimoto, T. Hasegawa, and N. Ebuchi, 2016: Oceanic Rossby waves over eastern tropical Pacific of both hemispheres forced by anomalous surface winds after mature phase of ENSO. J. Phys. Oceanogr., 46, 3397–3414, https://doi.org/10.1175/JPO-D-15-0118.1.
Allen, T. L., and B. E. Mapes, 2017: The late spring Caribbean rain-belt: Climatology and dynamics. Int. J. Climatol., 37, 4981–4993, https://doi.org/10.1002/joc.5136.
Amador, J. A., 1998: A climatic feature of the tropical Americas: The trade wind easterly jet. Top. Meteor. Oceanogr., 5, 91–102.
Barnes, G., 2001: Severe local storms in the tropics. Severe Convective Storms, Meteor. Monogr., No. 50, Amer. Meteor. Soc., 359–432, https://doi.org/10.1175/0065-9401-28.50.359.
Boettcher, M., and H. Wernli, 2013: A 10-yr climatology of diabatic Rossby waves in the Northern Hemisphere. Mon. Wea. Rev., 141, 1139–1154, https://doi.org/10.1175/MWR-D-12-00012.1.
Chang, E. K. M., S. Lee, and K. L. Swanson, 2002: Storm track dynamics. J. Climate, 15, 2163–2183, https://doi.org/10.1175/1520-0442(2002)015<02163:STD>2.0.CO;2.
Chen, A. A., and M. Taylor, 2002: Investigating the link between early season Caribbean rainfall and the El Niño. Int. J. Climatol., 22, 87–106, https://doi.org/10.1002/joc.711.
Chen, T.-C., 2002: A North Pacific short-wave train during the extreme phases of ENSO. J. Climate, 15, 2359–2376, https://doi.org/10.1175/1520-0442(2002)015<2359:ANPSWT>2.0.CO;2.
CPC, 2021: AAO, AO, NAO, PNA. National Weather Service, accessed 7 May 2022, www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_ao_index/teleconnections.shtml.
Dai, A., and T. M. L. Wigley, 2000: Global patterns of ENSO induced precipitation. Geophys. Res. Lett., 27, 1283–1286, https://doi.org/10.1029/1999GL011140.
Duran-Quesada, A. M., 2012: Sources of moisture for Central America and transport based on a Lagrangian approach: Variability, contributions to precipitation and transport mechanisms. Ph.D. thesis, University of Vigo, 287 pp.
Fröhlich, L., P. Knippertz, A. H. Fink, and E. Hohberger, 2013: An objective climatology of tropical plumes. J. Climate, 26, 5044–5060, https://doi.org/10.1175/JCLI-D-12-00351.1.
Giannini, A., Y. Kushnir, and M. A. Cane, 2001: Seasonality in the impact of ENSO and the North Atlantic high on Caribbean rainfall. Phys. Chem. Earth, 26, 143–147, https://doi.org/10.1016/S1464-1909(00)00231-8.
Gidden, M. J., and Coauthors, 2019: Global emissions pathways under different socioeconomic scenarios for use in CMIP6: A dataset of harmonized emissions trajectories through the end of the century. Geosci. Model Dev., 12, 1443–1475, https://doi.org/10.5194/gmd-12-1443-2019.
Gill, A. E., 1980: Some simple solutions for heat-induced tropical circulations. Quart. J. Roy. Meteor. Soc., 106, 447–462, https://doi.org/10.1002/qj.49710644905.
Harrigan, S., and Coauthors, 2020: GloFAS-ERA5 operational global river discharge reanalysis 1979–present. Earth Sys. Sci. Data, 12, 2043–2060, https://doi.org/10.5194/essd-12-2043-2020.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Hoskins, B. J., and D. J. Karoly, 1981: The steady linear response of a spherical atmosphere to thermal and orographic forcing. J. Atmos. Sci., 38, 1179–1196, https://doi.org/10.1175/1520-0469(1981)038<1179:TSLROA>2.0.CO;2.
Jiménez-Esteve, B., and D. I. V. Domeisen, 2018: The tropospheric pathway of the ENSO–North Atlantic teleconnection. J. Climate, 31, 4563–4584, https://doi.org/10.1175/JCLI-D-17-0716.1.
Jury, M. R., 2020: MJO influence in the Caribbean. Theor. Appl. Climatol., 139, 1559–1567, https://doi.org/10.1007/s00704-019-03037-x.
Jury, M. R., 2022: Inter-comparison of past and projected climate trends in Puerto Rico: 1950–2100. J. Water Climate Change, 13, 2713–2724, https://doi.org/10.2166/wcc.2022.071.
Jury, M. R., and D. M. Sanchez, 2009: Composite meteorological forcing of Puerto Rican springtime flood events. Wea. Forecasting, 24, 262–271, https://doi.org/10.1175/2008WAF2222151.1.
Jury, M. R., A. Winter, and B. A. Malmgren, 2007: Sub-regional precipitation climate of the Caribbean and relationships with ENSO and NAO. J. Geophys. Res., 112, D16107, https://doi.org/10.1029/2006JD007541.
Kanamitsu, M., W. Ebisuzaki, J. Woollen, S.-K. Yang, J. J. Hnilo, M. Fiorino, and G. L. Potter, 2002: NCEP-DOE AMIP-2 reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 1631–1644, https://doi.org/10.1175/BAMS-83-11-1631.
Kao, H.-Y., and J.-Y. Yu, 2009: Contrasting eastern-Pacific and central-Pacific types of ENSO. J. Climate, 22, 615–632, https://doi.org/10.1175/2008JCLI2309.1.
Kosaka, Y., and H. Nakamura, 2010: Mechanisms of meridional teleconnection observed between a summer monsoon system and a sub-tropical anticyclone. Part II: A global survey. J. Climate, 23, 5109–5125, https://doi.org/10.1175/2010JCLI3414.1.
Laing, A. G., 2004: Cases of heavy precipitation and flash floods in the Caribbean during El Niño winters. J. Hydrometeor., 5, 577–594, https://doi.org/10.1175/1525-7541(2004)005<0577:COHPAF>2.0.CO;2.
Lee, H.-T., 2014: Climate algorithm theoretical basis document: Outgoing longwave radiation (OLR). NOAA Climate Data Record Program Doc. CDRP-ATBD-0526, 46 pp.
Lopez, H., and B. P. Kirtman, 2019: ENSO influence over the Pacific North American sector: Uncertainty due to atmospheric internal variability. Climate Dyn., 52, 6149–6172, https://doi.org/10.1007/s00382-018-4500-0.
Martineau, P., G. Chen, and D. A. Burrows, 2017: Wave events: Climatology, trends, and relationship to Northern Hemisphere winter blocking and weather extremes. J. Climate, 30, 5675–5697, https://doi.org/10.1175/JCLI-D-16-0692.1.
Martinez, C., L. Goddard, Y. Kushnir, and M. Ting, 2019: Seasonal climatology and dynamical mechanisms of rainfall in the Caribbean. Climate Dyn., 53, 825–846, https://doi.org/10.1007/s00382-019-04616-4.
Matsuno, T., 1966: Quasi-geostrophic motions in the equatorial area. J. Meteor. Soc. Japan, 44, 25–43, https://doi.org/10.2151/jmsj1965.44.1_25.
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, 1260–1276, https://doi.org/10.1175/2007JCLI1855.1.
Riboldi, J., E. Rousi, F. D’Andrea, G. Rivière, and F. Lott, 2022: Circumglobal Rossby wave patterns during boreal winter highlighted by space–time spectral analysis. Wea. Climate Dyn., 3, 449–469, https://doi.org/10.5194/wcd-3-449-2022.
Saha, S., and Coauthors, 2014: The NCEP Climate Forecast System version 2. J. Climate, 27, 2185–2208, https://doi.org/10.1175/JCLI-D-12-00823.1.
Stein, A. F., R. R. Draxler, G. D. Rolph, B. J. B. Stunder, M. D. Cohen, and F. Ngan, 2015: NOAA HYSPLIT atmospheric transport and dispersion modelling system. Bull. Amer. Meteor. Soc., 96, 2059–2077, https://doi.org/10.1175/BAMS-D-14-00110.1.
Steinschneider, S., and U. Lall, 2016: Spatiotemporal structure of precipitation related to tropical moisture exports over the eastern United States and its relation to climate teleconnections. J. Hydrometeor., 17, 897–913, https://doi.org/10.1175/JHM-D-15-0120.1.
Trenberth, K. E., G. W. Branstator, D. Karoly, A. Kumar, N. Lau, and C. Ropelewski, 1998: Progress during TOGA in understanding and modeling global teleconnections associated with tropical sea surface temperatures. J. Geophys. Res., 103, 14 291–14 324, https://doi.org/10.1029/97JC01444.
Voldoire, A., and Coauthors, 2019: Evaluation of CMIP6 experiments with CNRM-CM6-1. J. Adv. Model. Earth Syst., 11, 2177–2213, https://doi.org/10.1029/2019MS001683.
White, W. B., Y. M. Tourre, M. Barlow, and M. Dettinger, 2003: A delayed action oscillator shared by biennial, interannual, and decadal signals in the Pacific basin. J. Geophys. Res., 108, 3070, https://doi.org/10.1029/2002JC001490.
