1. Introduction
Central American gyres (CAGs) are broad lower-tropospheric cyclonic circulations occurring near Central America, and are similar to broad monsoonal low pressure systems (MLs) in other oceanic basins (e.g., Boos et al. 2015; Hurley and Boos 2015). While MLs have been studied the most extensively in the north Indian Ocean basin (e.g., Piddington 1876; Eliot 1900; Krishnamurti et al. 1975; Godbole 1977; Sanders 1984; Douglas 1992; Boos et al. 2015), similar cyclonic circulations have been identified in the southern Indian Ocean basin (e.g., Davidson and Holland 1987; Baray et al. 2010), and western Pacific basin (e.g., Lander 1994; Harr et al. 1996; Aldinger and Stapler 1998; Molinari and Vollaro 2012; Beattie and Elsberry 2013; Crandall et al. 2014). The purpose of this research is to document the climatological structure and frequency of CAGs (i.e., MLs near Central America), which has mostly been documented through case studies (e.g., Lawrence 1998; Pasch and Roberts 2006; Aiyyer and Molinari 2008; Brennan 2010; Blake 2011; Montgomery et al. 2012).
MLs are often placed in one of two categories: monsoon depressions (MDs) and monsoon gyres (MGs). The American Meteorological Society’s Glossary of Meteorology notes that MDs and MGs possess loosely organized deep convection while maintaining a circular, closed, vortex characterized by a broad low-level radius of maximum winds (RMW) (Aldinger and Stapler 1998; American Meteorological Society 2016a,b). These ML categories are distinct from tropical cyclones (TCs), which are characterized by smaller RMWs (Kimball and Mulekar 2004), and from monsoon troughs (MTs), which have cyclonic winds, but do not have a closed vortex (Lander 1994). In addition to their broad lower-tropospheric cyclonic circulation, MDs also possess a similar-scale upper-tropospheric anticyclonic circulation (i.e., low PV) (Boos et al. 2015; Hurley and Boos 2015). In contrast, MGs are associated with an upper-tropospheric trough that features a positive PV anomaly to the northwest of the lower-tropospheric cyclonic circulation (Lander 1994; Crandall 2012; Molinari and Vollaro 2012). This upper-tropospheric structure contributes to convective asymmetry, where convective activity is focused primarily south and east of the lower-tropospheric circulation center (e.g., Lander 1994; Holland 1995; Molinari and Vollaro 2012; Crandall 2012; Crandall et al. 2014). Presently, there is little insight into whether MLs over Central America (CAGs) bear greater resemblance to MDs or MGs.
Previous case studies have documented broad lower-tropospheric cyclonic circulations with CAG characteristics over Central America (e.g., Fernandez and Barrantes 1996; Lawrence 1998; Aiyyer and Molinari 2008; Blake 2011; Montgomery et al. 2012). Fernandez and Barrantes (1996) investigated the development of a broad lower-tropospheric cyclonic circulation over Central America that was associated with widespread convection and heavy rainfall in May 1982. Aiyyer and Molinari (2008) investigated the development of a broad lower-tropospheric cyclonic circulation in September 1998 that they described as a cyclonic gyre, which formed in response to the convectively active phase of the MJO (Madden and Julian 1971) over Central America, and later became a TC (Frances in 1998) over the Gulf of Mexico (Lawrence 1998; Pasch et al. 2001). During the 2010 Pre-Depression Investigation of Cloud-Systems in the Tropics (PREDICT; Montgomery et al. 2012) experiment, a similar broad lower-tropospheric cyclonic circulation developed over Central America and later became a TC (Nicole in 2010) at 1200 UTC 28 September (Blake 2011) over the Caribbean Sea. This system was associated with an unusually large RMW and multiple relative vorticity maxima within a broader cyclonic circulation. In addition, widespread heavy rainfall accompanied this system, with 
The existing literature has contradictions on the frequency and timing of MLs over Central America during the rainy season (May–November). Hastenrath (1985, 238–240) presented a climatology of Central American temporals, which are loosely defined as multiday periods of continuous rainfall over portions of Central America during the rainy season. While temporals are not directly related to TCs, they are associated with systems possessing light wind cores that resemble the broad lower-tropospheric circulation of MLs (e.g., Fernandez and Barrantes 1996). Hastenrath (1985) noted that temporals develop approximately once per year, occurring most frequently in June and September–October, when the intertropical convergence zone (ITCZ) migrates northward. By contrast, the global ML climatology constructed by Hurley and Boos (2015) identified 4–10 MLs per year in a region encompassing Central America, with a seasonal peak in frequency in August.
The goals of this study are to document the climatological characteristics of CAGs, which includes the composite synoptic evolution of different CAG categories, the frequency of heavy precipitation associated with CAGs, and the intraseasonal variations associated with CAG activity. CAGs are identified using a new algorithm, which is introduced in section 2. Section 3 describes the frequency, composite structure, rainfall, and intraseasonal variability of CAGs. Section 4 summarizes and provides general conclusions.
2. Data and methodology
a. Datasets
CAGs are identified using four times daily (0000, 0600, 1200, and 1800 UTC) analyses from the National Centers for Environmental Prediction (NCEP) 0.5° Climate Forecast System Reanalysis (CFSR) dataset (Saha et al. 2010), between 1980 and 2010. The presence of convection associated with CAGs is confirmed using the National Climatic Data Center’s (NCDCs) 8-km gridded geostationary satellite archive (GridSat; Knapp et al. 2011) sampled every 6 h. MJO phase is determined from the Wheeler and Hendon (2004) Real-time Multivariate MJO (RMM) index available from the Center for Australian Weather and Climate Research (http://www.bom.gov.au/climate/mjo/graphics/rmm.74toRealtime.txt). Rainfall associated with CAGs is obtained from the 0.25° Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks for Climate Data Record (PERSIANN-CDR; Ashouri et al. 2015), which provides daily rainfall information for the majority of the period (1983–2010). PERSIANN-CDR preforms well in cases of extreme precipitation (Ashouri et al. 2015), which combined with its long-term period of observations makes it a suitable dataset for this study.
b. Previous identification methods of MLs
Previous studies have used a wide variety of methods to identify MLs. Mooley and Shukla (1987) identified MLs in the north Indian Ocean using surface pressure anomalies, while Chen and Weng (1999) and Yoon and Chen (2005) employed combinations of the time-mean Earth-relative streamline and pressure tendency charts. Unfortunately, techniques that use surface pressure to identify MLs are problematic over Central America, because sea level pressure can be ill-defined over complex topography. In addition, a ML’s minimum pressure is often not collocated with the large-scale circulation center (Crandall 2012). Hurley and Boos (2015) developed a global climatology of MLs, using the Hodges (1995) TRACK algorithm that follows individual 850-hPa relative vorticity maxima. They define MLs as 850-hPa vorticity maxima that last 
c. Identification algorithm for CAGs
This study employs an automated algorithm in order to identify CAGs that possess broad, closed Earth-relative low-level cyclonic circulations with large RMWs. CAGs are identified as maxima in 850-hPa circulation using a radial mean (100-km intervals) from 500 to 1000 km in radius. This technique is similar to Crandall (2012) and Crandall et al. (2014) which identified the center of a MG using radial mean circulation, but at larger radii (900–1200 km). Circulation is useful for identifying large-scale features because it indicates macroscopic rotation and is equivalent to area-averaged vorticity via Stokes’s theorem (Holton 2004, 86–93). The radial mean (500–1000 km) of area-average vorticity (hereafter AAVORT) was chosen to be consistent with the size of MDs and MGs described in previous studies. To be characterized as a CAG, a candidate system must have a maximum AAVORT 
While AAVORT identifies broad low-level cyclonic systems, it alone cannot separate CAGs from other large systems near Central America; therefore, additional criteria are needed to classify CAGs that exist for 
CAG algorithm diagnostics for three different candidate systems: (left) 0600 UTC 28 Sep 2010, (middle) 0600 UTC 21 Aug 2007, and (right) 1200 UTC 13 Oct 2002. (a)–(c) Infrared satellite imagery depicting brightness temperature (°C, shaded). (d)–(f) 850-hPa cyclonic relative vorticity (shaded, 
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
CAG algorithm diagnostics for three different candidate systems: (left) 0600 UTC 28 Sep 2010, (middle) 0600 UTC 21 Aug 2007, and (right) 1200 UTC 13 Oct 2002. (a)–(c) Infrared satellite imagery depicting brightness temperature (°C, shaded). (d)–(f) 850-hPa cyclonic relative vorticity (shaded,
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
CAG algorithm diagnostics for three different candidate systems: (left) 0600 UTC 28 Sep 2010, (middle) 0600 UTC 21 Aug 2007, and (right) 1200 UTC 13 Oct 2002. (a)–(c) Infrared satellite imagery depicting brightness temperature (°C, shaded). (d)–(f) 850-hPa cyclonic relative vorticity (shaded,
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
One final criterion is used to separate CAGs from large-scale troughs, which do not have a closed Earth-relative circulation. The existence of a closed circulation is determined by computing average 850-hPa tangential wind in 60° arc annuli between 500 and 1000 km from the center (hereafter Ar
While these criteria could be applied to any geographical area, the focus of this study is on broad cyclonic circulations that occur near Central America, defined as 5°–30°N and 70°–100°W. This region covers the Gulf of Mexico, the western Caribbean Sea, and most of Central America. Moreover, only the warm season (1 May–30 November) is considered because it coincides with the Central American rainy season (Magaña et al. 1999). These criteria yield 51 CAG cases over the 31-yr period.
The number of CAG cases appear to be relatively insensitive to the threshold values employed here (Table 1). Decreasing AAVORT or 
Sensitivity tests of CAG algorithm using variations to threshold values.
Additional quality control measures were used to eliminate other undesired cases. Two identified cases (25 September 1980, 2 November 2001) were removed from the climatology because they represented cases where an earlier-identified CAG (20 September 1980, 28 October 2001) went below threshold criteria for a time (i.e., these cases contained two discrete 
The final number of CAGs identified by this algorithm (47) is much lower than the total number of MDs or stronger cases (145) identified by Hurley and Boos (2015) for the same period and geographical area; these differences are attributable to the algorithms used. For example, the AAVORT threshold (
d. Organization of CAG categories and composites
The CAGs identified in section 2c are further subdivided based on the distinction between MDs and MGs, which have different upper-tropospheric characteristics. Recall that MDs feature an upper-tropospheric anticyclone (i.e., low PV) (Boos et al. 2015; Hurley and Boos 2015), while MGs possess an upper-tropospheric cyclone (i.e., high PV) (Molinari and Vollaro 2012; Crandall et al. 2014). Here, CAGs are subdivided based on the nearby 350-K PV at CAG identification time. The 350-K isentropic surface is used because it is located near the tropical tropopause, and was used in western Pacific MG studies (Molinari and Vollaro 2012; Crandall 2012). Presence of a nearby trough is determined using a PV test similar to the Ar
3. Climatology of Central American gyres
a. General statistics
CAGs exhibit a distinct bimodal distribution by month, with peaks in CAG occurrence in May–June (13 CAGs) and September–November (34 CAGs), while no CAGs were identified in July or August (Fig. 2a). Interestingly, the period lacking CAG activity coincides with a reduction in temporal frequency as discussed in Hastenrath (1985), and a relative reduction in precipitation over Central America called the midsummer drought (Magaña et al. 1999). By contrast, Hurley and Boos (2015) find a peak in MD activity over the same region during August. These midseason differences in ML activity between this climatology and Hurley and Boos (2015) are likely related to how MLs are identified, where the latter study is more likely to identify smaller-scale vorticity features within the ITCZ/MT in the eastern Pacific. The lack of CAGs in July–August is likely due to the evolution of the seasonal flow; this idea will be explored in section 3d.
(a) Monthly CAG genesis frequency (1980–2010), where red and blue colors represent the number of CAGs classified as nonbaroclinic and baroclinic events, respectively. (b) CAG locations within domain, where shading denotes probability of a CAG passing within a 500-km radius of each grid point per season (May–November). Genesis location of individual nonbaroclinic and baroclinic CAGs are shown using red and blue dots, respectively.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
(a) Monthly CAG genesis frequency (1980–2010), where red and blue colors represent the number of CAGs classified as nonbaroclinic and baroclinic events, respectively. (b) CAG locations within domain, where shading denotes probability of a CAG passing within a 500-km radius of each grid point per season (May–November). Genesis location of individual nonbaroclinic and baroclinic CAGs are shown using red and blue dots, respectively.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
(a) Monthly CAG genesis frequency (1980–2010), where red and blue colors represent the number of CAGs classified as nonbaroclinic and baroclinic events, respectively. (b) CAG locations within domain, where shading denotes probability of a CAG passing within a 500-km radius of each grid point per season (May–November). Genesis location of individual nonbaroclinic and baroclinic CAGs are shown using red and blue dots, respectively.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
CAG formation locations (Fig. 2b) are distributed across the east Pacific basin (12 cases), Central America (12 cases), and the Atlantic basin (23 cases), but their track density is maximized over Central America, where there is a 
Mean CAG statistics over each set of cases. The range of values is given in parentheses.
b. Composite evolution of nonbaroclinic and baroclinic CAGs
Comparing the evolution of nonbaroclinic and baroclinic CAG composites illustrates a number of critical differences in both kinematic and thermodynamic fields. Figure 3 shows the evolution of 850-hPa geopotential height and wind anomalies. Prior to and during CAG development, both composites are characterized by an area of anomalous westerly wind south of 10°N in the eastern Pacific (Figs. 3a–d). These wind anomalies are associated with a height gradient that transitions from positive height anomalies (+0.5σ) in the east and central Pacific south of 10°N to negative height anomalies (−0.5σ) northeastward toward the composite CAG center. In the nonbaroclinic CAG category, these low-level westerly anomalies gradually strengthen over time in conjunction with a westward extension of statistically significant negative height anomalies (Figs. 3a,c,e). In the baroclinic CAG category, easterly wind anomalies are present poleward of the Greater Antilles, in association with a height gradient between the CAG and an anomalously strong (+0.5–1.0σ) subtropical ridge north over the eastern United States (Figs. 3b,d,f). This enhanced subtropical ridge is not present in the nonbaroclinic CAG category. Over time, both CAG categories develop a broad circular area of statistically significant negative height anomalies (< −1.0σ) and cyclonic wind anomalies surrounding the composite CAG center; similar to the composite lower-tropospheric structure of MGs and MDs in other basins (e.g., Wu et al. 2013; Hurley and Boos 2015).
CAG composite 850-hPa geopotential height (black contours, dam), standardized geopotential height anomaly (shaded, σ), and anomalous winds (vectors, m s−1). The nonbaroclinic CAG composite at (a) t0 − 48 h, (c) t0, and (e) t0 + 48 h. The baroclinic CAG composite at (b) t0 − 48 h, (d) t0, and (f) t0 + 48 h. Red and blue circles denote the genesis points of nonbaroclinic and baroclinic CAGs, respectively, in (a)–(d), and follow the composite center CAG at t0 + 48 h in (e),(f). Stippled regions denote where the 850-hPa height anomalies are statistically significant.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
CAG composite 850-hPa geopotential height (black contours, dam), standardized geopotential height anomaly (shaded, σ), and anomalous winds (vectors, m s−1). The nonbaroclinic CAG composite at (a) t0 − 48 h, (c) t0, and (e) t0 + 48 h. The baroclinic CAG composite at (b) t0 − 48 h, (d) t0, and (f) t0 + 48 h. Red and blue circles denote the genesis points of nonbaroclinic and baroclinic CAGs, respectively, in (a)–(d), and follow the composite center CAG at t0 + 48 h in (e),(f). Stippled regions denote where the 850-hPa height anomalies are statistically significant.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
CAG composite 850-hPa geopotential height (black contours, dam), standardized geopotential height anomaly (shaded, σ), and anomalous winds (vectors, m s−1). The nonbaroclinic CAG composite at (a) t0 − 48 h, (c) t0, and (e) t0 + 48 h. The baroclinic CAG composite at (b) t0 − 48 h, (d) t0, and (f) t0 + 48 h. Red and blue circles denote the genesis points of nonbaroclinic and baroclinic CAGs, respectively, in (a)–(d), and follow the composite center CAG at t0 + 48 h in (e),(f). Stippled regions denote where the 850-hPa height anomalies are statistically significant.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
While composite nonbaroclinic and baroclinic CAGs have similar low-level mass and kinematic structures near the center, they exhibit vastly different moisture anomalies (Fig. 4). Prior to CAG development (Figs. 4a,b), statistically significant positive precipitable water anomalies (
As in Fig. 3, but for mean precipitable water (black contours, mm), and standardized anomalies (shaded, σ).
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
As in Fig. 3, but for mean precipitable water (black contours, mm), and standardized anomalies (shaded, σ).
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
As in Fig. 3, but for mean precipitable water (black contours, mm), and standardized anomalies (shaded, σ).
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
Differences in the precipitable water anomalies between nonbaroclinic and baroclinic CAG categories are hypothesized to result from differences in their upper-tropospheric PV structure (Fig. 5). At t0, nonbaroclinic CAGs have 200-hPa PV 
(a),(b) Composite 200-hPa PV (shaded, PVU), 200–850-hPa vertical wind shear (vectors, m s−1), 500-hPa ascent (purple contours with shading, 
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
(a),(b) Composite 200-hPa PV (shaded, PVU), 200–850-hPa vertical wind shear (vectors, m s−1), 500-hPa ascent (purple contours with shading,
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
(a),(b) Composite 200-hPa PV (shaded, PVU), 200–850-hPa vertical wind shear (vectors, m s−1), 500-hPa ascent (purple contours with shading,
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
Cross sections northwest–southeast through nonbaroclinic and baroclinic CAG composites also show clear vertical structural differences (Figs. 6–7). At t0 − 48 h, the composite environment of nonbaroclinic CAGs is characterized by PV 
Time evolution of northwest–southeast-oriented cross sections of the nonbaroclinic CAG composite at (a),(b) t0 − 48 h; (c),(d) t0; and (e),(f) t0 + 48 h. (left) Potential vorticity (shaded, PVU), zonal wind [dashed (solid) black contours for easterly (westerly) winds every 4 m s−1], and total wind (barbs, kt). (right) Temperature anomaly (shaded, K), potential temperature (black contours, every 5 K), and vertical ascent (purple contours with shading, 
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
Time evolution of northwest–southeast-oriented cross sections of the nonbaroclinic CAG composite at (a),(b) t0 − 48 h; (c),(d) t0; and (e),(f) t0 + 48 h. (left) Potential vorticity (shaded, PVU), zonal wind [dashed (solid) black contours for easterly (westerly) winds every 4 m s−1], and total wind (barbs, kt). (right) Temperature anomaly (shaded, K), potential temperature (black contours, every 5 K), and vertical ascent (purple contours with shading,
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
Time evolution of northwest–southeast-oriented cross sections of the nonbaroclinic CAG composite at (a),(b) t0 − 48 h; (c),(d) t0; and (e),(f) t0 + 48 h. (left) Potential vorticity (shaded, PVU), zonal wind [dashed (solid) black contours for easterly (westerly) winds every 4 m s−1], and total wind (barbs, kt). (right) Temperature anomaly (shaded, K), potential temperature (black contours, every 5 K), and vertical ascent (purple contours with shading,
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
As in Fig. 6, but for the baroclinic CAG composite, where blue circles denote the center of the baroclinic CAG composite at 850 hPa.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
As in Fig. 6, but for the baroclinic CAG composite, where blue circles denote the center of the baroclinic CAG composite at 850 hPa.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
As in Fig. 6, but for the baroclinic CAG composite, where blue circles denote the center of the baroclinic CAG composite at 850 hPa.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
There are notable differences in the vertical structure of baroclinic CAGs compared to nonbaroclinic CAGs (Fig. 7). At t0 − 48 h, the composite environment of baroclinic CAGs is characterized by a northwest–southeast upper-tropospheric PV gradient (Fig. 7a), deep tropospheric ascent at and to the southeast of the CAG center (Fig. 7b), and a northwest–southeast temperature gradient (Fig. 7b). Comparing the baroclinic CAG composite to the nonbaroclinic CAG composite, the most notable difference is the broad upper-tropospheric trough northwest of the CAG center point, which is associated with a large (
c. Rainfall
Individual CAG events are often accompanied by excessive precipitation that can result in flooding to portions of Central America and the Caribbean (Pasch and Roberts 2006; Brennan 2010; Blake 2011). Since widespread precipitation is also observed within the CAG composites (Figs. 5e–f), this section will compare composite CAG rainfall to the rainy season (May–November) climatology. In addition, CAG rainfall will be described in terms of extreme precipitation frequency and coverage.
Figure 8 compares the Earth-relative average daily rainfall of CAG events available in the PERSIANN-CDR dataset (N = 39) to the average daily rainfall during the rainy season. The Earth-relative framework is used here because climatological precipitation is strongly influenced by terrain over Central America; the drawback of this approach is that individual CAG rainfall areas do not always overlap, which in turn dilutes the signal. Composite daily rainfall during CAG events (
Average daily rainfall rate composite for CAGs (shaded, mm day−1) using rainfall observed within 10° of a CAG center compared to climatological rainfall rate during 1 May–30 Nov (black contours, every 5 mm day−1).
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
Average daily rainfall rate composite for CAGs (shaded, mm day−1) using rainfall observed within 10° of a CAG center compared to climatological rainfall rate during 1 May–30 Nov (black contours, every 5 mm day−1).
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
Average daily rainfall rate composite for CAGs (shaded, mm day−1) using rainfall observed within 10° of a CAG center compared to climatological rainfall rate during 1 May–30 Nov (black contours, every 5 mm day−1).
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
CAGs are frequently associated with intense daily rainfall rates. Because of large variations in the climatological daily precipitation rate across Central America and adjacent oceans, the definition of extreme precipitation must vary in space. Here, extreme precipitation is defined as the 95th percentile of all precipitating days in the rainy season, which is a threshold that has been used in past studies (e.g., Curtis et al. 2007). Next, the number of locations within 10° of the CAG center where the daily precipitation exceeds that threshold is counted in 0.25° boxes. Not surprisingly, locations characterized by the highest CAG mean rainfall rate (Central American coastal regions along the east Pacific and Caribbean) also experience the highest number of days with extreme precipitation (Fig. 9a). Within a 10° radius of a typical CAG event, approximately 25% of grid points experience at least one day of extreme precipitation (Fig. 9b). Each CAG event observed in the PERSIANN-CDR climatology has a location (0.25° box) with at least 2 days of extreme precipitation while a large minority of CAG cases (17 out of 39, 43.6%) possess a location with at least 4 days of extreme precipitation. These statistics strongly suggest that CAG events are tied to extreme precipitation occurrence, often for multiple days in a large area, and help explain why CAG events often feature catastrophic flooding (Pasch and Roberts 2006; Brennan 2010; Blake 2011).
(a) The number of extreme precipitation days attributed to CAG events (shaded, days) within 10° of a CAG center and the daily rainfall rate that represents the 95th percentile of all rainfall events (contours, every 10 mm day−1 with thickness denoting magnitude). (b) A bar graph showing the fraction of area within 10° of the CAG of extreme precipitation, where the number of extreme precipitation days is on the x axis and area percentage on the y axis. The number of CAGs that possess at least one point of extreme precipitation is labeled within.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
(a) The number of extreme precipitation days attributed to CAG events (shaded, days) within 10° of a CAG center and the daily rainfall rate that represents the 95th percentile of all rainfall events (contours, every 10 mm day−1 with thickness denoting magnitude). (b) A bar graph showing the fraction of area within 10° of the CAG of extreme precipitation, where the number of extreme precipitation days is on the x axis and area percentage on the y axis. The number of CAGs that possess at least one point of extreme precipitation is labeled within.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
(a) The number of extreme precipitation days attributed to CAG events (shaded, days) within 10° of a CAG center and the daily rainfall rate that represents the 95th percentile of all rainfall events (contours, every 10 mm day−1 with thickness denoting magnitude). (b) A bar graph showing the fraction of area within 10° of the CAG of extreme precipitation, where the number of extreme precipitation days is on the x axis and area percentage on the y axis. The number of CAGs that possess at least one point of extreme precipitation is labeled within.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
d. Seasonal variability
The remainder of this paper explores the bimodal seasonal peak in CAG activity and specifically the lack of CAGs during July and August (Fig. 2a). This period also coincides with a relative reduction of seasonal rainfall over Central America (i.e., the midsummer drought; Magaña et al. 1999). Small et al. (2007) noted that during the midsummer drought, low-level easterly wind anomalies are observed in the east Pacific equatorward of Central America, which results in moisture flux divergence near Central America. By contrast, CAGs are preceded with an extensive corridor of 850-hPa westerly wind anomalies in the same region (Fig. 3). As a consequence, one possible explanation for the lack of CAGs during July–August is that seasonal changes in the zonal wind in the eastern Pacific adjacent to Central America are not favorable for CAG formation.
Low-level zonal winds in regions adjacent to Central America experience large changes in magnitude during the rainy season and occur in conjunction with seasonal changes in CAG occurrence. Figure 10 depicts the frequency of zonal wind during 1 May–30 November for boxes in the eastern Pacific (0°–15°N, 85°–110°W, Fig. 10a) and Caribbean Sea (10°–20°N, 60°–85°W, Fig. 10b), encompassing the statistically significant anomalies observed in Fig. 3. Not surprisingly, mean easterly zonal winds are most frequently observed in the eastern Pacific and Caribbean regions, but are weakest in May–June and September–October (−1 m s−1 in the eastern Pacific and −6 m s−1 in the Caribbean, respectively). Large variability about the mean exists in the eastern Pacific region, which occasionally yields westerly zonal wind with a frequency between 10% and 20% (of 48-h periods) during May–June and September–October when the mean easterly flow is weaker. These time periods coincide with all CAG events, and a majority of CAGs are preceded by westerly zonal wind in the east Pacific region (Fig. 10a). Conversely, when mean easterly zonal wind is strongest in July–August, westerly zonal winds rarely occur (frequency of 
The 48-h running mean 850-hPa zonal wind frequency (shaded, %) for (a) averaged over the eastern Pacific domain (0°–15°N, 110°–80°W) and (b) averaged over the Caribbean Sea domain (10°–20°N, 85°–60°W) as a function of date. The black line in (a),(b) shows the climatological zonal wind for that time period. Red and blue circles denote the zonal wind averaged 48 h before genesis of nonbaroclinic and baroclinic CAGs, respectively.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
The 48-h running mean 850-hPa zonal wind frequency (shaded, %) for (a) averaged over the eastern Pacific domain (0°–15°N, 110°–80°W) and (b) averaged over the Caribbean Sea domain (10°–20°N, 85°–60°W) as a function of date. The black line in (a),(b) shows the climatological zonal wind for that time period. Red and blue circles denote the zonal wind averaged 48 h before genesis of nonbaroclinic and baroclinic CAGs, respectively.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
The 48-h running mean 850-hPa zonal wind frequency (shaded, %) for (a) averaged over the eastern Pacific domain (0°–15°N, 110°–80°W) and (b) averaged over the Caribbean Sea domain (10°–20°N, 85°–60°W) as a function of date. The black line in (a),(b) shows the climatological zonal wind for that time period. Red and blue circles denote the zonal wind averaged 48 h before genesis of nonbaroclinic and baroclinic CAGs, respectively.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
There are several hypotheses that may explain why westerly zonal winds in the eastern Pacific result in more favorable conditions for CAG development. One possibility is that the westerly winds provide a source of zonal kinetic energy that is converted to eddy kinetic energy via barotropic conversion (Krishnamurti et al. 1976; Aiyyer and Molinari 2008). This may manifest as the conversion of cyclonic shear vorticity into cyclonic curvature vorticity, as westerly winds equatorward of Central America (Fig. 10a) interact with easterly trade winds poleward of Central America (Fig. 10b). Another possibility is that westerly zonal winds in the eastern Pacific may trigger a series of processes that provide a source of eddy potential energy, which through diabatic processes is converted to eddy kinetic energy (Krishnamurti et al. 1976). First, the interaction of eastern Pacific westerly lower-tropospheric flow with Caribbean easterly trade winds can promote lower-tropospheric convergence over Central America. Then, if enhanced moisture preexisting near Central America (i.e., Fig. 4) overlaps with this lower-tropospheric convergence, deep moist convection can result (Fig. 5). Latent heat release associated with deep moist convection can diabatically redistribute the PV profile below, enhancing the lower-tropospheric cyclonic flow that characterizes the CAG circulation (Fig. 6). A definitive explanation for the role of the synoptic environment is beyond the scope of this study.
e. MJO
On the intraseasonal scale, an active MJO can enhance westerly zonal winds in the eastern Pacific and promote convection over Central America (Maloney and Hartmann 2000); both features that may be important to CAG development. Stratifying CAG cases by MJO RMM phase (using Wheeler and Hendon 2004) indicates a large majority of CAG cases (72.3%) occur in phases 1, 2, and 8 (Fig. 11a); which are typically associated with active convection over Central America (Wheeler and Hendon 2004). The proportion of CAG cases in these RMM phases remains consistent even when removing CAG cases associated with weak MJOs (as defined in Wheeler and Hendon 2004).
(a) Bar graph of the number of CAG events in each RMM phase, where nonbaroclinic and baroclinic CAGs denoted by red and blue, respectively. (b),(c) Composite 850-hPa geopotential height (black contours, dam), standardized geopotential height anomaly (shaded, σ), and anomalous winds (vectors, m s−1) for (b) MJO RMM phases 8, 1, and 2 (N = 3041 days) and (c) MJO RMM phases 3–7 (N = 4365 days) occurring during 1 May–30 Nov.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
(a) Bar graph of the number of CAG events in each RMM phase, where nonbaroclinic and baroclinic CAGs denoted by red and blue, respectively. (b),(c) Composite 850-hPa geopotential height (black contours, dam), standardized geopotential height anomaly (shaded, σ), and anomalous winds (vectors, m s−1) for (b) MJO RMM phases 8, 1, and 2 (N = 3041 days) and (c) MJO RMM phases 3–7 (N = 4365 days) occurring during 1 May–30 Nov.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
(a) Bar graph of the number of CAG events in each RMM phase, where nonbaroclinic and baroclinic CAGs denoted by red and blue, respectively. (b),(c) Composite 850-hPa geopotential height (black contours, dam), standardized geopotential height anomaly (shaded, σ), and anomalous winds (vectors, m s−1) for (b) MJO RMM phases 8, 1, and 2 (N = 3041 days) and (c) MJO RMM phases 3–7 (N = 4365 days) occurring during 1 May–30 Nov.
Citation: Monthly Weather Review 145, 5; 10.1175/MWR-D-16-0411.1
One potential reason for this MJO preference is that these MJO phases have similar planetary-scale anomalies to the CAG composites (i.e., Fig. 3). This possibility is evaluated by compositing the CFSR daily 850-hPa geopotential height and wind anomalies of MJO phases associated with enhanced CAG activity during May–November (phases 1, 2, 8; N = 3041 days; Fig. 11b) and comparing them to the remaining MJO phases (3, 4, 5, 6; N = 4365 days; Fig. 11c). MJO phases 1, 2, and 8, which coincide with enhanced CAG activity, are characterized by negative height anomalies over Central America, increased height anomalies equatorward in the eastern Pacific, and westerly wind anomalies maximized in the eastern Pacific along an anomalous geopotential height gradient (Fig. 11b). By contrast, the remaining MJO phases have composite easterly wind anomalies associated with the opposite geopotential height anomaly gradient in the eastern Pacific (Fig. 11c). The low-level geopotential height and winds observed in MJO phases with enhanced CAG activity are consistent with the eastern Pacific synoptic–planetary pattern observed in Fig. 3. The MJO composite structure is also consistent with Aiyyer and Molinari (2008), where an active MJO period was associated with lower-tropospheric westerly winds in the eastern Pacific. These westerly winds aid in convergence and cyclonic relative vorticity along Central America, and are favorable factors for the development of CAGs.
4. Conclusions
The purpose of this study was to document the climatological characteristics of CAGs, which are broad lower-tropospheric cyclonic circulations that feature light wind cores, closed Earth-relative circulations, and widespread heavy precipitation. These systems were identified via an algorithm that evaluated area-average relatively vorticity maxima with large RMWs and closed Earth-relative circulations, which distinguish these features from TCs and open troughs. The resulting CAG cases were further stratified by the nearby upper-tropospheric (350 K) PV structure to distinguish between different types of CAG.
CAGs were identified during the rainy season in May–June and September–November with an annual frequency of 1.5 cases per year (Fig. 2a). These peaks and annual frequency of CAG activity in the present study are similar to temporals previously discussed in Hastenrath (1985). By contrast, Hurley and Boos (2015) identified a larger number of MDs in the same domain, with a peak in August. These differences are likely due to algorithm differences in Hurley and Boos (2015), which is more prone to identifying TCs and nonclosed vorticity features in the eastern Pacific ITCZ/MT.
Separating CAGs into nonbaroclinic and baroclinic types yields dynamically distinct structures analogous to MDs and MGs. The composite evolution of nonbaroclinic CAGs shows the development of a broad lower-tropospheric circulation and upper-tropospheric warm core, as anomalous 850-hPa westerly flow from the east Pacific (Fig. 3) converges over enhanced precipitable water near Central America (Fig. 4), promoting deep tropospheric ascent and precipitation surrounding the circulation (Fig. 5). Implied convective activity from the vertical ascent is associated with warming the upper troposphere, enhancing lower-tropospheric PV and cyclonic winds below the level of maximum heating (Fig. 6). In contrast, baroclinic CAGs have a highly asymmetrical distribution of moisture and precipitation, where enhanced precipitable water and precipitation occur to the east of an amplifying upper-tropospheric trough (Figs. 4 and 5). Ultimately, the upper-tropospheric trough becomes superimposed onto the lower-tropospheric circulation, and resembles a tropospheric cold core system (Fig. 7). Both types of CAG feature extreme precipitation over large areas for several days (Fig. 9), which can result in flooding.
The bimodal seasonal frequency of CAGs is linked to changes in the seasonal zonal winds in the eastern Pacific and the Caribbean (Fig. 10). CAGs occur most frequently in May–June and September–November, when zonal 850-hPa westerly winds in the eastern Pacific are more common (Fig. 10a). Periods of westerly zonal winds in the eastern Pacific may result in lower-tropospheric cyclonic shear vorticity, which can then be converted to cyclonic curvature vorticity over Central America. The anomalous westerly zonal wind pattern observed with CAGs is consistent with phases 1, 2, and 8 of the MJO, which coincide with a majority of CAGs (Fig. 11).
There are a number of aspects of CAGs that were not addressed here, which merit further work. It is likely that the higher topography of Central America plays a critical role in the generation and organization of convection and precipitation associated with CAG events. Moreover, gaps within the elevated terrain of Central America (Chivela Pass, Gulf of Papagayo, and Gulf of Panama) are known to generate low-level vorticity (Holbach and Bourassa 2014), which may also influence CAG organization. The CAG climatology and algorithm presented here can be applied to any region of the tropics, and may yield a different climatology relative to Hurley and Boos (2015). Future work may also investigate CAG predictability in numerical weather prediction, especially whether nonbaroclinic or baroclinic CAG events have different predictability given their tropical (MJO) or extratropical (Rossby wave breaking) influences. Finally, ML studies in the west Pacific basin have documented their influence on the motion of nearby TCs (Lander 1994; Carr and Elsberry 1995; Crandall et al. 2014). Similar interactions have been noted between CAGs and TCs (Pasch and Roberts 2006; Montgomery et al. 2012), and documenting the nature by which CAGs influence TC motion could improve TC track forecasts near CAGs.
Acknowledgments
The authors wish to thank Ms. Alicia Bentley (University at Albany), Dr. Alan Brammer (University at Albany), Mr. Brian Crandall (Atmospheric Research Science Center), Dr. Kyle Griffin (Riskpulse), Dr. Matthew Janiga (Naval Postgraduate School), Dr. John Molinari (University at Albany), and Ms. Ajda Savarin (University of Miami) for helpful discussions and research assistance. This research was funded by NSF Grant ATM-0849491 and NOAA Grant NA14OAR4830172.
REFERENCES
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, 2691–2704, doi:10.1175/2007JAS2348.1.
Aldinger, W. T., and W. Stapler, 1998: 1998 annual tropical cyclone report. Tech. Rep. ADA381675, JTWC, U.S. Naval Pacific Meteorology and Oceanography Center, B, 224 pp.
American Meteorological Society, 2016a: Monsoon depression. Glossary of Meteorology. [Available online at http://glossary.ametsoc.org/wiki/Monsoon_depression.]
American Meteorological Society, 2016b: Monsoon gyre. Glossary of Meteorology. [Available online at http://glossary.ametsoc.org/wiki/Monsoon_gyre.]
Ashouri, H., K. L. Hsu, S. Sorooshian, D. K. Braithwaite, K. R. Knapp, L. D. Cecil, B. R. Nelson, and O. P. Prat, 2015: PERSIANN-CDR: Daily precipitation climate data record from multisatellite observations for hydrological and climate studies. Bull. Amer. Meteor. Soc., 96, 69–83, doi:10.1175/BAMS-D-13-00068.1.
Baray, J.-L., G. Clain, M. Plu, E. Feld, and P. Caroff, 2010: Occurrence of monsoon depressions in the Southwest Indian Ocean: Synoptic descriptions and stratosphere to troposphere exchange investigations. J. Geophys. Res., 115, D17108, doi:10.1029/2009JD013390.
Beattie, J. C., and R. L. Elsberry, 2013: Horizontal structure of monsoon depressions in the western North Pacific at formation time. Geophys. Res. Lett., 40, 983–987, doi:10.1002/grl.50198.
Blake, E. S., 2011: Tropical Cyclone Report: Tropical Storm Nicole (28–29 September 2010). Tech. Rep. AL162010, National Hurricane Center, 15 pp. [Available online at http://www.nhc.noaa.gov/data/tcr/AL162010_Nicole.pdf.]
Boos, W. R., J. V. Hurley, and V. S. Murthy, 2015: Adiabatic westward drift of Indian monsoon depressions. Quart. J. Roy. Meteor. Soc., 141, 1035–1048, doi:10.1002/qj.2454.
Brennan, M. J., 2010: Tropical Cyclone Report: Tropical Storm Matthew (23–26 September 2010). Tech. Rep. AL152010, National Hurricane Center, 15 pp. [Available online at http://www.nhc.noaa.gov/data/tcr/AL152010_Matthew.pdf.]
Carr, L. E., and R. L. Elsberry, 1995: Monsoonal interactions leading to sudden tropical cyclone track changes. Mon. Wea. Rev., 123, 265–290, doi:10.1175/1520-0493(1995)123<0265:MILTST>2.0.CO;2.
Chen, T.-C., and S.-P. Weng, 1999: Interannual and intraseasonal variations in monsoon depressions and their westward-propagating predecessors. Mon. Wea. Rev., 127, 1005–1020, doi:10.1175/1520-0493(1999)127<1005:IAIVIM>2.0.CO;2.
Chen, T.-C., S.-Y. Wang, M.-C. Yen, and W. A. Gallus, 2004: Role of the monsoon gyre in the interannual variation of tropical cyclone formation over the western North Pacific. Wea. Forecasting, 19, 776–785, doi:10.1175/1520-0434(2004)019<0776:ROTMGI>2.0.CO;2.
Crandall, B., 2012: An analysis of the formation and evolution of the 1989 western North Pacific subtropical gyre. M.S. thesis, Dept. of Atmospheric and Environmental Sciences, University at Albany, State University of New York, 91 pp.
Crandall, B., J. Molinari, and D. Vollaro, 2014: Forecasting challenges associated with tropical cyclones within subtropical gyres. Wea. Forecasting, 29, 99–114, doi:10.1175/WAF-D-13-00053.1.
Curtis, S., A. Salahuddin, R. F. Adler, G. J. Huffman, G. Gu, and Y. Hong, 2007: Precipitation extremes estimated by GPCP and TRMM: ENSO relationships. J. Hydrometeor., 8, 678–689, doi:10.1175/JHM601.1.
Davidson, N. E., and G. J. Holland, 1987: A diagnostic analysis of two intense monsoon depressions over Australia. Mon. Wea. Rev., 115, 380–392, doi:10.1175/1520-0493(1987)115<0380:ADAOTI>2.0.CO;2.
Davis, C. A., and L. F. Bosart, 2003: Baroclinically induced tropical cyclogenesis. Mon. Wea. Rev., 131, 2730–2747, doi:10.1175/1520-0493(2003)131<2730:BITC>2.0.CO;2.
Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553–597, doi:10.1002/qj.828.
Derbyshire, S. H., I. Beau, P. Bechtold, J. Grandpeix, J. Piriou, J. Redelsperger, and P. M. M. Soares, 2004: Sensitivity of moist convection to environmental humidity. Quart. J. Roy. Meteor. Soc., 130, 3055–3079, doi:10.1256/qj.03.130.
Douglas, M. W., 1992: Structure and dynamics of two monsoon depressions. Part I: Observed structure. Mon. Wea. Rev., 120, 1524–1547, doi:10.1175/1520-0493(1992)120<1524:SADOTM>2.0.CO;2.
Eliot, J., 1900: Hand Book of Cyclonic Storms in the Bay of Bengal for the Use of Sailors. 2nd ed. Meteorological Department of the Government of India, Calcutta, India, 316 pp.
Fernandez, W., and J. A. Barrantes, 1996: The Central American temporal: A long-lived tropical rain-producing system. Top. Meteor. Oceanogr., 3 (2), 73–88.
Galarneau, T. J., R. McTaggart-Cowan, L. F. Bosart, and C. A. Davis, 2015: Development of North Atlantic tropical disturbances near upper-level potential vorticity streamers. J. Atmos. Sci., 72, 572–597, doi:10.1175/JAS-D-14-0106.1.
Godbole, R. V., 1977: The composite structure of the monsoon depression. Tellus, 29A, 25–40, doi:10.3402/tellusa.v29i1.11327.
Grabowski, W. W., and M. W. Moncrieff, 2004: Moisture–convection feedback in the tropics. Quart. J. Roy. Meteor. Soc., 130, 3081–3104, doi:10.1256/qj.03.135.
Harr, P. A., R. L. Elsberry, and J. C. L. Chan, 1996: Transformation of a large monsoon depression to a tropical storm during TCM-93. Mon. Wea. Rev., 124, 2625–2643, doi:10.1175/1520-0493(1996)124<2625:TOALMD>2.0.CO;2.
Hastenrath, S., 1985: Climate and Circulation of the Tropics. 1st ed. D. Reidel Publishing Company, 455 pp.
Hawkins, H. F., and S. M. Imbembo, 1976: The structure of a small, intense hurricane—Inez 1966. Mon. Wea. Rev., 104, 418–442, doi:10.1175/1520-0493(1976)104<0418:TSOASI>2.0.CO;2.
Hodges, K. I., 1995: Feature tracking on the unit-sphere. Mon. Wea. Rev., 123, 3458–3465, doi:10.1175/1520-0493(1995)123<3458:FTOTUS>2.0.CO;2.
Holbach, H. M., and M. A. Bourassa, 2014: The effects of gap-wind-induced vorticity, the monsoon trough, and the ITCZ on east Pacific tropical cyclogenesis. Mon. Wea. Rev., 142, 1312–1325, doi:10.1175/MWR-D-13-00218.1.
Holland, G. J., 1995: Scale interaction in the western Pacific monsoon. Meteor. Atmos. Phys., 56, 57–79, doi:10.1007/BF01022521.
Holton, J. R., 2004: An Introduction to Dynamic Meteorology. 4th ed. Elsevier Academic Press, 535 pp.
Hurley, J. V., and W. R. Boos, 2015: A global climatology of monsoon low pressure systems. Quart. J. Roy. Meteor. Soc., 141, 1049–1064, doi:10.1002/qj.2447.
Kimball, S. K., and M. S. Mulekar, 2004: A 15-year climatology of North Atlantic tropical cyclones. Part I: Size parameters. J. Climate, 17, 3555–3575, doi:10.1175/1520-0442(2004)017<3555:AYCONA>2.0.CO;2.
Knapp, K. R., and Coauthors, 2011: Globally gridded satellite observations for climate studies. Bull. Amer. Meteor. Soc., 92, 893–907, doi:10.1175/2011BAMS3039.1.
Krishnamurti, T. N., M. Kanamitsu, R. Godbole, C.-B. Chang, F. Carr, and J. H. Chow, 1975: Study of a Monsoon Depression (I) Synoptic Structure. J. Meteor. Soc. Japan, 53, 227–240.
Krishnamurti, T. N., M. Kanamitsu, R. Godbole, C.-B. Chang, F. Carr, and J. H. Chow, 1976: Study of a Monsoon Depression (II) Dynamical Structure. J. Meteor. Soc. Japan, 54 (4), 208–225.
Lander, M. A., 1994: Description of a monsoon gyre and its effects on the tropical cyclones in the western North Pacific during August 1991. Wea. Forecasting, 9, 640–654, doi:10.1175/1520-0434(1994)009<0640:DOAMGA>2.0.CO;2.
Lawrence, M. B., 1998: Tropical Cyclone Report: Tropical Storm Frances. Tech. Rep. AL061998, National Hurricane Center, 8 pp. [Available online at http://www.nhc.noaa.gov/data/tcr/AL061998_Frances.pdf.]
Madden, R. A., and P. R. Julian, 1971: Detection of a 40–50-day oscillation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28, 702–708, doi:10.1175/1520-0469(1971)028<0702:DOADOI>2.0.CO;2.
Magaña, V., J. A. Amador, and S. Medina, 1999: The midsummer drought over Mexico and Central America. J. Climate, 12, 1577–1588, doi:10.1175/1520-0442(1999)012<1577:TMDOMA>2.0.CO;2.
Maloney, E. D., and D. L. Hartmann, 2000: Modulation of hurricane activity in the Gulf of Mexico by the Madden-Julian Oscillation. Science, 287, 2002–2004, doi:10.1126/science.287.5460.2002.
Molinari, J., and D. Vollaro, 2012: A subtropical cyclonic gyre associated with interactions of the MJO and the midlatitude jet. Mon. Wea. Rev., 140, 343–357, doi:10.1175/MWR-D-11-00049.1.
Montgomery, M. T., and Coauthors, 2012: The Pre-Depression Investigation of Cloud-Systems in the Tropics (PREDICT) Experiment: Scientific basis, new analysis tools, and some first results. Bull. Amer. Meteor. Soc., 93, 153–172, doi:10.1175/BAMS-D-11-00046.1.
Mooley, D., and J. Shukla, 1987: Characteristics of the westward moving summer monsoon low pressure systems over the Indian region and their relationship with the monsoon rainfall. Tech. Rep., Center for Ocean–Land–Atmosphere Interactions, University of Maryland, College Park, MD, 218 pp. [Available online at http://cola.gmu.edu/people/Shukla’s%20Articles/1988/Characteristics.pdf.]
Pasch, R. J., and D. P. Roberts, 2006: Tropical Cyclone Report: Hurricane Stan 1–5 October 2005. Tech. Rep. AL202005, National Hurricane Center, 12 pp. [Available online at http://www.nhc.noaa.gov/data/tcr/AL202005_Stan.pdf.]
Pasch, R. J., L. A. Avila, and J. L. Guiney, 2001: Atlantic hurricane season of 1998. Mon. Wea. Rev., 129, 3085–3123, doi:10.1175/1520-0493(2001)129<3085:AHSO>2.0.CO;2.
Piddington, H., 1876: Sailor’s Horn-Book for the Law of Storms. 6th ed. Williams & Norgate, London, England, 452 pp.
Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 1015–1057, doi:10.1175/2010BAMS3001.1.
Sanders, F., 1984: Quasi-geostrophic diagnosis of the monsoon depression of 5–8 July 1979. J. Atmos. Sci., 41, 538–552, doi:10.1175/1520-0469(1984)041<0538:QGDOTM>2.0.CO;2.
Schenkel, B. A., and R. E. Hart, 2012: An examination of tropical cyclone position, intensity, and intensity life cycle within atmospheric reanalysis datasets. J. Climate, 25, 3453–3475, doi:10.1175/2011JCLI4208.1.
Shieh, O. H., and S. J. Colucci, 2010: Local minimum of tropical cyclogenesis in the eastern Caribbean. Bull. Amer. Meteor. Soc., 91, 185–196, doi:10.1175/2009BAMS2822.1.
Small, R. J. O., S. P. De Szoeke, and S.-P. Xie, 2007: The Central American midsummer drought: Regional aspects and large-scale forcing. J. Climate, 20, 4853–4873, doi:10.1175/JCLI4261.1.
Wheeler, M. C., 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, 1917–1932, doi:10.1175/1520-0493(2004)132<1917:AARMMI>2.0.CO;2.
Wu, L., H. Zong, and J. Liang, 2013: Observational analysis of tropical cyclone formation associated with monsoon gyres. J. Atmos. Sci., 70, 1023–1034, doi:10.1175/JAS-D-12-0117.1.
Yoon, J.-H., and T.-C. Chen, 2005: Water vapor budget of the Indian monsoon depression. Tellus, 57A, 770–782, doi:10.1111/j.1600-0870.2005.00145.x.
Zuluaga, M. D., and R. A. Houze Jr., 2015: Extreme convection of the near-equatorial Americas, Africa, and adjoining oceans as seen by TRMM. Mon. Wea. Rev., 143, 298–316, doi:10.1175/MWR-D-14-00109.1.