1. Introduction
The environmental conditions favorable for tropical cyclogenesis have been well documented in several studies (Riehl 1954; Gray 1968, 1975; McBride and Zehr 1981). Qualitatively, the necessary conditions for tropical cyclogenesis are an environment containing a concentrated area of cyclonic relative vorticity in the low troposphere and divergence in the upper troposphere, high sea surface temperatures (SSTs), low vertical wind shear, a deep moist layer, and persistent deep convection (Lee 1989; Zehr 1992; Kerns and Zipser 2009; Fu et al. 2012; Kerns and Chen 2013).
While the aforementioned studies focus primarily on the large-scale preconditioning, the recent focus has been placed on the mesoscale organization that occurs during the genesis process. It is now well accepted that mesoscale convective systems (MCSs) and their associated dynamic and thermodynamic responses [e.g., spinup of the secondary circulation as a result of latent heating; Ooyama (1982)] must play a role in intensification processes such as genesis. Even though the mesoscale convergence associated with cloud systems within a single MCS have been observed to support, or even directly contribute to, tropical depression formation (Zipser and Gautier 1978; Molinari et al. 2004; Houze et al. 2009), any pathway to tropical cyclogenesis likely requires multiple convective periods.
Formation is rarely an instantaneous event in time; rather, the formation process typically involves “priming” or “incubation,” in which a series of convective pulses or events modify the kinematic and thermodynamic environments in a way that makes the formation of a tropical cyclone more likely. While some studies identify MCSs and the formation of mesoscale convective vortices (MCVs) at midlevels within the stratiform rain region as one pathway to genesis (Harr et al. 1996a,b; Ritchie and Holland 1997; Simpson et al. 1997), others identify a pathway that involves the aggregation or axisymmetrization of cyclonically rotating convective cores known as vortical hot towers (VHTs) within an already cyclonic vorticity-rich region, such as that associated with an MCV (Montgomery and Enagonio 1998; Möller and Montgomery 2000; Hendricks et al. 2004; Montgomery et al. 2006; Sippel et al. 2006; Bell and Montgomery 2010; Braun et al. 2010). Remnant convective-scale vortical circulations from previous convective updrafts have also been identified observationally within the stratiform region and could support cyclogenesis through their axisymmetrization into the larger-scale vortex (Houze et al. 2009). With respect to the thermodynamic evolution, a potentially important role of convection is to moisten the midlevels to near saturation (Bister and Emanuel 1997; Raymond et al. 1998; Nolan 2007; Montgomery and Smith 2011; Raymond et al. 2011; Wang 2012) and stabilize the low troposphere, which subsequently favors a bottom-heavy mass flux profile and low-level convergence (Raymond and Sessions 2007; Raymond and López-Carrillo 2011; Raymond et al. 2011; Davis and Ahijevych 2012; Zawislak and Zipser 2014).
Observational case studies of tropical cyclogenesis commonly rely on some combination of aircraft in situ data (flight-level winds, radar, and dropsondes) and coincident infrared (IR) satellite imagery to analyze genesis processes (Zipser and Gautier 1978; Harr et al. 1996a,b; Bister and Emanuel 1997; Ritchie and Holland 1997; Simpson et al. 1997; Reasor et al. 2005; Houze et al. 2009). Given that they are typically only available once or twice a day and confined to within a couple of days of genesis, analyses utilizing airborne observations often inadequately characterize the evolution of the convective systems throughout the pregenesis stage. Studies that attempt to characterize the evolution of deep convective properties in developing disturbances typically utilize IR (Zehr 1992; Hopsch et al. 2010; Kerns and Chen 2013) and/or lightning data (Chronis et al. 2007; Leary and Ritchie 2009) as proxies for convective areal coverage and intensity. Although IR provides one source for inferring convective intensity, widespread cold brightness temperatures in IR associated with cirrus outflow can be misinterpreted as active deep convection, or if convection is active, mask the convective organization underneath the cold cirrus shield. For this reason, passive microwave (PMW) satellite data, which give direct evidence of the precipitation rate and organization underneath the cirrus shield, are preferred for identifying convective properties. Leppert et al. (2013a,b) take a different approach from those studies by exclusively using IR and/or lightning by synthesizing those datasets with PMW information from the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI). Composite results from developing and nondeveloping easterly waves in the Atlantic, viewed within both the Eulerian (Leppert et al. 2013b) and Lagrangian [or wave following; Leppert et al. (2013a)] frameworks, indicate that the coverage of convection may be more important to tropical cyclogenesis than convective intensity. This conclusion contrasts with those from other studies, which observed enhanced lightning, and thus more intense convection, in developing disturbances than nondeveloping disturbances (Leary and Ritchie 2009; Chronis et al. 2007). Larger raining areas have also been found to be more important than convective intensity for rapidly intensifying (RI) tropical cyclones (Jiang and Ramirez 2013).
The objective of this study is to complement the efforts of the aforementioned studies by also quantifying the convective properties of developing and nondeveloping disturbances. However, whereas those studies offer composite results, in this study the convective evolution of developing and nondeveloping disturbances is also evaluated within the context of multiple case studies, in which data from multiple satellite-borne PMW instruments are supplemented with a comprehensive dataset of in situ aircraft observations. The continuity of observations from both aircraft and satellite platforms should help elucidate the properties of convective systems involved in organizing the circulation, as well as offer insight into how the timing of intensification relates to deep convective activity. Specifically, the following questions will be addressed:
Do convective properties exhibit a distinguishable trend as genesis nears?
Is there a credible connection between convection and the organization, or intensification, of the incipient circulation as seen by in situ observations?
Are pregenesis convective properties unique compared to those of nondeveloping disturbances?
Convective properties to be investigated include the area, intensity, frequency, and proximity of deep convection to the disturbance center. Question 2 will not only address the pregenesis trends of the areal coverage, intensity, and proximity of deep convection, but also identify whether any particular day during the pregenesis stage (e.g., the day of genesis) exhibits unique properties as compared to the other days. A specific emphasis will be placed on evaluating the properties of intense convection, as well as using available in situ data to determine whether there is a close relationship between the timing and location of intense convective events and tropical cyclogenesis. Although the relative proportion of stratiform and convective precipitation is critical for determining the vertical distribution of the latent heat release, given the difficulty in differentiating stratiform areas in PMW satellite data, stratiform rain will not be explicitly analyzed in this study. Instead, the evolution of the total raining area, which has contributions from both stratiform and convective precipitation, and the frequency of rainfall near the center, will be included in the analysis.
Unlike many previous observational genesis case studies, which usually describe a single case, this study will combine information from 12 developing and 3 nondeveloping cases. The majority of the cases will be those sampled during the coordinated triagency field deployment in the Atlantic during August and September of 2010. These programs include the National Aeronautics and Space Administration (NASA) Genesis and Rapid Intensification Processes (GRIP; Braun et al. 2013) experiment, the National Science Foundation/National Center for Atmospheric Research (NSF/NCAR) Pre-Depression Investigation of Cloud-Systems in the Tropics (PREDICT; Montgomery et al. 2012) experiment, and the National Oceanic and Atmospheric Administration (NOAA) Intensity Forecast Experiment (IFEX; Rogers et al. 2006, 2013). Given the unprecedented continuity of observations achieved during the PREDICT–GRIP–IFEX (PGI; triagency) field campaigns, as well as the availability of data from well-sampled cases from other recent (since 2005) field programs, these datasets provide arguably the best opportunity to date to synthesize observations to adequately describe the time evolution of the pregenesis disturbance.
2. Data and methodology
a. Case descriptions
Table 1 summarizes the cases included in the dataset. Many of the developing cases (Earl, Fiona, Karl, Matthew, and Nicole) and all of the nondeveloping cases examined (ex-Gaston, PGI-271 and -30) are from PGI. In this study, Gaston will be treated as both developing (pregenesis through 1200 UTC 2 September, after which Gaston is downgraded to a remnant low) and nondeveloping (from 1800 UTC 2 September to the end of the track); all PGI flights sampled Gaston after it had weakened. Other developing cases investigated include Nuri [the Tropical Cyclone Structure (TCS) 2008 experiment in the western Pacific], Gert [2005 NOAA IFEX and NASA’s Tropical Cloud Systems and Processes (TCSP) experiment; Halverson et al. (2007)], as well as storms exclusively investigated by NOAA and U.S. Air Force (USAF) aircraft: Bonnie (2010), Danny (2009), Fay (2008), and Kyle (2008). These cases are identical to the sample investigated in Zawislak and Zipser (2014, hereafter ZZ14), and are chosen since they represent the best-sampled (from dropsondes) genesis cases from recent field programs. The tracks of the disturbances are shown in Fig. 1. Genesis is defined as the tropical depression (TD) classification by the National Hurricane Center (NHC). Although genesis is often difficult to define, with the exception of Earl and Gaston, for the majority of the cases one can have more confidence in the genesis time given that in situ data were used by the NHC to upgrade an investigated disturbance to a TD.
Summary of cases examined by NASA, NSF/NCAR, NOAA, Naval Research Laboratory (NRL), and USAF aircraft.
Tracks of the developing and nondeveloping disturbances with type and intensity shaded.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
The large-scale conditions of each developing disturbance are evaluated using the National Centers for Environmental Prediction (NCEP) Final (FNL) model analyses. The necessary environmental conditions for tropical cyclogenesis appear to be met throughout the pregenesis stage of all developing disturbances examined: each has a concentrated area of cyclonic relative vorticity (average of 2 × 10−5 s−1 between 925 and 600 hPa within a 3° radial circle around the vorticity maximum center), high relative humidity (>70% within 3° at 700 and 600 hPa), low-to-moderate 850–200-hPa deep-layer vertical wind shear (2–11 m s−1 in a 2°–8° annulus), and sufficiently high SSTs (>26.5°C). For the nondeveloping disturbances, while all cases examined exhibit sufficient values of SST and deep-layer vertical wind shear, the relative humidities at 700 and 600 hPa typically fall below 70% and the relative vorticity does not increase over time and, at times, drops to nearly zero.
b. Vorticity maxima tracking
The disturbance is defined by the vorticity maximum (VM) manually tracked in the 1° × 1° NCEP FNL model analyses. Using a Gaussian smoother, the relative vorticity fields are spatially smoothed so as to prevent contamination from spurious grid-scale “bull’s-eye” features. VMs of at least 2 × 10−5 s−1 are tracked from when they are consistently seen in the analysis for at least 24 h, until they no longer exceed the required threshold, can no longer be consistently tracked, or make landfall (anywhere but in the Caribbean Islands) (see additional details in ZZ14). Although 925-, 700-, and 600-hPa VMs are also tracked, the analyses presented are based on the 850-hPa VM center.
The pouch center, defined as the center of the recirculating region within the comoving (or Lagrangian) framework, is also tracked (Dunkerton et al. 2009; Montgomery et al. 2010; Wang et al. 2010). The comoving field is computed by subtracting the zonal phase speed of the wave from the zonal (u) wind field. Because this study computes a daily phase speed, rather than one that is representative of the entire lifetime of the wave, disturbances often exhibit phase speeds outside of the “typical” range of 4–8 m s−1. In these situations a pouch is not always clearly defined. As a consequence, the pouch analysis can be less reliable since in many cases a coherent pouch is not tracked throughout the pregenesis period. Regardless, genesis is unlikely to occur in many of these scenarios since wave vorticity is primarily from shear rather than curvature. The results for the pouch, when actually present, do not differ significantly from those for the vorticity maxima and, therefore, are not shown.
c. Satellite datasets
The TRMM satellite package was launched in 1997 and contains a suite of instruments that includes the first spaceborne precipitation radar (PR), TMI, a Visible and Infrared Scanner (VIRS), as well as a Lightning Imaging Sensor (LIS). TMI (a conical scanner) has an 850-km-wide swath and, at 85 GHz, a footprint of 7 × 5 km2. SSM/I and the next-generation SSMIS are a suite of microwave instruments flown on the Defense Meteorological Satellite Program (DMSP) satellites (F-8–F-15 and F-16–F-17, respectively). SSM/I (SSMIS) is a 7 (24)-channel PMW radiometer with frequencies ranging from 19 to 85 (19 to 183) GHz that samples a swath width of approximately 1400 (1700) km. The footprint is 16 × 13 km2 at 85 and 91 GHz (12.5-km spatial sampling). AMSR-E is a conically scanning PMW radiometer on board the Aqua satellite. The swath width of AMSR-E is approximately 1500 km, and although it has fewer channels (12) than SSMIS, the instrument samples at greater spatial resolution; the footprints are comparable to that of TMI, 6 × 4 km2 at 85 GHz (mean spatial resolution of 5.4 km). This study utilizes AMSR-E data from the National Snow and Ice Data Center (NSIDC) level-2A product, resampled to 11 km [nearly matching the SSM/I(SSMIS)] (Ashcroft and Wentz 2013).
“Strong” convection is defined as convection with PMW PCT ≤ 210 K. This value is chosen as it represents the top 5% threshold for the minimum 85-GHz PCT in a 15-yr climatology of June–September TRMM precipitation features2 in the Atlantic basin (area bounded by 5°–30°N, 10°–100°W; ocean features only). “Intense” convection is defined by PCT ≤ 160 K, which is the threshold for the top 1% in the climatology. Despite the larger footprint size of SSM/I(SSMIS), which may make it less likely to detect depressed brightness temperatures associated with intense convection, and calibration differences, these thresholds have been similarly applied to all sensors. These differences are likely small compared to the larger separation (50 K) between the strong and intense PCT thresholds. Because three sensors (F-15, F-16, and F-17) are included, the contribution of samples from SSM/I(SSMIS) (423 overpasses, 62% of the total) in the dataset is substantially greater than either AMSR-E (153, 22%) or TMI (108, 16%).
Given the limited width, not all swaths from PMW instruments will completely cover the inner core. Overpasses are only included in the database when brightness temperature data are available within 0.5° of the VM center. As for the temporal sampling from PMW overpasses, most instruments will only observe the disturbance at most twice a day. Since TRMM is in a circular, non-sun-synchronous orbit, the local passage time over the convection, particularly at low latitudes, varies each day. In contrast, AMSR-E and SSM/I(SSMIS) are on polar-orbiting platforms and will pass at a similar local time each day. AMSR-E samples at approximately 0130 and 1330 local time (LT), while SSM/I(SSMIS) samples near 0600 and 1800 LT. These times are conveniently located around the early (AMSR-E) and later [SSM/I(SSMIS)] stages of the typical early morning oceanic rainfall maximum and afternoon diurnal minimum (Chang et al. 1995; Nesbitt and Zipser 2003). The least sampled time period by the PMW dataset is between 0800 and 1200 local time, which is during the transition between the typical oceanic diurnal maximum and minimum. Given the discontinuous sampling of the disturbance by PMW instruments, the convective evolution of each disturbance as seen from the PMW database is not necessarily complete. To help provide a more complete time evolution, PMW data are combined with rainfall statistics computed from the level-3 TRMM 3B42 TMI merged-IR product, which is available 3-hourly at 0.25° resolution (Huffman et al. 2007).
Finally, statistics will be computed for within both a 3° and a 1° radial circle of the center (defined as the inner core). Given that the VM centers are determined from a model analysis with a 1° resolution, one should heed caution when interpreting the statistics for the 1° inner core since a small change in the center location can significantly change the statistics. This is likely most prevalent early in the pregenesis stage where weak, or even multiple, VMs make center estimates more difficult to determine. As genesis nears, VMs are stronger and better defined. As a result, the 1° statistics are likely more representative of the convective properties occurring near the center.
3. Results
a. Overall statistics
1) Raining area
The rainfall statistics for the 3° and 1° inner cores are provided in Figs. 2 and 3, respectively. Note that in Figs. 2 and 3 (and all similar figures that follow), with the exception of ex-Gaston, nondeveloping samples are plotted for convenience on the abscissa as “before genesis” (negative). Although shown together, the nondeveloping time series is not intended to be directly compared with the developing time series. Gaston is plotted separately (dashed in Figs. 2 and 3; black symbols in Figs. 4 and 5); the pregenesis stage to 2.5 days after genesis is considered the developing part of Gaston, while the period from 2.5 days after genesis to the end of the track is considering the nondeveloping part of ex-Gaston.
Time series of (a) the total raining fraction, (b) fraction of all pixels with convective rain rate ≥ 5 mm h−1, and (c) fraction of raining pixels with convective rain rate within 3° of the 850-hPa VM center for developing disturbances (solid, light gray), nondeveloping disturbances (solid, red), and Gaston (dashed, red); the developing mean (black) and developing median (blue) are also shown.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
As in Fig. 2, but for within 1° of the 850-hPa VM center.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
Within 3° of the 850-hPa VM center, (a) mean 85–91-GHz PCT for all pixels with PCT ≤ 250 K, (b) minimum 85–91-GHz PCT, and (c) fraction of pixels with PCT ≤ 210 K for developing disturbances (light gray symbol), nondeveloping disturbances (red), and Gaston (black). The size of the symbol indicates the fractional coverage of the swath data within 3° of the 850-hPa VM center; the larger the symbol, the higher the fractional swath data coverage within the radial circle. The mean (black) and median (blue) of developing samples are also included.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
As in Fig. 4, but within 1° of the 850-hPa VM center.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
In Figs. 2a and 3a, the fractional area within the inner core that is raining (all pixels with a rain rate greater than zero) will be a proxy for the raining area. The raining area does not differentiate stratiform from convective regions. The results indicate that, on average, the total raining fraction within the 3° inner core of developing cases shows no noticeable trend prior to genesis; high raining fractions are just as likely to be observed 3 days prior to genesis as 1 day prior. For comparison, the time series for the 1° inner core, while exhibiting somewhat greater raining fractions overall, indicates a slight increasing trend in raining fraction from 4 days before genesis to 1 day after. The increasing trend is, however, not consistent among all developing disturbances. Overall, for both inner-core definitions, the maximum raining fraction is observed within 24 h of genesis in only 4 of the 12 developing cases (Table 2). For the majority of cases (5 out of the 12 for within 3°), the maximum raining fraction occurs between 1 and 2 days prior to genesis. Figures 2a and 3a also have one common feature; despite relatively persistent rainfall within the inner core 1–3 days before formation, in 3 (5 for 1°) of the 12 developing cases (Table 2), the minimum raining fraction of the pregenesis stage actually occurs within 1 day of formation.
Summary of rain statistics (from 3B42) for the 3° (1°) inner core for developing and nondeveloping (italics) disturbances; boldface indicates within 24 h of genesis, and duration is the total number of tracked hours before genesis (developing) or the end of the track (nondeveloping).
Figures 2b and 3b also show the time series of the proxy for the “convective” raining fraction, which is defined as the fractional area within the inner core with 3B42 rain rates ≥5 mm h−1.3 Although the mean convective rain fraction shows no noticeable trend prior to genesis for either the 3° or 1° inner-core analysis, Table 2 indicates that the maximum convective raining fraction within 3° (1°) of the center occurs within 24 h of genesis in 5 (3) of the 12 developing cases, and between 1–2 days prior to genesis in 4 (6) cases. However, the maximum convective raining fraction of the pregenesis period does not occur within 2 days of genesis for all cases; three of the disturbances exhibit the maximum convective raining fraction more than 2 days (and as many as 3.5 days) prior to genesis. In addition, despite the most impressive convective episodes (in terms of areal coverage) occurring within 1–2 days of formation for the majority of cases, in nearly every developing disturbance there is no convective rainfall within 1° of the center at some time within 12 h of genesis (Table 2). In fact, in half of the developing cases the inner core is completely void of convective rainfall even as far as 3° from the center at some time within a day of genesis. This result is reminiscent of results described by Zehr (1992), in which the author identifies two distinct stages to tropical cyclogenesis separated by a period of relatively inactive convection. This inactive period was found to exist between the initiation of the incipient mesoscale vortex (or even tropical depression; stage 1) and the formation of a minimal tropical storm (stage 2).
Figures 2a and 3a demonstrate another important result; for the cases included, the nondeveloping disturbances typically exhibit lower raining fractions than do developing disturbances, while the overall convective raining fraction near the center tends to be less (Figs. 2b and 3b). In addition, while the fraction of the raining area within 3° that is considered convective (Fig. 2c) can be just as high in nondeveloping cases as developing, when the inner core is confined to 1° (Fig. 3c), the fraction of the raining area that is convective tends to be larger in developing disturbances. Due to the sample size of nondeveloping cases, these results should only be taken as suggestive. The disturbances investigated likely do not represent the variability possible across all nondeveloping disturbances. Only a larger sample of nondeveloping disturbances will yield definitive conclusions regarding the differences in their convective properties compared to developing disturbances.
2) Frequency
Table 2 also presents the frequency of rainfall within 1° and 3° for all cases. The frequency is defined as the percentage of hours before genesis where the raining fraction within the inner core exceeds 40%. In cases where the pregenesis disturbance is tracked for more than 2 days (nine cases), the frequency within 3° (1°) varies from 56% to 88% (63% to 92%). Gaston is considered an exception at 13% (8%); pre-Gaston is nearly absent of precipitation and seems to actually develop after one convective episode, even though most of the intense rainfall is greater than 3° from the center. For the three developing disturbances that are tracked for 2 days or less—Gert, Danny, and TD2—the frequencies within 3° (1°) are 56%, 93%, and 83% (63%, 100%, and 92%), respectively. The nondeveloping disturbances—ex-Gaston, PGI-27, and PGI-30—have frequencies of 27%, 35%, and 6% (51%, 50%, and 19%), respectively. If the raining fraction threshold is increased from 40% to 75%, the frequency varies from 4% to 33% (4% to 73%) for developing disturbances, while nondeveloping disturbances typically have frequencies less than 7%. Overall, these results suggest that, although the frequency among the developing disturbances varies over a fairly large range, the pregenesis disturbances examined have distinguishably more frequent rainfall exceeding 40% and 75% coverage within the inner core than do the nondeveloping disturbances examined.
3) Intensity
The time series in Fig. 4 show statistics from 85- to 91-GHz frequencies for each PMW overpass of the 3° inner core for developing and nondeveloping disturbances. Figures 4a and 4b represent proxies for the convective intensity, as Fig. 4a shows the mean PCT of all pixels within the inner core that are less than or equal to 250 K, while Fig. 4b shows the overall minimum PCT. Figure 4c4 is the fraction of pixels within the inner core that are less than 210 K (i.e., a proxy for the fraction of strong convection). Although the mean PCT indicates no noticeable trend prior to formation, the minimum PCT does show a slight decreasing trend (i.e., increasing intensity) during the period from 1 day before to 1 day after genesis; prior to this period, there is no obvious trend in convective intensity. In fact, Figs. 2–5 (for areal fraction and intensity) all appear to indicate that genesis is consistently declared while disturbances are going from the convective minimum shortly before declaration to a maximum shortly after. Compared to the 3° inner core, the decreasing trend within 1 day of genesis is more noticeable when the analysis is confined to 1° (Fig. 5b). Strong convection is, however, not exclusive to this period; overpasses with strong convection are similarly observed as many as 2–4 days prior to genesis, even with the smaller sample of overpasses on those days. This result is verified in Table 3, which shows the “latest” (or closest) hour to genesis, as well as the earliest hour in the track, that an overpass with strong convection is observed within 3° and 1° of the center. Table 3 (the column marked 210 K latest) indicates that while all 12 developing cases exhibit strong convection within 1° of the center within 36 h of genesis, nearly every case also exhibits strong convection more than 36 h prior to formation (Table 3; the 210 K earliest column). In only 1 (Nuri) of the 12 cases is strong convection not observed within 1° until the day before genesis. Intense convection (defined as PCT ≤ 160 K) is consistently observed within 1°–3° of the center within a day of genesis (Table 3); however, within 1° only 6 of the 12 cases have overpasses that exhibit intense convection at any time in their pregenesis stage. Intense convection is observed within 1° within a day of genesis in only two of those six cases.
Summary of the “latest” (closest hour to genesis), and earliest (in the pregenesis track) hours in which 210 and 160 K are observed in a PMW overpass of the 3° (1°) center; boldface indicates within 24 h of genesis and duration is the total number of tracked hours before genesis. Dashed lines indicate that the PCT threshold was not observed within 1°.
Figures 4c and 5c also show the time series of the fractional area of strong convection within 3° and 1° of the center, respectively. Neither time series indicates any significant trend until the period within a day prior to, and after, genesis. In other words, there does not appear to be distinguishably greater fractions of strong convection on any particular day prior to formation; relatively high fractions of strong pixels (e.g., greater than 2% within 3°, 5% within 1°) are just as likely to be observed 1 day prior to formation as they are 4 days prior to formation (even with the smaller sample size this early in the pregenesis stage). In fact, when evaluated individually (Table 4), in only 4 (3) of the 12 developing cases is the maximum areal fraction observed within 3° (1°) of the center within 24 h of genesis. In comparison, the maximum within 3° (1°) is observed between 1 and 2 days of genesis in six (four) cases, and outside of 2 days in two (four) cases. Similar to the convective raining fraction (Table 2), another important result from Table 4 is that, in nearly every developing disturbance, there is no strong convection observed within 1° of the center at some time within a day of genesis. In fact, within a day of genesis nearly half of the cases do not exhibit strong convection even within 3° of the center (Table 4).
Summary of the fractional area of strong convection (PCT ≤ 210 K) for developing and nondeveloping (italics) cases for PMW overpasses of the 3° (1°) center; boldface indicates within 24 h of genesis and duration is the total number of tracked hours before genesis (developing) or the end of the track (nondeveloping). Dashed lines indicate that the PCT threshold was not observed within 1°.
Figures 4 and 5 also show overpasses of the nondeveloping VM centers. Although typically less intense, some overpasses indicate convection in nondeveloping disturbances that is comparable in intensity to that in developing disturbances. In fact, some overpasses of ex-Gaston indicate a convective intensity (mean and minimum PCT) that is comparable to the intensity seen in overpasses of mature tropical cyclones. The main distinction between developing and nondeveloping disturbances is that, even though the maximum fractions are comparable (Table 3), the nondeveloping cases examined exhibit relatively high fractions of strong and intense convection far less frequently than do developing disturbances. Although this result may be biased by the small sample size of nondeveloping PMW overpasses, this observation supports a similar result presented for the raining fraction (Figs. 2 and 3), and is consistent with the composite results from Leppert et al. (2013a,b), who concluded that convective area (from IR) is more important to tropical cyclogenesis. They did not find statistically significant differences in convective intensity between developing and nondeveloping disturbances in the Atlantic.
Overall, these results suggest that, although invariably observed in developing disturbances, there does not appear to be any unique attribute for strong convection during the pregenesis stage. The overall intensity and areal coverage do not indicate an obvious increasing trend and are not necessarily greater within a day of genesis than in days earlier, while intense convection (PCT ≤ 160 K) does not necessarily need to be in close proximity to the center (within 1°) for genesis to occur. So, while strong convection (PCT ≤ 210 K) is always observed within 1° of the center within 36 h of genesis, strong convection is not unique to that period. Although this result may not be surprising, a question remains: even though they are not directly involved in the formation of a TD, how are strong and intense convective events occurring as many as 4 days in advance of genesis (such as in Fay, Kyle, Karl, and Fiona) important to priming the disturbance for formation?
b. Individual case results
1) Karl and Matthew (2010)
Karl is first classified as a tropical depression at 1200 UTC on 14 September near 17.5°N, 82°W. Starting on 9 September, Karl’s low (925–850 hPa) and midlevel (700–600 hPa) VM tracks are shown in Fig. 6. While the midlevel vorticity appears to originate from a wave over the central Atlantic, the low-level vorticity originates in the monsoon region off the South American coast. The tracks indicate a substantial misalignment of the low- and midlevel vorticity maxima early in the track. By 13 September, low- and midlevel centers are better aligned and remain so throughout the formation period. Seventeen total flights by dropsonde-equipped PGI and USAF aircraft investigated Karl in the 5 days up to, and including, formation. The evolution of the low-level (925 hPa) and midlevel (600 hPa) wind fields from dropsonde observations, as well as accumulated rainfall, for the period between 10 and 15 September is shown in Fig. 7. Four days before genesis (10 September), a distinct low-level circulation is observed. Formation is, however, considered unlikely given the substantial misalignment of the low- and midlevel centers (Fig. 6). By 12 September, the disturbance appears less organized; the low-level circulation is no longer apparent, and the wave trough appears asymmetric since southeasterly winds on the east side of the trough axis are stronger than the northeasterly winds on the west side (the wave axis appears to have a southwest–northeast horizontal tilt). Only when the VM align, and the winds became more symmetric on 13 September, did the wave vorticity come predominately from curvature, rather than from shear, and thus become more favorable for formation. This evolution is verified observationally by Davis and Ahijevych (2012), who used dropsonde-derived circulation centers to show that the misalignments of the pre-Karl low- and midlevel circulation centers favorably decrease between 11 and 13 September.
VM locations at 925, 850, 700, and 600 hPa during Karl and Matthew at 0000 and 1200 UTC (day of the month labeled).
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
Select dropsonde observations at (a) low (925 hPa) and (b) midlevels (600 hPa) of Karl from 10 to 14 Sep 2010. Dropsonde locations are time–space adjusted, based on the zonal phase speed of Karl, to the time indicated. Accumulated rainfall (mm) from TRMM 3B42 is contoured (period is from 0000 UTC 10 Sep to 0000 UTC 15 Sep). The open circle is the genesis location at 1200 UTC 14 Sep [adapted from Braun et al. (2013)].
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
Figure 8a shows the azimuthally averaged rain rate around the 850-hPa VM center in Karl. Karl exhibits a distinct diurnal cycle during the pregenesis stage, which is also noted by Davis and Ahijevych (2012) using IR data and examined in a numerical simulation by Melhauser and Zhang (2014). Prior to formation, the most widespread convective rainfall occurs 3 days before genesis (maximum around 1800 UTC 11 September). Dropsonde observations (Fig. 7) from the PREDICT Gulfstream-V (G-V) flight (centered around 1200 UTC 12 September) just after this episode suggests that not only did convection fail to organize a circulation pattern at low or midlevels, but the pre-Karl wave is actually less organized. Each subsequent convective episode in the 3 days leading up to genesis after 11 September has less areal coverage and is farther from the VM center. Despite apparently less convection occurring near the center, dropsonde observations from flights on 13 and 14 September do indicate some organization at both low and midlevels; however, whether the convection is responsible for the organization of the circulation is unclear.
Time series of azimuthally averaged rain rate (from TRMM 3B42) for (a) Karl and (b) Matthew around the 850-hPa VM center. The right axis represents the day of month at 0000 UTC, and the black line shows the genesis time.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
To further characterize the intensity of each convective episode, PCT statistics from PMW overpasses within 1° of the VM center are shown in Fig. 9a. According to the minimum and mean PCT, no convective episode is noticeably more intense than any other prior to formation. In fact, the final convective episode before formation (around 1800 UTC 13 September) is characteristically the least intense; the minimum PCT is as much as 40 K warmer than previous episodes, while the fractional coverage of strong pixels within 1° is less than 1%. Considering that the most “favorable” convective episode for genesis5 is 3 days prior to formation, and that the total raining and convective area in each episode is decreasing in the days that follow, one may surmise that the formation of Karl is more closely tied to the low- and midlevel vorticity alignment than any single distinguishing characteristic of the convection. Except for the fact that it occurs in a more organized disturbance, compared to previous episodes, there does not appear to be anything “special” about the convective episode occurring in the day prior to genesis.
For (a) Karl and (b) Matthew, time series of mean PCT of pixels ≤ 250 K (filled black), minimum PCT (filled gray), and fraction of pixels within 1° that are ≤210 K (open symbols) within 1° of the 850-hPa VM center. Day of the month at 0000 UTC is labeled along the top abscissa with the longitude at that time.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
In contrast to Karl (Fig. 8a), Matthew (track in Fig. 6) does not exhibit a clear favorable convective episode prior to the formation (Fig. 8b). There is, however, a marked increase in organization in the day (22 September) following the convective episode 2–2.5 days prior to formation. According to the dropsondes from flights on 22 September (Fig. 10), both the low- (925 hPa) and midlevel (600 hPa) trough have well-defined curvature and cyclonic circulations. Whether the convective episode 2–2.5 days prior to formation is responsible for this organization is unclear, since the circulation seen in the G-V flight (approximately 15–18 h after the event) may simply not have been adequately sampled by the DC-8 and G-V on 21 September, and is also 2° south of the latitude where the most intense convection occurred the previous day. In the 36 h prior to genesis, the convective rainfall is primarily confined to within 1° of the VM center as Matthew approaches genesis (Fig. 8b); however, even though strong convection appears near the center, the areal coverage does not exceed 2% (Fig. 9b). Considering that the overall intensity and areal coverage of strong convection does not seem to be distinguishable in any particular episode before genesis, one may speculate that the proximity and persistence of (not necessarily intense) convection near the center are the most important properties within 2 days of formation of Matthew.
Select dropsonde observations at (a) low (925 hPa) and (b) midlevels (600 hPa) of Matthew from 20 to 24 Sep 2010. Dropsonde locations are time–space adjusted, based on the zonal phase speed of Matthew, to the time indicated. Accumulated rainfall (mm) from TRMM 3B42 is contoured (period is from 0000 UTC 20 Sep to 0000 UTC 24 Sep). The open circle is the genesis location at 1800 UTC 23 Sep [adapted from Braun et al. (2013)].
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
2) Fay (2008) and Fiona (2010)
Fay and Fiona (tracks in Fig. 11) exhibit two distinctly different convective evolutions prior to genesis. While Fay does not exhibit any particularly impressive periods of deep convection at any radius from the center (Fig. 12b), Fiona has two particularly impressive periods (Fig. 12a). Both episodes (around 0000 UTC on 28 and 29 August6) are considered favorable among all proxies; the rainfall is intense (azimuthal mean rain rates exceed 7 mm h−1), is in close proximity to the center (within 3°), and is widespread (raining fraction greater than 75% within 3°). Although few pregenesis in situ observations exist within 24 h to verify the direct impact of the convection on the intensification of the pre-Fiona circulation, observations from the most recent G-V flight (midday on 30 August; not shown) after the second episode indicates that at the time of formation, only a low-level circulation pattern is present; no well-defined midlevel circulation is apparent. Like Karl, Fiona does not develop immediately after either one of the apparently favorable convective episodes. In fact, according to 3B42 and PMW overpasses, strong convection appears to be nearly nonexistent within the inner core (Figs. 12a and 13a) within a day of genesis. This time evolution suggests that by 30 August, the circulation is already primed for formation; widespread, intense precipitation may not be required to further intensify the circulation to tropical cyclone strength. So a question remains: even though genesis is not directly attributed to the most favorable convective episode in Karl, Matthew, and Fiona, do changes in the thermodynamic environment as a result of this episode lessen the requirements (for area, intensity, proximity) of subsequent convective episodes to further intensify the vortex? This question will be addressed in section 4.
As in Fig. 6, but for Fiona and Fay.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
As in Fig. 8, but for (a) Fiona and (b) Fay.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
As in Fig. 9, but for (a) Fiona and (b) Fay.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
The inner core of Fay is experiencing frequent, though not necessarily intense (i.e., rain rate < 5 mm h−1), rainfall (Fig. 12b) as the raining fraction greater than 40% within 3° (1°) occurs during 77% (73%) of the pregenesis period (Table 2). Given the favorable environment (in terms of vertical wind shear, relative humidity, and moisture; not shown) and the lack of distinguishing convective episodes like those observed in Fiona (Figs. 12b and 13b), the persistence of rainfall near the center may be more critical to the formation of Fay than any widespread, strong convective episode near the center.
3) Gaston (2010)
Gaston originates as an easterly wave that left the coast of Africa around 28 August. Vorticity maxima tracking begins on 29 August (Fig. 14) and indicates two distinct tracks that merge on 1 September. A TD is declared at 0600 UTC. Although Gaston appears to form without any significant rainfall within 2° of the center (Fig. 15a), this evolution is likely not representative of the formation process as the 850-hPa VM originates in the dry northern track (Reed et al. 1977; Pytharoulis and Thorncroft 1999). Significant rainfall is, however, observed near the center of the more southern 700-hPa track (not shown). The southern track is typically more moist and is, thus, more favorable for genesis (Thorncroft and Hodges 2001). Gaston briefly reaches tropical storm strength, but by 1800 UTC on 2 September Gaston weakens and is no longer classified as a TD. The VMs are, however, consistently tracked after 2 September as ex-Gaston progresses westward, and ends when the VMs make landfall in Central America on 11 September.
As in Fig. 6, but for (top) Gaston and (bottom) PGI-27.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
As in Fig. 8, but for (a) Gaston and (b) PGI-27. The dashed black line indicates the time when Gaston was downgraded from a TD, after which the event was considered nondeveloping.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
Given the potential for redevelopment, PREDICT and GRIP focus their efforts on ex-Gaston between 3 and 8 September. During this period, low- (925 hPa) and midlevel (600 hPa) VM centers in the model analysis are vertically misaligned by 1°–2°. In comparison, using PREDICT dropsonde data, Davis and Ahijevych (2012) find the observed circulation centers (900 and 500 hPa in their study) to be misaligned by as much as 3°. As a result of the vortex misalignment, Davis and Ahijevych (2012) identify strong relative flow and a lateral transport of dry air at midlevels over the low-level circulation center. Coupled with the lack of deep convection (compared to Karl and Matthew), the authors conclude that the relatively dry air at midlevels is the primary reason for the nondevelopment of ex-Gaston. In contrast to Davis and Ahijevych (2012), Fritz and Wang (2013) conclude that, rather than being laterally entrained into the center, dry air is vertically transported downward from the upper levels (where a wave pouch does not exist) to the midlevels.
Both the model analysis and dropsonde observations verify that the midlevels are drying during the period between 3 and 8 September. The model analysis mean 600-hPa RH decreases from nearly 80% on 3 September to 45% on 8 September, while the dropsonde-derived 700–400-hPa mean mixing ratio drops below 4.5 g kg−1 (not shown). The drying at midlevels likely explains why the raining area within the inner core (for both 1° and 3° definitions) is decreasing (Fig. 15a), as well as the reason for the lower frequency (26% for the 40% raining fraction threshold; Table 2) of rainfall compared to the developing cases (frequencies typically above 50%). Despite less rainfall overall, the convective raining fraction (Figs. 2c and 3c) and fraction of strong convection (Fig. 16a) do indicate periods of convection during the nondevelopment stage that are comparable to developing disturbances (Fig. 5), although their radial distance from the center is often outside of 1° (Fig. 15a). The intensity of convection (minimum PCT) in ex-Gaston is also comparable to not only the pregenesis stage of developing disturbances, but also to developed tropical cyclones (Fig. 4). This is not surprising as an analysis from the PGI flights into ex-Gaston (Smith and Montgomery 2012; ZZ14) indicates both a decrease in θe at midlevels and an increase in the difference between the surface θe and midlevel θe minimum—an indication of increased convective instability. The convective characteristics of ex-Gaston identified using the PMW data in this study support the characteristics presented in Davis and Ahijevych (2012) using IR data. In that study, the authors also conclude that while intense convection persists through the life cycle of ex-Gaston, the areal coverage of convection is less than that observed in both Karl and Matthew.
As in Fig. 9, but for (a) Gaston and (b) PGI-27.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
4) PGI-27 (2010)
PREDICT flights on 17 and 18 August sampled a weak easterly wave (PGI-27) south of Puerto Rico and Hispaniola. Vorticity maxima at low (925–850 hPa) and mid- (700–600 hPa) levels are coherently tracked from east of the Lesser Antilles, all the way to landfall in Mexico after passing through the Bay of Campeche (Fig. 14). While the midlevel relative humidity (above 70%) and 850–200-hPa vertical wind shear (below 10 m s−1) are apparently favorable, and multiple periods of strong convection exist (Figs. 15b and 16b), the most obvious reason for nondevelopment appears to be related to the substantial misalignment of the low- and midlevel VMs (Fig. 14). Convection may not support intensification if the VMs are misaligned by as much as 4°, as observed in PGI-27. From 20 August until landfall, the alignment improves (whether the convection is responsible is unclear); however, now intensification may be limited by the fact that the disturbance is both interacting with land and that the inner core (within 2°) is mostly void of rainfall (Figs. 15b and 16b).
4. Discussion
After a careful synthesis of PMW satellite observations, results indicate that the most favorable convective episode (in terms of area of strong convection, intensity, and proximity to the center) in developing disturbances does not necessarily occur in the day prior to genesis. While a few cases (Fiona, Nuri, and TD2) exhibit the most favorable episode around 36 h prior to genesis, in Karl the episode is observed as many as 3 days prior to formation. Although episodes 2–4 days before formation do not directly result in an event developing, we hypothesize that they may create a more favorable environment for episodes that follow, perhaps even lessening the requirements (of area and proximity) of those later episodes to further intensify the disturbance. For example, early convective episodes may contribute to the genesis process through favorably moistening the midlevels (Bister and Emanuel 1997; Raymond et al. 1998; Nolan 2007; Montgomery and Smith 2011; Raymond et al. 2011; Wang 2012; ZZ14), stabilizing the troposphere through low-level cooling (Raymond and Sessions 2007; Raymond and López-Carrillo 2011; Raymond et al. 2011; Davis and Ahijevych 2012, 2013; Komaromi 2013; ZZ14), and warming the upper troposphere (latent heating in stratiform regions and/or convectively driven subsidence), which, through hydrostatic adjustment, will initiate surface pressure falls and low-level convergence (Zhang and Zhu 2012).
Even though all of the developing cases exhibit strong convection within 2 days of genesis, in all but four cases (Earl, TD2, Fiona, and Nuri) those areas are outside of 1°, even as far as 2.5°–3°, from the center as genesis nears. Although not in close proximity to the center, the convection may still contribute to the organization and intensification of the circulation, most notably as a pathway to vertically align the incipient vortex. Davis and Ahijevych (2012) emphasize the importance of vertical vortex alignment to the formation of Karl and Matthew, and the nondevelopment of ex-Gaston. Their analysis indicates that the difference between the 925- and 500-hPa circulation centers in developing Karl and Matthew is reduced from 2°–3° from 2 to 4 days prior to genesis to less than 1° at genesis, while in nondeveloping ex-Gaston the difference systematically increases to nearly 4°. Raymond and López-Carrillo (2011) also identify a 2°–3° difference between the surface and 5-km circulation centers in developing Nuri, but in that case they conclude the alignment to still be sufficient for formation.
To examine the importance of vertical alignment among a larger sample size, the radial difference between the 925–850-, 925–700-, and 925–600-hPa VM centers in the NCEP FNL analysis are presented for all disturbances examined in this study in Fig. 17. Nondeveloping disturbances are also provided for comparison. Figure 17 confirms that the vertical alignment of the low- and midlevel vorticity maxima is a necessary condition for formation. Consistent with the observed circulation centers identified in individual case studies by Davis and Ahijevych (2012), on average the 925–600-hPa difference is as high as 2.5° as many as 4 days prior to genesis, and is reduced to less than 1° by genesis. The same tendency is observed for the 925–700-hPa (reduced from 1.5° to 0.5°) and 925–850-hPa (reduced from 0.5° to less than 0.25°) differences. Although the low- and midlevel centers are usually misaligned, nondeveloping disturbances can, at times, exhibit an alignment within 1° (Fig. 17); thus, vertical alignment is necessary, but on its own not sufficient, for formation.
Time series of the radial difference between the VM location at (a) 925–850, (b) 925–700, and (c) 925–600 hPa for developing disturbances (solid, light gray), nondeveloping disturbances (solid, red), and Gaston (dashed, red); the developing mean (black) and developing median (blue) are also shown.
Citation: Monthly Weather Review 142, 12; 10.1175/MWR-D-14-00028.1
Some recent studies hypothesize how deep convection, in conjunction with vertical shear, may be important to alignment (Molinari et al. 2004, 2006; Davis and Ahijevych 2012; Rappin and Nolan 2012). In observational case studies of the formation of Danny (1997) and Gabrielle (2001), Molinari et al. (2004, 2006) suggest that deep convection in an environment with moderate (5–10 m s−1) to strong vertical wind shear (12–13 m s−1) can still provide a pathway to genesis. They speculate that persistent convection downshear of the vortex center can produce a new mesovortex that, if stronger, can ultimately absorb the original vortex and become the new, dominant vortex for formation, while also favorably moistening and stabilizing the troposphere (Bister and Emanuel 1997; Raymond et al. 1998). However, whether such a pathway can explain the formation of cases in this study—particularly those cases that do not exhibit deep convection in close proximity to the center (within 1°) within 2 days of genesis—is unclear. Despite the improved spatial and temporal coverage of in situ data of the PGI cases over cases from previous genesis-related field programs, the ability to link organization of a low- and/or midlevel circulation to convective events remains a challenge.
5. Conclusions
As deep convection is invariably identified as an important component of tropical cyclogenesis, the objective of this study was to quantify the time evolution of convection observed during the pregenesis stage of developing disturbances. The goals were to evaluate whether deep convective properties exhibit a distinguishable trend during the pregenesis stage and to determine whether convective properties of developing disturbances are distinguishably different as compared to nondeveloping disturbances. To address these main objectives, the convective properties from 12 developing and 3 nondeveloping disturbances in the Atlantic and northwest Pacific were investigated using an extensive collection of data from multiple PMW instruments [AMSR-E, TMI, SSM/I (SSMIS)].
Among all the properties examined (raining area, intensity, area of “strong” and “intense” convection, frequency, and proximity), the results suggest that the raining area (contributions from both convective and stratiform regions) and frequency of rainfall within 3° of the center may be two of the most important properties for formation. The raining area near the developing center is distinguishably greater when compared to nondeveloping disturbances examined, as almost all developing disturbances exhibit total raining fractions within the inner core (3°) above 40% for at least half of the pregenesis period. However, although important for development, the raining area within 3° of the center does not show an obvious trend (and only a slight increasing trend for 1°) during the pregenesis stage; large raining areas are just as likely to be observed multiple days prior to genesis than in the day prior to genesis. In fact, in only 3 (4 for 1° inner core) of the 12 developing cases is the maximum raining fraction of the pregenesis period observed within a day of formation. While the frequency of precipitation varies among the developing cases (in most cases, frequency is greater than 50% for the 3° inner core and 60% for 1° case), results strongly suggest that more frequent precipitation is a critical property of developing waves. Not only does it differentiate developers from the nondevelopers (in cases examined, frequencies are less than 35%), but it must be important for formation in cases such as Fay, Danny, and Matthew, which do not exhibit any distinguishing deep convective episodes such as those observed in Fiona prior to genesis.
Compared to previous days, there does not appear to be anything special about strong convection (defined as a PMW overpass with PCT ≤ 210 K and 3B42 rain rates ≥ 5 mm h−1) occurring in the day before genesis, except for the fact that it occurs in a more organized disturbance. Strong convection is observed throughout the pregenesis stage, while the overall intensity and fractional coverage of strong convection within the inner core does not necessarily increase as genesis nears. In fact, the most important period may be between 1 and 2 days of formation since, in the majority of cases (7 out of 12), the maximum areal fraction of strong convection (within 3°) of the pregenesis period occurs during that period; only 3 cases exhibit the maximum fraction within 24 h of genesis.
With respect to the most intense convection, even though every developing disturbance exhibits a PMW overpass with intense convection (PCT ≤ 160 K) within 1°–3° at some time within a day of genesis, in only half of the 12 disturbances investigated is that convection very near the center (within 1°) at any time during the pregenesis stage. Likewise, even though the fraction of strong and intense convection events is often distinguishably greater in developing disturbances (similar to the raining area), the intensity (minimum PCT) of convection does not necessarily differentiate the developing from the nondeveloping disturbances examined. This conclusion is consistent with the composite results from Leppert et al. (2013a,b).
Given the variability in the timing, location, and area of convection among the cases studied, its specific role in the formation process remains difficult to ascertain and supports the hypothesis that there are multiple pathways to tropical cyclogenesis. For example, the most “favorable” convective episodes do not necessarily occur within 24–36 h of genesis, such as in Karl and Matthew. Although not directly responsible for formation, as long as they are in close proximity to the center (within 3°), early convective events may modify the environment in such a way that convection in the days that follow is not required to be as widespread, or close to the center, to further organize the disturbance. For example, they may favorably moisten and stabilize the troposphere, intensify the midlevel circulation, initiate low-level convergence, or even act to vertically align the incipient vortex.
Dropsonde observations from research flights following convective episodes in Karl and Matthew (Figs. 7 and 10) suggest that convection may have played a role in organizing both low- (925 hPa) and midlevel (600 hPa) circulations; however, distinguishing this relationship from other, perhaps large scale, reasons for increased organization is still difficult. Tropical cyclogenesis involves a complex interplay between the large-scale environment and mesoscale precipitation; the environmental characteristics not only influence the initiation and attributes of the convective events, but the large-scale environment is subsequently modified by the convective events themselves. Despite improved the spatial and temporal coverage of in situ data of a number of the cases investigated in this study, the ability to use observations to further understand this complicated interaction remains a challenge. Given the small sample size of cases in this study, results are only suggestive—not definitive—that there are significant differences in the convective properties of developing and nondeveloping disturbances. In particular, one should heed caution when generalizing the conclusions for the null cases, as the three cases included in this study may not be representative of the variability possible across all nondeveloping disturbances.
Acknowledgments
This study is funded by NASA Grants NNX09AC44G and NNX11AB59G through the NASA Science Mission Directorate, and under the leadership and support of Dr. Ramesh Kakar, NASA Headquarters. This research is part of the lead author’s Ph.D. dissertation and has benefited from the comments from his committee: Drs. Steven Krueger, Zhaoxia Pu, Jim Steenburgh, and Jeffrey Halverson. The authors also thank Dr. Chuntao Liu and two anonymous reviewers for helpful comments that have improved the methodology and figures in the manuscript.
REFERENCES
Ashcroft, P., and F. J. Wentz, cited 2013: AMSR-E/Aqua L2A global swath spatially-resampled brightness temperatures, version 3. National Snow and Ice Data Center, NASA DAAC, Boulder, CO, doi:10.5067/AMSR-E/AE_L2A.003.
Bell, M. M., and M. T. Montgomery, 2010: Sheared deep vortical convection in pre-depression Hagupit during TCS08. Geophys. Res. Lett.,37, L06802, doi:10.1029/2009GL042313.
Bister, M., and K. A. Emanuel, 1997: The genesis of Hurricane Guillermo: TEXMEX analyses and a modeling study. Mon. Wea. Rev., 125, 2662–2682, doi:10.1175/1520-0493(1997)125<2662:TGOHGT>2.0.CO;2.
Braun, S. A., M. T. Montgomery, K. J. Mallen, and P. D. Reasor, 2010: Simulation and interpretation of the genesis of Tropical Storm Gert (2005) as part of the NASA Tropical Cloud Systems and Processes experiment. J. Atmos. Sci., 67, 999–1025, doi:10.1175/2009JAS3140.1.
Braun, S. A., and Coauthors, 2013: NASA’s Genesis and Rapid Intensification Processes (GRIP) field experiment. Bull. Amer. Meteor. Soc., 94, 345–363, doi:10.1175/BAMS-D-11-00232.1.
Chang, A. T. C., L. S. Chiu, and G. Yang, 1995: Diurnal cycle of oceanic precipitation from SSM/I data. Mon. Wea. Rev., 123, 3371–3380, doi:10.1175/1520-0493(1995)123<3371:DCOOPF>2.0.CO;2.
Chronis, T. G., E. R. Williams, E. N. Anagnostou, and W. A. Petersen, 2007: African lightning: Indicator of tropical Atlantic cyclone formation. Eos, Trans. Amer. Geophys. Union, 88, 397–398, doi:10.1029/2007EO400001.
Davis, C. A., and D. A. Ahijevych, 2012: Mesoscale structural evolution of three tropical weather systems observed during PREDICT. J. Atmos. Sci., 69, 1284–1305, doi:10.1175/JAS-D-11-0225.1.
Davis, C. A., and D. A. Ahijevych, 2013: Thermodynamic environments of deep convection in Atlantic tropical disturbances. J. Atmos. Sci., 70, 1912–1928, doi:10.1175/JAS-D-12-0278.1.
Dunkerton, T. J., M. T. Montgomery, and Z. Wang, 2009: Tropical cyclogenesis in a tropical wave critical layer: Easterly waves. Atmos. Chem. Phys., 9, 5587–5646, doi:10.5194/acp-9-5587-2009.
Fritz, C., and Z. Wang, 2013: A numerical study of the impacts of dry air on tropical cyclone formation: A development case and a nondevelopment case. J. Atmos. Sci., 70, 91–111, doi:10.1175/JAS-D-12-018.1.
Fu, B., M. S. Peng, T. Li, and D. E. Stevens, 2012: Developing versus nondeveloping disturbances for tropical cyclone formation. Part II: Western North Pacific. Mon. Wea. Rev., 140, 1067–1080, doi:10.1175/2011MWR3618.1.
Gray, W. M., 1968: Global view of the origin of tropical disturbance and storms. Mon. Wea. Rev., 96, 669–700, doi:10.1175/1520-0493(1968)096<0669:GVOTOO>2.0.CO;2.
Gray, W. M., 1975: Tropical cyclone genesis. Dept. of Atmos. Science Paper 232, Colorado State University, 121 pp.
Halverson, J., and Coauthors, 2007: NASA’s Tropical Cloud Systems and Processes experiment. Bull. Amer. Meteor. Soc., 88, 867–882, doi:10.1175/BAMS-88-6-867.
Harr, P. A., R. L. Elsberry, and J. C. L. Chan, 1996a: 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.
Harr, P. A., M. S. Kalafsky, and R. L. Elsberry, 1996b: Environmental conditions prior to formation of a midget tropical cyclone during TCM-93. Mon. Wea. Rev., 124, 1693–1710, doi:10.1175/1520-0493(1996)124<1693:ECPTFO>2.0.CO;2.
Hendricks, E. A., M. T. Montgomery, and C. A. Davis, 2004: The role of “vortical” hot towers in the formation of Tropical Cyclone Diana (1984). J. Atmos. Sci., 61, 1209–1232, doi:10.1175/1520-0469(2004)061<1209:TROVHT>2.0.CO;2.
Hopsch, S. B., C. D. Thorncroft, and K. R. Tyle, 2010: Analysis of African easterly wave structures and their role in influencing tropical cyclogenesis. Mon. Wea. Rev., 138, 1399–1419, doi:10.1175/2009MWR2760.1.
Houze, R. A., Jr., W.-C. Lee, and M. M. Bell, 2009: Convective contribution to the genesis of Hurricane Ophelia (2005). Mon. Wea. Rev., 137, 2778–2800, doi:10.1175/2009MWR2727.1.
Huffman, G. J., and Coauthors, 2007: The TRMM Multisatellite Precipitation Analysis (TMPA): Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J. Hydrometeor., 8, 38–55, doi:10.1175/JHM560.1.
Jiang, H., and E. M. Ramirez, 2013: Necessary conditions for tropical cyclone rapid intensification as derived from 11 years of TRMM data. J. Climate, 26, 6459–6470, doi:10.1175/JCLI-D-12-00432.1.
Jiang, H., C. Liu, and E. J. Zipser, 2011: A TRMM-based tropical cyclone cloud and precipitation feature database. J. Appl. Meteor. Climatol., 50, 1255–1274, doi:10.1175/2011JAMC2662.1.
Kerns, B., 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, 2576–2591, doi:10.1175/2008MWR2545.1.
Kerns, B., and S. S. Chen, 2013: Cloud clusters and tropical cyclogenesis: Developing and nondeveloping systems and their large-scale environment. Mon. Wea. Rev., 141, 192–210, doi:10.1175/MWR-D-11-00239.1.
Komaromi, W. A., 2013: An investigation of composite dropsonde profiles for developing and nondeveloping tropical waves during the 2010 PREDICT field campaign. J. Atmos. Sci., 70, 542–558, doi:10.1175/JAS-D-12-052.1.
Leary, L. A., and E. A. Ritchie, 2009: Lightning flash rates as an indicator of tropical cyclone genesis in the eastern North Pacific. Mon. Wea. Rev., 137, 3456–3470, doi:10.1175/2009MWR2822.1.
Lee, C. S., 1989: Observational analysis of tropical cyclogenesis in the western North Pacific. Part I: Structural evolution of cloud clusters. J. Atmos. Sci., 46, 2580–2598, doi:10.1175/1520-0469(1989)046<2580:OAOTCI>2.0.CO;2.
Leppert, K. D., D. J. Cecil, and W. A. Petersen, 2013a: Relation between tropical easterly waves, convection, and tropical cyclogenesis: A Lagrangian perspective. Mon. Wea. Rev., 141, 2649–2668, doi:10.1175/MWR-D-12-00217.1.
Leppert, K. D., W. A. Petersen, and D. J. Cecil, 2013b: Electrically active convection in tropical easterly waves and implications for tropical cyclogenesis in the Atlantic and east Pacific. Mon. Wea. Rev., 141, 542–556, doi:10.1175/MWR-D-12-00174.1.
Liu, C., E. J. Zipser, D. J. Cecil, S. W. Nesbitt, and S. Sherwood, 2008: A cloud and precipitation feature database from nine years of TRMM observations. J. Appl. Meteor. Climatol., 47, 2712–2728, doi:10.1175/2008JAMC1890.1.
McBride, J. L., and R. Zehr, 1981: Observational analysis of tropical cyclone formation. Part II: Comparison of non-developing versus developing systems. J. Atmos. Sci., 38, 1132–1151, doi:10.1175/1520-0469(1981)038<1132:OAOTCF>2.0.CO;2.
Melhauser, C., and F. Zhang, 2014: Diurnal radiation cycle impact on the pregenesis environment of Hurricane Karl (2010). J. Atmos. Sci., 71, 1241–1259, doi:10.1175/JAS-D-13-0116.1.
Molinari, J., D. Vollaro, and K. L. Corbosiero, 2004: Tropical cyclone formation in a sheared environment: A case study. J. Atmos. Sci., 61, 2493–2509, doi:10.1175/JAS3291.1.
Molinari, J., P. Dodge, D. Vollaro, K. L. Corbosiero, and F. Marks Jr., 2006: Mesoscale aspects of the downshear reformation of a tropical cyclone. J. Atmos. Sci., 63, 341–354, doi:10.1175/JAS3591.1.
Möller, J. D., and M. T. Montgomery, 2000: Tropical cyclone evolution via potential vorticity anomalies in a three-dimensional balance model. J. Atmos. Sci., 57, 3366–3387, doi:10.1175/1520-0469(2000)057<3366:TCEVPV>2.0.CO;2.
Montgomery, M. T., and J. Enagonio, 1998: Tropical cyclogenesis via convectively forced vortex Rossby waves in a three-dimensional quasigeostrophic model. J. Atmos. Sci., 55, 3176–3207, doi:10.1175/1520-0469(1998)055<3176:TCVCFV>2.0.CO;2.
Montgomery, M. T., and R. K. Smith, 2011: The genesis of Typhoon Nuri as observed during the Tropical Cyclone Structure 2008 (TCS08) field experiment—Part 2: Observations of the convective environment. Atmos. Chem. Phys., 11, 31 115–31 136, doi:10.5194/acpd-11-31115-2011.
Montgomery, M. T., M. E. Nicholls, T. A. Cram, and A. B. Saunders, 2006: A vortical hot tower route to tropical cyclogenesis. J. Atmos. Sci., 63, 355–386, doi:10.1175/JAS3604.1.
Montgomery, M. T., Z. Wang, and T. J. Dunkerton, 2010: Coarse, intermediate and high resolution numerical simulations of the transition of a tropical wave critical layer to a tropical storm. Atmos. Chem. Phys., 10, 10 803–10 827, doi:10.5194/acp-10-10803-2010.
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.
Nesbitt, S. W., and E. J. Zipser, 2003: The diurnal cycle of rainfall and convective intensity according to three years of TRMM measurements. J. Climate, 16, 1456–1475, doi:10.1175/1520-0442-16.10.1456.
Nolan, D., 2007: What is the trigger for tropical cyclogenesis? Aust. Meteor. Mag., 56, 241–266.
Ooyama, K. V., 1982: Conceptual evolution of the theory and modeling of the tropical cyclone. J. Meteor. Soc. Japan, 60, 369–379.
Pytharoulis, I., and C. D. Thorncroft, 1999: The low-level structure of African easterly waves in 1995. Mon. Wea. Rev., 127, 2266–2280, doi:10.1175/1520-0493(1999)127<2266:TLLSOA>2.0.CO;2.
Rappin, E. D., and D. S. Nolan, 2012: The effects of vertical shear orientation on tropical cyclogenesis. Quart. J. Roy. Meteor. Soc., 138, 1035–1054, doi:10.1002/qj.977.
Raymond, D. J., and S. L. Sessions, 2007: Evolution of convection during tropical cyclogenesis. Geophys. Res. Lett.,34, L06811, doi:10.1029/2006GL028607.
Raymond, D. J., and C. López Carrillo, 2011: The vorticity budget of developing Typhoon Nuri (2008). Atmos. Chem. Phys., 11, 147–163, doi:10.5194/acp-11-147-2011.
Raymond, D. J., C. López Carrillo, and L. López Cavazos, 1998: Case-studies of developing east Pacific easterly waves. Quart. J. Roy. Meteor. Soc., 124, 2005–2034, doi:10.1002/qj.49712455011.
Raymond, D. J., S. L. Sessions, and C. López Carrillo, 2011: Thermodynamics of tropical cyclogenesis in the northwest Pacific. J. Geophys. Res.,116, D18101, doi:10.1029/2011JD015624.
Reasor, P. D., M. T. Montgomery, and L. F. Bosart, 2005: Mesoscale observations of the genesis of Hurricane Dolly (1996). J. Atmos. Sci., 62, 3151–3171, doi:10.1175/JAS3540.1.
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, 317–333, doi:10.1175/1520-0493(1977)105<0317:TSAPOA>2.0.CO;2.
Riehl, H., 1954: Tropical Meteorology. McGraw-Hill, 392 pp.
Ritchie, E. A., and G. J. Holland, 1997: Scale interactions during the formation of Typhoon Irving. Mon. Wea. Rev., 125, 1377–1396, doi:10.1175/1520-0493(1997)125<1377:SIDTFO>2.0.CO;2.
Rogers, R., and Coauthors, 2006: The Intensity Forecast Experiment: A NOAA multiyear field program for improving tropical cyclone intensity forecasts. Bull. Amer. Meteor. Soc., 87, 1523–1537, doi:10.1175/BAMS-87-11-1523.
Rogers, R., and Coauthors, 2013: NOAA’s Hurricane Intensity Forecasting Experiment: A progress report. Bull. Amer. Meteor. Soc., 94, 859–882, doi:10.1175/BAMS-D-12-00089.1.
Simpson, J., E. Ritchie, G. J. Holland, J. Halverson, and S. Stewart, 1997: Mesoscale interactions in tropical cyclone genesis. Mon. Wea. Rev., 125, 2643–2661, doi:10.1175/1520-0493(1997)125<2643:MIITCG>2.0.CO;2.
Sippel, J. A., J. W. Nielsen-Gammon, and S. E. Allen, 2006: The multiple-vortex nature of tropical cyclogenesis. Mon. Wea. Rev., 134, 1796–1814, doi:10.1175/MWR3165.1.
Smith, R. K., and M. T. Montgomery, 2012: Observations of the convective environment in developing and non-developing tropical disturbances. Quart. J. Roy. Meteor. Soc., 138, 1721–1739, doi:10.1002/qj.1910.
Spencer, R. W., H. M. Goodman, and R. E. Hood, 1989: Precipitation retrieval over land and ocean with the SSM/I: Identification and characteristics of the scattering signal. J. Atmos. Oceanic Technol., 6, 254–273, doi:10.1175/1520-0426(1989)006<0254:PROLAO>2.0.CO;2.
Thorncroft, C. D., and K. I. Hodges, 2001: African easterly wave variability and its relationship to Atlantic tropical cyclone activity. J. Climate, 14, 1166–1179, doi:10.1175/1520-0442(2001)014<1166:AEWVAI>2.0.CO;2.
Wang, Z., 2012: Thermodynamic aspects of tropical cyclone formation. J. Atmos. Sci., 69, 2433–2451, doi:10.1175/JAS-D-11-0298.1.
Wang, Z., M. T. Montgomery, and T. J. Dunkerton, 2010: Genesis of pre–Hurricane Felix (2007). Part I: The role of the easterly wave critical layer. J. Atmos. Sci., 67, 1711–1729, doi:10.1175/2009JAS3420.1.
Zawislak, J., and E. J. Zipser, 2014: Analysis of the thermodynamic properties of developing and nondeveloping tropical disturbances using a comprehensive dropsonde dataset. Mon. Wea. Rev., 142, 1250–1264, doi:10.1175/MWR-D-13-00253.1.
Zehr, R. M., 1992: Tropical cyclogenesis in the western North Pacific. NOAA Tech. Rep. NESDIS 61, 181 pp.
Zhang, D.-L., and L. Zhu, 2012: Roles of upper-level processes in tropical cyclogenesis. Geophys. Res. Lett.,39, L17804, doi:10.1029/2012GL053140.
Zipser, E. J., and C. Gautier, 1978: Mesoscale events within a GATE tropical depression. Mon. Wea. Rev., 106, 789–805, doi:10.1175/1520-0493(1978)106<0789:MEWAGT>2.0.CO;2.
The PGI number was assigned to disturbances tracked in real time during the PREDICT field program (Montgomery et al. 2012).
The climatology is part of the University of Utah TRMM Precipitation Feature (PF) database [see Liu et al. (2008) for additional details].
This threshold is identical to the one used in the TRMM Tropical Cyclone Cloud and Precipitation Feature Database (Jiang et al. 2011), although the authors do not explicitly identify this rain rate as “convective.” Even though stratiform rain can exhibit rain rates exceeding 5 mm h−1, at the coarse resolution of 3B42 this rain rate is considered sufficiently intense to be classified as convective.
Only overpasses in which the fractional coverage of the swath data within the inner core is at least 90% are used for Figs. 4c, 5c, and Table 4.
The most “favorable” convective episode is identified as the episode that exhibits comparatively, to all other episodes during the pregenesis stage, the most impressive combination of raining and convective area, convective intensity, and proximity to the center.
PMW overpasses early on 29 August indicate that the convection is located exclusively between 2° and 3° of the VM center and, therefore, does not appear in the 1° statistics in Fig. 13a. Although 3B42 rain rates (Fig. 12a) indicate heavy rainfall from this event extending within 1°, this is possibly a consequence of the 3B42 resolution and/or the deficiencies associated with the IR rain-rate retrieval in 3B42, which is used exclusively in the absence of TMI-derived rain rates.