1. Introduction
Tropical cyclone (TC) intensity forecasting has been a challenging task owing to the interaction among multiscale processes (Marks and Shay 1998; Wang and Wu 2004). Among the environmental conditions, the background flow plays an important role in modulating TC intensity change, especially when the background flow contains deep-layer shear (DLS; usually defined as the horizontal wind difference between 200 and 850 hPa). Statistical analyses (DeMaria and Kaplan 1994; DeMaria 1996; Velden and Sears 2014; Wang et al. 2015; Finocchio and Majumdar 2017a; Rios-Berrios and Torn 2017) showed that TC intensification rate has a negative linear correlation (around −0.2 to −0.4) with the DLS magnitude, implying that other environmental factors would likely affect TC intensity change under sheared environments. Under moderate DLS (10–20 kt; 1 kt ≈ 0.51 m s−1), the operational intensity forecast error for intense TCs (maximum surface wind speed > 70 kt) is significantly greater than that in the regimes with weak (<10 kt) or strong (>20 kt) DLS (Bhatia and Nolan 2013). In addition, the predictability for the intensification of simulated TCs decreases under moderate-sheared environments (Zhang and Tao 2013; Finocchio and Majumdar 2017b). Therefore, the characteristics of the background flow profile need to be taken into account when investigating TC intensity change under moderate DLS.
Environmental DLS has been considered as a detrimental factor for intensification and a dominant cause for asymmetric TC structure. Several possible mechanisms have been proposed to explain how DLS impedes TC development, such as vortex tilting (Jones 1995), dilution of upper-level warm core (Frank and Ritchie 1999, 2001), and dry/low-entropy-air intrusion via the midtroposphere (Simpson and Riehl 1958; Tang and Emanuel 2010) or boundary layer (Tang and Emanuel 2010; Riemer and Laliberté 2015). In addition, inner-core asymmetries generated by the adiabatic response of thermal balance to the DLS-induced tilt (Jones 1995; DeMaria 1996) and the diabatic feedback (Frank and Ritchie 2001) can increase outward eddy fluxes of momentum, resulting in TC weakening (Wu and Braun 2004). Observational studies (e.g., Corbosiero and Molinari 2002; Chen et al. 2006; DeHart et al. 2014) have shown that the eyewall becomes more asymmetric when the DLS magnitude increases, with the maximum precipitation signature occurring in the downshear-left (DL) quadrant. Both idealized simulations (e.g., Nolan et al. 2007) and observational composites (e.g., Alvey et al. 2015) showed that the asymmetric eyewall structure limits the axisymmetric projection of diabatic heating, implying that the strong DLS is detrimental for TC intensification. Although DLS tilts and weakens TCs via the aforementioned mechanisms, TCs can withstand the negative impact of DLS to sustain its intensity or even intensify through a number of processes. Some of those processes include upshear precession (Jones 1995), vortex Rossby wave damping (Schecter et al. 2002; Reasor et al. 2004; Reasor and Montgomery 2015), downshear reformation (Molinari et al. 2004, 2006; Molinari and Vollaro 2010; Nguyen and Molinari 2012), and vortex restructuring (Rios-Berrios et al. 2018). All those processes facilitate tilt reduction, vortex realignment, and intensification.
In addition to the DLS, several studies have examined the impact of assorted characteristics of the sheared background flow profile on TC intensity. Idealized simulations, with 10 m s−1 DLS, have shown that shear height and depth (Finocchio et al. 2016) and tropical cyclone-relative environmental helicity (hereafter referred to as TCREH; Nolan 2011; Onderlinde and Nolan 2014, 2016; Gu et al. 2018, 2019) can modulate the intensity evolution and inner-core structure. The deep-sheared layer, upper-level shear, and positive TCREH are more favorable to the upshear precession process and intensification. Moreover, the relative orientation of the low-level flow (LLF) with respect to the DLS direction (Rappin and Nolan 2012; Onderlinde and Nolan 2016; Chen et al. 2019) can also affect intensification of simulated TCs. In the simulations of Rappin and Nolan (2012) and Onderlinde and Nolan (2016), upshear-pointing LLF is more favorable for TC intensification than downshear-pointing LLF. However, when the LLF points to the DL quadrant, the TC intensification rate is greater in the simulation of Chen et al. (2019). This result showed agreement with the observational composite in western North Pacific (Chen et al. 2018) which demonstrated that the left-of-shear-pointing LLF favors greater intensification rate. Both Rappin and Nolan (2012) and Chen et al. (2019) suggested that the imposition of the LLF on TC circulation enhances (reduces) the surface wind speed and surface heat fluxes on the right-hand side (left-hand side) of the center relative to the LLF direction. The change of surface heat fluxes can modulate the thermodynamic properties of the air parcels that propagate into the inner core.
Note that Rappin and Nolan (2012) and Chen et al. (2019) emphasized the importance of the LLF direction in different ways. Rappin and Nolan (2012) suggested that the upshear-pointing LLF favors intensification since the increased surface heat fluxes in the left-of-shear half can enhance the moist entropy in the boundary layer, after being ventilated by the midlevel dry air from the downstream (relative to the cyclonic circulation of TC) of the DL convection. As the moist entropy in the boundary layer can be maintained, it can support the maintenance of the convective complex in the DL quadrant and the rate of this convective complex to propagate into the upshear-left (UL) quadrant. Faster cyclonic propagation of the convective complex increases upshear precession rate which favors intensification. In contrast, Chen et al. (2019) pointed out that the DL-pointing LLF results in an enhancement of the downshear-right (DR) surface heat fluxes. Therefore, the air parcels, entering the inner core, can obtain more energy from the surface along trajectories, and thus benefiting the development of axisymmetric eyewall and intensification. The abovementioned discrepancies may arise from the model configuration, experimental design, and initial conditions. For instance, the intensity of initially dry vortex used in Rappin and Nolan (2012) and Chen et al. (2019) are 10 and 18 m s−1, respectively, which are relatively weak storms. Moreover, in their simulations, the value of DLS are of 5 and 10 m s−1, respectively.
Aside from the kinematic components, the precipitation pattern (e.g., Rogers et al. 2013; Alvey et al. 2015; Stevenson et al. 2018) and other thermodynamic factors (e.g., Rios-Berrios and Torn 2017; Nguyen et al. 2019) are also related to various intensification rates under moderate-sheared environments. Although several cases experience rapid intensification with an apparent asymmetric eyewall [e.g., Hurricane Earl (2010); Chen and Gopalakrishnan 2015], the intensifying TCs are generally accompanied by a nearly symmetric eyewall (Rogers et al. 2013; Alvey et al. 2015; Rios-Berrios et al. 2016; Stevenson et al. 2018). The symmetric eyewall provides the axisymmetric heating and reduces the dry-air intrusion from the upshear half. Both processes benefit TC intensification. In addition, the onset of rapid intensification becomes more sensitive to boundary layer moisture when DLS magnitude rises to moderate value since the distribution of convection is modulated by boundary layer moisture perturbations (Zhang and Tao 2013; Rios-Berrios et al. 2018; Rios-Berrios 2020). From a climatological perspective, Rios-Berrios and Torn (2017) demonstrated that precipitable water, sea surface temperature and relative humidity (RH) in the lower to middle troposphere of intensifying TCs are significantly higher than that of steady-state TCs under moderate DLS conditions. In addition, as shown from dropsonde observations, rapidly intensifying TCs are associated with higher values of low-level temperature, equivalent potential temperature (θe) and specific humidity in the upshear-right (UR) quadrant (Nguyen et al. 2019). Nguyen et al. (2019) showed that the enhanced UR low-level θe, resulting from the larger sea surface fluxes in the upshear half, contributes to the increased DR low-level instability, enhanced DL convection, reduced low-entropy-air intrusion and increased intensification rate.
While the aforementioned modeling studies have shown that the low-level thermodynamic structure and LLF can modulate TC intensification (Nolan 2011; Rappin and Nolan 2012; Onderlinde and Nolan 2014, 2016; Finocchio et al. 2016; Gu et al. 2018, 2019; Chen et al. 2019), all the simulations start with a weak initial vortex. Therefore, it is worth investigating the impact of the LLF on intense TCs. As the primary intensity forecast error occurs in the regime of intense storms (>70 kt) under moderate DLS (Bhatia and Nolan 2013), in this study, a set of idealized simulation is conducted to investigate the impact of the LLF direction on the intensity change of intense TCs under a moderate-sheared environment. Section 2 describes the model configuration and experimental design. The results of a simulation with no background flow and the LLF experiments (with background flow added) are presented in sections 3 and 4, respectively. A discussion and conclusions are provided in section 5.
2. Methodology
a. Model setting
The Advanced Research Weather Research and Forecasting (WRF-ARW) Model (version 3.9.1) is used to conduct idealized simulations. A two-way vortex-following nested grid is adopted in all simulations. The sizes of the three domains from the outermost to the innermost are 250 × 250, 202 × 202, and 352 × 352 grid points, with the grid spacing of 18, 6, and 2 km, respectively. The outermost domain uses doubly periodic lateral boundary conditions. All three domains have 41 vertical levels with more condensed levels in the lower level. The model top is at 25 km. All simulations are conducted on an f plane with the Coriolis parameter of 5.0 × 10−5 s−1 and fixed sea surface temperature of 28°C. The physical parameterization schemes include the WRF single-moment 6-class microphysics scheme (Hong and Lim 2006), the rapid radiative transfer model scheme (Mlawer et al. 1997) for longwave radiation, the Dudhia radiation parameterization scheme (Dudhia 1989) for shortwave radiation, the Yonsei University planetary boundary layer scheme (Hong et al. 2006), and the revised MM5 Monin–Obukhov scheme (Jiménez et al. 2012) for surface-layer scheme. No cumulus parameterization is employed owing to the convection-permitting resolution of the innermost domain.
b. The NBF and LLF experiments
To investigate the impact of the LLF direction on TC intensity under moderate-sheared environment, we first conduct a simulation under a static environment, referred to as the no background flow (NBF) experiment. The background thermodynamic properties are initialized with the Dunion moist tropical sounding (Dunion 2011). A Rankine vortex (Rotunno and Emanuel 1987) with a maximum tangential wind of 15 m s−1, a radius of maximum wind (RMW) of 82.5 km, a cutoff radius of 412.5 km, and a vortex depth of 20 km is placed at the center of the outermost domain. After the initialization, the NBF experiment is integrated for 168 h.
We aim to investigate the impact of LLF direction on TC intensity change. Therefore, a set of background flow profiles for the LLF experiments is created by systematically varying α with other fixed variables. Since we focus on the environment under moderate DLS, ush is set at 7 m s−1, the global mean of DLS according to Rios-Berrios and Torn (2017). Importantly, all profiles have the same shear magnitude. Based on the setting of Finocchio et al. (2016), pc and δp are designated at 430 and 350 hPa, respectively, yielding the sheared layer between pb = 600 hPa and pt = 250 hPa, not a sensitive regime for intensification according to Finocchio et al. (2016). The magnitude (A) of LLF is 3.5 m s−1, a reasonable value of the LLF (Chen et al. 2018). α is the controlled variable which varies between 0° and 360° in 30° increments. Figure 1 shows the hodographs and vertical profiles for the horizontal wind of LLF experiments. Each experiment name (from m00 to m33) is shown in Fig. 1a with the last two numbers representing the LLF direction divided by 10 (α/10). If α = 0° (180°), the LLF is 3.5 m s−1 westerly (easterly) which is oriented (counteroriented) to the westerly DLS (Fig. 1b). Note that changing the LLF direction results in the variation the upper-level flow in LLF experiment. However, as the zonal wind increasing rate is fixed in all simulations, the simulated TCs of the LLF experiment experience similar upper-level environmental flow if the translational speed is subtracted from the horizontal wind filed.
(a) The hodographs (250–600 hPa) of the horizontal environmental flow used in each member of the LLF experiment. The black thick arrow indicates the 7.0 m s−1 westerly deep-layer shear. The solid circle (diamond) symbol signifies the 600-hPa (250-hPa) level. (b) The vertical profile of the horizontal wind vectors for the corresponding LLF direction.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
(a) The hodographs (250–600 hPa) of the horizontal environmental flow used in each member of the LLF experiment. The black thick arrow indicates the 7.0 m s−1 westerly deep-layer shear. The solid circle (diamond) symbol signifies the 600-hPa (250-hPa) level. (b) The vertical profile of the horizontal wind vectors for the corresponding LLF direction.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
(a) The hodographs (250–600 hPa) of the horizontal environmental flow used in each member of the LLF experiment. The black thick arrow indicates the 7.0 m s−1 westerly deep-layer shear. The solid circle (diamond) symbol signifies the 600-hPa (250-hPa) level. (b) The vertical profile of the horizontal wind vectors for the corresponding LLF direction.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
The procedures for constructing the NBF and LLF experiments are outlined below:
A 7-day simulation (NBF) is simulated without any background flow.
At the time when the maximum surface wind speed of NBF reaches 70 kt, the set of designed background flow is imposed onto the vortex. To simplify the interpretation of the experiments, the shortwave and longwave radiation schemes are turned off after the background flow imposition.
Each experiment is integrated for 72 h. During the simulation, the horizontal wind beyond 300-km radius from the center of the innermost domain is nudged to the desired background flow every 30 min in order to keep the background flow unchanged throughout the simulation, while not modifying the TC structure. A 300-km radius is used since it is the separation of the environment and the area of the TC circulation according to Emanuel and Zhang (2017) and a similar value of the nudging range (approximately 360 km) in Onderlinde and Nolan (2017). Additional simulations with 200- and 400-km nudging radius yield similar intensity evolution for m00 (figure not shown); therefore, only the results with 300-km nudging radius are shown. Careful examination for the evolution of the background flow averaged in the innermost domain show that DLS and TCREH remain at the initial value during the simulation (figures not shown). Note that the environmental westerly DLS requires a meridional temperature gradient to satisfy thermal wind balance. Since the variation of the horizontal flow in the LLF experiment leads to an unrealistic pressure field with both open and closed lateral boundary condition, the doubly periodic boundary condition is applied in our simulation. However, the doubly periodic boundary condition cannot support the meridional temperature gradient. Instead of using the point-downscaling method (Nolan 2011) that modified the primitive equations of the WRF Model, nudging of the background flow is applied every 30 min to keep the background wind profile fixed during the simulation. Although the thermodynamic profile is not nudged, the environmental RH and temperature averaged within 200–800 km in radius are comparable among all experiments at all model levels with little evolution throughout the 72-h simulation (figures not shown). Moreover, since the thermodynamic properties are horizontally uniform, the role of the environmental horizontal temperature advection can be ignored.
3. Results—NBF experiment
The NBF experiment is discussed in this section. The black solid line in Fig. 2a shows the evolution of the maximum azimuthal-mean surface (at 10-m height) tangential wind speed in NBF. Since the minimum sea level pressure and the maximum surface wind speed could be largely affected by the asymmetric convection in a sheared storm, the maximum azimuthal-mean surface tangential wind speed is chosen to represent the mean component of storm intensity. The tangential and radial wind are calculated relative to the 850-hPa potential vorticity (PV) center, which represents the low-level center. The PV centers for each pressure level are defined as the weighted centroid of the squared value of PV within 50-km radius relative to the minimum sea level pressure. Note that the 50-km radius is chosen to avoid the influence of rainband-generated midlevel PV since it is 2.5 times of the RMW which is mentioned below. According to Ryglicki and Hart (2015), although different center-tracking algorithms cause a large spread of center position for weak TCs, the spread is relatively small for intense (hurricane-strength) TCs. Since we focus on intense TCs, the current definition of PV center should be suitable.
(a) Evolution of TC intensity (maximum azimuthal-mean surface tangential wind speed) for the NBF experiment (black solid line) and the LLF experiments. The left and right dashed lines denote the start (75 h) and end (147 h) time of the LLF experiments with respect to the simulation time of NBF. (b) As in (a), but for the LLF experiments with respect to the simulation time of the LLF experiments. (c) As in (b), but for the ensemble average (dark-colored thick line) of the FI (green; m15, m18, and m21) and SI (purple; m27, m30, and m33) groups. The unperturbed simulation and ensemble members are presented with a light-colored thin line. (d) The 24-h intensification rate for the LLF experiments during 12–24 h after the background flow is imposed. The abscissa is the low-level flow direction α for each member. The dark-colored (light-colored) dot represents the unperturbed simulation (five other ensemble members).
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
(a) Evolution of TC intensity (maximum azimuthal-mean surface tangential wind speed) for the NBF experiment (black solid line) and the LLF experiments. The left and right dashed lines denote the start (75 h) and end (147 h) time of the LLF experiments with respect to the simulation time of NBF. (b) As in (a), but for the LLF experiments with respect to the simulation time of the LLF experiments. (c) As in (b), but for the ensemble average (dark-colored thick line) of the FI (green; m15, m18, and m21) and SI (purple; m27, m30, and m33) groups. The unperturbed simulation and ensemble members are presented with a light-colored thin line. (d) The 24-h intensification rate for the LLF experiments during 12–24 h after the background flow is imposed. The abscissa is the low-level flow direction α for each member. The dark-colored (light-colored) dot represents the unperturbed simulation (five other ensemble members).
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
(a) Evolution of TC intensity (maximum azimuthal-mean surface tangential wind speed) for the NBF experiment (black solid line) and the LLF experiments. The left and right dashed lines denote the start (75 h) and end (147 h) time of the LLF experiments with respect to the simulation time of NBF. (b) As in (a), but for the LLF experiments with respect to the simulation time of the LLF experiments. (c) As in (b), but for the ensemble average (dark-colored thick line) of the FI (green; m15, m18, and m21) and SI (purple; m27, m30, and m33) groups. The unperturbed simulation and ensemble members are presented with a light-colored thin line. (d) The 24-h intensification rate for the LLF experiments during 12–24 h after the background flow is imposed. The abscissa is the low-level flow direction α for each member. The dark-colored (light-colored) dot represents the unperturbed simulation (five other ensemble members).
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
In NBF, the initial Rankine vortex gradually intensifies after 36 h (Fig. 2a). After reaching its maximum intensity at 108 h, the vortex slightly weakens. The maximum surface wind reaches 70 kt at 75 h (figure not shown) with the intensity of 31 m s−1 (Fig. 2a). Therefore, 75 h is chosen as the starting time for background flow imposition for the LLF experiments. During the next 48 h after 75 h, the low-level and midlevel PV rings (Figs. 3a,d) and the upper-level warm core (Fig. 3g) strengthen in the inner core of the NBF experiment. The PV ring is located inside of the surface RMW and reaches 65 potential vorticity units (PVU; 1 PVU = 10−6 K kg−1 m2 s−1) at both lower and middle levels (Figs. 3a,d). Furthermore, the upper-level warm core is concentrated near the center with the perturbation potential temperature (deviation from the averaged potential temperature beyond 500-km radius at the initial time of NBF) exceeding 12 K (Fig. 3g). The strengthening of the inner core is accompanied by the eyewall development where vigorous convection provides diabatic heating near the RMW (Fig. 3j).
The evolution of azimuthal-mean (a)–(c) 500- and (d)–(f) 850-hPa potential vorticity (PVU), (g)–(i) the perturbation of potential temperature (K) averaged between 300 and 500 hPa, and (j)–(l) the diabatic heating rate (shading; K h−1) and vertical velocity (contours; every 0.5 m s−1) averaged between 100 and 925 hPa for the (left) NBF, (center) the ensemble mean of the FI group, and (right) the ensemble mean of the SI group. The solid and dashed lines represent the 3-h running-mean RMW at the surface and the 500-hPa level, respectively.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
The evolution of azimuthal-mean (a)–(c) 500- and (d)–(f) 850-hPa potential vorticity (PVU), (g)–(i) the perturbation of potential temperature (K) averaged between 300 and 500 hPa, and (j)–(l) the diabatic heating rate (shading; K h−1) and vertical velocity (contours; every 0.5 m s−1) averaged between 100 and 925 hPa for the (left) NBF, (center) the ensemble mean of the FI group, and (right) the ensemble mean of the SI group. The solid and dashed lines represent the 3-h running-mean RMW at the surface and the 500-hPa level, respectively.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
The evolution of azimuthal-mean (a)–(c) 500- and (d)–(f) 850-hPa potential vorticity (PVU), (g)–(i) the perturbation of potential temperature (K) averaged between 300 and 500 hPa, and (j)–(l) the diabatic heating rate (shading; K h−1) and vertical velocity (contours; every 0.5 m s−1) averaged between 100 and 925 hPa for the (left) NBF, (center) the ensemble mean of the FI group, and (right) the ensemble mean of the SI group. The solid and dashed lines represent the 3-h running-mean RMW at the surface and the 500-hPa level, respectively.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
4. Results—LLF experiments
a. Intensity change and axisymmetric structure
To investigate the role of LLF direction on TC intensity change, the evolution of intensity and axisymmetric structure are examined. In Fig. 2b, all the LLF experiments weaken around 6–12 h after the imposition of the sheared background flow. After the weakening period, the LLF experiments experience a large variety of intensity evolutions. When the LLF points toward 90°–240°, the simulated TCs experience smaller intensity decrease during the weakening period and undergo faster intensification rate after 12 h (Figs. 2b,d). However, the intensity of these members (m09, m12, m15, m18, m21, and m24) remain at approximately 40 m s−1 after 24–30 h (Fig. 2b). In contrast, when the LLF points toward 0°–60° and 270°–330°, these members (m00, m03, m06, m27, m30, and m33) experience larger intensity decrease during the weakening period and intensify at a slower rate after 12 h (Figs. 2b,d). The members with the DR-pointing LLF (m27, m30, m33, and m00), while intensifying relatively slower, also reach 40 m s−1 around 36–42 h (Fig. 2b). However, the members with DL-pointing LLF (m06 and m09) remain its intensity around 32 m s−1 and even decay after 42 h (Fig. 2b).
Rather than trying to explain the complexity of the intensity evolution during the entire simulation, we focus on examining the intensification rate during 12–24 h—the period of largest intensification rate differences between experiments. Among the LLF experiments, m12, m15, and m18 are chosen to represent the fast-intensifying (FI) group. Meanwhile, m27, m30, and m33 are selected to represent the slow-intensifying (SI) group. The members with the LLF direction at the intermediate zone of each intensification regimes are chosen to make these members more suitable in representing the FI and SI groups. Note that m06 and m09 are not categorized into a group because of the relatively lower number of members. The intensity change of these members could be largely sensitive to the small LLF direction change. In addition, since the intensity evolution under the influence of DLS is sensitive to the boundary layer moisture and convection (Zhang and Tao 2013; Rios-Berrios et al. 2018; Rios-Berrios 2020), ensemble simulations are conducted to examine the robustness of the intensification rate difference. Based on the methodology of Rios-Berrios et al. (2018), the ensemble simulations are conducted by adding random perturbation to the water vapor mixing ratio (within ±5 g kg−1) below 950 hPa and only within the innermost domain at 0 h of the LLF experiment. Five ensemble members are generated in addition to each unperturbed member of both FI and SI groups, yielding a total 18 members for each group. The intensity evolutions are shown in Fig. 2c, demonstrating the apparent difference of intensity evolution before 24 h between the ensemble mean. However, the evolution of the SI group after 30 h is relatively sensitive to the boundary layer moisture perturbation, leading to insignificant intensification rate difference during 12–30 h between the FI and SI groups. The ensemble mean is next examined to elucidate the relationship between the LLF direction and intensification rate during 12–24 h. During this period, the FI (SI) group members have intensification rates ranging from 9 to 18 (2 to 11) m s−1 (24 h)−1 (Fig. 2d). Note that the 6 unperturbed members (m12, m15, m18, m27, m30, and m33) are further simulated with the radiation schemes turned on to validate the robustness of the differences between the FI and SI groups (figure not shown). The intensity evolution shows a qualitatively consistent difference between the FI and SI groups with slightly earlier reintensification, lower intensification rate, and smaller spread among the group members. These features are also demonstrated in Rios-Berrios (2020) when the DLS is of 5 m s−1.
The intensity evolution (Fig. 2c) indicates that the DLS-relative LLF direction can modulate the intensification rate for intense TCs. The UL-pointing LLF favors faster intensification rate while the DR-pointing LLF leads to slower intensification rate within 12 h after the weakening period. Rappin and Nolan (2012) demonstrated that the upshear-pointing LLF is more favorable to TC intensification than the downshear-pointing LLF. Chen et al. (2019) showed that simulated TCs tend to have higher intensification rate with DL-pointing LLF than UR-pointing LLF, yet this study shows that the DL-pointing LLF leads to slower intensification rate, followed by a decaying stage.
While the upper-level warm core and PV structure of the FI and SI groups can still develop, both groups experience a relatively slower development of these inner-core structures as compared to the NBF (Figs. 3a–i). Likewise, the deep-layer averaged diabatic heating and upward motion around the RMW are relatively weaker than that of the NBF (Figs. 3k,l). Note that apparent differences are also identified between the FI and SI groups. The low-level PV develops after 15 h in the FI group (Fig. 3e), while it weakens until around 18 h in the SI group (Fig. 3f). The midlevel PV (Figs. 3b,c) and upper-level warm core (Figs. 3h,i) of the FI group develop faster than that of the SI group. The heating in the eyewall of the FI group develops earlier at 12 h than that of the SI group as well (Figs. 3k,l).
A well-established eyewall can provide larger diabatic heating near the RMW which can support the intensification and development of the inner-core structure of a TC (Miyamoto and Takemi 2013). Moreover, the inner-core axisymmetry also increases during intensification (Miyamoto and Takemi 2013). The axisymmetric component of diabatic heating plays a dominant role on the TC intensification over the asymmetric component (Nolan et al. 2007). As in Miyamoto and Takemi (2013), the axisymmetry for a physical variable Φ is
(a) The evolution of the averaged axisymmetry of 500-hPa potential vorticity (solid line) and 300–500-hPa perturbation potential temperature (dashed line) within a 30-km radius. (b) As in (a), but for the axisymmetry of diabatic heating rate (dotted line) averaged between 100 and 925 hPa within 15–30-km radius. (c) Averaged diabatic heating rate between 100 and 925 hPa within 15–30-km radius. The green (purple) line represents the ensemble mean of the FI (SI) group.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
(a) The evolution of the averaged axisymmetry of 500-hPa potential vorticity (solid line) and 300–500-hPa perturbation potential temperature (dashed line) within a 30-km radius. (b) As in (a), but for the axisymmetry of diabatic heating rate (dotted line) averaged between 100 and 925 hPa within 15–30-km radius. (c) Averaged diabatic heating rate between 100 and 925 hPa within 15–30-km radius. The green (purple) line represents the ensemble mean of the FI (SI) group.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
(a) The evolution of the averaged axisymmetry of 500-hPa potential vorticity (solid line) and 300–500-hPa perturbation potential temperature (dashed line) within a 30-km radius. (b) As in (a), but for the axisymmetry of diabatic heating rate (dotted line) averaged between 100 and 925 hPa within 15–30-km radius. (c) Averaged diabatic heating rate between 100 and 925 hPa within 15–30-km radius. The green (purple) line represents the ensemble mean of the FI (SI) group.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
The ensemble mean of the azimuthal-mean secondary circulation for the (left) FI and (right) SI groups. The time series is shown in 6-h averaged values from 0 to 24 h. The contour (shading) denotes the vertical (radial) velocity in m s−1.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
The ensemble mean of the azimuthal-mean secondary circulation for the (left) FI and (right) SI groups. The time series is shown in 6-h averaged values from 0 to 24 h. The contour (shading) denotes the vertical (radial) velocity in m s−1.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
The ensemble mean of the azimuthal-mean secondary circulation for the (left) FI and (right) SI groups. The time series is shown in 6-h averaged values from 0 to 24 h. The contour (shading) denotes the vertical (radial) velocity in m s−1.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Radial profiles of the azimuthal-mean moist entropy (J K−1 m−2) averaged between 200 and 850 hPa for the ensemble mean of the (a) FI and (b) SI group. The time series of radial profile is shown in 3-h-averaged value from 0 to 30 h.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Radial profiles of the azimuthal-mean moist entropy (J K−1 m−2) averaged between 200 and 850 hPa for the ensemble mean of the (a) FI and (b) SI group. The time series of radial profile is shown in 3-h-averaged value from 0 to 30 h.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Radial profiles of the azimuthal-mean moist entropy (J K−1 m−2) averaged between 200 and 850 hPa for the ensemble mean of the (a) FI and (b) SI group. The time series of radial profile is shown in 3-h-averaged value from 0 to 30 h.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
b. Asymmetric structure
In previous studies (e.g., Jones 1995; Rappin and Nolan 2012; Finocchio et al. 2016; Rios-Berrios et al. 2018), the evolution of the inner-core tilt has been shown as an important feature for the TC intensification under sheared environment. The 500–850 hPa tilt magnitude evolution (Fig. 7a) shows that both FI and SI group experience an increase of tilt during the beginning of the weakening period from 6 to 9 h and tilt magnitude decrease after 9 h. The tilt of the FI group remains at a smaller value (around 4–7 km) than that of the SI group (around 6–8 km). Tilt angle also demonstrates substantial difference in the first 12 h when the midlevel center in the FI group is located at the cyclonic downwind side of midlevel center in the SI group (Fig. 7b). On the other hand, although both groups show the preference of DL tilt, there is no signal of sudden change of the position of low-level center (figure not shown) and rapid decrease of tilt (Fig. 7) as suggested by the process of downshear reformation for weak TCs (Molinari et al. 2004, 2006; Molinari and Vollaro 2010; Nguyen and Molinari 2012).
(a) Evolution of the 500–850-hPa tilt magnitude (km) of the potential vorticity centroid for the FI (green) and SI (purple) groups during 0–24 h. (b) As in (a), but for the tilt angle (°). The dark-colored thick line represents the ensemble mean, while the light-colored thin lines represent the evolution of each ensemble member.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
(a) Evolution of the 500–850-hPa tilt magnitude (km) of the potential vorticity centroid for the FI (green) and SI (purple) groups during 0–24 h. (b) As in (a), but for the tilt angle (°). The dark-colored thick line represents the ensemble mean, while the light-colored thin lines represent the evolution of each ensemble member.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
(a) Evolution of the 500–850-hPa tilt magnitude (km) of the potential vorticity centroid for the FI (green) and SI (purple) groups during 0–24 h. (b) As in (a), but for the tilt angle (°). The dark-colored thick line represents the ensemble mean, while the light-colored thin lines represent the evolution of each ensemble member.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
To examine the tilt reduction, the 500-hPa PV evolution and eyewall development are further investigated herein. A series of studies have shown that intense convective events in the inner core usually occur in the DL quadrant (e.g., Corbosiero and Molinari 2002; Chen et al. 2006; DeHart et al. 2014). However, intense convection can also develop in other quadrants. According to Alvey et al. (2015), the TCs with higher intensification rate are accompanied by greater axisymmetry of precipitation. When the eyewall precipitation extends azimuthally from the downshear half to the UL quadrant, it can enhance the axisymmetric diabatic heating and benefit intensification. Several numerical studies also showed that the UL convection can benefit the onset of intensification (e.g., Rappin and Nolan 2012; Rios-Berrios et al. 2018). In the LLF experiments, the increase of axisymmetric diabatic heating after 12 h leads to the higher intensification rate for the FI group (Figs. 4b,c). Therefore, it is worth investigating which quadrant contributes to the increase of axisymmetric diabatic heating.
Figure 8 shows the ensemble-mean simulated reflectivity and 1 m s−1 upward motion at 700 hPa and distribution of the 500-hPa PV. At the beginning (Figs. 8a,b), the eyewall convection develops in all shear-relative quadrants with numerous midlevel PV patches. As the simulation proceeds, the DLS starts to modulate the structure. At 6 h, the intense updrafts are confined in the downshear quadrants, with the maximum of simulated reflectivity at 700 hPa at the downwind side of the updrafts relative to the primary storm circulation (Figs. 8c,d). At this time, the midlevel PV forms a weak ring structure in both groups. In addition, a rainband outside the eyewall is located in the downshear quadrants. Substantial differences in convective activity occur after 12 h when the axisymmetric diabatic heating between the two groups starts to diverge (Fig. 4c). The intense updrafts of the FI group can extend from the DL quadrant to the UL quadrant (Fig. 8e), while the intense updrafts of the SI group is located in the downshear quadrants (Fig. 8f). At 12 h, the FI group shows a ring structure of midlevel PV, while the SI group has a weaker midlevel PV at the center. After 18 h, the eyewall and the midlevel PV of the FI group is well developed, showing more axisymmetric eyewall convection and intense PV ring (Figs. 8g,i). In contrast, the eyewall of the SI group is more asymmetric and sometimes open in the right-of-shear half (Figs. 8h,j) with weaker PV at the center. The axisymmetric heating has been shown as a positive factor for TC intensification (Frank and Ritchie 2001; Nolan et al. 2007) by reducing the ventilation of the inner core, especially the mid- to upper-level structure. The earlier development of PV ring in the FI group results in a more upright inner-core PV structure accompanied by a faster intensification rate during 12–24 h.
Plan view of the ensemble mean of the simulated reflectivity (shading; dBZ) and 1 m s−1 upward motion (contours) at 700 hPa and the 500-hPa potential vorticity (tiled shading; PVU) relative to the 850-hPa vortex center (origin). The black arrow denotes the DLS direction. (left) The FI group; (right) the SI group. The snapshots are shown every 6 h from 0 to 24 h.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Plan view of the ensemble mean of the simulated reflectivity (shading; dBZ) and 1 m s−1 upward motion (contours) at 700 hPa and the 500-hPa potential vorticity (tiled shading; PVU) relative to the 850-hPa vortex center (origin). The black arrow denotes the DLS direction. (left) The FI group; (right) the SI group. The snapshots are shown every 6 h from 0 to 24 h.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Plan view of the ensemble mean of the simulated reflectivity (shading; dBZ) and 1 m s−1 upward motion (contours) at 700 hPa and the 500-hPa potential vorticity (tiled shading; PVU) relative to the 850-hPa vortex center (origin). The black arrow denotes the DLS direction. (left) The FI group; (right) the SI group. The snapshots are shown every 6 h from 0 to 24 h.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
The deep-layer-mean diabatic heating averaged within the eyewall region (between 15 and 30 km in radius) is examined hereafter (Fig. 9). During the first 12 h, in agreement with previous studies (Corbosiero and Molinari 2002; Chen et al. 2006; DeHart et al. 2014; Alvey et al. 2015), the DL (UR) heating is the largest (smallest) among all quadrants in both groups (Figs. 9b,c). However, there is no substantial group difference for DL and UR heating in the first 12 h. Corresponding to the UL eyewall development (Fig. 8), the FI group have a higher UL heating after 6 h (Fig. 9a). In contrast, the SI group experience a higher DR heating in the first 12 h (Fig. 9d). After 12 h, the diabatic heating in the downshear quadrants of the FI group increases (Figs. 9b,d). Such an increase of downshear and UL diabatic heating in the FI group is related to the early eyewall development after 12 h (Fig. 8). The faster intensification rate of the FI group between 12 and 24 h is accompanied by the larger diabatic heating in all quadrants except the UR quadrant (Fig. 9). This configuration of diabatic heating is considered as a favorable factor for intensification by contributing to the increase of axisymmetric heating of the FI group after 12 h (Fig. 4c). On the other hand, the stronger DR convection and weaker UL convection result in a weaker PV tower and larger downshear tilt of the SI group.
Averaged diabatic heating rate averaged between 100 and 925 hPa within 15–30-km radius in each shear-relative quadrants in the first 24 h. The green (purple) line represents the ensemble mean of the FI (SI) group.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Averaged diabatic heating rate averaged between 100 and 925 hPa within 15–30-km radius in each shear-relative quadrants in the first 24 h. The green (purple) line represents the ensemble mean of the FI (SI) group.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Averaged diabatic heating rate averaged between 100 and 925 hPa within 15–30-km radius in each shear-relative quadrants in the first 24 h. The green (purple) line represents the ensemble mean of the FI (SI) group.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
The UL convection has been shown to benefit TC intensification in a sheared environment both in numerical simulations (e.g., Rappin and Nolan 2012; Rios-Berrios et al. 2018) and observational studies (e.g., Alvey et al. 2015; Rios-Berrios and Torn 2017). To examine the role of UL convection in the first 24 h, the evolution of UL-quadrant-averaged secondary circulation is examined (Fig. 10). After the imposition of sheared background flow, a midlevel inflow intrudes into the UL eyewall. The FI group experiences a slower decay of the UL eyewall than the SI group (Figs. 10a–d). At 6–12 h, for the UL quadrant of the SI group, the 1.5 m s−1 contour of downdraft extends from 450 to 650 hPa and the boundary layer inflow almost vanishes (Fig. 10d). In contrast, the downward motion in the UL eyewall of the FI group only reaches 0.9 m s−1 (Fig. 10c). The differences of UL eyewall evolution between the two groups (Fig. 10) correspond well to the differences of the diabatic heating before 12 h (Fig. 9a). After 12 h, the UL eyewall of FI group is reestablished with a weaker midlevel inflow and robust secondary circulation (Figs. 10e,g). Moreover, the 500-hPa RH averaged around the RMW indicates that the midlevel inflow intrudes into the eyewall convection region with notable decrease of RH in the upshear side (Fig. 11). The RH and downward motion differences in the UL quadrant between the two groups indicate that the weaker UL downward motion of the FI group during 6–12 h (Fig. 10c) results in higher upshear RH (Fig. 11c). Since the UL ventilation is weaker during the first 12 h, the boundary layer is not intruded by the low-entropy air. Afterward, the air parcels in the UL boundary layer can maintain their energy, propagate azimuthally into the DR quadrant and support the downshear convection, as suggested by Rappin and Nolan (2012) and Nguyen et al. (2019). Consequently, the enhanced eyewall organization of the FI group after 12 h leads to the higher intensification rate. However, the stronger ventilation in the SI group leads to the greater decrease of midlevel RH (Fig. 11b) which is an unfavorable configuration for intensification (Rappin and Nolan 2012; Alvey et al. 2015; Rios-Berrios and Torn 2017; Stevenson et al. 2018). The ventilation also reduces the boundary layer moisture and makes the recovery of boundary layer moisture more difficult, as shown in section 4c. In all, since the UL convection cannot only increase the axisymmetric diabatic heating in the inner core but also reduce the upshear ventilation effect, the higher UL heating in the first 12 h provides a favorable condition for the eyewall development in downshear quadrants after 12 h. The evolution of the axisymmetric and asymmetric structure indicates that the higher intensification rate of the FI group during 12–24 h mainly results from the well-constructed eyewall after 12 h. In addition, the slower decay of the UL convection in the first 12 h leads to a closed eyewall and well-developed PV ring after 12 h by enhancing the axisymmetric diabatic heating and reducing the UL ventilation effect. Therefore, it is worth examining the relationship between the LLF direction and the UL convective activity.
As in Fig. 5, but for the azimuthal mean in the UL quadrant.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
As in Fig. 5, but for the azimuthal mean in the UL quadrant.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
As in Fig. 5, but for the azimuthal mean in the UL quadrant.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Azimuth–time evolution of ensemble-mean relative humidity (%; shading) and vertical motion (tiled shading; m s−1) averaged within 15–30-km radius at 500 hPa for the (a) FI and (b) SI group and (c) the difference between the FI and SI groups. Azimuths are defined with respect to the westerly DLS and represent the following: right-of-shear (RS; −90°), downshear (DS; 0°), left-of-shear (LS; 90°), and upshear (US; −180° and 180°).
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Azimuth–time evolution of ensemble-mean relative humidity (%; shading) and vertical motion (tiled shading; m s−1) averaged within 15–30-km radius at 500 hPa for the (a) FI and (b) SI group and (c) the difference between the FI and SI groups. Azimuths are defined with respect to the westerly DLS and represent the following: right-of-shear (RS; −90°), downshear (DS; 0°), left-of-shear (LS; 90°), and upshear (US; −180° and 180°).
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Azimuth–time evolution of ensemble-mean relative humidity (%; shading) and vertical motion (tiled shading; m s−1) averaged within 15–30-km radius at 500 hPa for the (a) FI and (b) SI group and (c) the difference between the FI and SI groups. Azimuths are defined with respect to the westerly DLS and represent the following: right-of-shear (RS; −90°), downshear (DS; 0°), left-of-shear (LS; 90°), and upshear (US; −180° and 180°).
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
c. Backward trajectory analyses
To further examine the relationship between the LLF direction and the intensification rate during 12–24 h, the association between the surface heat fluxes and surface wind speed needs to be identified. The value difference of surface heat flux and the percentage change of surface wind speed between the ensemble-mean and the mean value of all the 12 unperturbed LLF experiments averaged during the first 12 h are shown in Fig. 12. At the first 12 h, the percentage change of surface wind speed is well collocated with the value difference of surface heat fluxes for both FI and SI groups, especially beyond 50-km radius where the surface heat flux changes around 10–30 W m−2. This result indicates that the LLF direction is responsible for the change of surface heat fluxes at the beginning of the LLF simulations. The FI group experiences enhanced surface heat fluxes in the downshear half (Fig. 12a), while the SI group undergoes enhanced surface heat fluxes in the upshear half (Fig. 12b). Moreover, a strong signature of surface heat fluxes enhancement occurs at 10–30 km in radius in the UL quadrant of the SI group (Fig. 12b), indicating a stronger ventilation effect in the SI group, since the surface heat fluxes also depend on the water vapor disequilibrium at the air–sea interface.
Mean surface heat flux of all the unperturbed LLF experiments (gray contours; W m−2) and the difference of surface heat flux (shading; W m−2) and the percentage change of surface wind speed (black contours; m s−1) between the ensemble mean and the mean value of all the unperturbed LLF experiments. All the variables are averaged between 0 and 12 h. (left) The result of the FI group; (right) the result of the SI group. Surface heat flux is the sum of the latent heat flux and sensible heat flux. The black arrow denotes the DLS direction.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Mean surface heat flux of all the unperturbed LLF experiments (gray contours; W m−2) and the difference of surface heat flux (shading; W m−2) and the percentage change of surface wind speed (black contours; m s−1) between the ensemble mean and the mean value of all the unperturbed LLF experiments. All the variables are averaged between 0 and 12 h. (left) The result of the FI group; (right) the result of the SI group. Surface heat flux is the sum of the latent heat flux and sensible heat flux. The black arrow denotes the DLS direction.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Mean surface heat flux of all the unperturbed LLF experiments (gray contours; W m−2) and the difference of surface heat flux (shading; W m−2) and the percentage change of surface wind speed (black contours; m s−1) between the ensemble mean and the mean value of all the unperturbed LLF experiments. All the variables are averaged between 0 and 12 h. (left) The result of the FI group; (right) the result of the SI group. Surface heat flux is the sum of the latent heat flux and sensible heat flux. The black arrow denotes the DLS direction.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
In addition to the change of surface heat fluxes, the thermodynamic properties of the boundary layer are examined. The θe averaged below 900 hPa at 12 h is chosen to represent the overall favorableness of the boundary layer for convection (Fig. 13). Both the FI and SI groups demonstrate a high θe at the center and a local minimum of θe spiraled from the downshear side to the left-of-shear side and then the upshear side at the area mainly beyond 50-km radius (Figs. 13a,b). The banded structure of the local minimum implies that the intrusion of the midlevel air is associated with the downshear rainband, DL convection, and the dynamically forced ventilation by the DLS in the upshear half. The θe difference between the FI and SI groups (Fig. 13c) illustrates that the ventilation in the left-of-shear to UL quadrant of the SI group is much stronger than that of the FI group. The θe of the FI group is several degrees Celsius higher at the downshear half beyond 50 km to the left-of-shear half in the inner core (within 30 km), providing more favorable thermodynamic conditions for inner-core convection, especially the UL convection which is addressed below.
Plan view of the ensemble-mean equivalent potential temperature (K) averaged below 900 hPa at 12 h for the (a) FI and (b) SI group and (c) the difference between the two groups (FI − SI). The black dots indicate the initial points for the backward trajectory analysis at 700 hPa. The black arrow denotes the DLS direction.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Plan view of the ensemble-mean equivalent potential temperature (K) averaged below 900 hPa at 12 h for the (a) FI and (b) SI group and (c) the difference between the two groups (FI − SI). The black dots indicate the initial points for the backward trajectory analysis at 700 hPa. The black arrow denotes the DLS direction.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Plan view of the ensemble-mean equivalent potential temperature (K) averaged below 900 hPa at 12 h for the (a) FI and (b) SI group and (c) the difference between the two groups (FI − SI). The black dots indicate the initial points for the backward trajectory analysis at 700 hPa. The black arrow denotes the DLS direction.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Rappin and Nolan (2012) indicated that the enhanced surface heat fluxes in the downwind side of the DL convective complex can support the warming and moistening of the boundary layer in the inner core. Therefore, the DL convective complex can sustain its intensity and propagate into the UL quadrant, leading to upshear precession and intensification. Although the enhanced surface heat fluxes are shown at the downstream of the DL convection in the UL quadrant (Fig. 12), the upshear precession process does not occur in both groups. Chen et al. (2019) showed that the enhanced DR surface heat fluxes can support the envelop of high θe in the boundary layer at the downshear side, leading to the axisymmetric eyewall development and intensification. Though, in this study, the major group difference of the θe in the boundary layer occurs at the region with low θe rather than the high θe envelop in the downshear side (Fig. 13c).
Since the UL convection during 6–12 h contributes to the different intensification rates, the air parcels, entering the UL convection from the boundary layer, are investigated with a backward trajectory analysis based on the Read/Interpolate/Plot 4 (RIP4; Stoelinga 2009) software. The air parcels for this analysis are initially placed at 700 hPa in the UL quadrant at 16–30 km in radius (Fig. 13), representing the place of eyewall convection with large group difference of the boundary layer θe in the inner core (Fig. 13c). The radial and azimuthal resolution for these parcels are 2 km and 5°, respectively, which yields 152 air parcels for each LLF simulation (Fig. 13). These parcels are traced backward from 12 to 0 h with a time resolution of 10 min.
Since the purpose of the trajectory analysis is the examination of the relationship between the surface heat fluxes and the properties of the air parcels that enter UL convection, two criteria are set to choose the parcels for further statistical analysis: the parcels which are (i) experiencing upward motion at 12 h at 700 hPa and (ii) located below 700 hPa at each time step of the backward trajectory. The selected parcel at each time is referred to an event. Afterward, all events from the 18 experiments in each group are analyzed collectively to represent the overall features and to obtain a large-enough sample size for reliable statistical analyses.
The results of the backward trajectory analyses are shown in Fig. 14. The mean value of moist static energy and θe of air parcels, entering the UL upward motion of the FI group at 12 h, are significantly greater than that of the SI group (Figs. 14a,c). The time evolution of moist static energy and θe show that the air parcels of the FI group maintain their energy during the propagating process in the boundary layer, while the air parcels of the SI group experience a decline of energy likely due to stronger ventilation. Moreover, the mean value of moist static energy and θe of both groups are not significantly different until 3 h (Figs. 14a,c). The height of these air parcels mainly remain around 1–2 km before ascending in the UL quadrant (Fig. 14b). The SI group shows slightly but significantly larger height of these air parcels. Therefore, the moist static energy evolution difference is mainly associated with water vapor and temperature changes. Both the water vapor mixing ratio (Fig. 14d) and temperature (Fig. 14e) show similar evolution as the moist static energy and θe (Figs. 14a,c). Overall, the air parcels that propagate from the boundary layer into the UL convection are more humid and with higher temperature, leading to greater support to the UL convection in the FI group.
Time evolution for the mean value of (a) moist static energy, (b) height, (c) equivalent potential temperature, (d) mixing ratio, (e) temperature, and (f) accumulated heating from the latent heat flux (solid line) and sensible heat flux (dotted line) for the selected events in the backward trajectory analysis started from t = 12 h. (g) As in (a), but for the total number of selected events. The selected air parcels are chosen using the criteria listed in the main text. The green (purple) line denotes the FI (SI) group. The black line in (g) shows the sum of selected events of FI and SI groups. The diamond sign indicates that the mean between FI and SI groups are significantly different at 99% confidence level based on two-tailed Student’s t test. The time resolution for backward trajectory analysis is 10 min.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Time evolution for the mean value of (a) moist static energy, (b) height, (c) equivalent potential temperature, (d) mixing ratio, (e) temperature, and (f) accumulated heating from the latent heat flux (solid line) and sensible heat flux (dotted line) for the selected events in the backward trajectory analysis started from t = 12 h. (g) As in (a), but for the total number of selected events. The selected air parcels are chosen using the criteria listed in the main text. The green (purple) line denotes the FI (SI) group. The black line in (g) shows the sum of selected events of FI and SI groups. The diamond sign indicates that the mean between FI and SI groups are significantly different at 99% confidence level based on two-tailed Student’s t test. The time resolution for backward trajectory analysis is 10 min.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Time evolution for the mean value of (a) moist static energy, (b) height, (c) equivalent potential temperature, (d) mixing ratio, (e) temperature, and (f) accumulated heating from the latent heat flux (solid line) and sensible heat flux (dotted line) for the selected events in the backward trajectory analysis started from t = 12 h. (g) As in (a), but for the total number of selected events. The selected air parcels are chosen using the criteria listed in the main text. The green (purple) line denotes the FI (SI) group. The black line in (g) shows the sum of selected events of FI and SI groups. The diamond sign indicates that the mean between FI and SI groups are significantly different at 99% confidence level based on two-tailed Student’s t test. The time resolution for backward trajectory analysis is 10 min.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
To investigate the relationship between these properties and the surface heat fluxes, the accumulated latent and sensible heating along the parcel trajectories are examined. The accumulated heating is calculated only when the air parcels are located below 950 hPa. The accumulated heating from the latent heat flux (Fig. 14f) shows significant differences between the FI and SI groups around 3.5 h, followed by a significant difference for water vapor around 4.5 h (Fig. 14d). On the other hand, although a significant difference for temperature occurs around 4.5 h (Fig. 14e), the accumulated heating from the sensible heat flux (Fig. 14f) shows substantial differences after 9 h with the value being one order of magnitude smaller than that of the latent heat flux. These selected events usually encircle the center several times before ascending in the UL quadrant, resulting in difficulties to show trajectories in a plan view. Therefore, the grid-averaged properties along the trajectories of the selected events are examined (Fig. 15). The FI group demonstrates broader horizontal range of the source of air parcels from the boundary layer than the SI group. The moist static energy and θe both show a local minimum around 20–40 km in radius (Figs. 15a–d) and inward-increasing trend of averaged simulation time (Figs. 15i,j). These features indicate that the moist static energy and θe in the boundary layer are reduced by ventilation around 20–40 km, and recovered by the surface heat fluxes afterward. Note that the overall moist static energy and θe are lower in the SI group, especially within 40-km radius (Figs. 15a–d). The averaged accumulated heating also shows a local minimum around 20–40-km radius (Figs. 15e,f), resulting from the ventilation of midlevel dry air, where the FI group shows lower influence of the ventilation effect on the accumulated heating. The distribution of grid-averaged simulation time (Figs. 15i,j) and occurrence frequency (Figs. 15m,n) also suggest that the selected air parcels of the SI group mainly comes from the ventilation in the UL quadrant. The distribution of surface heat flux experienced by the air parcels (Figs. 15g,h) implies that the selected air parcels in the FI group experience greater surface heat flux in the downshear side while those in the SI group undergo greater surface heat flux in the upshear side. Since the timing of the significant difference for the thermodynamic properties is around 3–4.5 h (Figs. 14a,c–f), the surface heat fluxes in the downshear side beyond 40-km radius play an important role for the maintenance of energy in the boundary layer in the FI group (Fig. 15g). Despite the surface heat flux is enhanced in the UL quadrant of the SI group (Fig. 12b), the UL convection is not supported by greater energy (Fig. 14a). Backward trajectory analyses starting from 6 or 9 h also show similar results (figures not shown). Therefore, in the FI group, the boundary layer air parcels provide greater energy to the UL convection during 6–12 h, thus contributing to the increased axisymmetric heating and reduced ventilation. Both of these processes favor the higher intensification rate for the FI group after 12 h. In summary, changing the LLF direction can modulate the surface heat fluxes during the first 12 h by altering the surface wind through the imposition between the TC circulation and the environmental wind. The deviation among the members of LLF experiments can result in different surface heat fluxes experienced by the selected air parcels.
Plan view of the grid-averaged (a),(b) moist static energy (kJ), (c),(d) equivalent potential temperature (K), (e),(f) accumulated heating (kJ), (g),(h) surface heat flux (W m−2), and (i),(j) simulation time (hour) for the selected events. (m),(n) The occurrence frequency (%) is shown with the number denoting the total number the selected events. The black arrow denotes the DLS direction.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Plan view of the grid-averaged (a),(b) moist static energy (kJ), (c),(d) equivalent potential temperature (K), (e),(f) accumulated heating (kJ), (g),(h) surface heat flux (W m−2), and (i),(j) simulation time (hour) for the selected events. (m),(n) The occurrence frequency (%) is shown with the number denoting the total number the selected events. The black arrow denotes the DLS direction.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
Plan view of the grid-averaged (a),(b) moist static energy (kJ), (c),(d) equivalent potential temperature (K), (e),(f) accumulated heating (kJ), (g),(h) surface heat flux (W m−2), and (i),(j) simulation time (hour) for the selected events. (m),(n) The occurrence frequency (%) is shown with the number denoting the total number the selected events. The black arrow denotes the DLS direction.
Citation: Journal of the Atmospheric Sciences 78, 9; 10.1175/JAS-D-20-0360.1
5. Discussion and conclusions
In this study, the impact of the LLF direction on the intensification of intense TCs under moderate DLS is investigated by using idealized numerical experiments based on the WRF Model. A set of background flows is constructed by varying the LLF direction with fixed moderate DLS, shear depth, shear height, LLF magnitude and TCREH. The NBF experiment is conducted under a quiescent environment. Afterward, the background flow profiles for the LLF experiments are imposed when the maximum surface wind speed of the NBF first reaches 70 kt. The intensity evolution of the LLF experiments demonstrates a weakening period during the first 12 h, followed by an intensification period. A large variety of intensity evolutions indicates that the LLF direction can have substantial impact on the intensity change of intense TCs. The UL-pointing LLF favors fast intensification, while the DR-pointing LLF leads to slow intensification. To further investigate the role of the LLF direction on the intensity change, a subset of members is selected to represent the FI and SI group based on the intensification rate during 12–24 h. Moreover, a set of ensemble simulations is conducted by perturbing the moisture below 950 hPa to test the sensitivity of the intensity change of each group to boundary layer moisture perturbation.
By analyzing the ensemble mean of axisymmetric structure, the FI group shows an earlier development of the inner-core PV rings and warm core after 12 h. Both the value and axisymmetry of the inner-core PV and warm core of the FI group increase more rapidly than that of the SI group. Corresponding to the earlier development of the PV and warm core, the eyewall convection of the FI group enhances and becomes relatively axisymmetric after 12 h. In contrast, the eyewall convection of SI group gradually develops after 24 h. In previous studies (Nolan et al. 2007; Miyamoto and Takemi 2013), the enhancement of axisymmetric diabatic heating favors the construction of the inner core structure, secondary circulation as well as the intensification. The radial profile of moist entropy also indicates that the earlier development of eyewall convection benefits the faster intensification rate of the FI group by promoting a stronger radial gradient of moist entropy.
The evolution of asymmetric structure is also investigated under the influence of DLS with different LLF direction. The evolution of tilt shows that the SI group experiences larger tilt reduction during 6–9 h while the tilt of FI group remain at a smaller value in the first 24 h. However, the upshear precession process (Jones 1995) and downshear reformation (Molinari et al. 2004, 2006; Molinari and Vollaro 2010; Nguyen and Molinari 2012), shown as important mechanisms for TC intensification in previous works, are not found in this study. After imposing the background flow, the eyewall convection becomes asymmetric in both groups. While the eyewall of the SI group is mainly active in the downshear half, the convection of the FI group can develop through the UL quadrant. The midlevel PV of FI group can develop and form a ring structure after 12 h, while the midlevel PV of SI group is weaker. The axisymmetric heating can reduce the tilt by benefitting the development of midlevel PV. The larger UL diabatic heating during 6–12 h is favorable for the subsequent increase of axisymmetric heating and tilt reduction after 12 h, since the UL heating can reduce the ventilation effect and benefit the convection development in the downshear quadrants.
A backward trajectory is applied to analyze the relationship among the LLF direction, the thermodynamic properties of the boundary layer, and the UL convection. The LLF direction is shown to be capable of modulating the surface wind speed and heat fluxes during the first 12 h when the enhanced surface heat fluxes occur mainly in the downshear and upshear half for the FI and SI groups, respectively. Moreover, the distribution of the θe within the boundary layer demonstrates that the high inner-core θe air is located under the UL eyewall of the FI group in the first 12 h. By analyzing the plan view of the grid-averaged thermodynamic properties along the air-parcel trajectories, the selected air parcels of FI group experience larger latent and sensible heat fluxes in the downshear half corresponded to the area of with enhanced surface heat flux. Since the selected air parcels of the FI group acquire higher energy supply from the surface heat fluxes in the downshear side, the UL convection is supported by the air parcels with larger moist static energy, θe, water vapor mixing ratio, and temperature. The maintenance of the UL convection then can enhance the axisymmetric heating and reduce the ventilation effect, both benefiting the following higher intensification rate.
In summary, this study emphasizes the role of LLF direction on the intensity and the structural change of intense TCs under moderate DLS. The results support that the change of surface heat fluxes generated by the imposition of the LLF and the TC circulation can lead to a variety of intensity changes. This study provides new insights in understanding the mechanism for the intensification of strong TCs under moderate DLS. Nonetheless, the sensitivity to other aspects of background flow (e.g., magnitude of the LLF, shear height, shear depth, and DLS) still needs to be addressed. Moreover, the intensity at the time for the imposition of the background can be extended to a value larger than 70 kt. In addition, the mechanism for the intensity evolution after 30 h is worth further investigation. Last, the impact of the entire structure of the background wind profile still needs to be carefully assessed in future studies.
Acknowledgments
This work is supported by the Ministry of Science and Technology of Taiwan under Grants MOST 107-2111-M-002-016-MY3 and MOST 109-2123-M-002-003-, and by the Office of Naval Research through Grant N00014-20-1-2467. RRB is supported by NCAR, which is a major facility sponsored by the U.S. National Science Foundation under Cooperative Agreement 1852977.
REFERENCES
Alvey, G. R., III, J. Zawislak, and E. Zipser, 2015: Precipitation properties observed during tropical cyclone intensity change. Mon. Wea. Rev., 143, 4476–4492, https://doi.org/10.1175/MWR-D-15-0065.1.
Bhatia, K. T., and D. S. Nolan, 2013: Relating the skill of tropical cyclone intensity forecasts to the synoptic environment. Wea. Forecasting, 28, 961–980, https://doi.org/10.1175/WAF-D-12-00110.1.
Chen, B., C. A. Davis, and Y. Kuo, 2018: Effects of low-level flow orientation and vertical shear on the structure and intensity of tropical cyclones. Mon. Wea. Rev., 146, 2447–2467, https://doi.org/10.1175/MWR-D-17-0379.1.
Chen, B., C. A. Davis, and Y. Kuo, 2019: An idealized numerical study of shear-relative low-level mean flow on tropical cyclone intensity and size. J. Atmos. Sci., 76, 2309–2334, https://doi.org/10.1175/JAS-D-18-0315.1.
Chen, H., and S. G. Gopalakrishnan, 2015: A study on the asymmetric rapid intensification of Hurricane Earl (2010) using the HWRF system. J. Atmos. Sci., 72, 531–550, https://doi.org/10.1175/JAS-D-14-0097.1.
Chen, S. S., J. A. Knaff, and F. D. Marks, 2006: Effects of vertical wind shear and storm motion on tropical cyclone rainfall asymmetries deduced from TRMM. Mon. Wea. Rev., 134, 3190–3208, https://doi.org/10.1175/MWR3245.1.
Corbosiero, K. L., and J. Molinari, 2002: The effects of vertical wind shear on the distribution of convection in tropical cyclones. Mon. Wea. Rev., 130, 2110–2123, https://doi.org/10.1175/1520-0493(2002)130<2110:TEOVWS>2.0.CO;2.
DeHart, J. C., R. A. Houze Jr., and R. F. Rogers, 2014: Quadrant distribution of tropical cyclone inner-core kinematics in relation to environmental shear. J. Atmos. Sci., 71, 2713–2732, https://doi.org/10.1175/JAS-D-13-0298.1.
DeMaria, M., 1996: The effect of vertical shear on tropical cyclone intensity change. J. Atmos. Sci., 53, 2076–2088, https://doi.org/10.1175/1520-0469(1996)053<2076:TEOVSO>2.0.CO;2.
DeMaria, M., and J. Kaplan, 1994: A Statistical Hurricane Intensity Prediction Scheme (SHIPS) for the Atlantic basin. Wea. Forecasting, 9, 209–220, https://doi.org/10.1175/1520-0434(1994)009<0209:ASHIPS>2.0.CO;2.
Dudhia, J., 1989: Numerical study of convection observed during the Winter Monsoon Experiment using a mesoscale two-dimensional model. J. Atmos. Sci., 46, 3077–3107, https://doi.org/10.1175/1520-0469(1989)046<3077:NSOCOD>2.0.CO;2.
Dunion, J. P., 2011: Rewriting the climatology of the tropical North Atlantic and Caribbean Sea atmosphere. J. Climate, 24, 893–908, https://doi.org/10.1175/2010JCLI3496.1.
Emanuel, K., and F. Zhang, 2017: The role of inner-core moisture in tropical cyclone predictability and practical forecast skill. J. Atmos. Sci., 74, 2315–2324, https://doi.org/10.1175/JAS-D-17-0008.1.
Finocchio, P. M., and S. J. Majumdar, 2017a: A statistical perspective on wind profiles and vertical wind shear in tropical cyclone environments of the Northern Hemisphere. Mon. Wea. Rev., 145, 361–378, https://doi.org/10.1175/MWR-D-16-0221.1.
Finocchio, P. M., and S. J. Majumdar, 2017b: The predictability of idealized tropical cyclones in environments with time-varying vertical wind shear. J. Adv. Model. Earth Syst., 9, 2836–2862, https://doi.org/10.1002/2017MS001168.
Finocchio, P. M., D. S. Nolan, and M. Iskandarani, 2016: Idealized tropical cyclone responses to the height and depth of environmental vertical wind shear. Mon. Wea. Rev., 144, 2155–2175, https://doi.org/10.1175/MWR-D-15-0320.1.
Frank, W. M., and E. A. Ritchie, 1999: Effects of environmental flow upon tropical cyclone structure. Mon. Wea. Rev., 127, 2044–2061, https://doi.org/10.1175/1520-0493(1999)127<2044:EOEFUT>2.0.CO;2.
Frank, W. M., and E. A. Ritchie, 2001: Effects of vertical wind shear on the intensity and structure of numerically simulated hurricanes. Mon. Wea. Rev., 129, 2249–2269, https://doi.org/10.1175/1520-0493(2001)129<2249:EOVWSO>2.0.CO;2.
Gu, J., Z. Tan, and X. Qiu, 2015: Effects of vertical wind shear on inner-core thermodynamics of an idealized simulated tropical cyclone. J. Atmos. Sci., 72, 511–530, https://doi.org/10.1175/JAS-D-14-0050.1.
Gu, J., Z. Tan, and X. Qiu, 2018: The evolution of vortex tilt and vertical motion of tropical cyclones in directional shear flows. J. Atmos. Sci., 75, 3565–3578, https://doi.org/10.1175/JAS-D-18-0024.1.
Gu, J., Z. Tan, and X. Qiu, 2019: Intensification variability of tropical cyclones in directional shear flows: Vortex tilt–convection coupling. J. Atmos. Sci., 76, 1827–1844, https://doi.org/10.1175/JAS-D-18-0282.1.
Hong, S.-Y., and J.-O. J. Lim, 2006: The WRF single-moment 6-class microphysics scheme (WSM6). J. Korean Meteor. Soc., 42, 129–151.
Hong, S.-Y., Y. Noh, and J. Dudhia, 2006: A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev., 134, 2318–2341, https://doi.org/10.1175/MWR3199.1.
Jiménez, P. A., J. Dudhia, J. F. González-Rouco, J. Navarro, J. P. Montávez, and E. García-Bustamante, 2012: A Revised scheme for the WRF surface layer formulation. Mon. Wea. Rev., 140, 898–918, https://doi.org/10.1175/MWR-D-11-00056.1.
Jones, S. C., 1995: The evolution of vortices in vertical shear. I: Initially barotropic vortices. Quart. J. Roy. Meteor. Soc., 121, 821–851, https://doi.org/10.1002/qj.49712152406.
Marks, F. D., and L. K. Shay, 1998: Landfalling tropical cyclones: Forecast problems and associated research opportunities. Bull. Amer. Meteor. Soc., 79, 305–323, https://doi.org/10.1175/1520-0477(1998)079<0305:LTCFPA>2.0.CO;2.
Miyamoto, Y., and T. Takemi, 2013: A transition mechanism for the spontaneous axisymmetric intensification of tropical cyclones. J. Atmos. Sci., 70, 112–129, https://doi.org/10.1175/JAS-D-11-0285.1.
Mlawer, E. J., S. J. Taubman, P. D. Brown, M. J. Iacono, and S. A. Clough, 1997: Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. J. Geophys. Res., 102, 16 663–16 682, https://doi.org/10.1029/97JD00237.
Molinari, J., and D. Vollaro, 2010: Rapid intensification of a sheared tropical storm. Mon. Wea. Rev., 138, 3869–3885, https://doi.org/10.1175/2010MWR3378.1.
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, https://doi.org/10.1175/JAS3291.1.
Molinari, J., P. Dodge, D. Vollaro, K. L. Corbosiero, and F. Marks, 2006: Mesoscale aspects of the downshear reformation of a tropical cyclone. J. Atmos. Sci., 63, 341–354, https://doi.org/10.1175/JAS3591.1.
Nguyen, L. T., and J. Molinari, 2012: Rapid intensification of a sheared, fast-moving hurricane over the Gulf Stream. Mon. Wea. Rev., 140, 3361–3378, https://doi.org/10.1175/MWR-D-11-00293.1.
Nguyen, L. T., R. Rogers, J. Zawislak, and J. A. Zhang, 2019: Assessing the influence of convective downdrafts and surface enthalpy fluxes on tropical cyclone intensity change in moderate vertical wind shear. Mon. Wea. Rev., 147, 3519–3534, https://doi.org/10.1175/MWR-D-18-0461.1.
Nolan, D. S., 2011: Evaluating environmental favorableness for tropical cyclone development with the method of point-downscaling. J. Adv. Model. Earth Syst., 3, M08001, https://doi.org/10.1029/2011MS000063.
Nolan, D. S., Y. Moon, and D. P. Stern, 2007: Tropical cyclone intensification from asymmetric convection: Energetics and efficiency. J. Atmos. Sci., 64, 3377–3405, https://doi.org/10.1175/JAS3988.1.
Onderlinde, M. J., and D. S. Nolan, 2014: Environmental helicity and its effects on development and intensification of tropical cyclones. J. Atmos. Sci., 71, 4308–4320, https://doi.org/10.1175/JAS-D-14-0085.1.
Onderlinde, M. J., and D. S. Nolan, 2016: Tropical cyclone–relative environmental helicity and the pathways to intensification in shear. J. Atmos. Sci., 73, 869–890, https://doi.org/10.1175/JAS-D-15-0261.1.
Onderlinde, M. J., and D. S. Nolan, 2017: The tropical cyclone response to changing wind shear using the method of time-varying point-downscaling. J. Adv. Model. Earth Syst., 9, 908–931, https://doi.org/10.1002/2016MS000796.
Rappin, E. D., and D. S. Nolan, 2012: The effect of vertical shear orientation on tropical cyclogenesis. Quart. J. Roy. Meteor. Soc., 138, 1035–1054, https://doi.org/10.1002/qj.977.
Reasor, P. D., and M. T. Montgomery, 2015: Evaluation of a heuristic model for tropical cyclone resilience. J. Atmos. Sci., 72, 1765–1782, https://doi.org/10.1175/JAS-D-14-0318.1.
Reasor, P. D., M. T. Montgomery, and L. D. Grasso, 2004: A new look at the problem of tropical cyclones in vertical shear flow: Vortex resiliency. J. Atmos. Sci., 61, 3–22, https://doi.org/10.1175/1520-0469(2004)061<0003:ANLATP>2.0.CO;2.
Riemer, M., and F. Laliberté, 2015: Secondary circulation of tropical cyclones in vertical wind shear: Lagrangian diagnostic and pathways of environmental interaction. J. Atmos. Sci., 72, 3517–3536, https://doi.org/10.1175/JAS-D-14-0350.1.
Rios-Berrios, R., 2020: Impacts of radiation and cold pools on the intensity and vortex tilt of weak tropical cyclones interacting with vertical wind shear. J. Atmos. Sci., 77, 669–689, https://doi.org/10.1175/JAS-D-19-0159.1.
Rios-Berrios, R., and R. D. Torn, 2017: Climatological analysis of tropical cyclone intensity changes under moderate vertical wind shear. Mon. Wea. Rev., 145, 1717–1738, https://doi.org/10.1175/MWR-D-16-0350.1.
Rios-Berrios, R., R. D. Torn, and C. A. Davis, 2016: An ensemble approach to investigate tropical cyclone intensification in sheared environments. Part II: Ophelia (2011). J. Atmos. Sci., 73, 1555–1575, https://doi.org/10.1175/JAS-D-15-0245.1.
Rios-Berrios, R., C. A. Davis, and R. D. Torn, 2018: A hypothesis for the intensification of tropical cyclones under moderate vertical wind shear. J. Atmos. Sci., 75, 4149–4173, https://doi.org/10.1175/JAS-D-18-0070.1.
Rogers, R., P. Reasor, and S. Lorsolo, 2013: Airborne Doppler observations of the inner-core structural differences between intensifying and steady-state tropical cyclones. Mon. Wea. Rev., 141, 2970–2991, https://doi.org/10.1175/MWR-D-12-00357.1.
Rotunno, R., and K. A. Emanuel, 1987: An air–sea interaction theory for tropical cyclones. Part II: Evolutionary study using a nonhydrostatic axisymmetric numerical model. J. Atmos. Sci., 44, 542–561, https://doi.org/10.1175/1520-0469(1987)044<0542:AAITFT>2.0.CO;2.
Ryglicki, D. R., and R. E. Hart, 2015: An investigation of center-finding techniques for tropical cyclones in mesoscale models. J. Appl. Meteor. Climatol., 54, 825–846, https://doi.org/10.1175/JAMC-D-14-0106.1.
Schecter, D. A., M. T. Montgomery, and P. D. Reasor, 2002: A theory for the vertical alignment of a quasigeostrophic vortex. J. Atmos. Sci., 59, 150–168, https://doi.org/10.1175/1520-0469(2002)059<0150:ATFTVA>2.0.CO;2.
Simpson, R. H., and H. Riehl, 1958: Mid-tropospheric ventilation as a constraint on hurricane development and maintenance. Proc. Tech. Conf. on Hurricanes, Miami, FL, Amer. Meteor. Soc., D4.1–D4.10.
Stevenson, S. N., K. L. Corbosiero, M. DeMaria, and J. L. Vigh, 2018: A 10-year survey of tropical cyclone inner-core lightning bursts and their relationship to intensity change. Wea. Forecasting, 33, 23–36, https://doi.org/10.1175/WAF-D-17-0096.1.
Stoelinga, M. T., 2009: A users’ guide to RIP version 4.5: A program for visualizing mesoscale model output. UCAR, accessed 30 July 2015, http://www2.mmm.ucar.edu/wrf/users/docs/ripug.htm.
Tang, B., and K. Emanuel, 2010: Midlevel ventilation’s constraint on tropical cyclone intensity. J. Atmos. Sci., 67, 1817–1830, https://doi.org/10.1175/2010JAS3318.1.
Velden, C. S., and J. Sears, 2014: Computing deep-tropospheric vertical wind shear analyses for tropical cyclone applications: Does the methodology matter? Wea. Forecasting, 29, 1169–1180, https://doi.org/10.1175/WAF-D-13-00147.1.
Wang, Y., and C. Wu, 2004: Current understanding of tropical cyclone structure and intensity changes—A review. Meteor. Atmos. Phys., 87, 257–278, https://doi.org/10.1007/s00703-003-0055-6.
Wang, Y., Y. Rao, Z. Tan, and D. Schönemann, 2015: A statistical analysis of the effects of vertical wind shear on tropical cyclone intensity change over the western North Pacific. Mon. Wea. Rev., 143, 3434–3453, https://doi.org/10.1175/MWR-D-15-0049.1.
Wu, L., and S. A. Braun, 2004: Effects of environmentally induced asymmetries on hurricane intensity: A numerical study. J. Atmos. Sci., 61, 3065–3081, https://doi.org/10.1175/JAS-3343.1.
Zhang, F., and D. Tao, 2013: Effects of vertical wind shear on the predictability of tropical cyclones. J. Atmos. Sci., 70, 975–983, https://doi.org/10.1175/JAS-D-12-0133.1.
