Volume 49, Issue 19 e2022GL100579
Research Letter
Open Access

Microphysical Differences in the Concentric Eyewalls of Typhoon Lekima (2019)

Xuwei Bao

Corresponding Author

Xuwei Bao

Key Laboratory of Numerical Modeling for Tropical Cyclones, and Shanghai Typhoon Institute, China Meteorological Administration, Shanghai, China

Correspondence to:

X. Bao,

[email protected]

Contribution: Conceptualization, Methodology, ​Investigation, Writing - original draft, Writing - review & editing, Project administration, Funding acquisition

Search for more papers by this author
Shuai Zhang

Shuai Zhang

Key Laboratory of Numerical Modeling for Tropical Cyclones, and Shanghai Typhoon Institute, China Meteorological Administration, Shanghai, China

Contribution: Methodology, Software, Validation

Search for more papers by this author
Yan Liang

Yan Liang

Wenzhou Meteorological Bureau, China Meteorological Administration, Wenzhou, China

Contribution: Resources, Data curation

Search for more papers by this author
First published: 06 October 2022
Citations: 2

Abstract

Using polarimetric radar observations, different microphysical characteristics were found in the asymmetric concentric eyewalls of Typhoon Lekima (2019). In addition to the differential reflectivity (ZDR) column, a typical size-sorting radar signature was evident in the upright inner eyewall: a cyclonically upwind offset of the ZDR maximum relative to the horizontal reflectivity (ZH) and specific differential phase (KDP) maxima. In contrast, the large values of ZDR were located above the melting layer along the upper edge of the more tilted outer eyewall and did not extend downward to form a ZDR column. In the cyclonic direction, moreover, the large values of the three polarimetric variables in the outer eyewall almost overlapped, without the cyclonic offset of typical size-sorting radar signature. Thus, more active warm-cloud (cold-cloud) processes dominate in the outer (inner) eyewall, particularly for the production of large drops, closely associated with the intrinsic dynamical structure of the concentric eyewalls.

Key Points

  • Different polarimetric radar signatures were present in the concentric eyewalls of Typhoon Lekima (2019)

  • Large drops are produced dominantly by the warm-cloud (cold-cloud) processes in the outer (inner) eyewall

  • Different dynamical structures in the concentric eyewalls of Lekima are responsible for the variation in their microphysics

Plain Language Summary

Intensity and precipitation forecasts of tropical cyclones (TCs) remain challenging, particularly for those with concentric eyewalls. The secondary eyewall formation and following eyewall replacement cycles frequently result in significant changes of TC structure and intensity, but this evolution is rarely successfully predicted by current numerical models. Insufficient understanding of microphysics has been a major source of uncertainties in representing these concentric eyewall cases. Typhoon Lekima (2019) was the first TC case with concentric eyewalls that made landfall in Eastern China and was completely captured by coastal dual-polarization radars, which provided an opportunity to investigate the microphysical characteristics in the concentric eyewalls. It is shown that the microphysical characteristics or processes in the secondary (primary) eyewall of Lekima were different from (similar to) typical microphysical characteristics or processes as documented in previous studies of single eyewall cases. Therefore, different model microphysical settings are suggested to represent the primary and secondary eyewalls of TCs.

1 Introduction

Over the past two decades, the lack of clear progress in the forecast of tropical cyclone (TC) intensity compared with their tracks has led to challenges in the quantitative precipitation forecast of TCs (Cangialosi et al., 2020; Gall et al., 2013; Ray et al., 2022). Secondary eyewall formation (SEF) is a common phenomenon in intense TCs, and its associated eyewall replacement cycles frequently result in structural modulation and intensity change of TCs (Houze et al., 2006; Kuo et al., 2009; Sitkowski et al., 2011; Yang et al., 2013). However, current numerical models have poor ability to predict SEF and the corresponding intensity fluctuations (Kossin & Sitkowski, 2009; F. Zhang et al., 2017; Y. Wang & Tan, 2020). Cloud or precipitation microphysics have been known to play a critical role in the intensity evolution of TCs (Khain et al., 2015; McFarquhar & Black, 2004; Pattnaik et al., 2010). Insufficient microphysical observations in TCs constrain the improvement of TC forecasts, particularly within the concentric eyewalls of a TC (Brown et al., 2016; Didlake & Kumjian, 2018; Houze et al., 2007). This study attempts to advance the understanding of microphysical characteristics and processes within the concentric eyewalls of a TC using dual-polarization radar observations.

A variety of different dynamical mechanisms have been proposed to elucidate the formation of SEF, which are mostly vortex Rossby waves, axisymmetrization process, unbalanced boundary dynamics, and so on (C. Wu et al., 2012; H. Wang et al., 2019; Kepert, 2013; Kuo et al., 2008; Montgomery & Kallenbach, 1997; Nong & Emanuel, 2003; Qiu et al., 2010; Terwey & Montgomery, 2008). Nevertheless, a consensus among these mechanisms is the crucial importance of high ambient relative humidity to the formation of the outer (secondary) eyewall (Hill & Lackmann, 2009; Y. Wang & Tan, 2020). The outer eyewall often transitions from spiral rainbands, and sometimes still retains the features of outer rainbands at upper levels, suggesting a linkage between the SEF and outer rainbands (Didlake et al., 2018; Zhu & Zhu, 2014). The different responses of the inner and outer eyewalls to environmental wind shear were found to be responsible for the variation in quadrant distribution of their precipitation asymmetries (Hence & Houze, 2012; Wunsch & Didlake, 2018; Yu et al., 2021). Hence and Houze (2012) also identified an azimuthally downwind offset (positional deviation) between the reflectivity maxima in the concentric eyewalls. However, whether there are also microphysical differences in the concentric eyewalls has not yet been well investigated.

Unique dynamics in TC eyewalls ultimately result in a wide variety of hydrometeors (Black & Hallett, 1986; Houze, 2010; Houze et al., 1992). Larger mean raindrop diameter and lower mean concentration were measured by disdrometers in the eyewall than in the spiral rainbands of a TC, as a result of stronger and deeper convection in TC eyewalls (Bao et al., 2020). As the hydrometeors with various sizes fall out, their differential terminal fall speeds result in diverse sedimentation and advection behaviors, referred to as hydrometeor size sorting (Kumjian & Ryzhkov, 2012; Rosenfeld & Ulbrich, 2003). Based on years of aircraft observations, Houze (2010) reported that heavier ice particles like graupel are generally produced and immediately fall out near the strong updrafts at the inner edge of TC eyewalls, whereas small particles can be advected radially outward and cyclonically downwind to a greater distance. The size-sorting signature in single eyewall cases was also revealed by ground-based dual-polarization radars (Didlake & Kumjian, 2018; Feng & Bell, 2019; Laurencin et al., 2020). However, whether both concentric eyewalls have the same size-sorting signature remains unclear.

Dual-polarization radars have been extensively used to investigate precipitation microphysics, because they can provide additional information like shape, type, and phase of hydrometeors (Doviak et al., 2000; G. F. Zhang, 2016; Zrnić & Aydin, 1992). The Doppler radars along the eastern coastal area of China have recently been or are being upgraded with dual-polarization capability (G. F. Zhang et al., 2019). Typhoon Lekima (2019) was fortunately captured as the first concentric eyewall case (with an inner and outer radius of maximum wind of about 15 and 50 km, respectively), as the radar in Wenzhou had completed the dual-polarization upgrade (Dai et al., 2021; Huang et al., 2022). Thus, this study will analyze the features of polarimetric variables observed in the concentric eyewalls of Lekima and discuss the relevant microphysical processes. Data and methods are introduced in Section 2. Section 3 presents the observational analysis results and related discussion. Finally, conclusions are summarized in Section 4.

2 Data and Methods

2.1 Data

As in S. Zhang et al. (2022, hereafter ZB22), the radar data were derived from the S-band dual-polarization radar in Wenzhou, Zhejiang Province of China (Figure 1a). Wenzhou radar is one of the China New Generation Doppler Weather Radars (CINRAD) upgraded with dual-polarization capability (CINRAD-98DP, Zhao et al., 2019). Besides the conventional horizontal reflectivity (ZH, dBZ), polarimetric radars can provide other parameters, such as differential reflectivity (ZDR, dB) and differential phase (ΦDP, °) or specific differential phase (KDP, ° km−1). For brevity, ZDR is the ratio of horizontal reflectivity and vertical reflectivity, as a measure of raindrop shape and size; larger values denote larger or more oblate drops. As a measure of the mass of nonspherical hydrometeors within the sampling volume, KDP is normally proportional to the number concentration of mid-sized drops (1–3 mm; Berne & Krajewski, 2013; Feng & Bell, 2019).

Details are in the caption following the image

(a) Radar horizontal reflectivity ZH (dBZ) at 0.5° elevation from the Wenzhou (WZ) radar at 1857 LST, 9 August 2019. The black triangle and red plus signs denote the locations of the radar station and the center of Typhoon Lekima (2019), respectively; the black dashed circle has a 150-km radius from the radar station; the red dashed circle presents a range of 60 km from the TC center; and the black arrow indicates the direction of environmental wind shear. (b) ZH (dBZ) of Typhoon Lekima at 0.5° elevation but altitudes ≤3 km. Black lines represent radii of 60 km extending from the typhoon center at 30° and 195°, with 0° pointing due north. The four circles have radii of 5, 27, 38, and 60 km from the tropical cyclone center, denoting the ranges of inner (black solid circles) and outer (red dashed circles) eyewalls. (c) Average ZH (dBZ) at 3-km altitude from 1857 LST 9 August to 2203 LST 10 August 2019. Panels (d and e) same as panel (c), but for average differential reflectivity ZDR (dB) and specific differential phase KDP (° km–1).

The 3-hourly track of Typhoon Lekima is from the best track data set issued by the China Meteorological Administration (CMA, Ying et al., 2014). As in ZB22, the radar data utilized in this study starts when the eye of Typhoon Lekima was within approximately 150 km of the Wenzhou radar at 1857 local standard time (LST) 9 August 2019 (Figure 1a), and ends at 2203 LST 9 August 2019 before the TC center was 60 km from the coast (not shown). This ensures that both concentric eyewalls of Lekima were over the ocean during the entire analysis period (around 3 hr). As in Dai et al. (2021), the National Centers for Environmental Prediction Final Operational Global Analysis data is also used to calculate the vertical wind shear (VWS).

2.2 Data Processing

The same quality control (QC) procedures for S-band polarimetric radar data (PRD) as in ZB22 are first applied to exclude the impact of low signal-to-noise ratios (Schuur et al., 2003), non-meteorological echoes (Giangrande & Ryzhkov, 2008; Park et al., 2009), and ground-clutter blockage (Chu et al., 2018). The volume PRD after QC is then interpolated into storm-centered cylindrical coordinates with radial, vertical, and azimuthal intervals of 1 km, 0.5 km, and 0.5°, respectively. The storm center at each radar scan time is interpolated from the best track data.

The operational volume coverage pattern makes it difficult to detect the lower atmosphere, particularly far away from radar stations. In this study, although the TC center had been within the 150-km range of the Wenzhou radar at 1857 LST 9 August, the radar detection below 3-km height (at this altitude the ice particles from the upper levels have mostly melted into raindrops) only covered the western hemisphere of the inner core of Lekima (Figure 1b). This leads to the lack of observations in the eastern hemisphere of the inner core of Lekima (particularly at lower levels) in the early stages of radar observation. In order to obtain as much microphysical information at low levels as possible, this study investigates the characteristics of polarimetric variables just in the sectors counterclockwise from 30° to 195° (with 0° pointing due north), as shown in Figures 1b–1e. In fact, these sectors cover approximately half of the downshear-left (DL) quadrant, the upshear-left (UL), and almost the entire upshear-right (UR) quadrants of the concentric eyewalls, where the intense convection and heavy precipitation asymmetries are mainly concentrated (Figures 1c–1e).

3 Results and Discussion

3.1 Structure of Concentric Eyewalls

Similar to Dai et al. (2021), the environmental wind shear points almost due east, with an average magnitude of about 10 m s−1 during the analysis period of this study (Figure 1a). Both concentric eyewalls of Lekima exhibit evident wavenumber-1 ZH asymmetry, but with clear differences (Figures 1a–1b). The ZH enhancement zone is located in the left-of-shear quadrants of the inner eyewall (5–27 km from the TC center), whereas it is concentrated in the upshear quadrants of the outer eyewall (38–60 km). The 3-hr average ZH retains this asymmetric pattern (Figure 1c). The ZH maximum exceeding 40 dBZ is concentrated in the UL quadrant of the inner eyewall, while it spans almost the entire UR quadrant of the outer eyewall. Moreover, the outer eyewall generally has larger areal coverage of ZH ≥ 40 dBZ than the inner eyewall. A more downwind distribution of ZH maxima is found in the concentric eyewalls of Lekima with respect to the typical shear-induced eyewall asymmetry (DeHart et al., 2014; Hence & Houze, 2012; Reasor et al., 2013). In addition to the dominant effect of VWS on the asymmetry, the impact of storm motion and interaction with coastal land and spiral rainbands may be responsible for the slight deviation between the asymmetric ZH pattern in Lekima and that in a typical TC (Corbosiero & Molinari, 2003; Didlake & Kumjian, 2018). Note also that an identifiable rainband surrounded the outer eyewall, and ultimately spiraled inward to the outer eyewall in the UR quadrant (Figures 1a–1c).

The ZDR and KDP fields also present a similar wavenumber-1 asymmetric pattern in the concentric eyewalls (Figures 1d–1e). In the inner eyewall, however, there is a cyclonically upwind offset of the ZDR maximum with respect to the ZH maximum, while the higher values of KDP and ZH almost overlap. This is a polarimetric radar signature indicating size-sorting of various drops with different sedimentation and advection velocities in the TC eyewalls (Didlake & Kumjian, 2018; Feng & Bell, 2019; Laurencin et al., 2020). In contrast, the ZDR maximum appears at the downwind edge of the ZH enhancement zone in the outer eyewall, which seems an opposite offset to the typical size-sorting signature. Thus, whether there is a microphysical process in the outer eyewall that differs from the typical size-sorting process will be discussed in the following subsection.

In addition to the azimuthal distribution of polarimetric variables, it is also worth analyzing their radial distribution. The higher values of ZDR (≥1.2 dB) are located mainly at the radially inner edge of the inner eyewall, and the ZH and KDP maxima lie immediately outside these high values of ZDR (Figures 1c–1e). This radial distribution of polarimetric variables is fully consistent with the conceptual model depicted by Houze et al. (1992), namely that larger drops are closer to the inner edge of the TC eyewall. This distribution of larger ZDR along the inner edge of the TC eyewall resembles the “ZDR arc” signature at the gradient edge of enhanced ZH in supercell storms (Kumjian & Ryzhkov, 2008). In contrast, there are a few sporadic snapshots of ZDR ≥ 1.2 dB near the inner edge of the outer eyewall, indicated by smaller values of KDP (Figure 1d). This may also suggest the presence of large drops at the inner edge of the outer eyewall, but with a lower concentration than that in the inner eyewall. Other high values of ZDR (≥0.8 dB, as denoted by yellow shading in Figure 1d) are located mainly in the outer zone of the outer eyewall, coincidentally overlapped by high values of ZH and KDP in the UR quadrant (Figures 1c and 1e).

Figure 2 shows radius-height cross sections of polarimetric variables, azimuthally averaged counterclockwise from 30° to 195° (as displayed in Figure 1). The inner eyewall appears to be more upright, whereas the outer eyewall is tilted, as also demonstrated in previous study (Dai et al., 2021). A visible ZDR column extending up to 7-km height is found at the inner edge of the inner eyewall (Figure 2b), consistent with the “ZDR arc”-like distribution along the inner edge in Figure 1d. This suggests that large ice drops are being produced through top-and-down growth processes, such as riming and deposition growth in the upright updrafts of the inner eyewall, and falling out to the ground (Kumjian & Prat, 2014). Meanwhile, the small values of KDP and ZH at the inner edge are representative of a low concentration of these large drops (Figures 2a and 2c). This agrees with the results of aircraft observations that large ice or rain particles are often produced within the updraft at the inner edge of TC eyewalls and immediately fall out without long radial advection (Figure 25 in Houze, 2010). Other medium to small particles with high concentration may be advected radially outward, as suggested by the large values of KDP and ZH at the outer edge of the ZDR column (Figure 2).

Details are in the caption following the image

Radius-height cross sections of (a) ZH (dBZ), (b) ZDR (dB), and (c) KDP (° km−1) azimuthally averaged counterclockwise from 30° to 195°, as shown in Figures 1c–1e, and temporally averaged from 1857 LST 9 August to 2203 LST 10 August 2019. Vertical solid and dashed lines denote the ranges of the inner and outer eyewalls of Typhoon Lekima, respectively.

In contrast, there is no such ZDR column in the more tilted outer eyewall. Instead, a band of ZDR approximately ≥1 dB is distributed along the upper edge of the outer eyewall, mostly above the melting layer (about 5.5 km from Z. Wu et al. (2021)). This ZDR band is usually indicative of pristine or branched ice crystals or aggregations (Kumjian & Prat, 2014; Kumjian & Ryzhkov, 2012). Some of these ice crystals or aggregations may be ultimately advected outward to the upper levels of spiral rainbands by the TC outflow, as denoted by large values of KDP and ZDR at upper levels about 60–80 km from the TC center. As ice particles fell through the melting layer, a well-defined band of stratiform rain, visible as a bright band of ZDR, was produced at the outer edge of the outer eyewall and in the active zone of inner rainbands (Figure 2b). Huang et al. (2022) demonstrated stronger vertical motion in the outer eyewall than in the inner eyewall, which may be responsible for the stronger upward and outward transport of hydrometeors, as shown in Figure 2. It is of interest that the ZDR column indicating strong updraft (Kumjian & Prat, 2014) does not appear in the outer eyewall with stronger updraft. Although stronger updraft in the outer eyewall makes it easier to lift hydrometeors upward to upper levels, the excessive tilting of the eyewall is not conducive to the top-and-down growth of large drops and may explain the absence of a ZDR column. As a result, the outer eyewall is characterized by larger ZH and KDP along with smaller ZDR, indicating a high concentration of medium and small raindrops, rather than a ZDR column as seen in the inner eyewall (Figure 2).

Larger average ZH is also evident in the outer eyewall from contoured frequency by altitude diagrams of polarimetric variables (Figures 3a and 3d). The larger average ZDR above 7 km denotes more planar ice-phase hydrometeors in the outer eyewall (Figures 3b and 3e), consistent with the result in Figure 2b. Below 7 km, however, the average ZDR in the inner eyewall increases sharply with decreasing height, and even exceeds the value in the outer eyewall at warm-cloud altitudes. This behavior is largely attributable to the ZDR column in the inner eyewall (Figure 2b). Figures 3c and 3f show larger average KDP at all altitudes in the outer eyewall, suggesting higher concentration of mid-sized drops, in agreement with Figure 2c. Despite the absence of the ZDR column, therefore, more intense ZH and KDP towers below the melting layer suggest stronger warm-cloud processes in the outer eyewall, also indicated by the more rapid increase of the three polarimetric variables with decreasing height below the frozen level (Figure 3). Note that here the pronounced decrease of ZDR and KDP below 2 km in the inner eyewall may be due to the lack of radar observations, which may obscure real microphysical processes at these altitudes.

Details are in the caption following the image

Contoured frequency by altitude diagrams of (a) ZH (dBZ), (b) ZDR (dB), and (c) KDP (° km−1) within the inner eyewall. Panels (d–f) same as panels (a–c), but within the outer eyewall. The contour interval is 5%, and solid and dashed lines denote the average profiles within the inner and outer eyewalls, respectively.

3.2 Size-Sorting Behavior

As a fundamental microphysical process, the size sorting of hydrometeors in TC eyewalls has been noticed by previous studies (Feng & Bell, 2019; Didlake & Kumjian, 2018; Laurencin et al., 2020). The sedimentation and advection variations of hydrometeors with different sizes tend to result in an azimuthally upwind offset of the ZDR maximum relative to the ZH and KDP maxima.

Figure 4 illustrates temporally averaged azimuth-height distributions of polarimetric variables in the concentric eyewalls. The highest altitude of the 30-dBZ ZH contour is found in the DL quadrant of the inner eyewall, and this contour slopes downward cyclonically (Figure 4a). Therefore, the high values of ZH are strongly related to the melting of large ice particles in the inner eyewall. A ZDR column extends upward beyond the melting layer at the upwind edge of the ZH enhancement zone, consistent with large raindrops formed by the melting of ice particles falling to the ground (Figure 4b). Meanwhile, two other cyclonic advection paths of ice particles are found in the left-of-shear (convection-dominated) quadrants (Figure 4b). The relatively small ice crystals drift directly downwind, before dropping through the frozen level to produce a bright band. Other medium-sized ice particles first fall through the melting layer, but they are then advected downwind by the stronger cyclonic winds at low levels, indicated by the downwind-sloping ZDR pattern with decreasing height. The vertical structure of the KDP field is mostly collocated with the ZH field, with the largest values at lower altitudes (Figure 4c). In general, the azimuth-height structures of polarimetric variables in the inner eyewall of Lekima are similar to the typical size-sorting radar signature that is illustrated in Figure 3 of Feng and Bell (2019).

Details are in the caption following the image

Azimuth-height distributions of (a) ZH (dBZ), (b) ZDR (dB), and (c) KDP (° km−1) averaged radially within the inner eyewall and temporally averaged from 1857 LST 9 August to 2203 LST 10 August 2019. Panels (d–f) same as panels (a–c), but within the outer eyewall. UR, UL, and DL represent upshear-right, upshear-left, and downshear-left quadrants. SW, W, NW, and N indicate the southwest, west, northwest, and north directions, respectively.

In contrast, the outer eyewall exhibits entirely different azimuth-height structures of polarimetric variables (Figures 4d–4f). Specifically, there is no ZDR column upwind of the ZH enhancement zone (west flank of the outer eyewall) in Figure 4e. Instead, ZDR ≥ 1 dB is only present at upper levels, rather than extending downward to the ground (Figure 4e). This may be caused by the two following reasons. First, the more slantwise updraft hinders the production of large hydrometeors via top-and-down processes, leading to more columnar or dendritic ice crystals just suspended at the upper edge of the outer eyewall (Figure 2b). Second, the stronger cyclonic circulation at lower altitudes in the outer eyewall than in the inner eyewall, as demonstrated by Huang et al. (2022), can advect raindrops farther downwind. Consequently, the bright band of ZDR in the melting layer extends upwind in the outer eyewall (Figure 4e). A similar feature was actually represented in Figure 4 of Laurencin et al. (2020) but was not discussed in their study. As these raindrops from the melting of ice particles fall down to warm-cloud altitudes, they are able to grow via collision-coalescence, as indicated by the rapid increase of polarimetric variables with decreasing height below the frozen level in Figures 4d–4e.

In addition to the rate of increase with decreasing height below the melting layer, stronger warm-cloud processes in the outer eyewall are also indicated by the ZH and KDP maxima at lower altitudes (Figures 4d and 4f). Note that the warm-cloud processes are coincidently associated with the inward spiral of an inner rainband to the outer eyewall in the UR quadrant (Figure 1). Strong dynamical convergence at low levels was found near the outer edge of the TC eyewall in previous studies, caused by the interaction of the environmental inflow toward the TC eyewall, the mesoscale descending inflow jet along the inner rainband, and the cyclonic circulation within the eyewall (Cai & Tang, 2019; Didlake & Houze, 2013a2013b; Xiao et al., 2019). The updrafts forced by this lower-level convergence may provide favorable dynamical conditions for the stronger warm-cloud processes in the outer eyewall.

The azimuth-time distributions of polarimetric variables in the concentric eyewalls also demonstrate a pronounced typical size-sorting signature in the inner eyewall (Figure 5). Namely, ZDR ≥ 1.4 dB is concentrated in the DL quadrant, whereas the larger values of ZH and KDP are mainly in the UL quadrant (downwind of the largest ZDR). In contrast, the zones of enhanced ZH, ZDR, and KDP almost overlap in the outer eyewall, although the largest ZDR (≥1.4 dB) is slightly downwind of the largest ZH. Thus, the larger raindrops in the outer eyewall should be produced mainly via the collision-coalescence process in the warm-cloud layer, as discussed above (Figures 4d–4f), rather than by the dominant contribution of the melting of large ice particles in the inner eyewall.

Details are in the caption following the image

Azimuth-time plots of (a) ZH (dBZ), (b) ZDR (dB), and (c) KDP (° km−1) at 3-km altitude, averaged radially within the inner eyewall. Panels (d–f) same as panels (a–c), but within the outer eyewall.

4 Summary

Using the polarimetric radar in Wenzhou, Zhejiang Province of China, this study investigated the differences in microphysical characteristics between the concentric eyewalls of Typhoon Lekima (2019). A different microphysical process from the typical size-sorting process was introduced to account for the cyclonic distribution of microphysics observed in the outer eyewall, as well as the absence of a ZDR column.

Both concentric eyewalls of Lekima exhibited evident wavenumber-1 asymmetry: large polarimetric variables were concentrated mainly in the left-of-shear (upshear) quadrants of the inner (outer) eyewall. A ZDR column from the height of 7 km to the ground was measured at the inner edge of the upright inner eyewall. In addition, a typical size-sorting radar signature was also found in the inner eyewall: a cyclonically upwind offset of the ZDR maximum relative to the ZH and KDP maxima as reported in Feng and Bell (2019). These typical radar signatures as documented by previous studies support the dominant contribution of the melting of large ice particles to produce large raindrops in the inner eyewall.

In contrast, the large values of ZDR were located above the melting layer along the upper edge of the more tilted outer eyewall, rather than extending downward to the ground to form a ZDR column. Despite the absence of a ZDR column, the larger values of ZH and KDP as well as their rapid rate of increase with decreasing height were found at warm-cloud altitudes in the outer eyewall. In the cyclonic direction, moreover, the zones of enhanced values of the three polarimetric variables almost overlapped at warm-cloud altitudes, and the ZDR maximum was even downwind of the ZH maximum. All these atypical radar signatures, which differ from those in the inner eyewall and have not been investigated in previous studies of single eyewall cases, suggest more active warm-cloud processes for the production of large raindrops in the outer eyewall.

It has also been discussed that the variation in the microphysical characteristics or processes should be closely associated with different dynamical structures in the concentric eyewalls of Typhoon Lekima. However, different development stages of the concentric eyewalls have been documented by previous studies, leading to the difference of microphysical characteristics and processes. Therefore, whether the results at other stages are similar to those in this study will need to be examined in the future. In addition, whether the atypical radar signatures in the outer eyewall of Lekima can be present in single eyewall cases also requires more validation work.

Acknowledgments

This study was supported jointly by the Natural Science Foundation of Shanghai (22ZR1482000), the National Natural Science Foundation of China (41975069), and Typhoon Scientific and Technological Innovation Group of Shanghai Meteorological Service and the Research Program from Science and Technology Committee of Shanghai (19dz1200101). The authors thank Dr. Guanghua Chen for insightfuly discussion on this study early. We also thank the Editor, Associate Editor, and anonymous reviewers for their critical comments and constructive suggestions that greatly improved the manuscript.

    Data Availability Statement

    The dual-polarization radar data and relevant description used in this study can be downloaded in a public repository (https://doi.org/10.6084/m9.figshare.16599341.v1). The best-track data of Lekima are publicly available from https://tcdata.typhoon.org.cn/en/index.html.