11B-01 POLYGONAL EYEWALL ASYMMETRIES DURING THE OF HURRICANE MICHAEL (2018)

Ting-Yu Cha1*, Michael M. Bell1, Wen-Chau Lee2, Alexander J. DesRosiers1 1Colorado State University, Fort Collins, Colorado 2National Center for Atmospheric Research, Boulder, Colorado

1. INTRODUCTION Hurricane Michael (2018) was the first axisymmetric and asymmetric tangential winds Category 5 hurricane to make landfall in the (Jou et al. 2008; Cha and Bell 2019). United States since Hurricane Andrew (1992) and Analyses of Hurricane Michael (2018) caused extensive damage in and Georgia demonstrate the first observation of high-order (Beven et al. 2019). Satellite and radar imagery wave propagation using tangential wind showed evidence of an evolving polygonal asymmetries as a proxy for the PV signal. The eyewall as Michael underwent rapid results show that the propagation speeds of the intensification (RI) during its approach to Florida. waves are consistent with linear wave theory on Polygonal eyewalls are hypothesized to be the a vortex and help to provide new insight into result of asymmetric vorticity dynamics internal to physical mechanisms contributing to TC rapid the storm that can modulate TC structure and intensification. intensity through counter-propagating vortex Rossby waves (VRWs) that redistribute eyewall 2. DATA AND METHODOLOGY potential vorticity (PV) and angular momentum Hurricane Michael was within the NEXRAD and form (Hendricks et al. 2012; KEVX ground-based radar surveillance range Kuo et al. 1999,2016; Muramatsu 1986; Schubert during its rapid intensification (RI), and its et al. 1999, Lee and Wu 2019). TC internal intensity reached category 5 (140 kt) at the time dynamics are a primary factor impacting the rate of landfall around 1730 UTC (Fig. 1). The deep of intensification (Hendricks et al. 2010), layer 200 - 850 hPa vertical wind shear (VWS) suggesting that a better understanding of eyewall vector was about 5 m s-1 during the analysis structure may help to improve TC intensity period, and the VWS direction transitioned from forecasts. the southeast to the north-northeast. Previous observational studies have used Data from the operational Weather radar data to investigate polygonal eyewalls in Surveillance Doppler (WSR-88D) radar in Eglin TCs, but have been limited in their ability to AFB, Florida (KEVX) were analyzed from 1000- examine the dynamics of this phenomenon. Multi- 1930 UTC 10 October, using the Vortex Objective Doppler airborne radar analyses (Reasor et al. Radar Tracking and Circulation (VORTRAC) 2000) can retrieve the full wind field, but have software and the GVTD technique to retrieve the temporal sampling and aliasing limitations (Cha axisymmetric and asymmetric components of and Bell 2019). Studies analyzing the shape of tangential and axisymmetric radial winds (Jou et the in reflectivity (Itano and Hosoya 2013) or al. 2008, Cha and Bell 2019). One of the key using reflectivity as a proxy for PV (Corbosiero et assumptions of the GVTD retrieval is that the al. 2006) have limitations due to the reflectivity magnitude of asymmetric components of radial not being fully coupled with the PV (Moon and winds is much smaller than the magnitude of Nolan 2015). asymmetric components of tangential winds. The U.S. coastal Next Generation Weather More details can be found in Cha and Bell (2019). Radar (NEXRAD) system network provides Here, we use the combination of GVTD objective continuous surveillance of TCs near coastlines centers from 1000 to 1530 UTC and dynamic with high temporal and spatial resolution. In this centers from 1530 to 1930 UTC for the GVTD study, we perform an analysis using a novel analysis (Fig. 1a). technique by utilizing the Generalized Velocity Track Display (GVTD) technique with single 3. RESULTS ground-based Doppler radar data to retrieve the Figure 1c displays a sequence of reflectivity and spectral analysis of the radar data on * Corresponding author address: Ting-Yu Cha, October 10. Radar observations of Hurricane Colorado State University, Dept. of Atmospheric Michael show that the intensification was Science, Fort Collins, CO 80521; e-mail: accompanied by evolving reflectivity asymmetry. [email protected]. Michael's eyewall was initially characterized by 1 Figure 1: (a) Three tracks of Hurricane Michael on 10 October 2018 derived from NHC best track (NHC, blue), GVTD objective centers (red), and dynamic centers (green) from NOAA HRD (Willoughby and Chelmow 1982). (b) NHC best track wind intensity (black) and minimum central pressure (blue). The black arrows on the top row is the vertical wind shear direction from the Statistical Hurricane Intensity Prediction Scheme (SHIPS) database (DeMaria et al. 2005). Yellow shading indicates the analysis period. (c) KEVX radar reflecitivity scans at 1057, 1151, 1239, 1420, 1532, and 1728 UTC. The sequence of figures is from left to right and from top to bottom. an elliptical shape with deep convection located the increasing impact of the surface friction and at the ends of the major ellipse axis at 1100 UTC. shear as the shear was pointed to the northeast. An hour later, the elliptical eyewall evolved into a The observed reflectivity variability in triangle shape associated with strong Michael's inner core suggests that the wavenumber-1 reflectivity concentrated on the asymmetric VRWs dynamics play an important eastern quadrants. The wavy features continued role in modulating Michael's structure and rotating cyclonically along the eyewall, and intensity changes. However, up until now, there is Michael's eyewall was dominated by still no clear observational analysis to document wavenumber-1 asymmetry associated with spiral the high-wavenumber propagation of PV due to bands circulating at 1239 UTC. Afterwards, lack of both high spatial and temporal resolution Michael's eyewall transitioned to a square shape data. The GVTD technique can retrieve the at 1420 UTC, a triangle at 1532 UTC, and asymmetric tangential winds which can be used eventually axisymmetrized to a circular shape. as a proxy for vorticity with high temporal After Michael made landfall at 1730 UTC, all deep resolution and overcome past temporal convection was eroded on the southeast corner limitations. Therefore, the power spectrum of the eye, while the northwestern quadrant had radar reflectivity greater than 37 dBZ, suggesting

2 (a) (c) 1.0 1.0

0.8 0.8 ) ) 2 2 k k 2 2 C C 0.6 0.6

0.4 0.4 Power ( Power Power ( Power

0.2 0.2

0.0 0.0 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 Frequency (s 1) Frequency (s 1)

WN=1 WN=2 WN=3 WN=4 WN=5 WN=6 Nyquist frequency

(b) (d) 70 70 Theoretical wave speed for each wavenumber Theoretical wave speed for each wavenumber ) ) 60 60 1 1

50 50

40 40

30 30

20 20 Speed of the wave (ms wave the of Speed Speed of the wave (ms wave the of Speed 10 10

0 0 1 2 3 4 5 6 1 2 3 4 5 6 Wavenumber Wavenumber Figure 2: Spectral time decomposition of m = 1 – 6 (a) tangential wind components (c) reflectivity components Wavenumbers 4 to 6 are unfolded in a and c. The derived propagation speeds of each (b) tangential wavenumbers and (d) reflectivity wavenumber and the black dots are theoretical VRWs propagation speed. analyses in frequency domain of asymmetric these discrepancies are due to limitations in the tangential and reflectivity components were analysis, deviations from the theory, or performed in order to track the propagation speed uncoupling of the vorticity and divergence fields objectively (Fig. 2 a and c). Even with 5-minute is unclear. time resolution from the ground-based radar, the azimuthal frequency of high-order wavenumbers 4. SUMMARY can be aliased by the radar sampling. The The dynamical structure and evolution of maximum power of wavenumber-4 tangential Hurricane Michael during its rapid intensity wind is the close to the Nyquist frequency and change were examined using single Doppler both wavenumbers 5 and 6 appear to propagate radar observations, providing the first anticyclonically, suggesting that their propagation observational evidence of the evolving wind field speeds were aliased by the time interval of radar of a polygonal eyewall during RI to Category 5 scanning. Hence, the true propagation speed of intensity. Quantitative evidence of growing high- wavenumbers 4, 5 and 6 were unfolded using the order wavenumber structures in the tangential Nyquist velocity. wind field that suggest the presence of rapidly- Figure 2b and 2d display the theoretical speed evolving VRWs. A spectral time decomposition calculated from Eq. 1 and the observed speed. analysis of the GVTD winds indicates that the Peak power of each wavenumber is picked as the propagation speeds of different VRWs are observed VRW propagation velocity. Linearized consistent with linear wave theory on an wave theory estimates the azimuthal propagation amplifying vorticity gradient. The analysis has also demonstrated that the GVTD technique can speed of VRWs as � = � × (1 − ) where be a powerful observational tool to retrieve the TC � is the maximum tangential speed of the mean flow and m is the tangential wavenumber structural evolution during rapid intensification (Guinn and Schubert 1993, Kuo et al. 1999). The and landfall. The results highlight the value of observed values are generally consistent with the coastal radar observations to investigate physical linear wave theory, suggesting that the observed mechanisms of TC’s intensity, structure evolution asymmetries are well-described by VRW theory. with high temporal and spatial resolution, and will The propagation speeds for the reflectivity signal help to improve TC forecasts of RI in the future. are also generally consistent with the theory, but More results, including analysis of the symmetric both the wind and reflectivity deviate somewhat wind field and supporting airborne radar from the theory at above wavenumber 4. Whether observations, will be presented in the conference.

3 Vortex Kinematic Structure. Part I: 5. REFERENCES Formulation and Error Analysis. Mon. Wea. Bell, M. M., and W.-C. Lee, 2012: Objective Tropical Rev., 136, 995–1012, Cyclone Center Tracking Using Single- https://doi.org/10.1175/2007MWR2116.1. Doppler Radar. J. Appl. Meteor. Climatol., 51, 878–896, https://doi.org/10.1175/JAMC-D- Kuo, H.-C., R. T. Williams, and J.-H. Chen, 1999: A 11-0167.1. Possible Mechanism for the Eye Rotation of Typhoon Herb. J. Atmos. Sci., 56, 15. ——, M. T. Montgomery, and K. A. Emanuel, 2012: Air–Sea Enthalpy and Momentum Exchange ——, W.-Y. Cheng, Y.-T. Yang, E. A. Hendricks, and at Major Hurricane Wind Speeds Observed M. S. Peng, 2016: Deep convection in during CBLAST. J. Atmos. Sci., 69, 3197– elliptical and polygonal eyewalls of tropical 3222, https://doi.org/10.1175/JAS-D-11- cyclones: Deep Convection of Polygonal 0276.1. Eyewalls. J. Geophys. Res. Atmos., 121, 14,456-14,468, Beven II, L. J., R. Berg, and A. Hagen, 2019: https://doi.org/10.1002/2016JD025317. Hurricane Michael (2018). National Hurricane Center Tropical Cyclone Rep., 86 Lee, W.-C., and F. D. M. Jr, 2000: Tropical Cyclone pp. Kinematic Structure Retrieved from Single- Doppler Radar Observations. Part II: The Cha, T.-Y., M. M. Bell, 2019: The Comparison of GBVTD-Simplex Center Finding Algorithm. Single Doppler and Multiple Doppler Mon. Wea. Rev., 128, 12. Analyses, To be submitted. Lee, J.D. and Wu, C.C., 2018: The role of Corbosiero, K. L., J. Molinari, A. R. Aiyyer, and M. L. polygonal eyewalls in rapid intensification of Black, 2006: The Structure and Evolution of Typhoon Megi (2010). J. Atmos. Sci., 75(12), Hurricane Elena (1985). Part II: Convective 4175-4199. Asymmetries and Evidence for Vortex Rossby Waves. Mon. Wea. Rev., 134, 3073–3091, Moon, Y., and D. S. Nolan, 2015: Spiral in https://doi.org/10.1175/MWR3250.1. a Numerical Simulation of Hurricane Bill (2009). Part II: Propagation of Inner DeMaria, M., Mainelli, M., Shay, L.K., Knaff, J.A. Rainbands. J. Atmos. Sci., 72, 191–215, and Kaplan, J., 2005. Further improvements https://doi.org/10.1175/JAS-D-14-0056.1. to the statistical hurricane intensity prediction scheme (SHIPS). Wea. Forecasting, 20, 531- Muramatsu, T., 1986: The Structure of Polygonal Eye 543. of a Typhoon. Journal of the Meteorological Society of Japan, 64, 913–921, Guinn, T.A. and Schubert, W.H., 1993. https://doi.org/10.2151/jmsj1965.64.6_913. Hurricane spiral bands. J. Atmos. Sci., 50, 3380-3403. Reasor, P. D., M. T. Montgomery, F. D. M. Jr, and J. F. Gamache, 2000: Low-Wavenumber Structure Hendricks, E. A., M. S. Peng, B. Fu, and T. Li, 2010: and Evolution of the Hurricane Inner Core Quantifying Environmental Control on Observed by Airborne Dual-Doppler Radar. Tropical Cyclone Intensity Change. Mon. Wea. Mon. Wea. Rev., 128, 28. Rev., 138, 3243–3271, https://doi.org/10.1175/2010MWR3185.1. Schubert, W. H., M. T. Montgomery, R. K. Taft, T. A. Guinn, S. R. Fulton, J. P. Kossin, and J. P. ——, B. D. McNoldy, and W. H. Schubert, 2012: Edwards, 1999: Polygonal Eyewalls, Observed Inner-Core Structural Variability in Asymmetric Eye Contraction, and Potential Hurricane Dolly (2008)*. Mon. Wea. Rev., 140, Vorticity Mixing in Hurricanes. J. Atmos. Sci., 4066–4077, https://doi.org/10.1175/MWR-D- 56, 27. 12-00018.1. Willoughby, H.E. and Chelmow, M.B., 1982. Itano, T., and M. Hosoya, 2013: Spectral Analyses of Objective determination of hurricane tracks the Polygonal Eye of Typhoon Sinlaku. Mon. from aircraft observations. Mon. Wea. Wea. Rev., 141, 987–996, Rev., 110(9), pp.1298-1305. https://doi.org/10.1175/MWR-D-12-00122.1.

Jou, B. J.-D., W.-C. Lee, S.-P. Liu, and Y.-C. Kao, 2008: Generalized VTD Retrieval of Atmospheric

4