1. Introduction
The inner-core regions of tropical cyclones (TCs) are characterized by high wind speeds, turbulent activity, and convection. When an eye is present, the radius of maximum wind speed (RMW) is typically located within the eyewall. Standard measures of TC intensity are therefore sensitive to processes within the eyewall. Complex phenomena such as eyewall vorticity maxima and eyewall replacement cycles (Kossin 2015) have been observed within this region and are likely to be important factors in the development and maintenance of TCs. The turbulent activity in the inner-core boundary layer means that vertical fluxes of momentum, heat, and moisture have a significant effect on TC development (Ooyama 1969; Emanuel 1986; Rotunno et al. 2009; Smith and Montgomery 2010; Bryan 2012). Recently studies (J. A. Zhang et al. 2017; Lu and Wang 2019) showed that turbulent mixing in a numerical weather prediction (NWP) model was important in forecasting both rapid intensification events and the properties of landfalling TCs (F. Zhang et al. 2017; Zhang and Pu 2017).
In NWP models turbulent transport is parameterized as a subgrid-scale flux-gradient process with an effective turbulent diffusivity. Simulated TCs have been shown to be sensitive to the choice of planetary boundary layer (PBL) parameterizations (Braun and Tao 2000; Nolan et al. 2009; Smith and Thomsen 2010; Kepert 2012). The PBL schemes employed in NWP are usually calibrated to observations of both fluxes and diffusivities directly, and to the resulting vertical profiles of, for example, wind speed, temperature, and moisture. There remains a scarcity of observed turbulence data in the eyewall of the TC boundary layer because the very conditions which require attention—high wind speeds and turbulent intensity—make the gathering of boundary layer, inner-core data very challenging and safety concerns have largely prohibited the required research.
The most significant existing data comes from research aircraft flights in the 1980s when two hurricane boundary layer eyewall penetrations were made. Over 20 years later, Zhang et al. (2011, hereafter Z11) presented an analysis of the collected data in which turbulent fluxes and diffusivities were derived. These observations have been used by, for example, Zhang et al. (2012), as a means of calibrating an NWP PBL scheme. However, some uncertainty in these existing data due to a relatively low sampling frequency of 1 Hz and only two data points at wind speeds above 50 m s−1 prompted the authors to appeal for further analyses of the turbulent characteristics in the high-wind region of TCs. Zhang and Drennan (2012) undertook an observational study of turbulent fluxes and diffusivity in the hurricane boundary layer restricted to relatively low wind speeds outside the core but found a dependence of flux and diffusivity on height.
Here we present new vertical turbulent flux and diffusivity estimates based on data obtained from research flights in TC boundary layers comprising four eyewall penetrations over the South China Sea (SCS) jointly carried out by the Hong Kong Observatory (HKO) and the Government Flying Service of Hong Kong. These data contain more observations at wind speeds above 50 m s−1 than in the previous study and are the first over the SCS and indeed the western North Pacific.
2. Data overview
The HKO has worked in partnership with the Government Flying Service of the Hong Kong Special Administration Region since 2009 in equipping a fixed-wing aircraft, the BAe Jetstream 4100 (JS41), with an Aircraft Integrated Meteorological Measuring System 20 Hz (AIMMS20). The collaboration has resulted in research flights into TCs in the northern SCS (e.g., Chan et al. 2014). The AIMMS20 system provides measurements of the three components of the wind, temperature, relative humidity, and pressure at a frequency of 20 Hz. An overview of the AIMMS20 can be found in Beswick et al. (2008) and more details of the aircraft-AIMMS20 setup are reported in Chan et al. (2011).
The JS41 was routinely used for search and rescue (SAR) missions over the SCS region often including low-altitude surveillance flights under adverse weather conditions (including but not limited to TCs). The aircraft and its onboard instrument underwent routine inspection and maintenance by qualified engineers. No evidence was found that the data probe might be contaminated or degraded by sea salt. Furthermore, the JS41 performed routinely data comparison flights (up to three times every week) where the wind measurements by the AIMMS-20 probe were validated against concurrent Doppler lidar measurements at the Hong Kong International Airport.
Four flights through four TCs (Kalmaegi, Linfa, Mujigae, and Nida) over the period 2014–16 were chosen for this analysis as they contain legs in the boundary layer in the TC inner core. In Fig. 1 we show the evolution of the minimum pressure Pmin of the TCs observed as estimated by four TC agencies and Pmin at the time of the flight derived from flight-level data. Flights 1 and 4 coincide with the approximate lifetime Pmin of the respective cyclones, before they make landfall and dissipate. Linfa is in the process of making landfall and has possibly started to dissipate at the time of flight 2. Mujigae is undergoing a period of rapid intensification that continues for some time after flight 3.
The evolution of minimum pressure of the four TCs studies as estimated by four agencies. The red square is the minimum flight-level pressure observation extrapolated to sea level.
Citation: Journal of the Atmospheric Sciences 76, 12; 10.1175/JAS-D-19-0128.1
The flight paths, dates, and times and observed wind vectors are shown in Fig. 2. All flights begin in Hong Kong and head out into the SCS with typical airspeeds of 100 m s−1.
Path of flights segments below 1 km (dark blue) overlaid on satellite imagery. Orange symbols are arbitrarily scaled observed mean wind vectors, turquoise segments starting with black dots show the flight legs selected by quality control filters used in the analysis, and magenta is the trajectory of the cyclone during the flight interpolated from best track data. The plot titles give the cyclone name, flight segment start date and time (UTC), and duration (min).
Citation: Journal of the Atmospheric Sciences 76, 12; 10.1175/JAS-D-19-0128.1
Turbulent characteristics were calculated for selected legs (or flux runs) of each flight chosen according to quality control criteria detailed in section 4. Radial profiles of the total wind speed for each of these selected runs from the four flights are shown in Fig. 3a with radius normalized by the observed RMW. Figure 3b shows the azimuthal and radial location of each run relative to the translation of the cyclone center. It shows sampling of all quadrants with a small bias to the right front.
(a) Mean observed wind speed U against mean radius normalized by RMW (R/RMW) for each selected leg and (b) spatial distribution of legs relative to storm center with direction of storm translation at top and radial distance normalized by RMW.
Citation: Journal of the Atmospheric Sciences 76, 12; 10.1175/JAS-D-19-0128.1
3. Data analysis method
This technique implicitly assumes the Kolmogorov −5/3 power-law relationship holds over some frequency range. Bounds on the inertial range vary in the literature and are a function of the airspeed and scale of the largest eddies amongst other things. Instead of attempting to calculate a universal ideal inertial range for our data, we find the range of frequencies over which a least squares fit best matches the theoretical −5/3 gradient for each flight leg in the range 0.1–10 Hz with a width of half a decade. Deviations in the power spectrum fit gradient, merr = m + 5/3, are typically smaller than 0.02.
The position of the cyclone center is required to calculate the radial distance, R, and to transform rectilinear wind velocity components u and υ into tangential and radial components ut and ur. We found best track estimates insufficiently accurate for this purpose and so a method based on Willoughby and Chelmow (1982) was used to find the circulation center at the time of closest approach. A longer sampling window of 500 s was used to reduce the effect of asymmetric flow in the eye biasing the center location estimate. We derived the translation velocity from best track position estimates and combined it with our estimated center location at time of closest approach to give the center location as a function of time during the flight. The RMW was then estimated using the observed 1-min average wind speed.
4. Leg selection and quality control
We used an algorithm which scans through the flight-level data identifying legs which satisfy a set of quality control criteria. As momentum flux and diffusivity are known to vary with height, we constrain the analysis to legs with altitudes between 500 and 700 m. Aircraft attitude data are examined to ensure legs are level and straight. Following Zhang et al. (2009) absolute pitch and roll are required to be less than 5°. Legs with altitude variation greater than 30 m and heading variation exceeding 30° were rejected. We excluded legs with a mean wind speed below 10 m s−1. We also reject legs where an inertial range could not be found according to |merr| > αm with αm set to 0.1.
We apply two flux stationarity tests to the legs. First, a linear fit to the cumulative sum of
The above filtering resulted in 34 successful legs, corresponding to 33% of the flight time in the prescribed altitude range and includes five observations at wind speeds above 50 m s−1. This compared to the Z11 analysis where 24 legs were identified for turbulence analysis, two of which were above 50 m s−1.
Figure 4 shows analysis of a selections of legs with both high and low wind speeds passing the quality control criteria. Power spectra for
Example analysis of (a),(b) two high- and (c),(d) two low-wind-speed legs. (top to bottom) Power spectra of υt, υr, and w are in black, blue, and orange respectively, and cospectra, ogives (O), and cumulative sums (CS) of
Citation: Journal of the Atmospheric Sciences 76, 12; 10.1175/JAS-D-19-0128.1
Summary data and calculations for flight legs. Columns show flight number (1 = Kalmaegi, 2 = Linfa, 3 = Mujigae, 4 = Nida), time and date, duration (s), altitude z (m), radius R (km), total wind speed U (m s−1), Ft (m2 s−2), Fr (m2 s−2), EDR ε (cm2 s−3), σw (m s−1), TKE e (m2 s−2), Km (Hanna) (m2 s−1), and Km (TKE )(m2 s−1).
5. Results and discussion
a. Momentum flux
The vertical flux of total horizontal momentum |F| is shown as a function of total wind speed U in Fig. 5a. Momentum flux tends to increase with wind speed. The maximum flux observed here is approximately 2.5 m2 s−2 in the eyewall of Linfa at a speed of 55 m s−1. This compares to a value of 4.5 m2 s−2 observed in hurricane Hugo at a similar speed as reported in Z11. The vertical flux of tangential momentum Ft is plotted against distance from the storm center normalized by RMW in Fig. 5b. Values range from approximately −0.79 to 1.35 m2 s−2. This is a similar range to that observed by Zhang and Drennan (2012) over a range of heights in the boundary layer. There is a strong dependency on normalized radial distance from the storm center with fluxes beyond twice the RMW all positive (downward) and relatively small (less than 0.5 m2 s−2). Near the RMW fluxes exhibit a larger range with a tendency toward large negative values (upward fluxes). We have no observations of the vertical variation in υt for these cyclones but we propose the the change in sign of Ft around the RMW must coincide with a change in sign of the vertical gradient of υt at flight level (~600 m). This is consistent with the height of maximum tangential wind speed reducing from above flight level to below flight level as RMW is approached. This is in agreement with previously observed wind structure as observed, for example, by Zhang et al. (2013) in a dropsonde composite study (their Fig. 3) where the jet height lowers from above 1 km to below 500 m as RMW is approached from outside the inner core.
Vertical flux of (a) total momentum against wind speed, (b) tangential momentum against radius normalized by RMW, and (c) radial momentum against radius normalized by RMW.
Citation: Journal of the Atmospheric Sciences 76, 12; 10.1175/JAS-D-19-0128.1
Figure 5c shows the vertical flux of radial momentum Fr against R/RMW. Fluxes are mostly positive (downward) varying from 0 to 2 m2 s−2. There is a dependence on R/RMW, with observations in the eyewall tending to be much larger. This is consistent with increased vertical momentum diffusivity Km in the high wind speeds in the eyewall. Steepening of the vertical gradient of υr at flight level in the eyewall may also contribute to this effect and, again, this agrees with the dropsonde composite study of Zhang et al. (2013, their Fig. 4.).
b. Vertical momentum diffusivity
Vertical momentum diffusivity Km calculated using the Hanna method against wind speed found by (a) Z11 and (b) this study. Coefficients refer to Km = aUb determined through linear regression of log10Km onto log10U.
Citation: Journal of the Atmospheric Sciences 76, 12; 10.1175/JAS-D-19-0128.1
Since Km is known to vary with height, comparison of the two datasets should be qualified by the broader range of heights spanned by the Z11 data (422–847 m) than the new HKO data (583–653 m). However, most of the measurements from both sets are contained within the height range 450–650 m and limiting the data to this range gives similar power-law exponents (Z11: b = 1.06 ± 0.28; HKO: b = 1.65 ± 0.28). We also examined the residuals of the Km–U model and found no significant dependence on height for either the Z11 or the HKO observations. This is likely due the relatively narrow range of altitudes sampled and gives some confidence that the difference in exponents between datasets is not purely an artifact of height difference.
There is a large scatter in the Km–U plot reflected by an r2 value of 0.43. Values of Km at around 30 m s−1 vary by over a factor of 10 suggesting the importance of influences other than U in determining Km. As the scatter around the Km–U model is not explained by variation in altitude we expect that other factors (e.g., stability, wind shear) which can influence turbulent mixing are responsible and acknowledge that U is only one of many potential covariates of Km.
The TKE method of calculating Km yields a closer agreement with the Z11 data. We find b coefficient of 0.91 ± 0.30 with the corresponding value in the Z11 data 1.13 ± 0.21. However, we have lower confidence in using this method as the scatter was much larger (r2 = 0.16) and the b value was very sensitive to quality control filters applied. This is possibly because the stationary tests used here do not directly test for the stationarity of the components of TKE (
The power-law formulation and exponent provides a method for comparing observed diffusivity behavior with wind speed to those generated by parameterizations in numerical models. To demonstrate this, we present preliminary results from idealized TC simulations using the Weather Research and Forecasting (WRF) model (version 3.7.1; Skamarock et al. 2008) configured as described in Wang and Toumi (2019). The two most widely used parameterizations are examined: 1) the Yonsei University (YSU; Hong et al. 2006) nonlocal PBL scheme coupled with the MM5 similarity surface layer scheme (Zhang and Anthes 1982) and 2) the local Mellor–Yamada–Janjić (MYJ) PBL scheme with the Janjić Eta Monin–Obukhov surface-layer scheme (Janjić 2002). The simulated TCs were allowed to mature, then for 2 days, hourly instantaneous values of Km and U at approximately 570 m were azimuthally averaged (Zhang et al. 2012). Figure 7 shows the simulated U and Km values for the two simulations with power-law fit coefficients. The exponents for the YSU and MYJ simulations were 0.89 ± 0.01 and 4.25 ± 0.02, respectively. Our observed exponent of 1.73 ± 0.20 sits between these values and is compatible with neither to within standard error. We note that at, for example, 40 m s−1, the diffusivities are similar between the parameterizations but the variation with wind speed is dramatically different making the case for determining the exponent. A thorough examination is beyond the scope of this paper instead we present this analysis to demonstrate the value of the power-law exponent as a means of comparison between observations and models. A full analysis could examine spatial and temporal scales, TC development stages, the height, and the role of the surface layer.
Azimuthally averaged vertical momentum diffusivity Km against wind speed U for idealized TC simulations using the YSU (blue squares) and MYJ (orange circles) PBL schemes. Estimates of coefficients a and b refer to the power-law model Km = aUb. Data from within the eye and outside 10 times the RMW were excluded.
Citation: Journal of the Atmospheric Sciences 76, 12; 10.1175/JAS-D-19-0128.1
c. Power law relations of EDR, TKE, and σw to U
The Km–U relationship can be thought of in terms of the power-law behavior of the components of both the Hanna and TKE Km formulas. In Fig. 8 we show the log–log scatters of EDR, σw, and TKE against U revealing their power-law nature. Applying the same power-law model as above for ε–U gives bε = 2.77. Similarly, for σw–U we find
Log-scale plots of (a) EDR, (b) σw, and (c) TKE against U. The parameters a and b refer to y = axb, which were calculated using linear regression as described in the text. Markers are as in Fig. 5.
Citation: Journal of the Atmospheric Sciences 76, 12; 10.1175/JAS-D-19-0128.1
6. Conclusions
Observations of vertical momentum flux and diffusivity in the high-wind boundary layer region of TCs are scarce. We present new data from four eyewall penetration flights into TC boundary layers in the SCS jointly carried out by the HKO and Government Flying Service of Hong Kong. This paper shows the first direct observations of negative (upward) vertical fluxes of tangential momentum in the eyewall of the TC boundary layer. These are consistent with the current knowledge of TC wind structure.
We also demonstrate a robust, approximately superlinear dependence of Km on U across the range of observed wind speeds. This finding disagrees with the only previous example of diffusivity measurements at very high wind speeds in the Hurricane boundary layer which found the relationship to be approximately linear. However, the new data has more points at very high wind speeds, increasing confidence that the Km–U relationship we present holds in the eyewall of TCs.
The new observational data presented here may be useful in helping constrain the parameterization schemes in numerical models dealing with subgrid-scale mixing in extreme wind regimes, potentially improving their forecast skill.
Acknowledgments
This work and its contributors from Imperial College London were supported by the U.K.–China Research and Innovation Partnership Fund through the Met Office Climate Science for Service Partnership China as part of the Newton Fund. The Government Flying Service conducted the reconnaissance flights and the Hong Kong Observatory contributed the flight data for this study. Shuai Wang is also funded by the National Natural Science Foundation of China (Grant 41706007).
REFERENCES
Affre, C., A. Lopez, A. Carrara, A. Druilhet, and J. Fontan, 2000: The analysis of energy and ozone flux data from the LANDES 94 experiment. Atmos. Environ., 34, 803–821, https://doi.org/10.1016/S1352-2310(99)00295-2.
Beswick, K. M., M. W. Gallagher, A. R. Webb, E. G. Norton, and F. Perry, 2008: Application of the Aventech AIMMS20AQ airborne probe for turbulence measurements during the Convective Storm Initiation Project. Atmos. Chem. Phys., 8, 5449–5463, https://doi.org/10.5194/acp-8-5449-2008.
Braun, S. A., and W.-K. Tao, 2000: Sensitivity of high-resolution simulations of Hurricane Bob (1991) to planetary boundary layer parameterizations. Mon. Wea. Rev., 128, 3941–3961, https://doi.org/10.1175/1520-0493(2000)129<3941:SOHRSO>2.0.CO;2.
Bryan, G. H., 2012: Effects of surface exchange coefficients and turbulence length scales on the intensity and structure of numerically simulated hurricanes. Mon. Wea. Rev., 140, 1125–1143, https://doi.org/10.1175/MWR-D-11-00231.1.
Chan, P. W., K. K. Hon, and S. Foster, 2011: Wind data collected by a fixed-wing aircraft in the vicinity of a tropical cyclone over the south China coastal waters. Meteor. Z., 20, 313–321, https://doi.org/10.1127/0941-2948/2011/0505.
Chan, P. W., W. K. Wong, and K. K. Hon, 2014: Weather observations by aircraft reconnaissance inside Severe Typhoon Utor. Weather, 69, 199–203, https://doi.org/10.1002/wea.2315.
Emanuel, K. A., 1986: An air-sea interaction theory for tropical cyclones. Part I: Steady-state maintenance. J. Atmos. Sci., 43, 585–605, https://doi.org/10.1175/1520-0469(1986)043<0585:AASITF>2.0.CO;2.
Hanna, S. R., 1968: A method of estimating vertical eddy transport in the planetary boundary layer using characteristics of the vertical velocity spectrum. J. Atmos. Sci., 25, 1026–1033, https://doi.org/10.1175/1520-0469(1968)025<1026:AMOEVE>2.0.CO;2.
Hong, S., 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.
Janjić, Z., 2002: Nonsingular implementation of the Mellor–Yamada level 2.5 scheme in the NCEP Meso model. NCEP Office Note 437, 61 pp., http://www.emc.ncep.noaa.gov/officenotes/newernotes/on437.pdf.
Kepert, J. D., 2012: Choosing a boundary layer parameterization for tropical cyclone modeling. Mon. Wea. Rev., 140, 1427–1445, https://doi.org/10.1175/MWR-D-11-00217.1.
Kossin, J. P., 2015: Hurricane wind–pressure relationship and eyewall replacement cycles. Wea. Forecasting, 30, 177–181, https://doi.org/10.1175/WAF-D-14-00121.1.
Lu, X., and X. Wang, 2019: Improving hurricane analyses and predictions with TCI, IFEX field campaign observations, and CIMSS AMVs using the advanced hybrid data assimilation system for HWRF. Part I: What is missing to capture the rapid intensification of Hurricane Patricia (2015). Mon. Wea. Rev., 147, 1351–1373, https://doi.org/10.1175/MWR-D-18-0202.1.
Nolan, D. S., J. A. Zhang, D. P. Stern, D. S. Nolan, J. A. Zhang, and D. P. Stern, 2009: Evaluation of planetary boundary layer parameterizations in tropical cyclones by comparison of in situ observations and high-resolution simulations of Hurricane Isabel (2003). Part I: Initialization, maximum winds, and the outer-core boundary layer. Mon. Wea. Rev., 137, 3651–3674, https://doi.org/10.1175/2009MWR2785.1.
Ooyama, K., 1969: Numerical simulation of the life cycle of tropical cyclones. J. Atmos. Sci., 26, 3–40, https://doi.org/10.1175/1520-0469(1969)026<0003:NSOTLC>2.0.CO;2.
Rotunno, R., Y. Chen, W. Wang, C. Davis, J. Dudhia, and G. J. Holland, 2009: Large-eddy simulation of an idealized tropical cyclone. Bull. Amer. Meteor. Soc., 90, 1783–1788, https://doi.org/10.1175/2009BAMS2884.1.
Skamarock, W., and Coauthors, 2008: A description of the Advanced Research WRF version 3. NCAR Tech. Note NCAR/TN-475+STR, 113 pp., https://doi.org/10.5065/D68S4MVH.
Smith, R. K., and M. T. Montgomery, 2010: Hurricane boundary-layer theory. Quart. J. Roy. Meteor. Soc., 136, 1665–1670, https://doi.org/10.1002/qj.679.
Smith, R. K., and G. L. Thomsen, 2010: Dependence of tropical-cyclone intensification on the boundary-layer representation in a numerical model. Quart. J. Roy. Meteor. Soc., 136, 1671–1685, https://doi.org/10.1002/qj.687.
Sreenivasan, K. R., 1995: On the universality of the Kolmogorov constant. Phys. Fluids, 7, 2778–2784, https://doi.org/10.1063/1.868656.
Večenaj, Ž., D. Belušić, and B. Grisogono, 2010: Characteristics of the near-surface turbulence during a bora event. Ann. Geophys., 28, 155–163, https://doi.org/10.5194/angeo-28-155-2010.
Wang, S., and R. Toumi, 2019: Impact of dry midlevel air on the tropical cyclone outer circulation. J. Atmos. Sci., 76, 1809–1826, https://doi.org/10.1175/JAS-D-18-0302.1.
Willoughby, H. E., and M. B. Chelmow, 1982: Objective determination of hurricane tracks from aircraft observations. Mon. Wea. Rev., 110, 1298–1305, https://doi.org/10.1175/1520-0493(1982)110<1298:ODOHTF>2.0.CO;2.
Zhang, D., and R. A. Anthes, 1982: A high-resolution model of the planetary boundary layer—Sensitivity tests and comparisons with SESAME-79 data. J. Appl. Meteor., 21, 1594–1609, https://doi.org/10.1175/1520-0450(1982)021<1594:AHRMOT>2.0.CO;2.
Zhang, F., and Z. Pu, 2017: Effects of vertical eddy diffusivity parameterization on the evolution of landfalling hurricanes. J. Atmos. Sci., 74, 1879–1905, https://doi.org/10.1175/JAS-D-16-0214.1.
Zhang, F., Z. Pu, and C. Wang, 2017: Effects of boundary layer vertical mixing on the evolution of hurricanes over land. Mon. Wea. Rev., 145, 2343–2361, https://doi.org/10.1175/MWR-D-16-0421.1.
Zhang, J. A., 2010: Estimation of dissipative heating using low-level in situ aircraft observations in the hurricane boundary layer. J. Atmos. Sci., 67, 1853–1862, https://doi.org/10.1175/2010JAS3397.1.
Zhang, J. A., and W. M. Drennan, 2012: An observational study of vertical eddy diffusivity in the hurricane boundary layer. J. Atmos. Sci., 69, 3223–3236, https://doi.org/10.1175/JAS-D-11-0348.1.
Zhang, J. A., W. M. Drennan, P. G. Black, and J. R. French, 2009: Turbulence structure of the hurricane boundary layer between the outer rainbands. J. Atmos. Sci., 66, 2455–2467, https://doi.org/10.1175/2009JAS2954.1.
Zhang, J. A., F. D. Marks, M. T. Montgomery, and S. Lorsolo, 2011: An estimation of turbulent characteristics in the low-level region of intense Hurricanes Allen (1980) and Hugo (1989). Mon. Wea. Rev., 139, 1447–1462, https://doi.org/10.1175/2010MWR3435.1.
Zhang, J. A., S. Gopalakrishnan, F. D. Marks, R. F. Rogers, and V. Tallapragada, 2012: A developmental framework for improving hurricane model physical parameterizations using aircraft observations. Trop. Cyclone Res. Rev., 1 (4), 419–429.
Zhang, J. A., R. F. Rogers, P. D. Reasor, E. W. Uhlhorn, and F. D. Marks, 2013: Asymmetric hurricane boundary layer structure from dropsonde composites in relation to the environmental vertical wind shear. Mon. Wea. Rev., 141, 3968–3984, https://doi.org/10.1175/MWR-D-12-00335.1.
Zhang, J. A., R. F. Rogers, and V. Tallapragada, 2017: Impact of parameterized boundary layer structure on tropical cyclone rapid intensification forecasts in HWRF. Mon. Wea. Rev., 145, 1413–1426, https://doi.org/10.1175/MWR-D-16-0129.1.