• Andreas, E. L., 1995: The temperature of evaporating sea spray droplets. J. Atmos. Sci., 52 , 852862.

  • Atlas, D., , K. R. Hardy, , R. Wexler, , and R. J. Boucher, 1963: On the origin of hurricane spiral bands. Geofis. Int., 3 , 123132.

  • Barnes, G. M., , and M. D. Powell, 1995: Evolution of the inflow boundary layer of Hurricane Gilbert (1988). Mon. Wea. Rev., 123 , 23482368.

    • Search Google Scholar
    • Export Citation
  • Barnes, G. M., , and P. B. Bogner, 2001: Comments on “Surface observations in the hurricane environment.”. Mon. Wea. Rev., 129 , 12671269.

    • Search Google Scholar
    • Export Citation
  • Barnes, G. M., , G. D. Emmitt, , B. Brummer, , M. A. LeMone, , and S. Nicholls, 1980: The structure of a fair weather boundary layer based on the results of several measurement strategies. Mon. Wea. Rev., 108 , 349364.

    • Search Google Scholar
    • Export Citation
  • Barnes, G. M., , E. J. Zipser, , D. Jorgensen, , and F. D. Marks, 1983: Mesoscale and convective structure of a hurricane rainband. J. Atmos. Sci., 40 , 21252137.

    • Search Google Scholar
    • Export Citation
  • Bender, M. A., , R. E. Tuleya, , and Y. Kurihara, 1985: A numerical study of the effect of a mountain range on a landfalling tropical cyclone. Mon. Wea. Rev., 113 , 567582.

    • Search Google Scholar
    • Export Citation
  • Bender, M. A., , R. E. Tuleya, , and Y. Kurihara, 1987: A numerical study of the effect of island terrain on tropical cyclones. Mon. Wea. Rev., 115 , 130155.

    • Search Google Scholar
    • Export Citation
  • Black, P. G., , and G. J. Holland, 1995: The boundary layer of Tropical Cyclone Kerry (1979). Mon. Wea. Rev., 123 , 20072028.

  • Blackwell, K. G., 2000: The evolution of Hurricane Danny (1997) at landfall: Doppler-observed eyewall replacement, vortex contraction/intensification, and low-level wind maxima. Mon. Wea. Rev., 128 , 40024016.

    • Search Google Scholar
    • Export Citation
  • Bogner, P. B., , G. M. Barnes, , and J. L. Franklin, 2000: Conditional instability and shear for six hurricanes over the Atlantic Ocean. Wea. Forecasting, 15 , 192207.

    • Search Google Scholar
    • Export Citation
  • Bolton, D., 1980: The computation of equivalent potential temperature. Mon. Wea. Rev., 108 , 10461053.

  • Brand, S., , and J. W. Blelloch, 1973: Changes in the characteristics of typhoons crossing the Philippines. J. Appl. Meteor., 12 , 104109.

    • Search Google Scholar
    • Export Citation
  • Brand, S., , and J. W. Blelloch, 1974: Changes in the characteristics of typhoons crossing the island of Taiwan. Mon. Wea. Rev., 102 , 708713.

    • Search Google Scholar
    • Export Citation
  • Chan, J. C. L., , and X. Liang, 2003: Convective asymmetries associated with tropical cyclone landfall. Part I: f-plane simulations. J. Atmos. Sci., 60 , 15601576.

    • Search Google Scholar
    • Export Citation
  • Chen, Y., , and M. K. Yau, 2003: Asymmetric structures in a simulated landfalling hurricane. J. Atmos. Sci., 60 , 22942312.

  • Cione, J. J., , P. G. Black, , and S. H. Houston, 2000: Surface observations in the hurricane environment. Mon. Wea. Rev., 128 , 15501561.

    • Search Google Scholar
    • Export Citation
  • Daida, S. K., , and G. M. Barnes, 2003: Hurricane Paine (1986) grazes the high terrain of the Baja California peninsula. Wea. Forecasting, 18 , 981990.

    • Search Google Scholar
    • Export Citation
  • Eastin, M. D., 2003: Buoyancy of convective vertical motions in the inner core of intense hurricanes. Ph.D. dissertation, Dept. of Atmospheric Science, Colorado State University, 152 pp.

  • Emanuel, K. A., 1986: An air-sea interaction theory for tropical cyclones. Part I: Steady-state maintenance. J. Atmos. Sci., 43 , 585605.

    • Search Google Scholar
    • Export Citation
  • Hanley, D., , J. Molinari, , and D. Keyser, 2001: A composite study of the interactions between tropical cyclones and upper-tropospheric troughs. Mon. Wea. Rev., 129 , 25702584.

    • Search Google Scholar
    • Export Citation
  • Hawkins, H. F., , and S. M. Imbembo, 1976: The structure of a small, intense hurricane, Inez (1966). Mon. Wea. Rev., 104 , 418442.

  • Hock, T. F., , and J. L. Franklin, 1999: The NCAR GPS dropwindsonde. Bull. Amer. Meteor. Soc., 80 , 407420.

  • Howard, J. R., , A. L. Doggett, , R. E. Peterson, , P. G. Black, , J. L. Schroeder, , D. A. Smith, , and J. P. Dunyak, 2004: Transition in onshore hurricane boundary layer winds during the landfall of Hurricane Lili (2002). Preprints, 26th Conf. on Hurricanes and Tropical Meteorology, Miami, FL, Amer. Meteor. Soc., 526–527.

  • Jordan, C. L., 1958: Mean soundings for the West Indies area. J. Meteor., 15 , 9197.

  • Jorgensen, D. P., 1984: Mesoscale and convective-scale characteristics of mature hurricanes. Part II: Inner core structure of Hurricane Allen (1980). J. Atmos. Sci., 41 , 12871311.

    • Search Google Scholar
    • Export Citation
  • Jorgensen, D. P., , and P. T. Willis, 1982: A ZR relationship for hurricanes. J. Appl. Meteor., 21 , 356366.

  • Kepert, J. D., 2002: The wind-field structure of the tropical cyclone boundary layer. Ph.D. thesis, Dept. of Mathematics and Statistics, Monash University, 350 pp. [Available from Dept. of Mathematics and Statistics, Monash University, Vic 3800, Australia.].

  • Knupp, K. R., , J. Walters, , and M. Biggerstaff, 2004a: Doppler profiler and radar observations of boundary layer variability during the landfall of TS Gabreille. Preprints, 26th Conf. on Hurricanes and Tropical Meteorology, Miami, FL, Amer. Meteor. Soc., 568–569.

  • Knupp, K. R., , J. Walters, , and M. Biggerstaff, 2004b: Mesoscale and precipitation processes observed by Doppler profiler and radar during the landfall of Tropical Storm Gabrielle. Preprints, 26th Conf. on Hurricanes and Tropical Meteorology, Miami, FL, Amer. Meteor. Soc., 570–571.

  • Marks Jr., F. D., 1985: Evolution of the structure of precipitation in Hurricane Allen (1980). Mon. Wea. Rev., 113 , 909930.

  • Marks Jr., F. D., , R. A. Houze, , and J. F. Gamache, 1992: Dual-aircraft investigation of the inner core of Hurricane Norbert. Part I: Kinematic structure. J. Atmos. Sci., 49 , 919942.

    • Search Google Scholar
    • Export Citation
  • Miller, B. I., 1964: A study of the filling of Hurricane Donna (1960) over land. Mon. Wea. Rev., 92 , 389406.

  • Molinari, J., , P. K. Moore, , and V. P. Idone, 1999: Convective structure of hurricanes as revealed by lightning locations. Mon. Wea. Rev., 127 , 520534.

    • Search Google Scholar
    • Export Citation
  • Powell, M. D., 1982: The transition of the Hurricane Frederic boundary layer wind field of the open Gulf of Mexico to landfall. Mon. Wea. Rev., 110 , 19121932.

    • Search Google Scholar
    • Export Citation
  • Powell, M. D., 1987: Changes in the low-level kinematic and thermodynamic structure of Hurricane Alicia (1983) at landfall. Mon. Wea. Rev., 115 , 7599.

    • Search Google Scholar
    • Export Citation
  • Powell, M. D., 1990: Boundary layer structure and dynamics in outer hurricane rainbands. Mon. Wea. Rev., 118 , 891917.

  • Powell, M. D., , and S. H. Houston, 1996: Hurricane Andrew’s landfall in south Florida. Part II: Surface wind fields and potential real-time applications. Wea. Forecasting, 11 , 329349.

    • Search Google Scholar
    • Export Citation
  • Powell, M. D., , P. P. Dodge, , and M. L. Black, 1991: The landfall of Hurricane Hugo in the Carolinas: Surface wind distribution. Wea. Forecasting, 6 , 379399.

    • Search Google Scholar
    • Export Citation
  • Raman, S., 1979: Atmospheric turbulence and storm surge due to Hurricane Belle (1976). Mon. Wea. Rev., 107 , 314321.

  • Reasor, P. D., , M. T. Montgomery, , F. D. Marks Jr., , and J. F. Gamache, 2000: Low-wavenumber structure and evolution of the hurricane inner core observed by airborne dual-Doppler radar. Mon. Wea. Rev., 128 , 16531680.

    • Search Google Scholar
    • Export Citation
  • Rodgers, R., , S. Chen, , J. Tenerelli, , and H. Willoughby, 2003: A numerical study of the impact of vertical shear on the distribution of rainfall in Hurricane Bonnie (1998). Mon. Wea. Rev., 131 , 15771599.

    • Search Google Scholar
    • Export Citation
  • Rotunno, R., , and K. A. Emanuel, 1987: An air-sea interaction theory for tropical cyclones. Part II: Evolutionary study using a non-hydrostatic axisymmetric numerical model. J. Atmos. Sci., 44 , 542561.

    • Search Google Scholar
    • Export Citation
  • Shapiro, L. J., 1983: The asymmetric boundary layer flow under a translating hurricane. J. Atmos. Sci., 40 , 19841998.

  • Shea, D. J., , and W. M. Gray, 1973: The hurricane’s inner core region. Part I: Symmetric and asymmetric structure. J. Atmos. Sci., 30 , 15441564.

    • Search Google Scholar
    • Export Citation
  • Spratt, S. M., , F. D. Marks, , P. P. Dodge, , and D. W. Sharp, 2000: Examining the pre-landfall environment of mesovortices within a Hurricane Bonnie (1998) outer rainband. Preprints, 24th Conf. on Hurricanes and Tropical Meteorology, Ft. Lauderdale, FL, Amer. Meteor. Soc., 300–301.

  • Tuleya, R. E., 1994: Tropical storm development and decay: Sensitivity to surface boundary conditions. Mon. Wea. Rev., 122 , 291304.

  • Tuleya, R. E., , and Y. Kurihara, 1978: Numerical simulation of the landfall of tropical cyclones. J. Atmos. Sci., 35 , 242257.

  • Willoughby, H. E., , and M. B. Chelmow, 1982: Objective determination of hurricane tracks from aircraft observations. Mon. Wea. Rev., 110 , 12981305.

    • Search Google Scholar
    • Export Citation
  • Willoughby, H. E., , F. D. Marks Jr., , and R. J. Feinberg, 1984: Stationary and moving convective bands in hurricanes. J. Atmos. Sci., 41 , 31893211.

    • Search Google Scholar
    • Export Citation
  • Wroe, D. R., , and G. M. Barnes, 2003: Inflow layer energetics of Hurricane Bonnie (1998) near landfall. Mon. Wea. Rev., 131 , 16001612.

    • Search Google Scholar
    • Export Citation
  • Wu, C-C., , and Y. H. Kuo, 1999: Typhoons affecting Taiwan: Current understanding and future challenges. Bull. Amer. Meteor. Soc., 80 , 6780.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    (a) Earth-relative location of the 85 GPS dropwindsondes (diamonds) deployed from 1220 to 2300 UTC on 26 Aug 1998. The hurricane track is noted by the bold line and the WSR-88D locations (Wilmington and Morehead City) are denoted by stars. (b) The locations of the 64 sondes that provided data near the sea surface, plotted relative to the circulation center marked by a hurricane symbol.

  • View in gallery

    Single-sweep radar images (440 km × 440 km) from the WSR-88Ds at Wilmington (KLTX) at (a) 1300:06 and (b) 15:00:02 UTC and Morehead City (KMHX) at (c) 1659:55 and (d) 1957:07 UTC on 26 Aug 1998. Reflectivity values (dBZ) are defined in the color tables for each panel.

  • View in gallery

    Bonnie’s eyewall radius (km), as indicated by highest reflectivity features from 1000 to 2100 UTC on 26 Aug 1998. Measurements come from KLTX, KMHX, and the 42 RF lower fuselage radar (LF).

  • View in gallery

    Difference in reflectivity (dBZ) as a function of radial distance, northeast half of the transect shown in Fig. 2a minus the southwest half of the transect. The zero point is the circulation center.

  • View in gallery

    Storm-relative composite of Bonnie’s temperature (°C) at 10 m. Contours are every 1°C (bold dashed lines) except that the 25.5°C isotherm (dotted line) is denoted for the cool air found in the southwest eyewall region. The hurricane symbol denotes the storm center and the ring is the approximate eyewall location at 1730 UTC, the central time of this composite. The black arrows represent the wind speed and direction at 10 m. The scale is indicated to the lower-right side of the image.

  • View in gallery

    Storm-relative composite of Bonnie’s temperature (°C) at 2 km. Contours are every 1°C (bold dashed lines). The hurricane symbol denotes the storm center and the ring is the approximate eyewall location at 1730 UTC.

  • View in gallery

    Same as in Fig. 5, but for specific humidity (g kg−1) with 1 g kg−1 contour intervals (bold dashed lines). The 21.5 g kg−1 contour around the circulation center is depicted by a dotted line.

  • View in gallery

    Same as in Fig. 6, but for specific humidity (g kg−1) with 1 g kg−1 contour intervals (bold dashed lines).

  • View in gallery

    The LCL (m) with a contour interval of 200 m (bold dashed lines). Other figure aspects follow Fig. 6.

  • View in gallery

    The LI at 800 hPa (°C) with a contour interval of 0.5°C (bold dashed lines). Other figure aspects follow Fig. 6.

  • View in gallery

    Same as in Fig. 5, but for equivalent potentialtemperature (K) with contour intervals of 4 K (bold dashed lines).

  • View in gallery

    Equivalent potential temperature (K) at 2 km depicted with a contour interval of 4 K (bold dashed lines). Other figure aspects follow Fig. 6.

  • View in gallery

    Storm-relative tangential winds (m s−1) at 10 m with a contour interval of 5 m s−1 (bold dashed lines). Winds less than 15 m s−1 are tightly packed around the circulation center and are not drawn for clarity. Other figure aspects follow Fig. 6.

  • View in gallery

    Same as in Fig. 13, but at 2 km.

  • View in gallery

    Storm-relative radial winds (m s−1) at 10 m with a contour interval of 5 m s−1 (bold dashed lines). Negative values are inflow. Other figure aspects follow Fig. 6.

  • View in gallery

    Same as in Fig. 15, but at 2 km.

  • View in gallery

    Storm-relative inflow depth (km) with contour intervals of every 0.5 km (bold dashed lines). Shaded region depicts area of outflow.

  • View in gallery

    Surface analysis for 1200 UTC 26 Aug 1998. Thin black lines are isobars, thicker blue lines depict cold fronts, red and blue lines are stationary fronts, surface lows positions and central pressure are labeled in red, and selected stations show temperature (°F), dewpoint temperature (°F), pressure (mb), and winds following standard meteorological convention. (More information available online at http://weather.unisys.com)

  • View in gallery

    Geopotential heights (m) at 500 hPa at 1800 UTC 26 Aug 1998. Contour intervals are every 10 m (solid lines).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 13 13 1
PDF Downloads 6 6 2

Low-Level Kinematic, Thermodynamic, and Reflectivity Fields Associated with Hurricane Bonnie (1998) at Landfall

View More View Less
  • 1 University of Hawaii at Manoa, Honolulu, Hawaii
© Get Permissions
Full access

Abstract

During 11 h on 26 August 1998, two NOAA WP-3D aircraft deployed 85 Global Positioning System (GPS) dropwindsondes within 2° of latitude of the circulation center of Hurricane Bonnie as it made landfall in North Carolina. About 75% of the sondes successfully collected data, which were used to create a series of storm-relative horizontal maps of kinematic and thermodynamic variables from 10 m to 2 km. Reflectivity fields were analyzed from the Weather Surveillance Radar-1988 Dopplers (WSR-88Ds) located at Wilmington and Morehead City, North Carolina, and the tail and lower fuselage radars aboard the WP-3Ds.

GPS sonde performance and deployment spacing is adequate to identify several aspects of the vortex. These include 1) warm, dry, stable air in the offshore flow that results in reduced equivalent potential temperatures entering the southern portion of the eyewall, 2) cooler air collocated with the upwelled water in the right-rear quadrant and under the eyewall, and 3) an atypical radial wind pattern with strong inflow southwest of the circulation center and outflow northeast of the center. The strongly asymmetric structure found at 10 m becomes much more homogeneous by 2-km altitude.

No intense rainbands developed over land in the onshore flow nor did the bands in the onshore flow undergo any significant changes once they made landfall. Beyond the eyewall the offshore flow contained much less precipitation than the onshore portion of the storm.

Characteristics beyond the eyewall appear to have been modulated by the proximity to land but hurricane intensity did not vary. The authors infer that the lower energy content of the inflow was offset by the contraction of the eyewall.

Corresponding author address: G. M. Barnes, Dept. of Meteorology, University of Hawaii at Manoa, 2525 Correa Rd., Honolulu, HI 96822. Email: gbarnes@hawaii.edu

Abstract

During 11 h on 26 August 1998, two NOAA WP-3D aircraft deployed 85 Global Positioning System (GPS) dropwindsondes within 2° of latitude of the circulation center of Hurricane Bonnie as it made landfall in North Carolina. About 75% of the sondes successfully collected data, which were used to create a series of storm-relative horizontal maps of kinematic and thermodynamic variables from 10 m to 2 km. Reflectivity fields were analyzed from the Weather Surveillance Radar-1988 Dopplers (WSR-88Ds) located at Wilmington and Morehead City, North Carolina, and the tail and lower fuselage radars aboard the WP-3Ds.

GPS sonde performance and deployment spacing is adequate to identify several aspects of the vortex. These include 1) warm, dry, stable air in the offshore flow that results in reduced equivalent potential temperatures entering the southern portion of the eyewall, 2) cooler air collocated with the upwelled water in the right-rear quadrant and under the eyewall, and 3) an atypical radial wind pattern with strong inflow southwest of the circulation center and outflow northeast of the center. The strongly asymmetric structure found at 10 m becomes much more homogeneous by 2-km altitude.

No intense rainbands developed over land in the onshore flow nor did the bands in the onshore flow undergo any significant changes once they made landfall. Beyond the eyewall the offshore flow contained much less precipitation than the onshore portion of the storm.

Characteristics beyond the eyewall appear to have been modulated by the proximity to land but hurricane intensity did not vary. The authors infer that the lower energy content of the inflow was offset by the contraction of the eyewall.

Corresponding author address: G. M. Barnes, Dept. of Meteorology, University of Hawaii at Manoa, 2525 Correa Rd., Honolulu, HI 96822. Email: gbarnes@hawaii.edu

1. Introduction

As a tropical cyclone (TC) nears landfall an increasing portion of its circulation will pass over land. In this growing sector, there will be changes at the surface that include a reduction of latent and sensible heat flux, and enhanced roughness. These departures from ocean conditions may substantially modify the boundary layer that, in turn, could suppress convection in some sectors and enhance convection in others. Ultimately, TC traits such as intensity, strength, and the precipitation field could be affected prior to the eye making landfall.

Instruments near the coast are frequently disabled by the high winds, heavy rains, and storm surge of a landfalling TC. Research aircraft are currently restricted from flying over land in a TC and do not fly below 500 m in high-category TCs. Offsetting this data shortfall are the Global Positioning System (GPS) dropwindsondes, developed by the National Center for Atmospheric Research (NCAR), the National Oceanic and Atmospheric Administration (NOAA), and the German Aerospace Research Establishment. These sondes collect data at 2 Hz and provide 6–7-m vertical resolution down to the ocean surface. At present, the GPS sondes may not be deployed over land, but they can be used to sample the onshore and offshore flow over the adjacent coastal seas.

We use the GPS sondes to examine the vortex-scale thermodynamic and kinematic structure of Hurricane Bonnie on 26 August 1998, just prior to its landfall in North Carolina. The deployment of the sondes provides us with the opportunity to produce horizontal maps that can reveal the extent of the asymmetry in the TC, from the circulation center to about 2° of latitude radius, and observe if the nearness of land has any impact on the structure of Bonnie. A change of TC track due to the proximity of land, though important, is not our focus.

a. TC structure changes when near land

Both observation (Miller 1964) and simulation studies (Tuleya 1994) have implicated the loss of evaporation as the primary mechanism that causes a TC to decay over land. We expect that air originating from land, or air that was originally part of the vortex but has spent several hours over land, will be drier by the time it flows offshore. Drier air may reduce cloud and rain in the offshore portion of the TC. Simulations of landfalling TCs exhibit a modification of the stability and developing asymmetries in the rain field (Chan and Liang 2003). If this drier air reaches the eyewall, its reduced energy content may ultimately reduce storm intensity.

Numerous studies have identified a discontinuity in the wind at the coast due to increased roughness. Raman (1979) and Howard et al. (2004) used surface stations to identify an inverse relationship between wind speed and distance from the coast. Powell et al. (1991) and Powell and Houston (1996) reduced data largely from flight level to 10-m altitude to infer strong reductions in wind speed at the coast and a resulting increase in convergence. This reduction in the near-surface winds leads to an increase in the vertical shear of the horizontal wind in the lowest 500 m, enhanced convergence to the right of track, and divergence to the left of track as observed by Powell (1982) for Hurricane Frederic (1979). The simulation of a landfalling TC by Tuleya and Kurihara (1978) showed a strengthening of the radial winds and a reduction of the tangential winds through nearly a kilometer depth.

In at least some situations, these effects alter the flow field well offshore. Kepert (2002) analyzed 30 dropwindsondes released in Hurricane Mitch (1998) when the circulation center was 85 km from the Honduran coast. Mitch displayed large asymmetries in the near-eyewall region. The strongest inflow existed in the left-rear quadrant in the lowest 1.2 km, in contrast to the typical situation (Shea and Gray 1973; Shapiro 1983). The strongest tangential winds at any level were found one quadrant downstream of the strongest inflow. Kepert (2002) concluded that these asymmetries were largely due to the proximity of land. He verified this with a model run whereby the inclusion of land with its increased friction led to the wind asymmetries similar to those observed in Mitch.

Advection of continental air into the TC circulation is another important process at landfall (Wu and Kuo 1999). Knupp et al. (2004a, b) showed that stratiform rain over land produced a mesoscale region of cool air ahead of landfalling Gabrielle (2001). This cool air, maintained over land because of the much-reduced sensible heat flux, which acted as a cold front confined to the boundary layer as it was entrained into the TC circulation. The cool air was partly responsible for the reduction in cloud tops in the core of the TC.

A particularly strong land influence is possible when there are mountains present. Intensities for simulated TCs diminish chiefly in response to the drier air that subsides and enters the inner circulation (e.g., Bender et al. 1985, 1987). Occasionally an actual TC does react in a similar fashion, though the magnitude of the change is usually less in simulated TCs (Brand and Blelloch 1973, 1974).

The observed versus the probable response to the proximity of land can differ simply because of variations in the environmental flow. Powell (1987) found asymmetries in the surface wind field of Hurricane Alicia (1983) during landfall in the Galveston area. Alicia exhibited characteristics that differed from the expected with an inflow maximum, surface convergence and a rainfall maximum in the offshore flow. The location of the inflow maximum was partially explained by the southwesterly environmental flow, which paralleled the coastline. Studies of Hurricane Danny (1997) by Blackwell (2000) revealed precipitation and wind maxima in the offshore flow much like that seen in Alicia (1983). Daida and Barnes (2003) searched for a response in Hurricane Paine as it passed just south of the Baja Peninsula. The Baja ridge had less impact than expected partially because of the orientation of the environmental flow relative to the ridge, which limited the amount of subsiding air entering the TC inner core.

The proximity of a large landmass generally increases the likelihood that the TC will develop more asymmetric kinematic and thermodynamic fields (Wu and Kuo 1999; Brand and Blelloch 1974; Tuleya 1994). Mountainous islands (Hispaniola, Taiwan) have their chief impact via the development of up- and downslope flows (Bender et al. 1987). The diurnal cycle may further accentuate the differences between air of oceanic and continental origins.

b. Goals

The GPS sondes deployed over 11 h will be used to produce vortex-scale maps of kinematic and thermodynamic variables from 10-m to 2-km altitude. These maps will provide an estimate of the asymmetries in the inflow layer, produced in part by the nearness of land, which may impact intensity, strength, and the precipitation field of Bonnie. The Weather Surveillance Radar-1988 Doppler (WSR-88D) coastal radars at Wilmington and Morehead City, North Carolina, and the lower fuselage radar of the NOAA WP-3Ds will be used to determine the reflectivity patterns. The radars, besides revealing the steady structures during the analysis period, will also reveal the evolution of the precipitation field during landfall.

This study marks the first time that multiple levels of thermodynamic and kinematic variables of a TC at landfall with the GPS sonde have been produced. We view the development of these maps as a test of sonde quality. The maps that are presented here will be used in future papers to examine a variety of issues including energy and mass flow to the eyewall, and the role of spray in the energizing of the inflow.

2. Experimental design

a. GPS dropwindsonde

The GPS dropwindsonde has several advantages over the previous generation Omega dropwindsonde (ODW). The GPS sonde provides winds down to the surface while the ODW generally did not collect wind data below 400 m (Hock and Franklin 1999). The GPS sonde provides measurements every 7 m while the ODW was best interpreted with a vertical resolution of 150 m. Wind errors for the GPS sonde are 0.5 m s−1. Typical errors for pressure and temperature measurements are 1.0 hPa and 0.2°C, respectively (Hock and Franklin 1999). The GPS sonde tends to dry out upon exiting cloud much more readily than the ODW (Bogner et al. 2000) and errors for relative humidity average <5%.

b. Dropwindsonde quality control

The Atmospheric Sounding Processing Environment (ASPEN) program was used to process the dropwindsonde data. This software, developed at NCAR, accepts raw data from the dropwindsonde in the form of Airborne Vertical Atmosphere Profiling System (AVAPS) files. ASPEN was evaluated next to Editsonde, the Hurricane Research Division (HRD) dropwindsonde processor. The quality-controlled data from ASPEN were similar to those from Editsonde.

Like Editsonde, ASPEN also produces some questionable postprocessed data. One such error can be clearly seen in the lower levels of a skew T–log p diagram where the air is saturated yet the lapse rate is dry adiabatic. We have corrected this error, which is likely due to the humidity sensor remaining wet after passing through clouds following the scheme adopted by Bogner et al. (2000). Data has been accepted for analysis only after it has shown to be consistent in both the vertical for a particular drop, and in the horizontal when compared to nearby drops.

c. Flight plans

The spatial distribution of the GPS sondes is the result of two different experiments. The first aircraft (42RF) released sondes in four curved paths that mimic inflow trajectories to the eyewall (Wroe and Barnes 2003). Forty-five of the 54 dropwindsondes released from 1220 to 1759 UTC provided either thermodynamic, kinematic, or both types of data for a portion of the descent. The flight pattern for 43RF was designed to capture the boundary layer wind structure for both the onshore and offshore flow, and therefore consisted of several legs roughly parallel to the shoreline. Forty dropwindsondes released from 43RF from 1628 to 2259 UTC yielded some form of data during the descent. Note that NOAA was restricted from dropping sondes over land. Figure 1a displays the total 85 dropwindsondes used in this study as well as the hurricane’s track, which was established from flight level data of nine aircraft fixes from 0050 UTC on 26 August to 0325 UTC the following day.

d. Creation of the fields

During the 11 h of dropwindsonde deployment, Bonnie’s MSLP did not vary more than 3 hPa, averaging 964 hPa. We will assume that the hurricane is in a steady state and will composite the sondes with respect to storm center at 1730 UTC, which is the temporal midpoint for the sonde releases. At this time, Bonnie was centered at 33.26°N, 77.90°W, approximately 70 km from Cape Fear.

Nine center fixes obtained by the two WP-3Ds are used to estimate storm speed and direction. The technique developed by Willoughby and Chelmow (1982) is used to determine each circulation center position or fix. During the experiment, Bonnie curved to the north-northeast at about 4.4 m s−1 (Fig. 1).

Storm motion is subtracted from the earth-relative winds to obtain storm-relative radial and tangential winds. Positive radial wind (υr) is away from storm center (outflow) and positive tangential wind (υt) is cyclonic. The thermodynamic variables are computed following the recommendations by Bolton (1980).

The horizontal maps at 10, 50, 100, 200, 500, 1000, 1500, and 2000 m for each of the variables were initially drawn separately by each of the authors. These maps were compared and the reasons for a differing interpretation were discussed. We expected to see some strongly contrasting fields but were pleasantly surprised to see that our fields were similar. We did eliminate the rare bull’s-eye contours due to the data from a single sonde. Sonde data were reviewed based on where the sonde fell with respect to reflectivity features as revealed by the lower fuselage radar of the aircraft. We wanted to see if the data from a particular sonde could have been affected by strong convective-scale processes and thus might be less representative of the vortex scale. Field contours underwent another minor adjustment around the eyewall where we assumed the gradients would be concentrated. Typical spacing between GPS drops is 10–20 km near the eyewall and 20–40 km by 200-km radial distance from the circulation center. Features such as the eye and eyewall, and major rainbands are identified, but the gradients associated with these features must be interpreted more cautiously.

Of the 85 drops, approximately 88% provide temperature and pressure data to splash point, and about 84% provide relative humidity to that level. Losses are much greater for the wind data with only 64% of the sondes providing winds to splash point. A total of 64 sondes provided either thermodynamic, kinematic, or complete data in the surface layer. These are shown in their storm-relative positions in Fig. 1b. In the composite figures, the coastline is omitted to avoid the impression of analyzing the fields over land. For reference, the coast is oriented southwest to northeast.

e. Land-based and airborne radars

Two WSR-88Ds, at Wilmington (KLTX) and Morehead City (KMHX; Fig. 1a) collected data as Bonnie neared landfall. These radars have a wavelength of 11.1 cm (S band) and a beamwidth of 0.93°. The WSR-88Ds provided a detailed history of the reflectivity field of Bonnie as she made landfall from 1115 to 1957 UTC, until power loss at each radar site. Temporal resolution was approximately 6 min between baseline (0.5° elevation) scans. The data were interpreted quantitatively to a 150–175-km range, and qualitatively beyond that range due to the usual radar problems of beam spreading, increasing beam elevation, and beam attenuation. For much of the study the circulation center was within 75 km of Wilmington and 175 km of Morehead City.

The NOAA WP-3Ds are equipped with both lower fuselage and tail radars. Individual scans have been used to determine if each GPS drop was made in a particular feature such as the eye, eyewall, or rainband. These radars provided reliable quantitative estimates within about 30 and 70 km for the tail and lower fuselage radars, respectively (Marks 1985). Beyond those ranges, interpretation is best kept to qualitative assessments.

3. Results

a. Reflectivity

1) Eyewall reflectivity patterns

Observations prior to landfall have shown that Bonnie was a concentric eyewall system (Rodgers et al. 2003). The WSR-88Ds (Figs. 2a–d) reveal an outer eyewall with a radius of approximately 80 km. We use the sharp horizontal gradient of reflectivity and the aircraft in situ winds to define the outer eyewall. Within the broad eye, there are cells at 30-km radius that reach as high as 16 km, though most only reach 8–10-km altitude. These cells are the remains of a decaying inner eyewall that persist through 1957 UTC. A group of cells in the inner eyewall undergoes continual regeneration and can be tracked for the 7 h that the WSR-88D data are available. These cells maintain their distance of 25–30 km to the circulation center.

Analysis of the outer eyewall, which is the primary reflectivity feature of Bonnie, cannot rely on the entire period of observation from either radar because of range limitations, and to a lesser extent, because of attenuation. Beyond 220 km the center of the 0.5° elevation beam is near the freezing level and would miss the heavy precipitation cores and suffer contamination from the bright band. The far side or southern portion of the eyewall is within this range for the Wilmington radar after about 1400 UTC, and by 1630 UTC for Morehead City. Interpretation of rain-rate patterns on the leading edge or northern side of the eyewall can be done for either radar when the data are first available (∼1115 UTC), but for the far side of the eyewall we use Wilmington from 1400 to 1610 UTC and then use the Morehead City radar afterward. Comparison of the two radars for the northern portion of the eyewall reveals similar interpretations, leading us to believe that the calibration differences between the two are less than a few dBZ. Because of attenuation caused by strong intervening rainbands north of the eyewall, and because of heavy rain occasionally falling at the radar sites, we will discuss only the primary variations we observe around the eyewall.

Initially reflectivities are high in the northwest quadrant of the eyewall. Over 90% of the scans from the Wilmington radar from 1120 to 1500 UTC contain reflectivities of 43 dBZ or higher in this quadrant. After 1430 UTC radar returns start to increase in the southern quadrant of the eyewall, and by 1530 UTC this portion of the eyewall contains the highest reflectivity in the eyewall. Over 80% of the scans show high reflectivities in the south and southeast sectors between 1430 and 1957 UTC.

Weaker reflectivities persist in the northeast quadrant. Examination of the reflectivity found in the eyewall from 1120 to 1957 UTC reveals that over 80% of the scans have the lowest reflectivities in this quadrant. This persistent region of lighter rain will be of interest when we discuss the radial wind component field.

Estimates of rain rate are possible by employing the ZR relationship derived for hurricanes (Jorgensen and Willis 1982). The more intense portions, initially in the northwest and later in the south and southeast quadrants, typically contain rain rates that exceed 20 mm h−1 while the weaker sectors have rain rates closer to 4 mm h−1 (northeast quadrant of the eyewall).

Examination of the WSR-88Ds reveals that the outer edge of the main eyewall crosses Cape Fear at about 1400 UTC. About 5% of the eyewall is ashore by 1500 UTC, by 1700 UTC nearly 20% is over land, and by ∼2000 UTC almost 33% is over North Carolina. The intensity of the eyewall over land changes somewhat slowly, perhaps partly due to the strong advection of hydrometeors. The more intense portion of the eyewall remains in the south and southeast sectors, which are still over water. By 1800 UTC convective cells that were associated with the inner eyewall remnant are the strongest feature apparent in the northwest quadrant; these cells are over the coast.

2) Evolution of the eyewall radius

As Bonnie approached land, the outer eyewall radius steadily decreased (Fig. 3). Statistically, the rate is well approximated by a linear fit (R2 = 0.87). Besides the two WSR-88Ds, the lower fuselage radar from 42RF also provides data from approximately 1820 to 2100 UTC. All three radars reveal a similar trend, as the eyewall decreases from approximately 92 km at 1115 UTC to 64 km by 2037 UTC.

3) Rainbands

Four convective bands, two of which that are more than 200 km in length, could be identified from 1120 to 1957 UTC (Fig. 2). These bands tended to form beyond the eyewall on the east and northeast side, rotate cyclonically, and blanket the leading edge side (north) of the vortex. The bands also moved radially outward. Newer bands formed between the eyewall and the older band, and generally within 50 km of the eyewall. These bands are similar to those described by Atlas et al. (1963), Barnes et al. (1983), and Powell (1990) where the upwind side is convective and the downwind side is more stratiform. A couple of the cells in these bands produced mesovortices that may have spawned tornadoes (Spratt et al. 2000).

One band was monitored from 1400 to 1800 UTC with the Morehead City radar to estimate how the area of the 39 and 43 dBZ returns varied as the band comes ashore. The total area of the offshore and onshore reflectivity for either threshold changed only slightly (39 dBZ increased by 10%, 43 dBZ decreased by 15%) as the band shifted from almost entirely offshore to over 70% onshore during the 4 h of study. The results lead us to believe that the band came ashore without much enhancement or suppression.

There is probably little difference in the thermodynamics on either side of the coast since the air has been coming ashore for some time. One difference is the increased roughness over land that may cause a decrease in the winds, and perhaps enhance convergence. The behaviors of the cells suggest that this alteration of the flow is not deep enough to have an effect on the reflectivity field. Modeling studies (e.g., Chen and Yau 2003) that have cells blossom once they come ashore should closely examine their initial conditions, or how the friction and turbulence that develop over land is treated in the model. At this stage there is no moisture front coming ashore that some model runs may have emphasized. The only change near the coast is the development of a line of small cells that appear inland on the west side of the TC (Fig. 2a) but fails to rival any of the bands that existed previously in the onshore flow.

As will be shown, instability to the north of the circulation center and beyond the eyewall is present. The rainbands are less dependent on surface fluxes than the eyewall, where conditions approach moist neutral.

4) Asymmetry of rain beyond the main eyewall

Both the lower fuselage composites and the ground based radars reveal that the west and southwest portion of the TC, beyond the eyewall, has much less precipitation (Fig. 2b;1500:02 UTC in KLTX). This portion of the TC is over land, or immediately offshore and downwind of the Carolinas. The southwest to northeast transect delineated on the 1300:06 UTC image (Fig. 2a) is used to highlight the difference in reflectivity between the onshore (northeast) and offshore flow (southwest). The mean of seven transects taken every hour from 1300 to 1957 UTC from both WSR-88Ds (Fig. 4) reveals that from 140 to 250 km from the storm center the offshore flow contains much less precipitation than the onshore flow. Beyond 140 km from the circulation center, reflectivity is about 25 dBZ lower to the southwest compared to the same radial distance to the northeast. Given that maximum reflectivities are typically 45 dBZ to the northeast, a reduction to 20 dBZ reduces the rain rates from more than 30 to less than 2.0 mm h−1, based on the ZR relationship for hurricanes developed by Jorgensen and Willis (1982). Closer to the eyewall radius the difference between the onshore and offshore flow falls to approximately 10 dBZ as the southwestern half of the transect begins to contain rain. Only at and within the outer eyewall are the reflectivities at a given radius higher to the southwest than to the northeast.

A history of the reflectivity fields of Bonnie for several days prior to landfall shows that there is a correlation between the deep layer vertical shear of the horizontal wind and the distribution of precipitation (Rodgers et al. 2003). However, by 26 August the large-scale vertical shear is nearly along track and quite weak, which would favor a symmetric precipitation field. The fact that the outer rain field is not symmetric suggests that other factors, shortly to be discussed, contribute to the observed precipitation differences.

b. Thermodynamic structure

1) Temperature

The storm-relative temperature field at 10 m (Fig. 5) reveals a large warm eye, with temperatures of at least 27°C. Outside of the eye there is an annulus of cooler air, especially apparent to the southwest of the circulation center, where temperatures fall below 25.5°C. This cool annulus is collocated with the eyewall and rainbands, as identified with the WSR-88Ds, and is where downdrafts and resulting outflows are likely present.

The majority of the cooling to the southwest of the circulation center is located well beyond the eyewall. This suggests that adiabatic cooling, collocated with the pressure gradient that is under and just outward of the eyewall, is actually contributing little to the reduction in temperature. The greatest cooling occurs about 1.5° from the center, which is similar to the findings based on buoy (Cione et al. 2000) and dropwindsondes (Barnes and Bogner 2001).

Based on the total wind vectors in Fig. 5 the low-level inflow is not isothermal. This cooling is especially apparent in the air streaming offshore where temperatures fall from 28° to 26°C (Fig. 5). The WSR-88Ds detect no convective or stratiform precipitation in this region; this eliminates the possibility that downdrafts are responsible for the cooling. Relative humidities in this region range from 65% to 85%. These low relative humidities coupled with wind speeds that exceed 20 m s−1 make the evaporation of sea spray a possible cooling mechanism. The nonisothermal nature of the inflow all around the storm confirms the multistorm composite fields derived by Cione et al. (2000) and Barnes and Bogner (2001).

The evaporation of sea spray is believed to play a minor role for any cooling in and under the eyewall given that the average relative humidity in this region exceeds 90%. Andreas (1995) performed model calculations of the evaporating temperature of a spray droplet as a function of its initial radius, the surface-water salinity, and the ambient temperature and relative humidity. He demonstrates that the evaporating temperature of the spray droplet is very close to the ambient temperature at high relative humidities.

Bonnie also features an atmospheric wake that is apparent in the right-rear quadrant of the storm where temperatures fall below 25°C. This region is collocated with the cool, upwelled ocean (see Wroe and Barnes 2003, their Fig. 2), and where the air–sea temperature difference becomes small or even reverses sign. The 50-m temperature plot (not shown) supports the observations at 10 m, but by 500 m the cool regions start to lose their identity, and by 1 km the cool regions have disappeared.

The offshore flow at 10 m (southwest of the storm center) is 1°C warmer than the onshore flow at 1.5° radius from the storm center. The central dense overcast inhibits daytime warming over land within 150 km of the circulation center and keeps the differences between the on- and offshore temperatures small.

A warm concentric core is the primary feature at 2 km (Fig. 6). The warmest air is constrained in a small region within the eye due to the existence of the decaying inner eyewall. Onshore flow is cooler than the offshore flow at greater radial distances. The temperature difference between offshore and onshore that had decreased to near zero by 1-km altitude now starts to manifest increasing contrast. This is evidence of the intrusion of warmer air to the southwest of the center. The coherency in the various horizontal fields and the consistency from level to level are gratifying and demonstrate that possible noise sources such as calibration drift, thermistor wetting, and cloud-scale processes are not masking the mesoscale fields.

2) Specific humidity

The specific humidity field at 10 m exhibits the expected moist core with values equal to or greater than 21.5 g kg−1 (Fig. 7). The air is noticeably drier to the southwest of the circulation center (<15 g kg−1) than to the northeast. The 10-m total wind vectors on this plot indicate that as this air moves offshore and toward the eyewall there is a rapid moistening from ∼16 to ∼20 g kg−1. There also appears to be some moistening of the onshore flow, noted by the 20 g kg−1 contour north of the circulation center. The higher specific humidity found just outside and to the west of the eyewall is associated with outflows from rainbands. The 50-m plot of specific humidity (not shown) is consistent with the 10-m plot.

There is a considerable difference in specific humidity at 10 m between the onshore and offshore flow, with the offshore flow approximately 4 g kg−1 drier at a radius of nearly 200 km and about 2 g kg−1 drier at a radius of just over 100 km from the storm center. The offshore flow continues to be drier than the onshore flow until a height of 2 km (Fig. 8) where the specific humidity field is nearly symmetric about the circulation center.

3) Lifting condensation level

The lifting condensation level (LCL) is estimated using an average of T and q in the lowest 50 m. The highest LCL (Fig. 9) is found in the offshore flow. In the onshore flow, to the northeast of the center the LCLs barely exceed 400 m. As one approaches the eyewall the LCLs fall to less than 300 m; in the eye the LCL is at or less than 200 m. A lower LCL is also associated with the rainbands to the north and northwest. Saturated ascent plays an increasing role in the vertical transport of energy within the inflow layer with decreasing radial distance. Examination of photographs taken from the WP-3Ds verifies that layers of thin cloud become more prevalent as one nears the eyewall. The low cloud base and the near-saturated conditions through much of the lower troposphere limit the potential cooling possible from downdrafts. Maximum difference between a parcel that descended moist adiabatically and a dry adiabatic lapse rate [difference of ∼0.45°C (100 m)−1] from 300 m is about 1.6°C, which is just slightly larger than the magnitude of the cooling observed under the eyewall. The vertical structure of the boundary layer (the subject of another paper) departs from the conditions normally seen in the fair weather tropical boundary layer (e.g., Barnes et al. 1980).

4) Atmospheric stability

A modified lifted index (LI) was computed to determine the stability of the low-level air. This index was calculated at 800 hPa, versus the standard 500 hPa, since dropwindsondes were released from 700 hPa or lower. Mean values of temperature and humidity in the lowest 500 m were used to calculate LI. A LI800 of −1.0°C is typical for the Caribbean during hurricane season (July–October) based on the mean soundings presented by Jordan (1958).

Based on the LI800 (Fig. 10), stability in the lower troposphere is asymmetrical with the most stable air found in the offshore flow, south-southwest of the circulation center. Radar shows this region to be devoid of rain-producing cells. As this air flows toward the eyewall it destabilizes, but it never achieves an unstable (negative) value along the southern half of the eyewall. A smaller region of stable air is also found approximately 80 km northeast of the circulation center, collocated with the northeast eyewall and radially inward of a strong convective rainband as seen in the WSR-88D radar images (Fig. 2). The stability in this region may be attributed to convective downdrafts transporting cooler air to lower levels. The portion nearer the eyewall also features outflow at 800 hPa where warm air is being drawn out over the cooler lower levels of the atmosphere. Farther to the north of the storm center, where there are rainbands, the air is more unstable than the typical Caribbean situation described by Jordan (1958). The offshore flow destabilizes with decreasing radius while the onshore flow generally stabilizes as it approaches the eyewall.

Within a degree of latitude of the center, LI800 is higher (more stable) than the typical tropical LI800 during the hurricane season. The tendency to evolve toward moist neutral stability supports the conjecture by Emanuel (1986) and simulations by Rotunno and Emanuel (1987), the lightning observations by Molinari et al. (1999), and the sounding analyses by Bogner et al. (2000).

5) Equivalent potential energy

There is a maximum of 370 K in equivalent potential temperature (θE) at 10 m (Fig. 11) located in the eye. Values decrease with increasing radius, but are not symmetric. At the eyewall the difference in θE is about 4 K between the flow that originated over land (∼358 K) and the onshore flow (∼362 K). Farther to the southwest of the storm center (just over 1.5° latitude) θE decreases to below 346 K, while to the northeast of the storm center values only decrease to a little less than 358 K. The offshore flow is increasing 10–12 K over the 100–150-km distance it covers from the coast to the mean location of the eyewall. Energy content to the north of the eyewall increases only 4–5 K over a similar distance. The convective exchanges within the rainbands in this sector work to inhibit the increase in θE in this portion of the TC. By 2 km (Fig. 12) the highest values are still seen in the storm’s core, slightly to the east of the circulation center. At this level the asymmetry seen from the southwest to northeast at lower altitudes is no longer apparent.

The highest values of θE were not found in the eyewall, but in the eye of the tropical cyclone. This agrees with the observations by Hawkins and Imbembo (1976), Willoughby et al. (1984), Jorgensen (1984), and Eastin (2003). Given that one expects to see the highest θE air ascend in the eyewall the observation is intriguing. Air may either enter the eye without ascending in the eyewall or it may be detrained from the eyewall into the eye. For Bonnie there is the additional complication of the presence of the remnants of an inner eyewall. Winds in Bonnie’s eye are sufficiently strong (>10 m s−1) to allow surface fluxes to continue supplying energy into this air, and given the appropriate residence time, θE could easily become higher. Subsidence in the eye would inhibit this air from rising and lengthen residence time near the sea.

c. Kinematic structure

1) Tangential winds

At 10 m the strongest tangential winds (Fig. 13) oscillate in position around the mean location of the outer eyewall. There is evidence of a higher wavenumber structure, but it must be remembered that the observations were not gathered simultaneously. Scrutiny of the reflectivity field (Fig. 2b) demonstrates that the eyewall shape is occasionally elliptical with the major axis in the north–south direction. The timing of the drops south of the circulation center is biased toward the period when the eyewall appeared more elliptical and thus may be the reason why the fields bow radially beyond the mean position of the eyewall. The inner eyewall, still producing at least one group of deep cells that reach 16-km altitude, alters the radial gradient of the tangential winds away from the expected solid body rotation. Winds near the circulation center do approach zero, but the spatial extent of this region is so small that the contours would be densely packed and collocated with the circulation center symbol; thus, no contours were analyzed for winds <15 m s−1.

Offshore tangential winds are about 5 m s−1 weaker than the onshore flow from about 1° to 2° latitude from the circulation center. Increased roughness over land is reducing the tangential winds, and increasing the radial flow in the offshore flow.

At 2 km (Fig. 14) the region of strong winds (≥40 m s−1) is collocated with the eyewall and also expands to the northwest. The area to the northwest contains strong rainbands during the early portion of the experiment and is when this region was sampled. From 1° to 2° from the circulation center tangential winds are 5–10 m s−1 lower to the south and southwest than they are to the north and northeast. Beyond the eyewall, the tangential wind field remains asymmetric at 2 km, in contrast to the thermodynamic fields that shift toward a more symmetric appearance at that level.

2) Radial winds

The inflow to the eyewall is highly asymmetric at 10-m height (Fig. 15). The radial wind component varies from inflow over 20 m s−1 (negative values) coincident with the outer edge of the southwest eyewall to outflow of 2–3 m s−1 along the east-northeast arc of the eyewall. The northeast portion of the eyewall was also a region with much lower reflectivity compared to the northwest or south sectors. From 50 to 500 m (not shown) the maximum inflow region rotates anticyclonically with height.

Estimates of divergence along the southwest portion of the eyewall reach −5 × 10−4 s−1 based on the two-dimensional assumption that ignores changes in the tangential wind. Along the north portion of the eyewall divergence is only ¼ this value. The maximum value estimated for Bonnie is about 25%–50% of what Jorgensen (1984) found for high-category Hurricane Allen (1980), but similar to what Marks et al. (1992) found for Hurricane Norbert (1984). These comparisons appear favorable and suggest that our fields are capturing many of the key aspects of the TC. A three-dimensional estimate will be included in future work.

By 2 km (Fig. 16) outflow dominates the storm with the highest values to the east of the storm center. The remaining inflow is weak (∼5 m s−1 or less) and located in two regions south and northwest of the circulation center.

Strong radial flow asymmetry has been related to a variety of causes. Shapiro (1983) simulated the effect of storm translation speed on radial flow in a slab boundary layer and found that the greatest inflow in slow-moving hurricanes (translation speed of 5 m s−1) was located in the right-front quadrant. Composite observations from many TCs obtained a similar result (Shea and Gray 1973). While Bonnie’s speed (∼4 m s−1) is similar to Shapiro’s slow-moving storm, the maximum inflow is found to the left rear of the storm.

The interaction of the environmental flow with the vortex may create radial wind asymmetries. Hurricane Alicia’s (1983) inflow maximum in the offshore flow (left-front quadrant) is partially explained by the southwesterly environmental flow that paralleled the coastline (Powell 1987). However, Bonnie’s low-level environmental flow, as depicted by the (National Centers for Environmental Prediction) NCEP–NCAR reanalysis, was negligible in magnitude and likely not responsible for the strong radial inflow in the offshore flow.

The organization of convection can also promote radial wind asymmetries. Powell’s (1982) analysis of Hurricane Frederic (1979) at landfall revealed maximum inflow angles in the right-front quadrant of the storm where there was the strongest convection. Bonnie’s highest reflectivities around the eyewall (Fig. 2) were generally found to the southwest of the storm center, which is coincident with the strongest inflow at 10 m. Gaps or weaknesses in the reflectivity field of the eyewall were seen in the east-northeast sector, where there was weak outflow. Other studies concerned with vortex Rossby waves (e.g., Reasor et al. 2000) also show that strong convective cells alter the inflow. Finally, rainbands consisting of strong convective cells may modify the inflow (e.g., Barnes and Powell 1995).

The storm’s proximity to land may also create asymmetries in the radial flow. Kepert (2002) analyzed Hurricane Mitch (1998) and found the strongest inflow in the eyewall region to the left rear of the storm, similar to Hurricane Bonnie. Kepert concluded that even though Mitch was 85 km offshore at the analysis time, the maximum inflow was due to the nearness of land. The enhanced friction over land altered the speed and direction of the flow. This effect may be greater for Mitch since Honduras is mountainous while the coastal region of North Carolina contains only small hills. Kepert also found an anticyclonic rotation with height of Hurricane Mitch’s (1998) maximum inflow.

The GPS sondes deliver a detailed view of the inflow depth to the eyewall (Fig. 17). Depths of 1300–1700 m are found along the western half of the eyewall, as one moves cyclonically from north to south. To the east and east-northeast there is no inflow. The deepest and strongest inflow is correlated with the lowest energy content (see Fig. 11) around the eyewall.

d. Analysis over land

The 1200 UTC surface analysis on 26 August 1998 (Fig. 18) indicates that as Bonnie approached the coast it was not interacting with any strong surface weather features. At this time a weak high of 1018 hPa is centered west of Chicago. Farther east, stretching from New England to Arkansas, is a very weak cold front. The cold front is located on the western side of the Appalachian Mountains over 500 km from Bonnie’s center. The flow within 500 km of the TC is essentially dictated by the storm itself.

The NCEP–NCAR reanalysis fields at various levels were also examined. The geopotential at 500 hPa reveals a weak trough located at 40°N (Fig. 19). This is approximately 8° latitude north of the storm at the 1730 UTC composite time. In a potential vorticity study, Hanley et al. (2001) composited 121 Atlantic TCs and found that troughs at this extended range did not have a predictable impact on hurricane intensity. Given the weakness of the trough and its distance to the storm, we feel that it is unlikely that the inner-core region is being affected.

The NCEP–NCAR reanalyses of surface temperature and specific humidity at 1800 UTC on 26 August broadly match the independently derived GPS analyses. The onshore air has temperatures near 27°C while the offshore air is at 28°C by 1.5° from the circulation center, which is similar to the GPS sonde analysis (Fig. 5). In the reanalysis, the surface specific humidity increases from about 2 g kg−1 from the southwest to the northeast. This matches the sign but not the magnitude of the GPS analysis (increase of q is about 5 g kg−1). The difference is that the NCEP–NCAR reanalysis places the 15 g kg−1 contour farther onshore than what the sondes portray. Temperature and moisture at 850 hPa in the NCEP–NCAR reanalysis are more symmetric. This agrees with the dropwindsonde analyses, where asymmetries in the thermodynamic fields weaken from the surface to 1500-m height.

4. Conclusions

a. Caveats

The resolvable scales of this analysis preclude discussion of the sharp gradients within the eyewall and rainbands that have been detected routinely with 1-Hz sampling with research aircraft. The GPS sonde analysis cannot be used to address how conditions change in the onshore flow after it crosses the coast since NOAA is currently prohibited from dropping sondes over land. We assume steadiness over 11 h, which enables a depiction of the vortex scale, but de-emphasizes shorter-lived features.

b. Summary

The GPS sonde data yield coherent mesoscale structures that are consistent from level to level, down to 10-m height. This did not have to be so if many of the sondes were compromised by high winds and rain, drifted from calibration, or if the hurricane structure was dominated by randomly spaced, convective-scale features.

Bonnie entrains warm and dry air that has been over land. This air is stable, and inhibits precipitation in a wedge-shaped region bordering the land that narrows toward the eyewall in the offshore flow. This is at a time when the influence of the shear vector is small (Rodgers et al. 2003).

Initially the offshore flow is 10–12 K lower in θE than air at a corresponding radial distance in the onshore flow, but nearer the eyewall, these differences diminish. The reduction in the heterogeneity of the θE field is partly due to high evaporation from the sea into the dry offshore flow and partly because air nearer the circulation center has spent less time over land.

Cool air is collocated with the eyewall, and with the upwelled water in the right-rear quadrant. Maximum variability in the temperature field is found at the lowest levels. Causes include advection of slightly warmer air from land, the evaporation of spray into low relative humidity air, and downdrafts. Based on the relative humidity field, it is likely that spray evaporation is an important process initially in the offshore flow. Near the eyewall where relative humidity is near 95%, downdrafts play a larger role in cooling the inflow. While it is quite possible to observe strong asymmetries well away from land (e.g., Black and Holland 1995) we believe that the nearness of land contributed to the thermodynamic asymmetries observed in Bonnie, which are largest at 10-m altitude and become more homogeneous by 2-km altitude.

Bonnie manifests asymmetries in the kinematic fields that are similar to those noted by Kepert (2002) for Hurricane Mitch (1998) as it neared land. The strongest inflow is found in the left-rear quadrant, in contrast to the typical pattern found for TCs well out to sea. There is substantial evidence from prior TC studies (e.g., Powell and Houston 1996; Knupp et al. 2004a, b) that the onshore winds near the surface are reduced as they respond to the increased roughness over land. This produces a zone of convergence near the coast and we expected a similar response for Bonnie. However, we witnessed no changes in the intensity or alignment of the rainbands coming ashore. Apparently, the changes in the winds are not substantial enough to produce an enhancement of the convergence to the magnitude that results in the generation of new convective clouds.

Intensity did not vary during the sampling despite the ingestion of lower θE by the eyewall. This may be due to a lagged response, or to counteracting effects such as the contraction of the eyewall.

When will hurricanes react most strongly to the ingestion of air that has originated over land? We surmise that a slow-moving high-category TC, making landfall into an air mass that strongly contrasts with typical tropical oceanic conditions, and at a time in the diurnal cycle that maximizes these differences, will be the most likely TC candidate that will manifest a reduction in intensity prior to landfall.

Acknowledgments

This work is supported by NSF Grant ATM-0239648. The expertise and dedication by the NOAA Aircraft Operations Center and by the NOAA/AOML/Hurricane Research Division was indispensable in the collection of these observations. We especially thank Peter Dodge for making the WSR-88D data available.

REFERENCES

  • Andreas, E. L., 1995: The temperature of evaporating sea spray droplets. J. Atmos. Sci., 52 , 852862.

  • Atlas, D., , K. R. Hardy, , R. Wexler, , and R. J. Boucher, 1963: On the origin of hurricane spiral bands. Geofis. Int., 3 , 123132.

  • Barnes, G. M., , and M. D. Powell, 1995: Evolution of the inflow boundary layer of Hurricane Gilbert (1988). Mon. Wea. Rev., 123 , 23482368.

    • Search Google Scholar
    • Export Citation
  • Barnes, G. M., , and P. B. Bogner, 2001: Comments on “Surface observations in the hurricane environment.”. Mon. Wea. Rev., 129 , 12671269.

    • Search Google Scholar
    • Export Citation
  • Barnes, G. M., , G. D. Emmitt, , B. Brummer, , M. A. LeMone, , and S. Nicholls, 1980: The structure of a fair weather boundary layer based on the results of several measurement strategies. Mon. Wea. Rev., 108 , 349364.

    • Search Google Scholar
    • Export Citation
  • Barnes, G. M., , E. J. Zipser, , D. Jorgensen, , and F. D. Marks, 1983: Mesoscale and convective structure of a hurricane rainband. J. Atmos. Sci., 40 , 21252137.

    • Search Google Scholar
    • Export Citation
  • Bender, M. A., , R. E. Tuleya, , and Y. Kurihara, 1985: A numerical study of the effect of a mountain range on a landfalling tropical cyclone. Mon. Wea. Rev., 113 , 567582.

    • Search Google Scholar
    • Export Citation
  • Bender, M. A., , R. E. Tuleya, , and Y. Kurihara, 1987: A numerical study of the effect of island terrain on tropical cyclones. Mon. Wea. Rev., 115 , 130155.

    • Search Google Scholar
    • Export Citation
  • Black, P. G., , and G. J. Holland, 1995: The boundary layer of Tropical Cyclone Kerry (1979). Mon. Wea. Rev., 123 , 20072028.

  • Blackwell, K. G., 2000: The evolution of Hurricane Danny (1997) at landfall: Doppler-observed eyewall replacement, vortex contraction/intensification, and low-level wind maxima. Mon. Wea. Rev., 128 , 40024016.

    • Search Google Scholar
    • Export Citation
  • Bogner, P. B., , G. M. Barnes, , and J. L. Franklin, 2000: Conditional instability and shear for six hurricanes over the Atlantic Ocean. Wea. Forecasting, 15 , 192207.

    • Search Google Scholar
    • Export Citation
  • Bolton, D., 1980: The computation of equivalent potential temperature. Mon. Wea. Rev., 108 , 10461053.

  • Brand, S., , and J. W. Blelloch, 1973: Changes in the characteristics of typhoons crossing the Philippines. J. Appl. Meteor., 12 , 104109.

    • Search Google Scholar
    • Export Citation
  • Brand, S., , and J. W. Blelloch, 1974: Changes in the characteristics of typhoons crossing the island of Taiwan. Mon. Wea. Rev., 102 , 708713.

    • Search Google Scholar
    • Export Citation
  • Chan, J. C. L., , and X. Liang, 2003: Convective asymmetries associated with tropical cyclone landfall. Part I: f-plane simulations. J. Atmos. Sci., 60 , 15601576.

    • Search Google Scholar
    • Export Citation
  • Chen, Y., , and M. K. Yau, 2003: Asymmetric structures in a simulated landfalling hurricane. J. Atmos. Sci., 60 , 22942312.

  • Cione, J. J., , P. G. Black, , and S. H. Houston, 2000: Surface observations in the hurricane environment. Mon. Wea. Rev., 128 , 15501561.

    • Search Google Scholar
    • Export Citation
  • Daida, S. K., , and G. M. Barnes, 2003: Hurricane Paine (1986) grazes the high terrain of the Baja California peninsula. Wea. Forecasting, 18 , 981990.

    • Search Google Scholar
    • Export Citation
  • Eastin, M. D., 2003: Buoyancy of convective vertical motions in the inner core of intense hurricanes. Ph.D. dissertation, Dept. of Atmospheric Science, Colorado State University, 152 pp.

  • Emanuel, K. A., 1986: An air-sea interaction theory for tropical cyclones. Part I: Steady-state maintenance. J. Atmos. Sci., 43 , 585605.

    • Search Google Scholar
    • Export Citation
  • Hanley, D., , J. Molinari, , and D. Keyser, 2001: A composite study of the interactions between tropical cyclones and upper-tropospheric troughs. Mon. Wea. Rev., 129 , 25702584.

    • Search Google Scholar
    • Export Citation
  • Hawkins, H. F., , and S. M. Imbembo, 1976: The structure of a small, intense hurricane, Inez (1966). Mon. Wea. Rev., 104 , 418442.

  • Hock, T. F., , and J. L. Franklin, 1999: The NCAR GPS dropwindsonde. Bull. Amer. Meteor. Soc., 80 , 407420.

  • Howard, J. R., , A. L. Doggett, , R. E. Peterson, , P. G. Black, , J. L. Schroeder, , D. A. Smith, , and J. P. Dunyak, 2004: Transition in onshore hurricane boundary layer winds during the landfall of Hurricane Lili (2002). Preprints, 26th Conf. on Hurricanes and Tropical Meteorology, Miami, FL, Amer. Meteor. Soc., 526–527.

  • Jordan, C. L., 1958: Mean soundings for the West Indies area. J. Meteor., 15 , 9197.

  • Jorgensen, D. P., 1984: Mesoscale and convective-scale characteristics of mature hurricanes. Part II: Inner core structure of Hurricane Allen (1980). J. Atmos. Sci., 41 , 12871311.

    • Search Google Scholar
    • Export Citation
  • Jorgensen, D. P., , and P. T. Willis, 1982: A ZR relationship for hurricanes. J. Appl. Meteor., 21 , 356366.

  • Kepert, J. D., 2002: The wind-field structure of the tropical cyclone boundary layer. Ph.D. thesis, Dept. of Mathematics and Statistics, Monash University, 350 pp. [Available from Dept. of Mathematics and Statistics, Monash University, Vic 3800, Australia.].

  • Knupp, K. R., , J. Walters, , and M. Biggerstaff, 2004a: Doppler profiler and radar observations of boundary layer variability during the landfall of TS Gabreille. Preprints, 26th Conf. on Hurricanes and Tropical Meteorology, Miami, FL, Amer. Meteor. Soc., 568–569.

  • Knupp, K. R., , J. Walters, , and M. Biggerstaff, 2004b: Mesoscale and precipitation processes observed by Doppler profiler and radar during the landfall of Tropical Storm Gabrielle. Preprints, 26th Conf. on Hurricanes and Tropical Meteorology, Miami, FL, Amer. Meteor. Soc., 570–571.

  • Marks Jr., F. D., 1985: Evolution of the structure of precipitation in Hurricane Allen (1980). Mon. Wea. Rev., 113 , 909930.

  • Marks Jr., F. D., , R. A. Houze, , and J. F. Gamache, 1992: Dual-aircraft investigation of the inner core of Hurricane Norbert. Part I: Kinematic structure. J. Atmos. Sci., 49 , 919942.

    • Search Google Scholar
    • Export Citation
  • Miller, B. I., 1964: A study of the filling of Hurricane Donna (1960) over land. Mon. Wea. Rev., 92 , 389406.

  • Molinari, J., , P. K. Moore, , and V. P. Idone, 1999: Convective structure of hurricanes as revealed by lightning locations. Mon. Wea. Rev., 127 , 520534.

    • Search Google Scholar
    • Export Citation
  • Powell, M. D., 1982: The transition of the Hurricane Frederic boundary layer wind field of the open Gulf of Mexico to landfall. Mon. Wea. Rev., 110 , 19121932.

    • Search Google Scholar
    • Export Citation
  • Powell, M. D., 1987: Changes in the low-level kinematic and thermodynamic structure of Hurricane Alicia (1983) at landfall. Mon. Wea. Rev., 115 , 7599.

    • Search Google Scholar
    • Export Citation
  • Powell, M. D., 1990: Boundary layer structure and dynamics in outer hurricane rainbands. Mon. Wea. Rev., 118 , 891917.

  • Powell, M. D., , and S. H. Houston, 1996: Hurricane Andrew’s landfall in south Florida. Part II: Surface wind fields and potential real-time applications. Wea. Forecasting, 11 , 329349.

    • Search Google Scholar
    • Export Citation
  • Powell, M. D., , P. P. Dodge, , and M. L. Black, 1991: The landfall of Hurricane Hugo in the Carolinas: Surface wind distribution. Wea. Forecasting, 6 , 379399.

    • Search Google Scholar
    • Export Citation
  • Raman, S., 1979: Atmospheric turbulence and storm surge due to Hurricane Belle (1976). Mon. Wea. Rev., 107 , 314321.

  • Reasor, P. D., , M. T. Montgomery, , F. D. Marks Jr., , and J. F. Gamache, 2000: Low-wavenumber structure and evolution of the hurricane inner core observed by airborne dual-Doppler radar. Mon. Wea. Rev., 128 , 16531680.

    • Search Google Scholar
    • Export Citation
  • Rodgers, R., , S. Chen, , J. Tenerelli, , and H. Willoughby, 2003: A numerical study of the impact of vertical shear on the distribution of rainfall in Hurricane Bonnie (1998). Mon. Wea. Rev., 131 , 15771599.

    • Search Google Scholar
    • Export Citation
  • Rotunno, R., , and K. A. Emanuel, 1987: An air-sea interaction theory for tropical cyclones. Part II: Evolutionary study using a non-hydrostatic axisymmetric numerical model. J. Atmos. Sci., 44 , 542561.

    • Search Google Scholar
    • Export Citation
  • Shapiro, L. J., 1983: The asymmetric boundary layer flow under a translating hurricane. J. Atmos. Sci., 40 , 19841998.

  • Shea, D. J., , and W. M. Gray, 1973: The hurricane’s inner core region. Part I: Symmetric and asymmetric structure. J. Atmos. Sci., 30 , 15441564.

    • Search Google Scholar
    • Export Citation
  • Spratt, S. M., , F. D. Marks, , P. P. Dodge, , and D. W. Sharp, 2000: Examining the pre-landfall environment of mesovortices within a Hurricane Bonnie (1998) outer rainband. Preprints, 24th Conf. on Hurricanes and Tropical Meteorology, Ft. Lauderdale, FL, Amer. Meteor. Soc., 300–301.

  • Tuleya, R. E., 1994: Tropical storm development and decay: Sensitivity to surface boundary conditions. Mon. Wea. Rev., 122 , 291304.

  • Tuleya, R. E., , and Y. Kurihara, 1978: Numerical simulation of the landfall of tropical cyclones. J. Atmos. Sci., 35 , 242257.

  • Willoughby, H. E., , and M. B. Chelmow, 1982: Objective determination of hurricane tracks from aircraft observations. Mon. Wea. Rev., 110 , 12981305.

    • Search Google Scholar
    • Export Citation
  • Willoughby, H. E., , F. D. Marks Jr., , and R. J. Feinberg, 1984: Stationary and moving convective bands in hurricanes. J. Atmos. Sci., 41 , 31893211.

    • Search Google Scholar
    • Export Citation
  • Wroe, D. R., , and G. M. Barnes, 2003: Inflow layer energetics of Hurricane Bonnie (1998) near landfall. Mon. Wea. Rev., 131 , 16001612.

    • Search Google Scholar
    • Export Citation
  • Wu, C-C., , and Y. H. Kuo, 1999: Typhoons affecting Taiwan: Current understanding and future challenges. Bull. Amer. Meteor. Soc., 80 , 6780.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

(a) Earth-relative location of the 85 GPS dropwindsondes (diamonds) deployed from 1220 to 2300 UTC on 26 Aug 1998. The hurricane track is noted by the bold line and the WSR-88D locations (Wilmington and Morehead City) are denoted by stars. (b) The locations of the 64 sondes that provided data near the sea surface, plotted relative to the circulation center marked by a hurricane symbol.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 2.
Fig. 2.

Single-sweep radar images (440 km × 440 km) from the WSR-88Ds at Wilmington (KLTX) at (a) 1300:06 and (b) 15:00:02 UTC and Morehead City (KMHX) at (c) 1659:55 and (d) 1957:07 UTC on 26 Aug 1998. Reflectivity values (dBZ) are defined in the color tables for each panel.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 3.
Fig. 3.

Bonnie’s eyewall radius (km), as indicated by highest reflectivity features from 1000 to 2100 UTC on 26 Aug 1998. Measurements come from KLTX, KMHX, and the 42 RF lower fuselage radar (LF).

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 4.
Fig. 4.

Difference in reflectivity (dBZ) as a function of radial distance, northeast half of the transect shown in Fig. 2a minus the southwest half of the transect. The zero point is the circulation center.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 5.
Fig. 5.

Storm-relative composite of Bonnie’s temperature (°C) at 10 m. Contours are every 1°C (bold dashed lines) except that the 25.5°C isotherm (dotted line) is denoted for the cool air found in the southwest eyewall region. The hurricane symbol denotes the storm center and the ring is the approximate eyewall location at 1730 UTC, the central time of this composite. The black arrows represent the wind speed and direction at 10 m. The scale is indicated to the lower-right side of the image.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 6.
Fig. 6.

Storm-relative composite of Bonnie’s temperature (°C) at 2 km. Contours are every 1°C (bold dashed lines). The hurricane symbol denotes the storm center and the ring is the approximate eyewall location at 1730 UTC.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 7.
Fig. 7.

Same as in Fig. 5, but for specific humidity (g kg−1) with 1 g kg−1 contour intervals (bold dashed lines). The 21.5 g kg−1 contour around the circulation center is depicted by a dotted line.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 8.
Fig. 8.

Same as in Fig. 6, but for specific humidity (g kg−1) with 1 g kg−1 contour intervals (bold dashed lines).

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 9.
Fig. 9.

The LCL (m) with a contour interval of 200 m (bold dashed lines). Other figure aspects follow Fig. 6.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 10.
Fig. 10.

The LI at 800 hPa (°C) with a contour interval of 0.5°C (bold dashed lines). Other figure aspects follow Fig. 6.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 11.
Fig. 11.

Same as in Fig. 5, but for equivalent potentialtemperature (K) with contour intervals of 4 K (bold dashed lines).

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 12.
Fig. 12.

Equivalent potential temperature (K) at 2 km depicted with a contour interval of 4 K (bold dashed lines). Other figure aspects follow Fig. 6.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 13.
Fig. 13.

Storm-relative tangential winds (m s−1) at 10 m with a contour interval of 5 m s−1 (bold dashed lines). Winds less than 15 m s−1 are tightly packed around the circulation center and are not drawn for clarity. Other figure aspects follow Fig. 6.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 14.
Fig. 14.

Same as in Fig. 13, but at 2 km.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 15.
Fig. 15.

Storm-relative radial winds (m s−1) at 10 m with a contour interval of 5 m s−1 (bold dashed lines). Negative values are inflow. Other figure aspects follow Fig. 6.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 16.
Fig. 16.

Same as in Fig. 15, but at 2 km.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 17.
Fig. 17.

Storm-relative inflow depth (km) with contour intervals of every 0.5 km (bold dashed lines). Shaded region depicts area of outflow.

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 18.
Fig. 18.

Surface analysis for 1200 UTC 26 Aug 1998. Thin black lines are isobars, thicker blue lines depict cold fronts, red and blue lines are stationary fronts, surface lows positions and central pressure are labeled in red, and selected stations show temperature (°F), dewpoint temperature (°F), pressure (mb), and winds following standard meteorological convention. (More information available online at http://weather.unisys.com)

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Fig. 19.
Fig. 19.

Geopotential heights (m) at 500 hPa at 1800 UTC 26 Aug 1998. Contour intervals are every 10 m (solid lines).

Citation: Monthly Weather Review 133, 11; 10.1175/MWR3027.1

Save