• Bateman, M. G., , M. F. Stewart, , R. J. Blakeslee, , S. J. Podgorny, , H. J. Christian, , D. M. Mach, , J. C. Bailey, , and D. Daskar, 2007: A low-noise, microprocessor-controlled, internally digitizing rotating-vane electric field mill for airborne platforms. J. Atmos. Oceanic Technol., 24 , 12451255.

    • Search Google Scholar
    • Export Citation
  • Beven II, J. L., and Coauthors, 2008: Atlantic hurricane season of 2005. Mon. Wea. Rev., 136 , 11091173.

  • Black, M. L., , R. W. Burpee, , and F. D. Marks Jr., 1996: Vertical motion characteristics of tropical cyclones determined with airborne Doppler radial velocities. J. Atmos. Sci., 53 , 18871909.

    • Search Google Scholar
    • Export Citation
  • Cecil, D. J., , E. J. Zipser, , and S. W. Nesbitt, 2002: Reflectivity, ice scattering, and lightning characteristics of hurricane eyewalls and rainbands. Part I: Quantitative description. Mon. Wea. Rev., 130 , 769784.

    • Search Google Scholar
    • Export Citation
  • DeMaria, M., , M. Mainelli, , L. K. Shay, , J. A. Knaff, , and J. Kaplan, 2005: Further improvements to the Statistical Hurricane Intensity Prediction Scheme (SHIPS). Wea. Forecasting, 20 , 531543.

    • Search Google Scholar
    • Export Citation
  • Dodge, P., , R. W. Burpee, , and F. D. Marks Jr., 1999: The kinematic structure of a hurricane with sea level pressure less than 900 mb. Mon. Wea. Rev., 127 , 9871004.

    • Search Google Scholar
    • Export Citation
  • Dvorak, V. E., 1984: Tropical cyclone intensity analysis using satellite data. NOAA Tech. Rep. NESDIS 11, 47 pp.

  • Franklin, J. L., , and D. P. Brown, 2006: Tropical cyclone report: Hurricane Emily. National Hurricane Center, 18 pp. [Available online at http://www.nhc.noaa.gov/pdf/TCR-AL052005_Emily.pdf].

    • Search Google Scholar
    • Export Citation
  • Heymsfield, G. M., 1989: Accuracy of vertical air motions from nadir-viewing Doppler airborne radars. J. Atmos. Oceanic Technol., 6 , 10791082.

    • Search Google Scholar
    • Export Citation
  • Heymsfield, G. M., and Coauthors, 1996: The EDOP radar system on the high-altitude NASA ER-2 aircraft. J. Atmos. Oceanic Technol., 13 , 795809.

    • Search Google Scholar
    • Export Citation
  • Heymsfield, G. M., , J. B. Halverson, , and I. J. Caylor, 1999: A wintertime Gulf Coast squall line observed by EDOP airborne doppler radar. Mon. Wea. Rev., 127 , 29282950.

    • Search Google Scholar
    • Export Citation
  • Heymsfield, G. M., , J. B. Halverson, , J. Simpson, , L. Tian, , and T. P. Bui, 2001: ER-2 Doppler radar investigations of the eyewall of Hurricane Bonnie during the Convection and Moisture Experiment-3. J. Appl. Meteor., 40 , 13101330.

    • Search Google Scholar
    • Export Citation
  • Heymsfield, G. M., , L. Tian, , A. J. Heymsfield, , L. Li, , and S. Guimond, 2010: Characteristics of deep tropical and subtropical convection from nadir-viewing high-altitude airborne Doppler radar. J. Atmos. Sci., 67 , 285308.

    • Search Google Scholar
    • Export Citation
  • Hood, R. E., and Coauthors, 2006: Classification of tropical oceanic precipitation using high-altitude aircraft microwave and electric field measurements. J. Atmos. Sci., 63 , 218233.

    • Search Google Scholar
    • Export Citation
  • Iguchi, T., , and R. Meneghini, 1994: Intercomparison of single-frequency methods for retrieving a vertical rain profile from airborne or spaceborne radar data. J. Atmos. Oceanic Technol., 11 , 15071516.

    • Search Google Scholar
    • Export Citation
  • Kelley, O. A., , J. Stout, , and J. B. Halverson, 2004: Tall precipitation cells in tropical cyclone eyewalls are associated with tropical cyclone intensification. Geophys. Res. Lett., 31 , L24112. doi:10.1029/2004GL021616.

    • Search Google Scholar
    • Export Citation
  • Kelley, O. A., , J. Stout, , and J. B. Halverson, 2005: Hurricane intensification detected by continuously monitoring tall precipitation in the eyewall. Geophys. Res. Lett., 32 , L20819. doi:10.1029/2005GL023583.

    • Search Google Scholar
    • Export Citation
  • Kossin, J. P., 2002: Daily hurricane variability inferred from GOES infrared imagery. Mon. Wea. Rev., 130 , 22602270.

  • Lyons, W. A., , and C. S. Keen, 1994: Observations of lightning in convective supercells within tropical storms and hurricanes. Mon. Wea. Rev., 122 , 18971916.

    • Search Google Scholar
    • Export Citation
  • Mach, D. M., , and W. J. Koshak, 2007: General matrix inversion technique for the calibration of electric field sensor arrays on aircraft platforms. J. Atmos. Oceanic Technol., 24 , 15761587.

    • Search Google Scholar
    • Export Citation
  • Mach, D. M., , R. J. Blakeslee, , M. G. Bateman, , and J. C. Bailey, 2009: Electric fields, conductivity, and estimated currents from aircraft overflights of electrified clouds. J. Geophys. Res., 114 , D10204. doi:10.1029/2008JD011495.

    • Search Google Scholar
    • Export Citation
  • Molinari, J., , P. K. Moore, , V. P. Idone, , R. W. Henderson, , and A. B. Saljoughy, 1994: Cloud-to-ground lightning in Hurricane Andrew. J. Geophys. Res., 99 , 1666516676.

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

    • Search Google Scholar
    • Export Citation
  • Muramatsu, T., 1983: Diurnal variations of satellite-measured TBB areal distribution and eye diameter of mature typhoons. J. Meteor. Soc. Japan, 61 , 7790.

    • Search Google Scholar
    • Export Citation
  • Quinlan, K. R., 2008: Conditions influencing Hurricane Emily’s (2005) precipitation patterns, convection, and upper tropospheric outflow. M.S. thesis, Department of Atmospheric Science, University of Alabama in Huntsville, 116 pp.

  • Sato, K., 1993: Small-scale wind disturbances observed by the MU radar during the passage of Typhoon Kelly. J. Atmos. Sci., 50 , 518537.

    • Search Google Scholar
    • Export Citation
  • Spencer, R. W., , R. E. Hood, , F. J. Lafontaine, , E. A. Smith, , R. Platt, , J. Galliano, , V. L. Griffin, , and E. Lobl, 1994: High-resolution imaging of rain systems with the Advanced Microwave Precipitation Radiometer. J. Atmos. Oceanic Technol., 11 , 849857.

    • Search Google Scholar
    • Export Citation
  • Squires, K., , and S. Businger, 2008: The morphology of eyewall lightning outbreaks in two category 5 hurricanes. Mon. Wea. Rev., 136 , 17061726.

    • Search Google Scholar
    • Export Citation
  • Steranka, J., , E. B. Rodgers, , and R. C. Gentry, 1984: The diurnal variation of Atlantic Ocean tropical cyclone cloud distribution inferred from geostationary satellite infrared measurements. Mon. Wea. Rev., 112 , 23382344.

    • Search Google Scholar
    • Export Citation
  • Steranka, J., , E. B. Rodgers, , and R. C. Gentry, 1986: The relationship between satellite measured convective bursts and tropical cyclone intensification. Mon. Wea. Rev., 114 , 15391546.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Hovmöller diagrams as a function of distance (10-km bins) from Hurricane Emily’s center for (a) mean IR TB and (b) minimum IR TB. Contours are at 5-K intervals, but the shading differs for (a) and (b) because (a) covers a broader range of values. (c) Best-track maximum sustained wind (solid) and minimum surface pressure (dashed), with the time of the ER-2 observations marked.

  • View in gallery

    Hovmöller diagrams as a function of azimuth (30° bins) around Hurricane Emily’s center, within 50-km radius. (a) Mean IR TB using only pixels at or below 220 K (to avoid including the eye). (b) Minimum IR TB. (c) Best-track maximum sustained wind (solid) and minimum surface pressure (dashed), with the time of the ER-2 observations marked.

  • View in gallery

    Sequence of 85-GHz horizontal channel imagery [89 GHz for Advanced Microwave Scanning Radiometer for Earth Observing System (AMSR-E)] from the Navy Research Laboratory (NRL)—Monterey tropical cyclone Web page: (a) 1708 UTC 16 Jul TMI, (b) 1845 UTC 16 Jul AMSR-E, (c) 0119 UTC 17 Jul TMI, (d) 0159 UTC 17 Jul SSMI, (e) 0657 UTC 17 Jul AMSR-E, (f) 1236 UTC 17 Jul SSMI. Note that sensor resolution varies between instruments.

  • View in gallery

    GOES-11 IR temperatures at (a) 0745, (b) 0750, and (c) 0755 UTC 17 Jul with ER-2 flight track overlaid. Flight tracks are for five minutes, centered on the time of each satellite image.

  • View in gallery

    API (shaded) and LIP electric field vectors. The projections of the electric field are plotted as barbs originating at the aircraft location, pointing away from positive charge, for (a) 0735–0833; (b) 0833–0910; (c) 0936–1030; (d) 1030–1128 UTC.

  • View in gallery

    (a) Nadir EDOP reflectivity in an east-southeast–west-northwest direction across eye of Hurricane Emily from 0751 to 0755 UTC 17 Jul. Contour interval (CI) is 10 dB. (b) Nadir vertical velocity; positive values indicate upward motion. CI is 5, with 0 m s−1 (bold). Data have been removed from (a) and (b) for ER-2 roll variations greater than ±2°. (c) AMPR TB from middle of swath (includes rolls). For reference with Figs. 4, 5, the area extends from 17.76°N, 81.51°W to 17.97°N, 81.93°W.

  • View in gallery

    Trace of AMPR TB (K) along the aircraft flight track (solid gray line), normal to the aircraft (influenced by aircraft rolls, same as in Fig. 6c; dotted line), and 10° to the right of the flight track (across the center of the eye; dashed line) for (a) 10, (b) 19, (c) 37, (d) 85 GHz.

  • View in gallery

    GOES-11 IR temperatures at 0845 UTC Jul 17 with ER-2 flight path from 0842:30 to 0847:30 UTC overlaid.

  • View in gallery

    (a) Nadir EDOP reflectivity in a southwest–northeast direction across eye of Hurricane Emily from 0844 to 0848 UTC 17 Jul. CI is 10 dB. (b) Nadir EDOP vertical velocity; positive values indicate upward motion. CI is 5, with 0 m s−1 (bold). (c) AMPR brightness temperatures from middle of swath (includes rolls). For reference with Figs. 4, 8, the area extends from 17.86°N, 82.09°W to 18.22°N, 81.83°W.

  • View in gallery

    Trace of AMPR TBs (K) along the aircraft flight track (solid line), normal to the aircraft (influenced by aircraft rolls, same as in Fig. 9c; dotted line), and 20° to the right of the flight track (across the center of the eye; dashed line) for (a) 10, (b) 19, (c) 37, (d) 85 GHz.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 29 29 2
PDF Downloads 20 20 1

Intense Convection Observed by NASA ER-2 in Hurricane Emily (2005)

View More View Less
  • 1 University of Alabama in Huntsville, Huntsville, Alabama
© Get Permissions
Full access

Abstract

On 17 July, intense convection in the eyewall of Hurricane Emily (2005) was observed by the high-altitude (∼20 km) NASA ER-2 aircraft. Analysis of this convection is undertaken using downward-looking radar, passive microwave radiometer, electric field mills, and Geostationary Operational Environmental Satellite-11 (GOES-11) rapid-scan infrared imagery. Radar data show convection reaching more than 17 km, with reflectivity more than 40 dBZ and estimated updraft speeds greater than 20 m s−1 at ∼14-km altitude. All of the passive microwave frequencies (10, 19, 37, and 85 GHz) experienced scattering by large ice particles. Large electric fields with dozens of lightning flashes were recorded. Because of safety concerns arising from difficulties with the first two transects, the flight plan was modified to avoid passing above the eyewall again. These observations occurred 8–10 h after Emily’s peak 929-hPa intensity, with central pressures from reconnaissance aircraft having risen to 943 hPa immediately before the flight and 946 hPa immediately afterward (no such measurements available during the flight). Rapid-scan infrared imagery reveals that a period of episodic bursts of strong, deep convection was beginning just as the ER-2 arrived. The first leg across the eye coincided with a rapidly growing new cell along the flight track in the western eyewall. This strong convection may have been characteristic of Emily for the ∼24 h leading up to landfall in the Yucatan, but it does not appear to be a continuation of convective trends from the previous rapid intensification or peak intensity periods.

* Current affiliation: Nuclear Regulatory Commission, Washington, D.C

Corresponding author address: Dr. Daniel J. Cecil, Earth System Science Center, University of Alabama in Huntsville, 320 Sparkman Dr. NW, Huntsville, AL 35805. Email: cecild@uah.edu

Abstract

On 17 July, intense convection in the eyewall of Hurricane Emily (2005) was observed by the high-altitude (∼20 km) NASA ER-2 aircraft. Analysis of this convection is undertaken using downward-looking radar, passive microwave radiometer, electric field mills, and Geostationary Operational Environmental Satellite-11 (GOES-11) rapid-scan infrared imagery. Radar data show convection reaching more than 17 km, with reflectivity more than 40 dBZ and estimated updraft speeds greater than 20 m s−1 at ∼14-km altitude. All of the passive microwave frequencies (10, 19, 37, and 85 GHz) experienced scattering by large ice particles. Large electric fields with dozens of lightning flashes were recorded. Because of safety concerns arising from difficulties with the first two transects, the flight plan was modified to avoid passing above the eyewall again. These observations occurred 8–10 h after Emily’s peak 929-hPa intensity, with central pressures from reconnaissance aircraft having risen to 943 hPa immediately before the flight and 946 hPa immediately afterward (no such measurements available during the flight). Rapid-scan infrared imagery reveals that a period of episodic bursts of strong, deep convection was beginning just as the ER-2 arrived. The first leg across the eye coincided with a rapidly growing new cell along the flight track in the western eyewall. This strong convection may have been characteristic of Emily for the ∼24 h leading up to landfall in the Yucatan, but it does not appear to be a continuation of convective trends from the previous rapid intensification or peak intensity periods.

* Current affiliation: Nuclear Regulatory Commission, Washington, D.C

Corresponding author address: Dr. Daniel J. Cecil, Earth System Science Center, University of Alabama in Huntsville, 320 Sparkman Dr. NW, Huntsville, AL 35805. Email: cecild@uah.edu

1. Introduction and storm history

Hurricane Emily (2005) provided a rare opportunity to observe a strong category 4 hurricane with the NASA ER-2’s high-resolution downward-looking radar and radiometer, and electric field mills. It was the strongest hurricane to date observed by the ER-2. This paper analyzes intense convection observed during that flight, in the context of rapid-scan and normal-mode cloud-top temperature measurements from Geostationary Operational Environmental Satellite (GOES) satellites. There are two main motivations for this study: 1) Emily’s eyewall convective towers rank among the strongest seen in any ER-2 tropical cyclone (TC) missions, and 2) the intense convection was observed during a weakening period that continued through landfall a day later. In the absence of strong vertical wind shear, such strong eyewall convection is more commonly thought to be a trait of tropical cyclones that are either intensifying or maintaining high intensity.

Much attention in the literature, presentations, and less formal discussions is given to cases and hypotheses in which strong convective bursts accompany or precede intensification. The reasoning often involves increased latent heating in the strong convection, increased stretching of vorticity by the strong convection, and/or forced subsidence in response to the strong convection. There are statistical correlations between strong convection and tropical cyclone intensification. Steranka et al. (1986) found convective bursts with low mean infrared (IR) brightness temperature (TB) over a large area (∼200-km radius from the center) persisting for six or more hours and often followed by tropical cyclone intensification. Noteworthy for the current study, Steranka et al.’s results were less robust for the strongest hurricanes. Kelley et al. (2004, 2005) found intensification to be common if the eyewall has a 20-dBZ radar echo—even one covering a very small area—reaching at least 14.5-km height, especially if this condition has some temporal persistence. Many of the early papers on hurricane eyewall lightning noted outbreaks that occurred either during or preceding intensification periods (e.g., Lyons and Keen 1994; Molinari et al. 1994). As observations from more cases became available, Molinari et al. (1999) noted that lightning outbreaks may signify the onset, continuation, or end of deepening periods. The Hurricane Emily (2005) case in this study fits the profile of storms in which Molinari et al. (1999) describe the lightning outbreak as signaling the end (or even reversal) of an intensification trend. This paper demonstrates an example of a null case, with intense convection and a weakening tropical cyclone. Null cases seem to get much less attention in literature and are important to understand.

Hurricane Emily had strengthened to category 4 status early on 15 July, attaining a minimum pressure of 952 hPa and peak winds reaching 115 kt. Emily formed concentric eyewalls of 8 and 25 n mi later on 15 July and had weakened to a category 2 by 1800 UTC. Emily immediately began a period of restrengthening that continued through 16 July. As shown by the best track in Fig. 1c, a steady period of deepening occurred through the second half of 16 July, with the lowest aircraft pressure reading of 929 hPa occurring at 0000 UTC 17 July. At this point, Emily was approximately 100 n mi to the southwest of Jamaica. The ER-2 measurements discussed in this paper were taken around 0800 UTC 17 July. Emily began to weaken on 17 July and eventually made landfall near Tulum, Mexico, on the Yucatan Peninsula at 0630 UTC 18 July as a category 4 storm (955 hPa, 115 kt). Emily emerged into the Bay of Campeche later that day at about 1200 UTC as a weak category 1. Emily then went through its last intensification before making landfall near San Fernando, Mexico (about 120 km south-southwest of Brownsville, Texas).

Forecasts around the time of the flight generally called for neither strengthening nor weakening during this period. Internal structure changes were expected to determine intensity fluctuations before landfall. Eyewall replacement cycles or obvious external forcing were not observed to explain the weakening (Beven et al. 2008). Emily encountered moderate deep-layer vertical wind shear, with the Statistical Hurricane Intensity Prediction Scheme (SHIPS; DeMaria et al. 2005) diagnosing around 8 m s−1 from the west and southwest on 15–17 July (during both deepening and weakening episodes). This shear magnitude is near the SHIPS sample mean, contributing little to strengthening or weakening. Sea surface temperatures were around 29°C and gradually increasing, with a deep oceanic mixed layer.

In the Convection and Moisture Experiment-3 (CAMEX-3), the eyewall of Hurricane Bonnie (1998) was studied by ER-2 Doppler radar (EDOP; Heymsfield et al. 2001). That study used a variety of datasets including GOES, EDOP radar data, and in situ observations to document convective bursts as Bonnie was reaching its maximum intensity (954 hPa). Downdrafts originating at tropopause height within the eye of the storm as well as within the intense hot tower were key findings in their study. They inferred that up to 3°C of warming within the eye may have resulted from these convectively induced downdrafts, rather than from gradual subsidence. Bonnie’s intensification halted within a few hours after these measurements.

The Bonnie (1998) and Emily (2005) cases are included with other strong convection cases in a census of EDOP measurements by Heymsfield et al. (2010). Among tropical cyclones in four field programs, Emily had by far the strongest downdrafts and highest reflectivity towers seen by EDOP. Hurricane Georges (1998), Tropical Storm Chantal (2001), and a handful of nontropical cyclones cases had slightly stronger updrafts than those measured in Emily. Only one such case (from Florida sea-breeze convection) had 40-dBZ reflectivity heights comparable to those recorded in Emily.

Apart from these experiments, stronger hurricanes have been studied using lower altitude aircraft. Dodge et al. (1999) examined radar and flight-level (∼3 km) data from the National Oceanic and Atmospheric Administration (NOAA) WP-3D slightly after peak intensity for category 5 Hurricane Gilbert (1988), with the central pressure near 895 hPa. They noted that the inner eyewall was more erect than in weaker hurricanes, and the strong wind speeds extended through a deeper layer than expected. Updraft speeds approaching 25 m s−1 were retrieved from the vertical incidence Doppler data. A 14 m s−1 updraft was noted as high as 17 km. Squires and Businger (2008) synthesized lightning, radar, and passive microwave measurements for two category 5 hurricanes—Katrina (2005) and Rita (2005)—focusing on eyewall lightning. Katrina and Rita both featured eyewall lightning outbreaks during rapid intensification and again during the period of maximum intensity.

The NASA ER-2 aircraft was flown on 17 July 2005 to study the convective structure of an intense category 4–5 Hurricane Emily, particularly the eyewall. Emily had weakened to category 4 prior to the ER-2’s arrival at the storm, with central pressures of 943 and 946 hPa measured by Air Force Reserve reconnaissance flights just before and after the ER-2 mission. Unfortunately, the ER-2 was not equipped with dropsondes and the NOAA WP-3D aircraft were unavailable during this ER-2 flight, so any short time-scale fluctuations during this mission are not known. After two difficult passes across the center at ∼20-km altitude, the pilot deemed it unsafe to continue the mission plan with repeated eye crossings. An alternate plan was quickly devised with the ER-2 circumnavigating just outside the eyewall in a box pattern. This flight documents the eyewall and adjacent precipitation regions shortly after Emily’s peak intensity. We also examine satellite imagery to provide a spatial and temporal context for the ER-2 measurements. Infrared imagery shows episodic bursts of strong, deep convection beginning around the time of the ER-2 flight and continuing until landfall ∼24 h later.

2. Instruments, data, and methodology

a. EDOP

EDOP (Heymsfield et al. 1996) is a dual-beam (nonscanning), downward-looking, Doppler weather radar system that operates at X band (9.6 GHz). The EDOP uses two fixed-radar beams: one at nadir and the other at approximately 33.5° forward of nadir. This combination allows for retrieval of two-dimensional motions (vertical and along the direction of the aircraft heading). The forward beam is not shown in this study, however, because the flight tracks were not successfully aligned through the storm center. Flow along the plane of the quasi-radial cross sections in this case has a strong contribution from the hurricane’s tangential wind, making interpretations troublesome. The 3° beamwidth gives a ∼1.1-km footprint at the surface (∼20-km range) for the nadir beam, but the 0.5-s sampling interval gives profiles every ∼100 m for typical ER-2 speeds. The vertically pointing beam allows for higher vertical resolution (37.5-m gate spacing) when compared to horizontally scanning radars. Heymsfield (1989) explained in detail the uncertainties in using the nadir-viewing airborne radar, including complications from drift, groundspeed, aircraft vertical velocity, heading angle, pitch angle, roll angle, track angle, and reference frame variations (Earth-fixed frame and aircraft-fixed frame).

The EDOP radar was used primarily for this study to measure the reflectivity and vertical velocity values in the eyewall during the two passes across the center on 17 July 2005. Attenuation is corrected by using the method described in Iguchi and Meneghini (1994), with some values adjusted following G. Heymsfield (2009, personal communication). Vertical air motions are retrieved by approximating the hydrometeor fall speed based on radar reflectivity (related to particle size). This fall speed is then subtracted from the Doppler velocity, after aircraft motion has been removed. Fall speeds were estimated using the method described by Black et al. (1996) and Heymsfield et al. (1999). This was done by breaking the stratiform regions into three vertical regions: rain, snow, and the transition region, or melting layer that is identified by the bright band. Convective regions were defined as areas with reflectivities greater than 45 dBZ in the 0–6-km layer. Relationships for rain, snow, and ice were used to determine the fall speeds for both convective and stratiform regions. It is difficult to achieve precise measurements of vertical velocity (updrafts and downdrafts) due to uncertainties in hydrometeor fall speed and the effects of horizontal wind when the radar beam is not exactly at nadir. Heymsfield et al. (2010) used a different formulation for hydrometeor fall speed, but they reported peak updraft and downdraft speeds for this case that are very close to the values in this paper.

b. AMPR

The Advanced Microwave Precipitation Radiometer provides brightness temperatures at 10.7, 19.35, 37.0, and 85.5 GHz over a ±45° swath about the aircraft subtrack (Spencer et al. 1994). These frequencies are similar to those on several satellites. Unlike most satellite radiometers, the polarization varies from fully vertical on the left edge of the swath to fully horizontal on the right edge. The AMPR swath width at the surface is 40 km and the spatial sampling interval is 800 m for a typical ER-2 altitude of 20 km. However, the spatial resolution (field of view at the surface) of the 85.5-GHz channel is 640 m, with larger values for the lower frequencies (2.8 km for 19.35 and 10.7 GHz).

The AMPR data were used to determine the precipitation patterns in and around the eye of Hurricane Emily. The 10 GHz is used mostly to monitor the emission from liquid rain in the eyewall and the surrounding rainbands. The next frequency, 19 GHz, has a higher sensitivity to precipitation ice and becomes saturated more easily for the heaviest liquid rain within the storm. The 85-GHz channel is more sensitive to ice-scattering effects, which is particularly useful in diagnosing the deep cloud precipitation patterns in the eyewall where 85-GHz brightness temperatures are at their coldest. At this frequency, the ice-scattering effect is most likely caused by precipitating graupel.

Scattering at progressively lower frequencies suggests progressively larger particles. This principle is exploited in the AMPR precipitation index (API; Hood et al. 2006). API values 0–5 have no ice scattering clearly indicated in any channel. Values 6–10 have 85-GHz ice scattering, with 85-GHz brightness temperatures depressed below those from the other frequencies (and below a 275-K threshold). Values 11–15 also have 37-GHz ice scattering, with TB85 < TB37 < TB19. This requires larger/more dense ice particles than 85-GHz scattering alone. API values 16–18 additionally have scattering in the 19-GHz channel, with TB85 < TB37 < TB19 < TB10. This indicates still larger ice particles generated by stronger convection. Within each subrange of API values, a higher value indicates heavier liquid rain rates with greater emission in the 10-GHz channel. For example, an API value of six has ice scattering in the 85-GHz channel and passes an initial rain screen, it but has TB10 < 175 K, suggesting only light rain. Light rain from a stratiform anvil could fit this description. An API value of 10 requires TB10 > 250 K along with the 85-GHz scattering. This would be heavy rain, but lacking strong vertical development—perhaps on the edge of a deeper core.

c. LIP

The section of the ER-2 Lightning Instrument Package on the NASA ER-2 aircraft used in this study consists of the seven rotating-vane electric field mills. Each mill incorporates self-calibration capabilities that reduce the time required to obtain a full aircraft calibration. A full description of the mills is contained in Bateman et al. (2007). The instruments were calibrated in the laboratory and on the aircraft using the technique in Mach and Koshak (2007). Once calibrated on the aircraft, the output of the mills are used to produce the vector electric field over a wide dynamic range extending from fair weather electric fields (i.e., a few to tens of V m−1) to large thunderstorm fields (i.e., tens of kV m−1). The vector electric field record provides knowledge of the electrical structure of storms overflown by the ER-2. The field mills also provide a measurement of the electric charge Q on the aircraft and information on total lightning (i.e., cloud-to-ground plus intracloud lightning) from the abrupt electric field changes in the vector field data.

Electric field data and lightning flash counts are considered here in conjunction with API. The electric field information does not pinpoint a precise location for charge centers or lightning flashes, but it responds to electrification in the vicinity (within tens of kilometers) of the flight path. Lightning flash counts from LIP are considered to be lower limits on the true flash counts, because weak flashes can only be identified when they are near the airplane. Weak flashes near the plane cannot be readily distinguished from stronger flashes farther away. Along with the cross-scanning AMPR, these measurements provide information about storm structure in the vicinity of the flight track.

d. GOES infrared data

The GOES-11 satellite, which had been an on-orbit spare before becoming the operational GOES-West satellite in 2006, was operated in rapid-scan mode for the Caribbean and eastern North Pacific during the Tropical Cloud Systems and Processes (TCSP) project. This study used only the channel 4 10.7-μm IR data to produce imagery as well as time series of TB per quadrant. The IR data were plotted for each of the rapid-scan times, which occurred at 5-, 10-, or 15-min intervals. The plotted data include only brightness temperatures (loosely interpreted as cloud-top temperatures) of 210–190 K (−63°C to −83°C). This was done to allow the color table to enhance the coldest cloud-top temperatures and to view small changes in cloud-top temperatures in and near the eyewall.

The regular 30-min GOES-East IR was also used for context. Time series were constructed for the period 15–18 July to characterize Hurricane Emily during its rapid deepening and subsequent filling. Mean and minimum Tb were considered as a function of radius (10-km bins) out to 500 km (Fig. 1) and as a function of azimuth (30° bins) within 50 (Fig. 2) and 100 km (not shown) of the center. The time versus radius plots confirm that a 50-km radius is sufficient for characterizing the coldest eyewall cloud tops. Higher resolution time series from the rapid-scan data are shown by Quinlan (2008).

3. Hurricane Emily eyewall convection

a. Satellite-based time series

Our analysis begins at the end of Emily’s initial intensification to a category 4 hurricane early on 15 July. Emily briefly weakened back to category 2 by 1800 UTC 15 July and then resumed intensification to category 5 by 0000 UTC 17 July (Fig. 1c). Slow weakening followed until landfall at 0630 UTC 18 July.

The most prominent feature of the mean IR time series in Fig. 1a is the diurnal cycle beyond about 200 km from the center. Similar to diurnal cycles shown by Muramatsu (1983), Steranka et al. (1984), and Kossin (2002), the peak in cold cloudiness shifts later in the day (from afternoon to evening) with increasing radius. Muramatsu (1983) interpreted this as an increase in convection near the center, followed by an outward surge of cirrus to large radius. Kossin (2002) noted that the diurnal signal is much weaker near the center and implicated radiationally driven subsidence as a reason the cloud canopy shrinks at night (with stronger subsidence in the clearer regions at larger radius). In Fig. 1a, azimuthal mean IR temperatures are consistently below 205 K in the region of the eyewall until after landfall on 18 July, with no obvious diurnal signal. There are several episodes with mean IR temperature below 200 K and minimum IR temperature (the coldest 4-km pixel, Fig. 1b) below 195 K. The minimum IR temperatures were often colder in an outer rainband around 200 km from the center than in the eyewall, particularly on 16 July.

Considering the azimuthal distribution of inner-core IR temperatures (Fig. 2), the first episode of colder cloud tops on 15 July was particularly symmetric. Mean Tb were below 200 K and minimum Tb below 195 K in all azimuths. This occurred while the central pressure was filling (to 969 hPa) after Emily’s first peak intensity. The pattern became more asymmetric later in the day, with colder cloud tops on the west side until a brief data outage from 2000 to 2230 UTC. A symmetric pattern resumed by 0000 UTC 16 July. Minimum cloud-top temperatures slowly warmed to 200 K by 1200 UTC 16 July, but the pattern remained symmetric until then. An asymmetry developed with colder cloud tops toward the east after 1200 UTC, but intensification continued until 0000 UTC 17 July when Emily reached 929 hPa.

Emily abruptly filled between 0000 and 0600 UTC 17 July and then continued to fill more slowly until landfall. Another episode of colder cloud tops with mean TB below 200 K occurred between 0600 and 1200 UTC, coinciding with the ER-2 flight. The cold cloud tops were not quite as symmetric as on 15 and 16 July, with slightly warmer TB toward the north-northwest. This is also seen in section 3b. Cloud tops warmed briefly after 1200 UTC and became a bit more asymmetric (warmer on the north side), but another round of colder cloud tops began around 1800 UTC. The coldest cloud tops (187 K) of this 4-day period were early on 18 July, around the time of landfall.

The ER-2 launched at 0600 UTC 17 July and made its first pass across the eye at 0752 UTC. Previous reconnaissance reports indicated that Emily’s pressure had bottomed out at 929 hPa near 0000 UTC and subsequently risen to 940 hPa by 0600 UTC (Fig. 1). GOES-11 infrared imagery indicated warming cloud tops between 0000 and 0600 UTC, particularly in the southwest and northwest quadrants (to be shown later). Low-earth orbit passive microwave imagers (Fig. 3) displayed a solid ring of convection, with 85-GHz brightness temperatures below 225 K throughout the eyewall. The lowest brightness temperatures (below 190 K) wrapped cyclonically from east through south in imagery at 0119 and 0657 UTC, with a slight weakness to the southeast. There was more of a weakness in the southern part of the eyewall at 1236 UTC, although this is exaggerated by the coarser sensor resolution in Fig. 3f compared to Figs. 3c,e.

b. First eyewall transit

The NASA ER-2 made two transits across the eye of Hurricane Emily on the morning of 17 July. The first eyewall transit occurred at about 0752 UTC directed from the east-southeast to the west-northwest (Fig. 4), about an hour after the image depicted in Fig. 3e. Relative to the patterns seen in Fig. 3, the ER-2 first passed the relative weakness in the southeast part of the eyewall and exited across the stronger part of the eyewall on the western side. The AMPR data in Figs. 5a,b similarly show a weakness on the southeast side with stronger convection on the north and west sides. As shown in Figs. 5a, 4a–c, the flight track was to the left (south and west) of the center by a few kilometers. The flight track was intended to cross the center, based on extrapolation of the earlier reconnaissance aircraft fixes and IR satellite fixes (which may have suffered some errors from parallax or geolocation problems). This demonstrates one challenge of conducting the mission without a scanning radar, which was available from a NOAA WP-3D during most other TCSP missions.

During this first transit, the ER-2 experienced turbulence at an altitude of 20 km. This turbulence was surprising because there had been substantial filling of the storm immediately prior to the ER-2 flight. Prior ER-2 flights over hurricanes that had already ceased intensification (e.g., Bonnie in 1998, Erin in 2001) had not been problematic. Although the ER-2 did not penetrate convection at ∼20-km altitude, we speculate that the difficulties may have been related to upward-propagating gravity waves, as observed in the lower stratosphere above a typhoon by Sato (1993). As seen in Fig. 4, a new cell was developing on the inner edge of the western eyewall just as the ER-2 was arriving. Cloud tops with this new cell are indistinguishable from the background at 0745 UTC (Fig. 4a) but have cooled 13 K by 0750 UTC (Fig. 4b). This cooling rate implies an upward motion of ∼5 m s−1 or more for the cloud top. Although the ER-2 missed the center of the eye, it did pass almost directly over this cell (Fig. 4c).

Figure 6a shows the reflectivity from the EDOP radar as it passed over the eye and eyewall. Adjacent stratiform echoes with the typical bright band, weak ascent above, and weak descent below were observed before and after the segment that is shown in Fig. 6. The EDOP reflectivities and vertical velocities have been corrected for attenuation in the same manner presented in Iguchi and Meneghini (1994), although some attenuation is still noticeable in the lower levels of the eyewall. EDOP data from aircraft rolls greater than 2° have been omitted (white columns) because they are not aligned with the rest of the cross section and introduce a nonvertical component to the Doppler velocities. These rolls also cause some sharp gradients in AMPR brightness temperatures (Fig. 6c).

The prominent feature in Fig. 6a is the convective cell that was shown in the GOES IR data, beginning at the ∼30-km distance marker. The echo tops in this cell reached to slightly more than 17 km, and echo tops had a maximum reflectivity of more than 60 dBZ at an altitude of 4 km at 0752 UTC. EDOP observed reflectivities of 40 dBZ or greater up to an altitude of 14.5 km and reflectivities of 50 dBZ or greater up to an altitude of about 8.3 km. Among all other EDOP measurements of tropical cyclones (Heymsfield et al. 2010), the highest 40 (50)-dBZ heights had been ∼10 (∼7) km in Hurricane Georges 1998 (Hurricane Bonnie 1998). This portion of the eyewall appears to slope outward about 25° from the vertical, although this may be complicated by the flight track not being oriented normal to the eyewall. There are two distinct reflectivity towers, separated by only ∼3 km.

Vertical motions (Fig. 6b) are quite strong for this portion of the eyewall, although interpretation is complicated by uncertainty in the hydrometeor fall speed retrieval, components of the horizontal wind being included when the radar beam points slightly off nadir (as a result of aircraft pitch or roll), and aircraft vertical motion. Aircraft motion is accounted for in the initial EDOP processing, but vertical accelerations were particularly large in this case and may have caused a problem. The combination of large updraft and downdraft values aligned in the same vertical profile increases our confidence that very little aircraft motion is aliased into these measurements of the eyewall. The strongest portion of the updraft was measured at 14.25-km altitude and at the 35.0-km distance marker, and it is restricted to upper levels (entirely above 10 km) in this cross section. The retrieved vertical velocity is more than 20 m s−1, but this is most likely only the top of the spiraling convective updraft (assuming that the updraft is rooted at low levels far upstream, perhaps in the east or southeast eyewall where the ice-scattering maximum in Fig. 3e begins). Because the EDOP only measures a slice of the storm, it only captured a portion of this convection. As in Heymsfield et al. (2001)’s study of Hurricane Bonnie (1998), there are also strong upper-level downdrafts immediately inward from this updraft. The peak upper-level downdraft is 16 m s−1 at x = 33.8 and 15.1-km altitude, slightly more than a kilometer away from the peak 24 m s−1 updraft. Below the main updraft, there are strong deep-layer downdrafts measuring greater than 20 m s−1 below 10-km altitude. Considering the high reflectivity values, precipitation loading must have contributed to these downdrafts. Only two other tropical cyclones seen by EDOP (Heymsfield et al. 2010) had stronger updrafts than this cross section, and only the second eyewall pass of Hurricane Emily had a stronger downdraft.

API values are 17 (Fig. 5a) for this portion of the eyewall (on the west and northwest sides), indicative of scattering by large ice for the 85-, 37-, and 19-GHz channels (Fig. 6c). Scattering by large ice particles reduces the brightness temperatures to ∼100 (85), ∼160 (37), and ∼230 K (19 GHz) compared to background values that are normally ∼270 K as a result of emission from heavy liquid rain. Averaging over the scale of a Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI) footprint yields ∼110–115 K at 85 GHz, comparable to the lowest values in a 3-yr survey of eyewalls seen by TMI (Cecil et al. 2002) but not as strong as some continental convection. Close inspection (not shown) also reveals a slight scattering signature in the 10-GHz channel. The electric fields recorded by LIP are strong (∼10 kV m−1), with 28 lightning flashes identified during this pass. This is the most electrically active tropical cyclone measured by LIP from the ER-2 (∼20 flights), and it ranks near the 90th percentile of the 850 storm overflights (tropical and subtropical convective systems, not limited to tropical cyclones or oceanic systems) examined by Mach et al. (2009).

The southeastern portion of the eyewall (toward the left in Fig. 6, between 0–10 km on the x axis) is has a weaker reflectivity structure in this pass, although a deep-layer updraft is depicted in Fig. 6b. (This is not truly a two-dimensional updraft sloping outward as the figure would suggest, because the stronger tangential flow is normal to the plane of this figure.) The retrieved upward motion exceeds 20 m s−1 at 10–11-km altitude, coinciding with an elevated layer of 35–40-dBZ reflectivity. Reflectivities are mostly below 40 dBZ at all levels. LIP responds to electrification in this portion of the eyewall but with electric fields much weaker (∼0.5 kV m−1) than on the other side. API values are mostly 8–9, indicating moderate-to-heavy rain with 85-GHz ice scattering but not enough large particles to substantially scatter radiation for the lower frequencies. This region shows the common condition in rain where the 10-, 19-, and 37-GHz channels all respond to emission from liquid drops, with higher brightness temperatures at progressively higher frequencies (Fig. 6c) until ice scattering reverses the trend. The infrared cloud-top temperatures (Fig. 4b) are actually colder (194 versus 197 K) for this southeastern region than for the intense convective tower on the western side.

The AMPR traces in Fig. 6c respond to the lower portions of the southern eyewall during large aircraft rolls (up to ∼9°); the EDOP data is omitted from Figs. 6a,b during these rolls. The rolls are accounted for in the AMPR traces by comparing the dotted and solid lines in Fig. 7. Between these rolls, the eye is mostly devoid of measurable radar reflectivity above 5-km altitude and has reflectivities ∼0–20 dBZ in the lowest 5 km. There is a thin layer of weak reflectivity (∼0 dBZ) at around 10-km altitude. Strong downward motion (>10 m s−1) is indicated in the lowest 2–3 km of the eye. This seems suspicious, and it may result from inaccurate parameterization of particle fall speeds, aliasing of the horizontal wind if the radar beam is not quite vertical, or aircraft vertical motions. If the weak reflectivity is caused by a small contribution of large drops, the terminal fall speed would be underestimated and velocity retrievals would be biased negative. Variations in aircraft pitch could alias ∼2 m s−1 of horizontal wind into the radar beam, if pointed near the low-level wind maximum. The ER-2 experienced altitude changes up to ±12 m s−1 during this flight segment, which may have contributed to some error in vertical velocity.

AMPR’s side-to-side scanning provides more spatial context for the EDOP cross section. The brightness temperatures are plotted in Fig. 7 for an average of the two middle AMPR footprints (looking in the same direction as EDOP; same as in Fig. 6c) and also for scan angles 0° below the plane (after accounting for aircraft roll) and 10° to the right of the flight path. The middle footprints (dotted line) and the 0° view (solid line) are essentially the same, except during aircraft rolls. During the two large rolls just before and after 0752 UTC, the middle footprints receive emissions from the eyewall to the left of the flight track. The 0° view below the flight track reveals decreased brightness temperatures inside the eye instead. The 10° right trace (dashed line) is roughly across the center of the eye, which the ER-2 appears to have missed by ∼2–4 km (based on examination of the AMPR data). This reveals that the eye (with low brightness temperatures at 10, 19, and 37 GHz between ∼0752 and 0753) is much broader than it appears in Fig. 6c. The lower brightness temperatures to the right of the flight track in the eye also suggest that the eye may have been rain free at its center. There is a 12-km segment with 37-GHz brightness temperatures below 210 K in the eye; the brightness temperature maxima from eyewall rain are 19 km apart. (Maxima are further apart for the lower 10- and 19-GHz channels, but they have lower horizontal resolution that does not capture sharp gradients.) The 85-GHz minima from ice scattering aloft are 26 km apart.

c. Second eyewall transit

The ER-2 made its second transit above the eyewall at about 0845 UTC (Fig. 8), passing from southwest to northeast. The coldest cloud tops (∼193 K) were located south of the center (Fig. 8), with the ER-2 passing above ∼195-K infrared cloud tops on the southwest side. Close inspection of the AMPR data (Fig. 5b) indicates that the flight track was again slightly to the left (west and north) of the center of the eye. The EDOP reflectivity for the second transit of the eye is shown in Fig. 9a. As in Fig. 6, adjacent stratiform regions were observed but not shown here, and some problems with attenuation are apparent beneath the strongest echoes. The left side of the image is to the southwest side of the eye, and the right edge is to the northeast side where the coldest cloud tops were 199 K. The reflectivity tops (approximately −10 dBZ) are ∼1–2-km higher on the southwest side. Both the southwest and northeast sides of the eyewall feature strong convection but not as strong as that seen on the western side in Fig. 6a. Reflectivity of 30 dBZ reaches ∼11 km on both sides, with 40 dBZ reaching 6–7 km. Peak retrieved updraft speeds (Fig. 9b) approach 20 m s−1 at upper levels (∼11 km) on the southwest side (x = 10.4 km) and 15 m s−1 at midlevels (∼5 km) on the northeast side (x = 32.0 km). The 20 m s−1 updraft speed in this cross section is a bit more questionable than the others, as the entire column around x = 10 km has more positive vertical velocities than adjacent columns. Perhaps some fraction of aircraft vertical motion is slipping through the filters. Strong downdrafts are noted adjacent to both updrafts, both on the inward and outward sides. The AMPR brightness temperatures associated with this cross section are only scattered to ∼160 at 85 and ∼220 K at 37 GHz (Fig. 9c), suggestive of smaller ice than in the first pass.

A thick layer of 10–20-dBZ reflectivity is seen in the eye at 5–10-km altitude, separated from a shallow rain layer below 3 km. This is in contrast to the very thin layer of weak reflectivity in the upper portion of the eye during the previous pass. Animations of infrared imagery (not shown) depict the eye clouding over during the ER-2 flight, after being quite distinct for several hours prior. Most of that mid-to-upper-level echo layer has several meters-per-second downward motion, but the upper edges (with reflectivity generally less than 10 dBZ) has a few meters-per-second upward motion. Farther southwest from this cross section is a distinct outflow layer of thin cirrus clouds at about 15.5-km altitude (not shown). This has reflectivities above 0 dBZ for only a few hundred meters depth, above a ∼3-km gap of weaker reflectivity and no echo. The deep stratiform layer below this tops out at ∼12 km.

As with the first pass across the eye, AMPR swath data indicate the center of the eye was to the right (southeast) of the flight track. Eye brightness temperatures (20° to the right of the flight track; dashed lines in Fig. 10) are slightly lower in the low-frequency channels for the second pass than the first, suggesting a possibility of a rain-free center to the right of the flight track. The 37-GHz maxima are about 18 km apart and the 85-GHz minima about 28 km apart, similar to the diameters from the first pass. Without substantial aircraft rolls in this flight segment, the solid and dotted lines in Fig. 10 are nearly equal (i.e., the middle of the AMPR swath is almost directly beneath the aircraft when the roll is zero). The ice-scattering signatures are not as strong for the second pass, comparing the eyewall brightness temperature depressions in Fig. 10 to those in Fig. 7. This is consistent with the upper-level radar reflectivities in Fig. 9 not being as high as those in the western eyewall of Fig. 6. Lower AMPR brightness temperatures are seen to the sides (not shown here) but not quite as low as in the first pass. The first pass had 85-GHz brightness temperatures reduced to 100 K, but the lowest values during the second pass were ∼125 K (again in the western portion of the eyewall, in parts of the swath not shown in these figures). The electric fields measured by LIP are also not as strong in the second pass (2–5 kV m−1) as in the first (10 kV m−1), and fewer lightning flashes (20 versus 28) are detected.

d. Box patterns outside the eyewall

After the two passes across the eye, straight flight legs were set up adjacent to (but outside) the eyewall. Two of these box patterns were executed (Figs. 5c,d). Although the flight track was kept away from the eyewall for safety, LIP did record substantial electric fields when close enough to the eyewall (e.g., 0940 and 1010 UTC in Fig. 5c; 1040 and 1050 UTC in Fig. 5d.) ∼100 lightning flashes were detected by LIP during these two box patterns. During the second box pattern, the ER-2 pilot felt safe enough to shift the northern leg a bit closer to the eyewall. The AMPR brightness temperatures do reveal strong convection to the left of the plane in the northern eyewall, with the 85-GHz channel reduced below 145 K and API values of 17–18.

4. Synthesis and conclusions

It appears that the ER-2 flight in general, and the first leg across the eye in particular, are examples of being in the right place at the right time (or the wrong place at the wrong time, depending on one’s perspective). Cloud tops in the western portion of Emily’s inner core had been warming between the times of Emily’s peak intensity (∼0000 UTC 17 July) and the beginning of the ER-2 mission (∼0600 UTC 17 July). Episodic bursts of strong, deep convection began around the time the ER-2 mission began.

The northeast and southeast quadrants had been experiencing low infrared brightness temperatures during the hours leading up to the ER-2 flight, with minimum values typically around 198 K. Cloud tops were warming in the northwest and southwest quadrants, coincident with the filling surface pressures measured by Air Force Reserve reconnaissance. Near the time of the ER-2 launch (∼0600 UTC), episodes of deep convection with cold cloud tops ∼194–198 K began breaking out in the northwest quadrant and quickly advecting into the southwest quadrant. The first of these occurred around 0600 UTC and another around 0900 UTC (near the time of the second ER-2 pass). Although these cold cloud tops generally made their first appearance in the northwest quadrant, the cold cloud shield expanded while wrapping around the southwest and southeast quadrants. The areal-mean brightness temperatures were therefore lower for the southern than the northern parts of the storm (Fig. 2a) after these bursts had developed.

Passive microwave imagery, more responsive to midlevel precipitation ice than to cloud top, showed stronger signatures cyclonically wrapped from east through south (Figs. 3, 5). The strongest scattering signatures were on the north and west sides. Together with the infrared measurements, this suggests the updraft bursts were rooted toward the east but did not reach near the tropopause level until rotating around to the northwest and west sides. Their cold cloud tops expanded as they reached the southern side, giving the appearance in infrared imagery of stronger convection on the southern side of the storm.

The ER-2’s first pass was aligned slightly off center for Hurricane Emily’s eye but almost directly across a new cell growing rapidly in the western eyewall. GOES rapid-scan imagery revealed cloud tops 13 K cooler for this cell at 0750 UTC than at 0745 UTC, with the ER-2 topping the cell at 0753 UTC (Fig. 6). Considering the flight track and 5-min IR imagery in Figs. 4a–c, this strong convective tower might have been underestimated or missed entirely if the transect had occurred a few minutes earlier, and/or if the flight track had been more successfully aligned across the center of the eye. Strong updrafts were seen at upper levels (∼14 km), with strong downdrafts at midlevels. This is consistent with the updraft burst having originated farther upstream and rotating around the eyewall. The location of coldest infrared cloud tops can be rather misleading as an indicator of which part of the eyewall has the strongest convection.

The ER-2 measurements include radar reflectivity towers with 40 dBZ reaching 14-km altitude, 85-GHz brightness temperatures near 100 K, and 28 lightning flashes detected from the aircraft. In most regards, these were the strongest convective signatures seen by these instruments in ∼20 ER-2 tropical cyclone missions. Stronger signatures have occasionally been recorded in non-TC missions, generally convection over or near land. The measurements from Emily’s western eyewall also appear consistent with—but not more extreme than—some of the strongest hurricane eyewall cases seen by similar instruments on the TRMM satellite (e.g., 85-GHz brightness temperature at ∼110–115 K at TRMM footprint size). Convective signatures were much less noteworthy for the southeastern eyewall during this pass. The second pass at 0845 UTC (Fig. 9) also showed very strong convection (on the southwest and northeast sides), but signatures were not as strong as for the first pass.

The minimum IR cloud-top temperature time series (Figs. 1b, 2b) suggests that the episodic nature for deep, strong convection generally continued until Emily made landfall on the Yucatan Peninsula near 0600 UTC 18 July. Two particular episodes around 2100 UTC 17 July and 0300 UTC 18 July produced substantially colder cloud tops (∼190 K) than during the ER-2 mission. Earlier in Emily’s life cycle, similarly cold cloud tops were seen around 0900 UTC 15 July while Emily was filling after an initial 952-hPa peak intensity at 0600 UTC. The 1200 UTC 15 July reconnaissance mission reported concentric eyewalls, and the pressure had risen to 965 hPa (Franklin and Brown 2006). A key difference between 15–16 and 17–18 July is that the cold cloud tops were more symmetric between about 0900 UTC 15 July and 1200 UTC 16 July. Substantial deepening began during this period and continued until 0000 UTC 17 July. The episodes of strong convection on 17–18 July were more asymmetric, and Emily slowly weakened until landfall.

To an extent, the strong convection soon after Emily’s maximum intensity is reminiscent of the “maximum intensity” lightning outbreaks in category 5 Hurricanes Katrina and Rita later in 2005, described by Squires and Businger (2008). Molinari et al. (1999) suggested that eyewall lightning outbreaks in strong storms that have already gone through an intensification period may signal the end of intensification or the onset of weakening. However, the observations here were several hours after maximum intensity and after weakening had begun. The IR trends suggest that this strong convection was not continually occurring from the time of maximum intensity onward. Instead, there is a temporal gap with warming cloud-top temperatures on the western side for ∼6 h after maximum intensity. Some other observations of strong hurricanes [e.g., Dodge et al.’s (1999) study of category 5 Hurricane Gilbert] also show very strong updrafts after peak intensity. The bursts of strong convection during the ER-2 flight ushered in a ∼24-h period with cloud-top temperatures that were colder than during Emily’s maximum intensity.

Some prior studies link strong convection to tropical cyclone intensification regardless of symmetry or asymmetry (e.g., Kelley et al. 2004, 2005). Squires and Businger (2008) noted that some of their lightning outbreaks during rapid intensification and maximum intensity were symmetric, whereas others were asymmetric. Steranka et al. (1986) factored in symmetry by averaging infrared temperatures over a large area. Dvorak (1984) required a closed ring with a given brightness temperature for that temperature to be used in the intensity analysis. The strong convection in Hurricane Emily during this case study was asymmetric, consistent with moderate shear from the south and southwest. We suspect that this lack of symmetry in the eyewall convection contributed to Emily’s gradual weakening during this 17 July.

Acknowledgments

Besides acknowledging the entire TCSP team for successfully executing the project, we must particularly thank the ER-2 pilot (Dave Wright) and coordinator/communicator (Jan Nystrom) for completing this difficult flight. Gerry Heymsfield, Lin Tian, and Lihua Li (NASA GSFC; EDOP team) and Frank LaFontaine (NASA MSFC/Raytheon; AMPR team) provided the EDOP and AMPR data, and helpful discussions. Chris Velden and colleagues at the University of Wisconsin–CIMSS provided the GOES-11 rapid-scan data and helped with using it. Half-hourly GOES data were provided via the NASA TRMM program and NOAA Climate Prediction Center. Satellite microwave imagery is courtesy of the Naval Research Laboratory—Monterey tropical cyclone Web page. Reviews by Ed Zipser and anonymous reviewers helped improve earlier versions of this manuscript. This study was supported by the NASA TCSP program though Grant NNG05GR57G.

REFERENCES

  • Bateman, M. G., , M. F. Stewart, , R. J. Blakeslee, , S. J. Podgorny, , H. J. Christian, , D. M. Mach, , J. C. Bailey, , and D. Daskar, 2007: A low-noise, microprocessor-controlled, internally digitizing rotating-vane electric field mill for airborne platforms. J. Atmos. Oceanic Technol., 24 , 12451255.

    • Search Google Scholar
    • Export Citation
  • Beven II, J. L., and Coauthors, 2008: Atlantic hurricane season of 2005. Mon. Wea. Rev., 136 , 11091173.

  • Black, M. L., , R. W. Burpee, , and F. D. Marks Jr., 1996: Vertical motion characteristics of tropical cyclones determined with airborne Doppler radial velocities. J. Atmos. Sci., 53 , 18871909.

    • Search Google Scholar
    • Export Citation
  • Cecil, D. J., , E. J. Zipser, , and S. W. Nesbitt, 2002: Reflectivity, ice scattering, and lightning characteristics of hurricane eyewalls and rainbands. Part I: Quantitative description. Mon. Wea. Rev., 130 , 769784.

    • Search Google Scholar
    • Export Citation
  • DeMaria, M., , M. Mainelli, , L. K. Shay, , J. A. Knaff, , and J. Kaplan, 2005: Further improvements to the Statistical Hurricane Intensity Prediction Scheme (SHIPS). Wea. Forecasting, 20 , 531543.

    • Search Google Scholar
    • Export Citation
  • Dodge, P., , R. W. Burpee, , and F. D. Marks Jr., 1999: The kinematic structure of a hurricane with sea level pressure less than 900 mb. Mon. Wea. Rev., 127 , 9871004.

    • Search Google Scholar
    • Export Citation
  • Dvorak, V. E., 1984: Tropical cyclone intensity analysis using satellite data. NOAA Tech. Rep. NESDIS 11, 47 pp.

  • Franklin, J. L., , and D. P. Brown, 2006: Tropical cyclone report: Hurricane Emily. National Hurricane Center, 18 pp. [Available online at http://www.nhc.noaa.gov/pdf/TCR-AL052005_Emily.pdf].

    • Search Google Scholar
    • Export Citation
  • Heymsfield, G. M., 1989: Accuracy of vertical air motions from nadir-viewing Doppler airborne radars. J. Atmos. Oceanic Technol., 6 , 10791082.

    • Search Google Scholar
    • Export Citation
  • Heymsfield, G. M., and Coauthors, 1996: The EDOP radar system on the high-altitude NASA ER-2 aircraft. J. Atmos. Oceanic Technol., 13 , 795809.

    • Search Google Scholar
    • Export Citation
  • Heymsfield, G. M., , J. B. Halverson, , and I. J. Caylor, 1999: A wintertime Gulf Coast squall line observed by EDOP airborne doppler radar. Mon. Wea. Rev., 127 , 29282950.

    • Search Google Scholar
    • Export Citation
  • Heymsfield, G. M., , J. B. Halverson, , J. Simpson, , L. Tian, , and T. P. Bui, 2001: ER-2 Doppler radar investigations of the eyewall of Hurricane Bonnie during the Convection and Moisture Experiment-3. J. Appl. Meteor., 40 , 13101330.

    • Search Google Scholar
    • Export Citation
  • Heymsfield, G. M., , L. Tian, , A. J. Heymsfield, , L. Li, , and S. Guimond, 2010: Characteristics of deep tropical and subtropical convection from nadir-viewing high-altitude airborne Doppler radar. J. Atmos. Sci., 67 , 285308.

    • Search Google Scholar
    • Export Citation
  • Hood, R. E., and Coauthors, 2006: Classification of tropical oceanic precipitation using high-altitude aircraft microwave and electric field measurements. J. Atmos. Sci., 63 , 218233.

    • Search Google Scholar
    • Export Citation
  • Iguchi, T., , and R. Meneghini, 1994: Intercomparison of single-frequency methods for retrieving a vertical rain profile from airborne or spaceborne radar data. J. Atmos. Oceanic Technol., 11 , 15071516.

    • Search Google Scholar
    • Export Citation
  • Kelley, O. A., , J. Stout, , and J. B. Halverson, 2004: Tall precipitation cells in tropical cyclone eyewalls are associated with tropical cyclone intensification. Geophys. Res. Lett., 31 , L24112. doi:10.1029/2004GL021616.

    • Search Google Scholar
    • Export Citation
  • Kelley, O. A., , J. Stout, , and J. B. Halverson, 2005: Hurricane intensification detected by continuously monitoring tall precipitation in the eyewall. Geophys. Res. Lett., 32 , L20819. doi:10.1029/2005GL023583.

    • Search Google Scholar
    • Export Citation
  • Kossin, J. P., 2002: Daily hurricane variability inferred from GOES infrared imagery. Mon. Wea. Rev., 130 , 22602270.

  • Lyons, W. A., , and C. S. Keen, 1994: Observations of lightning in convective supercells within tropical storms and hurricanes. Mon. Wea. Rev., 122 , 18971916.

    • Search Google Scholar
    • Export Citation
  • Mach, D. M., , and W. J. Koshak, 2007: General matrix inversion technique for the calibration of electric field sensor arrays on aircraft platforms. J. Atmos. Oceanic Technol., 24 , 15761587.

    • Search Google Scholar
    • Export Citation
  • Mach, D. M., , R. J. Blakeslee, , M. G. Bateman, , and J. C. Bailey, 2009: Electric fields, conductivity, and estimated currents from aircraft overflights of electrified clouds. J. Geophys. Res., 114 , D10204. doi:10.1029/2008JD011495.

    • Search Google Scholar
    • Export Citation
  • Molinari, J., , P. K. Moore, , V. P. Idone, , R. W. Henderson, , and A. B. Saljoughy, 1994: Cloud-to-ground lightning in Hurricane Andrew. J. Geophys. Res., 99 , 1666516676.

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

    • Search Google Scholar
    • Export Citation
  • Muramatsu, T., 1983: Diurnal variations of satellite-measured TBB areal distribution and eye diameter of mature typhoons. J. Meteor. Soc. Japan, 61 , 7790.

    • Search Google Scholar
    • Export Citation
  • Quinlan, K. R., 2008: Conditions influencing Hurricane Emily’s (2005) precipitation patterns, convection, and upper tropospheric outflow. M.S. thesis, Department of Atmospheric Science, University of Alabama in Huntsville, 116 pp.

  • Sato, K., 1993: Small-scale wind disturbances observed by the MU radar during the passage of Typhoon Kelly. J. Atmos. Sci., 50 , 518537.

    • Search Google Scholar
    • Export Citation
  • Spencer, R. W., , R. E. Hood, , F. J. Lafontaine, , E. A. Smith, , R. Platt, , J. Galliano, , V. L. Griffin, , and E. Lobl, 1994: High-resolution imaging of rain systems with the Advanced Microwave Precipitation Radiometer. J. Atmos. Oceanic Technol., 11 , 849857.

    • Search Google Scholar
    • Export Citation
  • Squires, K., , and S. Businger, 2008: The morphology of eyewall lightning outbreaks in two category 5 hurricanes. Mon. Wea. Rev., 136 , 17061726.

    • Search Google Scholar
    • Export Citation
  • Steranka, J., , E. B. Rodgers, , and R. C. Gentry, 1984: The diurnal variation of Atlantic Ocean tropical cyclone cloud distribution inferred from geostationary satellite infrared measurements. Mon. Wea. Rev., 112 , 23382344.

    • Search Google Scholar
    • Export Citation
  • Steranka, J., , E. B. Rodgers, , and R. C. Gentry, 1986: The relationship between satellite measured convective bursts and tropical cyclone intensification. Mon. Wea. Rev., 114 , 15391546.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Hovmöller diagrams as a function of distance (10-km bins) from Hurricane Emily’s center for (a) mean IR TB and (b) minimum IR TB. Contours are at 5-K intervals, but the shading differs for (a) and (b) because (a) covers a broader range of values. (c) Best-track maximum sustained wind (solid) and minimum surface pressure (dashed), with the time of the ER-2 observations marked.

Citation: Monthly Weather Review 138, 3; 10.1175/2009MWR3063.1

Fig. 2.
Fig. 2.

Hovmöller diagrams as a function of azimuth (30° bins) around Hurricane Emily’s center, within 50-km radius. (a) Mean IR TB using only pixels at or below 220 K (to avoid including the eye). (b) Minimum IR TB. (c) Best-track maximum sustained wind (solid) and minimum surface pressure (dashed), with the time of the ER-2 observations marked.

Citation: Monthly Weather Review 138, 3; 10.1175/2009MWR3063.1

Fig. 3.
Fig. 3.

Sequence of 85-GHz horizontal channel imagery [89 GHz for Advanced Microwave Scanning Radiometer for Earth Observing System (AMSR-E)] from the Navy Research Laboratory (NRL)—Monterey tropical cyclone Web page: (a) 1708 UTC 16 Jul TMI, (b) 1845 UTC 16 Jul AMSR-E, (c) 0119 UTC 17 Jul TMI, (d) 0159 UTC 17 Jul SSMI, (e) 0657 UTC 17 Jul AMSR-E, (f) 1236 UTC 17 Jul SSMI. Note that sensor resolution varies between instruments.

Citation: Monthly Weather Review 138, 3; 10.1175/2009MWR3063.1

Fig. 4.
Fig. 4.

GOES-11 IR temperatures at (a) 0745, (b) 0750, and (c) 0755 UTC 17 Jul with ER-2 flight track overlaid. Flight tracks are for five minutes, centered on the time of each satellite image.

Citation: Monthly Weather Review 138, 3; 10.1175/2009MWR3063.1

Fig. 5.
Fig. 5.

API (shaded) and LIP electric field vectors. The projections of the electric field are plotted as barbs originating at the aircraft location, pointing away from positive charge, for (a) 0735–0833; (b) 0833–0910; (c) 0936–1030; (d) 1030–1128 UTC.

Citation: Monthly Weather Review 138, 3; 10.1175/2009MWR3063.1

Fig. 6.
Fig. 6.

(a) Nadir EDOP reflectivity in an east-southeast–west-northwest direction across eye of Hurricane Emily from 0751 to 0755 UTC 17 Jul. Contour interval (CI) is 10 dB. (b) Nadir vertical velocity; positive values indicate upward motion. CI is 5, with 0 m s−1 (bold). Data have been removed from (a) and (b) for ER-2 roll variations greater than ±2°. (c) AMPR TB from middle of swath (includes rolls). For reference with Figs. 4, 5, the area extends from 17.76°N, 81.51°W to 17.97°N, 81.93°W.

Citation: Monthly Weather Review 138, 3; 10.1175/2009MWR3063.1

Fig. 7.
Fig. 7.

Trace of AMPR TB (K) along the aircraft flight track (solid gray line), normal to the aircraft (influenced by aircraft rolls, same as in Fig. 6c; dotted line), and 10° to the right of the flight track (across the center of the eye; dashed line) for (a) 10, (b) 19, (c) 37, (d) 85 GHz.

Citation: Monthly Weather Review 138, 3; 10.1175/2009MWR3063.1

Fig. 8.
Fig. 8.

GOES-11 IR temperatures at 0845 UTC Jul 17 with ER-2 flight path from 0842:30 to 0847:30 UTC overlaid.

Citation: Monthly Weather Review 138, 3; 10.1175/2009MWR3063.1

Fig. 9.
Fig. 9.

(a) Nadir EDOP reflectivity in a southwest–northeast direction across eye of Hurricane Emily from 0844 to 0848 UTC 17 Jul. CI is 10 dB. (b) Nadir EDOP vertical velocity; positive values indicate upward motion. CI is 5, with 0 m s−1 (bold). (c) AMPR brightness temperatures from middle of swath (includes rolls). For reference with Figs. 4, 8, the area extends from 17.86°N, 82.09°W to 18.22°N, 81.83°W.

Citation: Monthly Weather Review 138, 3; 10.1175/2009MWR3063.1

Fig. 10.
Fig. 10.

Trace of AMPR TBs (K) along the aircraft flight track (solid line), normal to the aircraft (influenced by aircraft rolls, same as in Fig. 9c; dotted line), and 20° to the right of the flight track (across the center of the eye; dashed line) for (a) 10, (b) 19, (c) 37, (d) 85 GHz.

Citation: Monthly Weather Review 138, 3; 10.1175/2009MWR3063.1

Save