• Atlas, D., R. C. Srivastava, and R. S. Sekhon, 1973: Doppler radar characteristics of precipitation at vertical incidence. Rev. Geophys. Space Phys.,11, 1–35.

  • Balsley, B. B., and K. S. Gage, 1982: On the use of radars for operational wind profiling. Bull. Amer. Meteor. Soc.,63, 1009–1018.

  • Battan, L. J., 1973: Radar Observation of the Atmosphere. University of Chicago Press, 324 pp.

  • Carter, D. A., K. S. Gage, W. L. Ecklund, W. M. Angevine, P. E. Johnston, A. C. Riddle, J. S. Wilson, and C. R. Williams, 1995:Developments in UHF lower tropospheric wind profiling at NOAA’s Aeronomy Laboratory. Radio Sci.,30, 977–1001.

  • Crane, R. K., 1980: A review of radar observations of turbulence in the lower stratosphere. Radio Sci.,15, 177–193.

  • Currier, P. E., S. K. Avery, B. B. Balsley, K. S. Gage, and W. L. Ecklund, 1992: Combined use of the 50 MHz and 915 MHz wind profilers in the estimation of raindrop size distributions. Geophys. Res. Lett.,19, 1017–1020.

  • Ecklund, W. L., D. A. Carter, and B. B. Balsley, 1988: A UHF wind profiler for the boundary layer: Brief description and initial results. J. Atmos. Oceanic Technol.,5, 432–441.

  • ——, ——, ——, P. E. Currier, J. L. Green, B. L. Weber, and K. S. Gage, 1990: Field tests of a lower tropospheric wind profiler, Radio Sci.,25, 899–906.

  • ——, K. S. Gage, and C. R. Williams, 1995: Tropical precipitation studies using a 915-MHz wind profiler. Radio Sci.,30, 1055–1064.

  • ——, C. R. Williams, P. E. Johnston, and K. S. Gage, 1999: A 3-GHz Profiler for precipitating cloud studies. J. Atmos. Oceanic Technol.,16, 309–322.

  • Gage, K. S., 1990: Radar observation of the free atmosphere: Structure and dynamics. Radar in Meteorology, D. Atlas, Ed., Amer. Meteor. Soc., 534–565.

  • ——, and B. B. Balsley, 1978: Doppler radar probing of the clear atmosphere. Bull. Amer. Meteor. Soc.,59, 1074–1093.

  • ——, and ——, 1980: On the scattering and reflection mechanisms contributing to clear air radar echoes from the troposphere, stratosphere, and mesosphere. Radio Sci.,15, 243–257.

  • ——, C. R. Williams, and W. L. Ecklund, 1994: UHF wind profilers:A new tool for diagnosing tropical convective cloud systems. Bull. Amer. Meteor. Soc.,75, 2289–2294.

  • ——, ——, and ——, 1996: Application of the 915 MHz profiler for diagnosing and classifying tropical precipitating cloud systems. Meteor. Atmos. Phys.,59, 141–151.

  • Gossard, E. E., 1988: Measuring drop-size distributions in clouds with a clear-air-sensing Doppler radar. J. Atmos. Oceanic Technol.,5, 640–649.

  • ——, and R. G. Strauch, 1981: The refractive index spectra within clouds from forward-scatter radar observations. J. Appl. Meteor.,20, 170–183.

  • ——, R. B. Chadwick, T. R. Detman, and J. E. Gaynor, 1984a: Capability of surface-based clear-air Doppler radar for monitoring meteorological structure of elevated layers. J. Climate Appl. Meteor.,23, 474–485.

  • ——, W. D. Neff, R. J. Zamora, and J. E. Gaynor, 1984b: The fine structure of elevated refractive layers: Implications for over-the-horizon propagation and radar sounding systems. Radio Sci.,19, 1523–1533.

  • ——, R. G. Strauch, and R. R. Rogers, 1990: Evolution of dropsize distribution in liquid precipitation observed by ground-based Doppler radar. J. Atmos. Oceanic Technol.,7, 815–828.

  • ——, ——, D. C. Welsh, and S. Y. Matrosov, 1992: Cloud layers, particle identification, and rain-rate profiles from ZRVf measurements by clear-air Doppler radars. J. Atmos. Oceanic Technol.,9, 108–119.

  • Hardy, K. R., and K. S. Gage, 1990: History of radar studies of the clear atmosphere. Radar in Meteorology, D. Atlas, Ed., Amer. Meteor. Soc., 130–142.

  • Hill, R. J., 1978: Spectra of fluctuations in refractivity, temperature, humidity, and the temperature-humidity cospectrum in the inertial and dissipation ranges. Radio Sci.,13, 953–961.

  • ——, and S. F. Clifford, 1978: Modified spectrum of atmospheric temperature fluctuations and its application to optical propagation. J. Opt. Soc. Amer.,68, 892–899.

  • Houze, R. A., Jr., 1993: Cloud Dynamics. Academic Press, 573 pp.

  • Keenan, T., R. Carbone, S. Rutledge, J. Wilson, G. Holland, and P. May: 1996: The Maritime Continent Thunderstorm Experiment (MCTEX): Overview and initial results. Preprints, Seventh Conf. on Mesoscale Processes, Reading, United Kingdom, 326–328.

  • Knight, C. A., and L. J. Miller, 1993: First radar echoes from cumulus clouds. Bull. Amer. Meteor. Soc.,74, 179–188.

  • Ottersten, H., 1969: Atmospheric structure and radar backscattering in clear air. Radio Sci.,4, 1179–1193.

  • Parsons, D., and Coauthors, 1994: The integrated sounding system: Description and preliminary observations from TOGA COARE. Bull. Amer. Meteor. Soc.,75, 553–567.

  • Rajopadhyaya, D. K., P. T. May, R. C. Cifelli, S. K. Avery, C. R. Williams, W. L. Ecklund, and K. S. Gage, 1998: The effect of vertical air motions on rain rates and median volume diameter determined from combined UHF and VHF wind profiler measurements and comparisons with rain gauge measurements. J. Atmos. Oceanic Technol.,15, 1306–1319.

  • Ralph, F. M., 1995: Using radar-measured radial vertical velocities to distinguish precipitation scattering from clear-air scattering. J. Atmos. Oceanic Technol.,12, 257–267.

  • Rogers, R. R., and M. K. Yau, 1989: A Short Course in Cloud Physics. 3d ed. Butterworth-Heinemann, 290 pp.

  • ——, and W. O. J. Brown, 1997: Radar observations of a major industrial fire. Bull. Amer. Meteor. Soc.,78, 803–814.

  • ——, C. A. Knight, J. D. Tuttle, W. L. Ecklund, D. A. Carter, and S. A. Ethier, 1992: Radar reflectivity of the clear air at wavelengths of 5.5 and 33 cm. Radio Sci.,27, 645–659.

  • ——, W. L. Ecklund, D. A. Carter, K. S. Gage, and S. A. Ethier, 1993a: Research applications of a boundary-layer wind profiler. Bull. Amer. Meteor. Soc.,74, 567–580.

  • ——, D. Baumgardner, S. A. Ethier, D. A. Carter, and W. L. Ecklund, 1993b: Comparison of raindrop size distributions measured by radar wind profiler and by airplane. J. Appl. Meteor.,32, 694–699.

  • Tatarskii, V. I., 1971: The effects of the turbulent atmosphere on wave propagation. National Technical Information Service, U.S. Dept. of Commerce, TT 68-50464, 472 pp. [Available from NTIS, 5285 Port Royal Rd., Springfield, VA 22161.].

  • VanZandt, T. E., J. L. Green, K. S. Gage, and W. L. Clark, 1978: Vertical profiles of refractivity turbulence structure constant: Comparison of observations by the Sunset radar with a new theoretical model. Radio Sci.,13, 819–829.

  • Wakasugi, K., A. Mizutani, M. Matsuo, S. Fukao, and S. Kato, 1986:A direct method for deriving drop-size distributions and vertical air velocities from VHF Doppler radar spectra. J. Atmos. Oceanic Technol.,3, 623–629.

  • ——, ——, ——, ——, and ——, 1987a: Further discussion on deriving drop-size distribution and vertical air velocities directly from VHF Doppler radar spectra. J. Atmos. Oceanic Technol.,4, 170–179.

  • ——, B. B. Balsley, and T. L. Clark, 1987b: The VHF Doppler radar as a tool for cloud and precipitation studies. J. Atmos. Oceanic Technol.,4, 273–280.

  • Wilczak, J. M., and Coauthors, 1995: Contamination of wind profiler data by migrating birds: Characteristics of corrupted data and potential solutions. J. Atmos. Oceanic Technol.,12, 449–467.

  • Williams, C. R., W. L. Ecklund, and K. S. Gage, 1995: Classification of precipitating clouds in the tropics using 915-MHz wind profilers. J. Atmos. Oceanic Technol.,12, 996–1012.

  • Wilson, J. W., T. M. Weckwerth, J. Vivekanandan, R. M. Wakimoto, and R. W. Russell, 1994: Boundary layer clear-air radar echoes:Origin of echoes and accuracy of derived winds. J. Atmos. Oceanic Technol.,11, 1184–1206.

  • View in gallery

    Map of the Tiwi Islands in Australia, which were the site of the MCTEX Field Campaign.

  • View in gallery

    The 2835-MHz profiler shroud and antenna (left) and 920-MHz profiler antenna (right) at Garden Point, Australia, during MCTEX.

  • View in gallery

    Detection and saturation thresholds for the 2835- and 920-MHz profilers operated in the high height mode (500-m resolution).

  • View in gallery

    Sample spectra from 2 km recorded during MCTEX on 28 Nov 1995 for the 920-MHz profiler and the 2835-MHz profiler during clear conditions (top), stratiform rain (middle), and convective rain (bottom).

  • View in gallery

    Time–height sections of equivalent reflectivity for the 920-MHz profiler (top) and for the 2835-MHz profiler (middle). Time–height section of equivalent reflectivity difference (920–2835 MHz) (bottom) at Garden Point, Australia, on 3 Dec 1995.

  • View in gallery

    Frequency distributions of (a) 920-MHz equivalent reflectivity, (b) 2835-MHz equivalent reflectivity, and (c) equivalent reflectivity difference for observations at Garden Point, Australia, during MCTEX (13 Nov–7 Dec 1995). The panels on the left indicate altitude distribution of observations used in the distributions.

  • View in gallery

    Two-dimensional frequency distributions at 2 km of (a) 920-MHz equivalent reflectivity vs equivalent reflectivity difference (920–2835 MHz), (b) 920-MHz Doppler velocity vs equivalent reflectivity difference (920–2835 MHz), and (c) 920-MHz spectral width vs equivalent reflectivity difference (920–2835 MHz) for observations at Garden Point, Australia, during MCTEX (13 Nov–7 Dec 1995). The panels on the left show one-dimensional probability distributions of 920-MHz equivalent reflectivity, Doppler velocity, and spectral width, respectively; (bottom) the probability distribution of the equivalent reflectivity different.

  • View in gallery

    Scatterplot of 2835-MHz equivalent reflectivity at 2 km vs 920-MHz equivalent reflectivity at 2 km (left) and frequency distribution of 2835-MHz equivalent reflectivity at 2 km vs 920-MHz equivalent reflectivity at 2 km (right) for observations at Garden Point, Australia, during MCTEX (13 Nov–7 Dec 1995). The dashed lines indicate Rayleigh scatter and the solid lines indicate Bragg scatter.

  • View in gallery

    Scatterplot of 2835-MHz Doppler velocity at 2 km vs 920-MHz Doppler velocity at 2 km (left) and frequency distribution of 2835-MHz Doppler velocity at 2 km vs 920-MHz Doppler velocity at 2 km (right) for observations at Garden Point, Australia, during MCTEX (13 Nov–7 Dec 1995).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 153 153 7
PDF Downloads 29 29 7

Use of Two Profilers during MCTEX for Unambiguous Identification of Bragg Scattering and Rayleigh Scattering

View More View Less
  • 1 NOAA/ERL Aeronomy Laboratory, Boulder, Colorado
  • | 2 Cooperative Institute for Research in Environmental Sciences, University of Colorado, Boulder, Colorado
© Get Permissions
Full access

Abstract

A 2835-MHz (10.6-cm wavelength) profiler and a 920-MHz (32.6-cm wavelength) profiler were collocated by the NOAA Aeronomy Laboratory at Garden Point, Australia, in the Tiwi Islands during the Maritime Continent Thunderstorm Experiment (MCTEX) field campaign in November and December 1995. The two profilers were directed vertically and observed vertical velocities in the clear atmosphere and hydrometeor fall velocities in deep precipitating cloud systems. In the absence of Rayleigh scatterers, the profilers obtain backscattering from the refractive index irregularities created from atmospheric turbulence acting upon refractive index gradients. This kind of scattering is commonly referred to as Bragg scattering and is only weakly dependent on the radar wavelength provided the radar half-wavelength lies within the inertial subrange of homogeneous, isotropic turbulence. In the presence of hydrometeors the profilers observe Rayleigh backscattering from hydrometeors much as weather radars do and this backscatter is very dependent upon radar wavelength, strongly favoring the shorter wavelength profiler resulting in a 20-dB enhancement of the ability of the 2835-MHz profiler to observe hydrometeors. This paper presents observations of equivalent reflectivity, Doppler velocity, and spectral width made by the collocated profilers during MCTEX. Differential reflectivity is used to diagnose the type of echo observed by the profilers in the spectral moment data. When precipitation or other particulate backscatter is dominant, the equivalent reflectivities are essentially the same for both profilers. When Bragg scattering is the dominant process, equivalent reflectivity observed by the 1-GHz profiler exceeds the equivalent reflectivity observed by the 3-GHz profiler by approximately 18 dBZe. However, when the 3-GHz profiler half-wavelength is smaller than the inner scale of turbulence, the equivalent reflectivity difference exceeds 18 dBZe, and when both Rayleigh scattering and Bragg scattering are observed simultaneously, the equivalent reflectivity difference is less than 18 dBZe. The results obtained confirm the capability of two collocated profilers to unambiguously identify the type of echo being observed and hence enable the segregation of “clear air” and precipitation echoes for studies of atmospheric dynamics and precipitating cloud systems.

Corresponding author address: Dr. Kenneth S. Gage, NOAA/ERL Aeronomy Laboratory, R/E/AL3, 325 Broadway, Boulder, CO 80303.

Email: kgage@al.noaa.gov

Abstract

A 2835-MHz (10.6-cm wavelength) profiler and a 920-MHz (32.6-cm wavelength) profiler were collocated by the NOAA Aeronomy Laboratory at Garden Point, Australia, in the Tiwi Islands during the Maritime Continent Thunderstorm Experiment (MCTEX) field campaign in November and December 1995. The two profilers were directed vertically and observed vertical velocities in the clear atmosphere and hydrometeor fall velocities in deep precipitating cloud systems. In the absence of Rayleigh scatterers, the profilers obtain backscattering from the refractive index irregularities created from atmospheric turbulence acting upon refractive index gradients. This kind of scattering is commonly referred to as Bragg scattering and is only weakly dependent on the radar wavelength provided the radar half-wavelength lies within the inertial subrange of homogeneous, isotropic turbulence. In the presence of hydrometeors the profilers observe Rayleigh backscattering from hydrometeors much as weather radars do and this backscatter is very dependent upon radar wavelength, strongly favoring the shorter wavelength profiler resulting in a 20-dB enhancement of the ability of the 2835-MHz profiler to observe hydrometeors. This paper presents observations of equivalent reflectivity, Doppler velocity, and spectral width made by the collocated profilers during MCTEX. Differential reflectivity is used to diagnose the type of echo observed by the profilers in the spectral moment data. When precipitation or other particulate backscatter is dominant, the equivalent reflectivities are essentially the same for both profilers. When Bragg scattering is the dominant process, equivalent reflectivity observed by the 1-GHz profiler exceeds the equivalent reflectivity observed by the 3-GHz profiler by approximately 18 dBZe. However, when the 3-GHz profiler half-wavelength is smaller than the inner scale of turbulence, the equivalent reflectivity difference exceeds 18 dBZe, and when both Rayleigh scattering and Bragg scattering are observed simultaneously, the equivalent reflectivity difference is less than 18 dBZe. The results obtained confirm the capability of two collocated profilers to unambiguously identify the type of echo being observed and hence enable the segregation of “clear air” and precipitation echoes for studies of atmospheric dynamics and precipitating cloud systems.

Corresponding author address: Dr. Kenneth S. Gage, NOAA/ERL Aeronomy Laboratory, R/E/AL3, 325 Broadway, Boulder, CO 80303.

Email: kgage@al.noaa.gov

1. Introduction

Wind profilers were developed over a decade ago primarily to measure wind and other dynamical variables such as turbulence (see, e.g., Gage and Balsley 1978; Balsley and Gage 1982; Hardy and Gage 1990). During the past decade wind-profiling radars have become an accepted tool for meteorological research and for operational applications. The wind profilers usually operate at longer wavelengths than conventional weather radars and, at least in the clear atmosphere, the dominant scattering mechanism is Bragg scattering from refractive index inhomogeneities created by turbulence (e.g., Balsley and Gage 1982; Gage 1990).

Since the earliest profilers constructed to operate at very high frequency (VHF) were limited in their ability to observe the lower troposphere, the National Oceanic and Atmospheric Administration (NOAA) Aeronomy Laboratory developed the boundary layer wind profiler during the late 1980s and early 1990s, operating at a frequency near 1 GHz (30-cm wavelength), primarily for the measurement of lower-tropospheric winds in the deep tropics (Ecklund et al. 1988, 1990; Carter et al., 1995). This profiler is reliable, portable and was an important component of the TOGA COARE Integrated Sounding System (Parsons et al. 1994). During the 1990s the Forecast Systems Laboratory has developed a demonstration network of wind profilers operating in the meteorological frequency band near 400 MHz.

Both the 400-MHz (75-cm wavelength) NOAA demonstration network profilers and the 1-GHz lower-tropospheric wind profilers are quite sensitive to hydrometeors, insects, and birds (Wilson et al. 1994; Ralph 1995; Wilczak et al. 1995). As a consequence there is always some ambiguity when looking at profiler data concerning whether the echoes are truly due to Bragg scattering. If the echoes are not due to Bragg scattering, wind measurements can be compromised, especially if there is no correction for the fall velocities of hydrometeors, for example. Furthermore, turbulence measurements by wind profilers can be compromised if the echoes are not due to Bragg scattering as is usually assumed. Under these circumstances inferences concerning turbulence parameters such as the eddy dissipation rate ε or refractivity turbulence structure constant c2n are likely to be in error. It is therefore important in practice to be able to differentiate at least between Bragg scatter from refractivity turbulence and Rayleigh scatter from particulates and hydrometeors.

While wind profiling has been the major capability of profilers to be used in field campaigns and routine operations to date, there has been an increasing recognition of the utility of profilers for precipitation research (e.g., Wakasugi et al. 1986, 1987a,b; Gossard 1988; Gossard et al. 1990, 1992; Rogers et al. 1992, 1993a,b; Ralph 1995; Gage et al. 1996). The profilers provide a vertically looking continuous view of the evolution of precipitating cloud systems that can be used in a manner foreseen in the work of Atlas et al. (1973). The profilers can be used to study rain over a substantial height range (Gage et al. 1994, 1996; Ecklund et al. 1995; Williams et al. 1995) due to the relatively high height of the melting layer in the Tropics and to provide a continuous, height-resolved measurement of reflectivity from a few hundred meters above the surface.

A major issue in cloud physics and precipitation research has been the separation of vertical air motions from hydrometeor fall velocities. The separation is crucial to developing the capability of remote sensing of drop-size distributions using radar wind profilers. This issue has been a major roadblock in the progress of radar meteorology as recognized in the review of Atlas et al. (1973) and in Battan (1973). Several investigators have used profilers to determine drop-size distributions but the retrieval of the drop-size distribution requires either collocated 50- and 915-MHz profilers (Currier et al. 1992; Rajopadhyaya et al. 1998) or favorable conditions that enable the profiler to resolve both hydrometeor echoes and Bragg echoes (Wakasugi et al. 1986, 1987a; Gossard 1988; Gossard et al. 1990; Rogers et al. 1993b). While some progress has been made on this topic using stand-alone wind profilers, it is not yet possible under general conditions to separate vertical air motions and hydrometeor fall velocities using a single profiler.

The issue of which scattering mechanism dominates the profiler observations is central to the interpretation of the profiler observations. It is just as important to recognize hydrometeor contamination in wind measurements as “clear-air” Bragg scatter contamination in precipitation measurements since the presence of hydrometeors can compormise turbulence measurements and vertical velocity measurements. From a practical point of view the primary concern is being able to differentiate between echoes from hydrometeors and echoes from refractivity turbulence. This can be done routinely by comparing the equivalent reflectivities from two collocated profilers. While this has been done in a few case studies in the past (Knight and Miller 1993; Rogers and Brown 1997) it has not been used routinely.

In this paper we present the first results of the analysis of collocated profiler observations from the Tropics. The observations were obtained during the Maritime Continent Thunderstorm Experiment (MCTEX) in November–December 1995. For this campaign we operated a new 2835-MHz (10.6-cm wavelenth) profiler (Ecklund et al. 1999) alongside a standard 920-MHz (32.6-cm wavelength) profiler (Ecklund et al. 1988; Carter et al. 1995). Theoretical background for Bragg scattering and Rayleigh scattering is developed in section 2. Section 3 presents the instruments used in the MCTEX campaign and the methods of analysis. Section 4 contains the analysis of the two frequency profiler observations obtained during MCTEX. Section 5 contains a discussion of the results and section 6 presents our conclusions.

2. Theoretical background

a. Bragg scattering from the clear atmosphere

As mentioned in the introduction, the dominant scattering mechanism responsible for echoes observed by Doppler radar wind profilers is Bragg scattering (Balsley and Gage 1982). Bragg scattering arises from inhomogeneities in radio refractive index caused by atmospheric turbulence acting on gradients of radio refractive index (Ottersten 1969). Summaries of Bragg scattering as it relates to profiler observations can be found in Gage and Balsley (1980) and Gage (1990). According to the theory developed by Tatarskii (1971) the refractivity turbulence can be parameterized by C2n (m−2/3) and the radar volume reflectivity η (m−1) can be expressed in terms of C2n by
ηC2nλ−1/3
where λ (m) is the radar wavelength. Here C2n is the refractivity turbulence structure constant defined for locally homogeneous, isotropic turbulence in the inertial subrange by
nξ0ξnξ02C2nξ2/3
In Eq. (2), ξ (m) is an independent variable with dimensions of distance and n is the radio refractive index defined by
i1520-0469-56-21-3679-e3
where p (hPa) is the atmospheric pressure, e (hPa) is the partial pressure of water vapor, and T (Kelvins) is the absolute temperature. The wind profilers mentioned in the introduction are Doppler radars that are sensitive to refractive index inhomogeneities on the scale of half the wavelength of the probing radar. Thus a 50-MHz profiler is sensitive to 300-cm scale irregularities, a 400-MHz profiler is sensitive to 38-cm scale irregularities, a 1-GHz profiler is sensitive to 15-cm scale irregularities, and a 3-GHz profiler is sensitive to 5-cm scale irregularities. Provided that the Bragg scales to which the profilers are sensitive all lie within the inertial subrange, C2n measured by any of these profilers will be the same since it is independent of frequency.
The inertial subrange of homogeneous isotropic turbulence is bounded by an outer scale L0 and an inner scale l0. The intensity of atmospheric turbulence is typically expressed in terms of ε (m2 s−3), the eddy dissipation rate of turbulence. The outer scale of turbulence is often expressed in terms of the Ozmidov lengthscale as
L01/2;N−3/2
where N is the Brunt–Väisälä frequency defined by
N2gθθz
In Eq. (5), g (m s−2) is the acceleration due to gravity, θ is the potential temperature, and z is the vertical coordinate. Typical values of L0 in the free atmosphere are thought to be in the range of 10–100 m (Crane 1980). Since the Brunt–Väisälä frequency is directly related to the hydrostatic stability, L0 is largest in turbulent regions with low hydrostatic stability and smallest in weakly turbulent regions that are hydrostatically very stable.
The inner scale of the inertial subrange is given in terms of η0 (m), the Kolmogorov microscale
η0ν31/4;
where ν (m2 s−1) is kinematic viscosity (ν = μ/ρ, where μ is dynamic viscosity and ρ is atmospheric density). The inner scale is given by Hill and Clifford (1978) as
l0η0
Gossard et al. (1984a) have explored the critical radar wavelength λc (m) for the sensitivity of radars to turbulence in the lower troposphere in relation to the Kolmogorov microscale. They define λc = 8πη0 (or, equivalently, λc = 3.4l0) by choosing the wavenumber k (radians m−1) at which the spectral power is reduced by viscous effects to half that of the inertial subrange in Hill’s spectrum (Hill 1978). The inner scale of the inertial subrange of locally homogeneous, isotropic turbulence depends on the intensity of turbulence. It is small when turbulence intensity is large and largest when turbulence intensity is weak. Typical values of l0 in the lower atmosphere are on the order of several centimeters, consistent with a critical radar wavelength close to 10 cm. At times when the eddy dissipation rate is small, λc often becomes larger than 10 cm, as was shown by Gossard et al. (1984b). Note that l0 and λc increase with altitude in part because of the decreasein ρ.

b. Rayleigh scattering from hydrometeors

The volume reflectivity for backscattering from a volume filled with spherical raindrops is given by Battan (1973) as
ηπ5λ−4K2Z,
where |K|2 ≈ 0.93 for water (K is the complex index of refraction) and Z (mm6 m−3) is the equivalent radar reflectivity factor defined by
i1520-0469-56-21-3679-e9
In Eq. (9) D is drop diameter and N(D) is the drop-size distribution. Provided D < 0.1λ so that the backscattering from the raindrops is due to Rayleigh scattering, the reflectivity factor is independent of the radar wavelength. Since the volume reflectivity has a λ−4 dependence, shorter wavelength radars of similar power aperture products are much more sensitive than longer wavelength radars to Rayleigh scattering. In terms of backscattered power the radar equation that relates observed echo magnitude to radar reflectivity can be expressed as
PrCK2Zr−2
where Pr (W) is received power, r (m) is the range to the radar target, and C represents radar parameters that are constant (Battan 1973). Since Z varies over many orders of magnitude it is convenient (Rogers and Yau 1989) to express Eq. (10) so that
PrZrC.
The quantity 10 logZ is called the reflectivity factor in dBZ and it is conventional practice to express the backscattered echo magnitude from raindrops in terms of this quantity.
For generality in this paper the radar reflectivity expressed in dBZe is used regardless of whether the echoes arise from rain or snow or from refractivity turbulence structure. Following Battan (1973), Ze is defined by
λ4ηπ5K2
While |K|2 for ice differs from water, it is conventional to use the value of |K|2 for water in evaluating Eq. (12). Then, for snow, D in Eq. (9) is the melted diameter.
In the event that the echoes arise from Bragg scattering the relationship between C2n and the equivalent reflectivity in dBZe can be obtained from Eqs. (1) and (8). It is given by
i1520-0469-56-21-3679-e13
or
i1520-0469-56-21-3679-e15

c. Differential reflectivity for Bragg and Rayleigh scattering

For Bragg scattering the difference in dBZe observed by two profilers with wavelengths λ1 and λ2, respectively, is given by
(dBZe)λ1(dBZe)λ2λ1λ2
provided that both 0.5λ1 and 0.5λ2 lie within the inertial subrange of turbulence. The two wavelengths used in the MCTEX campaign were 32.6 cm (920 MHz) and 10.58 cm (2835 MHz). Using these values for the two profiler wavelengths λ1 and λ2 in Eq. (16) yields an equivalent reflectivity difference of 17.94 dBZe. The longer wavelength profiler has the greater equivalent reflectivity for Bragg backscatter as can be seen from Eq. (15).

3. Instruments and methods

During the MCTEX field campaign a 920-MHz profiler and 2835-MHz profiler were collocated at Garden Point on the Tiwi Islands, about 100 km northwest of Darwin, Australia, as shown in the map in Fig 1. The purpose of MCTEX was to study the life cycle of convective storms over the Tiwi Islands. The Tiwi Islands provide an ideal natural laboratory for studying tropical convection because deep convective systems develop over the islands on a daily basis during the rainy season owing to the organization of boundary layer convergence by the prevailing sea breeze circulation (Keenan et al. 1996).

a. The MCTEX profilers

The 920-MHz profiler used at MCTEX is essentially the same profiler described in Carter et al. (1995). The 2835-MHz profiler utilizes much of the same hardware and software as the 920-MHz profiler, and is described in Ecklund et al. (1999). The transmitter and antenna are necessarily different at the different operating frequencies. The transmitter is located in the transmit–receive module mounted on the edge of the shrouded dish antenna in a protective housing. Figure 2 shows the two profilers as they were set up side by side for the MCTEX campaign. The 2835-MHz antenna is fixed to point vertical. The perimeter shroud reduces the unwanted effects of feed spillover and ground clutter. The dish is underilluminated so that the beamwidth is 3.2° with a first sidelobe level about 25 dB below the main beam. The time series at each sampled range gate are time averaged and processed to produce 128 point Doppler power spectra at each sampled height as described in Carter et al. (1995). During typical 30-s dwell times at each sampled height about 120 spectra are produced at 3 GHz and these are averaged together and recorded on optical or magnetic media.

Calibrations of the two profilers used in this study are based on measured system parameters. Our experience is that reflectivities derived from profilers calibrated using measured parameters agree to within 1 or 2 dB. The long-term relative calibrations, however, agree to within less than a few tenths of a decibel over many months. For the purposes of this paper the very good long-term calibration stability is adequate. Absolute calibration of profilers with fixed vertically pointed antenna beams is more difficult and methods to obtain accurate absolute profiler calibration are currently under investigation.

Typical parameters for the 3-GHz profiler and the 1-GHz profiler are given in Table 1. Much of our work is in the Tropics where the tropopause and deep convective activity may extend to altitudes of over 18 km. In this environment we set the profiler interpulse period so that heights of up to about 22 km are not range aliased. We also set the maximum unaliased Doppler velocity to about plus or minus 20 m s−1 in order to capture the large updrafts and downdrafts often present in deep convection. Typical minimum and maximum sensitivities for both profilers in dBZe versus height are shown in Fig. 3. These values are calculated for 500-m height resolution using known system parameters and 30-s dwell times. The dynamic range of each profiler is limited to about 50 dB by the 8-bit digitizers.

b. Doppler spectra and spectral moments

The basic information content of the profiler observations is contained in the Doppler spectra. In order to illustrate the appearance of Doppler spectra under differing mixes of precipitation and nonprecipitation periods we have created Fig. 4. Figure 4 shows spectra observed at 2 km during MCTEX on 28 November 1995 derived from both profilers under clear-air, light stratiform rain, and convective rain, respectively. The Doppler spectra in the top panel of Fig. 4 are entirely due to Bragg scattering and, as expected, the reflectivity is greater from the 920-MHz profiler compared to the 2835-MHz profiler. During light stratiform rain, the Doppler spectra in the middle panel show both Bragg and Rayleigh components but the 920-MHz spectrum shows a more substantial Bragg component than does the 2835-MHz spectrum. The third peak on the right-hand side of the 2835-MHz spectrum is an “image” of the strong rain echo and should be disregarded. During heavier convective precipitation, the Bragg component is almost invisible in the 2835-MHz spectrum and is still quite small in the 920-MHz spectrum. Note that for the convective example in the bottom panel the 920-MHz spectrum indicates upward air motion of close to 6 m s−1 while the Doppler spectra from both profilers indicate hydrometeor fall velocities of about 2 m s−1.

In this paper we will only be using the standard spectral moments, which are routinely calculated in the profiler online program (POP; Carter et al. 1995). The spectral moments yield the radar reflectivity, the Doppler velocity, and the spectral width. For clear air (and nonprecipitating clouds), radar wind profiling this approach is entirely adequate. However, more generally, when hydrometeors are present, as in the spectra shown in the middle and bottom panels of Fig. 4, the three spectral moments are insufficient to describe the bimodal spectra. Multiple-target algorithms capable of describing more than the Bragg scattering component in the Doppler spectra are under development.

The current algorithms confuse the components of the Doppler spectra when more than one scattering mechanism is present. The zeroth moment will usually be a combination of the power under the spectrum from both sources. The first moment yields the Doppler velocity of the dominant mode so that it is entirely possible to get a small Doppler velocity associated with vertical air motions from the 920-MHz profiler and a larger Doppler velocity associated with hydrometeor fall velocities for the first moment of the spectrum for the 2835-MHz profiler. Under these circumstances the second spectral moment is likely to be broad.

4. Collocated profiler observations during MCTEX

a. Differential reflectivity observed at Garden Point, Australia

Ecklund et al. (1999) presented observations of deep convective cloud systems by the Garden Point profilers. Here we show the reflectivity difference between the two profiler observations of a deep convective anvil structure over Garden Point on the afternoon of 3 December 1995. On this day the initial deep convection missed Garden Point so that the hydrometeors observed over Garden Point were lofted into the upper troposphere by deep convection located near Garden Point.

Figure 5 contains the 2835- and 920-MHz observed equivalent reflectivity and reflectivity difference (920-MHz equivalent reflectivity minus 2835-MHz equivalent reflectivity) of an elevated anvil structure observed over Garden Point during 3 December 1995. The anvil structure above about 5 km is composed exclusively of frozen hydrometeors. At these altitudes neither profiler normally observes Bragg scatter. The equivalent reflectivity difference for this anvil structure is zero as expected since both profilers should observe the same equivalent reflectivity from hydrometeors. Below about 6 km, 920-MHz equivalent reflectivity exceeds the 2835-MHz equivalent reflectivity by more than 15 dBZe for the most part, excluding the period beginning at 1900 local time (LT) discussed below. Note that the values of differential reflectivity near the surface in the middle of the afternoon are close to 15–20 dBZe and that the values of differential reflectivity between 1700 and 1800 LT exceed 20 dBZe. The regions with large values of reflectivity difference decrease in altitude from about 5 km at 1600 LT until they reach the ground around 1730 LT. The periods of most active turbulence in the convective boundary layer are likely times in which the half-wavelength scales of both profilers are within the inertial subrange of turbulence. Later in the afternoon when surface heating is diminished, partly due to the presence of the thick cloud overhead, it is likely that the turbulence dissipates somewhat, the inner scale of turbulence increases, and the 5.5-cm scale to which the 2835-MHz profiler is sensitive lies outside the inertial subrange. As a consequence the equivalent reflectivity observed by the 2835-MHz profiler is substantially decreased and the differential reflectivity is correspondingly increased.

The period beginning about 1900 LT and ending about 2040 LT is of special interest. During this period we see the lowest 3 km possesses near-zero differential reflectivity implying Rayleigh scatter from particulates. This is a pattern that repeated itself at this time of day during MCTEX. It appears that these are echoes from bats and/or other night flyers. The differential reflectivity from the two profilers makes it easy to differentiate these echoes from the more common Bragg scattering. Note that in the case of backscattering from bats at 2835 MHz the type of scattering is more likely to be Mie than Rayleigh because of the large scattering cross section of bats.

b. Frequency distributions of observed equivalent reflectivity and differential reflectivity

Frequency distributions of equivalent reflectivity and differential reflectivity observed by the two profilers for the entire MCTEX campaign (13 November–7 December 1995) are shown in Figure 6. The panels on the left of Fig. 6 show the height distributions of echoes observed by the two profilers and on the bottom left the height distribution of echoes observed simultaneously by both profilers (which closely resembles the height distribution of echoes seen by the 920-MHz profiler). Note that the height distribution of the 2835-MHz echoes substantially exceeds the echoes observed by the 920-MHz profiler in the altitude range 6–12 km. The reason for this is the approximate 20-dB increased sensitivity of the 2835-MHz profiler to hydrometeors.

The top panels on the right of Fig. 6 show the frequency distribution of occurrence of equivalent reflectivity versus altitude for the 920- and the 2835-MHz profilers, respectively. The solid blue and red curves are the threshold values of detectability for the 920- and 2835-MHz profilers and the dashed blue and red lines are the threshold values for saturation for the 920- and 2835-MHz profilers, respectively. Note that the distribution of equivalent reflectivities seen by the 920-MHz profiler at low altitudes is typically 15–20 dBZe larger than the corresponding values observed in the same altitude range by the 2835-MHz profiler. These echoes are mostly from Bragg scatter. The echoes above 5–6 km are observed much less frequently and are more likely to be observed by the 2835-MHz profiler since these echoes arise exclusively from hydrometeors and the 2835-MHz profiler is about 20 dB more sensitive to these echoes.

The bottom right panel in Figure 6 shows the frequency distribution with altitude of reflectivity difference. Here two populations of echoes are clearly visible. The dominant echoes at the lowest altitudes present most of the time are due to Bragg scatter and the echoes clustered around zero reflectivity difference are due to Rayleigh scatter. It is interesting to note that most of the hydrometeor echoes are distributed in the range of 5–12 km.

c. Frequency distributions at 2.0 km

Frequency distributions for 920-MHz equivalent reflectivity, 920-MHz Doppler velocity, and 920-MHz spectral width versus equivalent reflectivity difference are shown in Fig. 7. These distributions shed further light on the interpretation of the observations. At 2 km, hydrometeor echoes and clear-air echoes can be observed by both profilers. Most of the time Bragg scattering is observed which means that the dominant population in each distribution is associated with this mechanism. The leftmost panels show the probability distributions of the 920-MHz reflectivity, the 920-MHz Doppler velocity, and the 920-MHz spectral width at 2 km. In each case these distributions are dominated by Bragg scatter. The bottom panel shows the distribution of equivalent reflectivity difference at 2 km. Most of the echoes are due to Bragg scatter but a small population of echoes are due to Rayleigh scatter. Note the large number of echoes with equivalent reflectivity difference exceeding 20 dBZe. These are from periods when the inner scale of the inertial subrange exceeds 5.5 cm as noted above.

The distribution of 920-MHz equivalent reflectivity versus equivalent reflectivity difference also shows two populations of echoes. The dominant population as before is due to Bragg scattering and the smaller population near-zero dBZe reflectivity difference is due to Rayleigh scatter. For the Bragg scatter echoes it is instructive to note that the echoes that lie in the 15–20 dBZe range of reflectivity difference tend to be associated with the stronger 920-MHz reflectivities while the echoes that exceed 20 dBZe of reflectivity difference tend to be associated with weaker 920-MHz reflectivities consistent with the idea that these echoes occur at times of weaker turbulence. The echoes that begin by clustering at 0-dBZe reflectivity difference and turn to larger reflectivity differences when the reflectivities exceed ∼18 dBZe are due to Rayleigh scatter. The nonzero reflectivity differences that occur with increasing reflectivity are a result of saturation. The 2835-MHz profiler saturates first so that the reflectivity differences are positive and increase with 920-MHz reflectivity.

The distribution of 920-MHz Doppler velocity shows that for Bragg scattering most of the vertical velocities lie in the range of ±0.5 m s−1. Note that the distribution narrows considerably with increasing reflectivity difference consistent with the interpretation that the larger values of reflectivity difference occur associated with weak turbulence. The smaller distribution of echoes from hydrometeors can be easily recognized by their nonzero fall velocities. However, not all the echoes with large Doppler velocities occur with near-zero differential reflectivity. Echoes with moderate Doppler velocities and moderate differential reflectivities likely occur at times when both Bragg and Rayleigh scattering components are present. The few echoes with large Doppler velocities and large differential reflectivities arise from the saturated echoes noted above.

The distribution of 920-MHz spectral width versus reflectivity difference is shown in the lowest color panel of Fig. 7. As before, the dominant population of echoes is associated with Bragg scattering and the broadest spectral widths (indicative of strong turbulence) have differential reflectivities in the range 15–20 dBZe while differential reflectivities that exceed 20 dBZe are associated with narrower spectral widths and weaker turbulence. The smaller population of hydrometeor echoes identified by near-zero differential reflectivity have spectral widths around 2 m s−1.

d. Scatterplots of equivalent reflectivity and Doppler velocity at 2.0 km

A scatterplot of 2835-MHz reflectivity versus 920-MHz reflectivity and its frequency distribution is shown in Fig. 8. The left panel in Fig. 8 contains the actual scatterplot with each symbol indicating an echo. The solid blue line indicates the expected locus of points for Bragg scatter and the dashed red line indicates the expected locus of points for Rayleigh scatter. The right panel contains the frequency distribution of points in the scatter diagram. Otherwise the panels are identical. As noted above, the dominant scatter mechanism at 2 km is Bragg scatter so that the vast majority of points are found in proximity to the blue line. At low equivalent reflectivities indicative of weak turbulence there is a clear indication of 2835-MHz reflectivities falling below where they should be if both profilers were observing turbulence within the inertial subrange. This is consistent with the assertion made above that when turbulence is weak, the inner scale of the inertial subrange exceeds the scale to which the 2835-MHz profiler is sensitive (5.5 cm). At high reflectivities there is an indication that the 2835-MHz reflectivities exceed the expected 920-MHz equivalent reflectivities. This feature is considered in the discussion (section 5). The smaller population of echoes that fall along the dashed red line are associated with Rayleigh scatter. As noted above, the observations depart systematically from the red dashed curve at high reflectivities because the 2835-MHz profiler saturates at 2 km when equivalent reflectivities exceed about 18 dBZe.

A scatterplot of Doppler velocities and their frequency distribution is contained in Fig. 9. In this figure the Doppler velocities associated with Bragg scatter are concentrated in the box outlined in red dashed lines. Rayleigh scattering from hydrometeors with substantial fall velocities make up the bulk of the remaining points in Fig. 9. Most of these show agreement and form a diagonal down to the left in Fig. 9. However, there are a number of points where the 2835-MHz Doppler velocity indicates a substantial fall velocity but the 920-MHz Doppler velocity is near zero. In these cases it is likely that the spectral moments derived from the 2835-MHz profiler are capturing the Rayleigh component of the spectrum and the spectral moments of the 920-MHz profiler are capturing the Bragg component of the spectrum. It is noteworthy that the converse is much less likely to happen as is reflected in the data. Nevertheless, there do appear to be a few examples where the 920-MHz profiler is giving a substantial negative Doppler velocity while the 2835-MHz profiler is indicating near-zero Doppler velocity. This could arise in some cases when the profilers are on the edge of precipitation because the observing volume of the 920-MHz radar substantially exceeds the observing volume of the 2835-MHz profiler.

5. Discussion

In this paper it has been shown how two profilers can be used to unambiguously differentiate Rayleigh scattering from particulates and Bragg scattering from refractivity turbulence. This is not the first time that multiple radars have been used to interpret scattering from the atmosphere. However, this is the first time the method has been applied to collocated profilers that operate unattended and observe routinely. It is this latter feature that promises to extend our understanding of atmospheric scattering processes and that will enable the rapid development of this methodology to make possible advances in turbulence observation and precipitation research using profilers.

The application to turbulence measurement is obvious considering that a major obstacle to accurate turbulence measurement especially at ultra high frequency is the ambiguity that arises from contaminating echoes arising from hydrometeors or biological targets. It is now possible to routinely eliminate this ambiguity by using differential reflectivity measurements from the two profilers. Thus, for example, it will now be possible to measure C2n unambiguously. However, the C2n measurements in the free troposphere using VHF profilers reported by VanZandt et al. (1978) may be only marginally improved since the 50-MHz profilers (with 6-m wavelength) are not very sensitive to Rayleigh scatter. Note that the profilers used here can observe to lower altitudes than the 50-MHz profiler used by VanZandt et al. (1978). Furthermore, the routine elimination of contaminating echoes using the methodology adopted here should enable the determination of eddy dissipation rate ε provided the profiler is equipped with a narrow enough antenna beam so that beam-broadening effects do not contaminate the spectral width. Note that, in view of the evidence presented above, the inner scale is often large enough to affect the performance of a profiler operating near 3 GHz. For turbulence measurements 1 GHz is probably a better choice.

The application to precipitation research follows from the need to separate Bragg scattering from Rayleigh scattering in profiler measurements of precipitating cloud systems. If the separation is not made the ambiguities in scattering type can compromise the results of the analysis. Now it is possible to routinely discard observations that do not satisfy the Rayleigh scatter criteria before analysis of precipitating cloud systems. This should lead to better classification of precipitating cloud systems.

While much of the potential for separation of scattering types can be achieved by use of spectral moment data, much more can be gained by applying spectral processing. What is needed is a multiple-target algorithm that recognizes and quantifies the moments of Rayleigh scatter and Bragg scatter in the same spectra. The development of this methodology should lead to improved estimates of turbulence, precipitation parameters, and ultimately the retrieval of drop-size distributions.

As pointed out in the introduction, it is a long-standing problem in cloud physics to separate air motions from hydrometeor fall velocities (Atlas et al. 1973). While it has proven possible to use a 50-MHz profiler to make this separation (Wakasugi et al. 1986, 1987a; Currier et al. 1992; Rajopadhyaya et al. 1998), it is not always possible to use a large antenna area 50-MHz profiler and a more portable tool is needed to accomplish the separation routinely. Once the hydrometeor fall velocity can be measured it is possible to derive the drop-size distribution and the rainfall rate. If the vertical air motions are not accounted for, the drop-size distribution cannot be determined and retrieved rainfall rates are grossly in error especially in convective rain (Rajopadhyaya et al. 1998).

A few remaining issues are worth discussing. In Fig. 8 we presented a scatterplot of 920-MHz equivalent reflectivity versus 2835-MHz equivalent reflectivity. While the overall results appear consistent with expectations for Bragg scatter there is one aspect that deserves further comment. In the discussion of Fig. 8 it was remarked that the large number of points with reflectivity difference exceeding the expected value of 18 dBZe for Bragg scattering was due to the half-wavelength of the 2835-MHz profiler lying outside the inertial subrange. When the inner scale of the inertial subrange exceeds 5.5 cm, the observed equivalent reflectivity will be less than expected for Bragg scatter and the reflectivity difference will be larger than expected for Bragg scatter. We also notice in Fig. 8 that the 2835-MHz reflectivity often exceeds the 920-MHz reflectivity expected for Bragg scatter especially in the limit of large equivalent reflectivity (see the red region just above the blue line in Fig. 8b). This effect is not expected and its possible causes are considered next.

A possible explanation for this effect is that it is an artifact of the instruments used. Briefly, the observing volumes of these two instruments are not well matched. The 920-MHz profiler used in MCTEX has a 9° beamwidth while the 2835-MHz profiler has a 3° beamwidth. Since the observing volume of the 920-MHz profiler greatly exceeds the observing volume of the 2835-MHz profiler, it is conceivable that under inhomogeneous conditions the localized strong 2835-MHz Bragg scatter echoes are smoothed substantially over the larger observing volume of the 920-MHz profiler. If this explanation is correct, then an experiment where the profiler observing volumes are matched should not show this effect. We plan an experiment to test this possibility in the near future.

Another possibility to explain this observation is related to some earlier work by Gossard and Strauch (1981). These authors explored the refractive index spectra in clouds and clear air in forward scatter. They found that in clouds, which have the highest reflectivities observed, the wavelength dependence of backscattering is reduced. This means that there is less of a difference in clouds than for clear air in the reflectivity difference expected for Bragg scatter. We intend to explore the implications of these observations further in future work.

Our initial motivation for developing the 3-GHz profiler was to observe the precipitating clouds in the Tropics above 5 km with better sensitivity (Ecklund et al. 1999). In order to obtain good sensitivity we have used a height interval of 500 m and an averaging time of about 30 s. The observations presented here were obtained with 8-bit digitizers, which have limited the dynamic range to about 50 dB. This limited range means that the 2835-MHz profiler saturates at a reflectivity of about 18 dBZe at a height of 2 km so that we have sacrificed quantitative observations of rain in the lower altitudes in favor of sensitivity to frozen hydrometeors at anvil altitudes. In future work, profiler dynamic range will be increased by over 25 dB to effectively eliminate profiler saturation effects.

6. Concluding remarks

In this paper we have compared the equivalent reflectivities obtained from two collocated profilers operating at Garden Point, Australia, during the Maritime Continent Thunderstorm Experiment that took place during November and December 1995. The two profilers operated at 920 MHz and 2835 GHz and observed the vertical structure of deep convective cloud systems on Melville Island during MCTEX.

This paper has focused attention on the utility of the difference in equivalent reflectivity as a diagnostic for differentiating Rayleigh scattering from hydrometeors and other particulates and Bragg scattering from refractivity turbulence. When expressed in terms of differential equivalent reflectivity in dBZe, we have found that Rayleigh scattering is associated with near-zero differential equivalent reflectivity and Bragg scattering is associated with an approximate 18-dBZe differential equivalent reflectivity. Differences that fall outside these theoretical values can occur under several circumstances.

The results presented here illustrate the general statistical properties of equivalent reflectivity, differential equivalent reflectivity, Doppler velocity, and spectral width observed in tropical environments with frequent active deep convection. Under these conditions most observations can be clearly identified as either associated with Rayleigh scattering from hydrometeors or with Bragg scattering from refractivity turbulence. Since only spectral moment data are used here, most observations that do not clearly fit theoretical expectations for Rayleigh scattering or Bragg scattering are the result of both types of scattering contributing to the Doppler spectra. Another common source of discrepancies is found to be associated with the failure of the assumption that for Bragg backscatter the half–radar wavelength for both profilers lies within the inertial subrange of atmospheric turbulence. Finally, remaining discrepancies are largely due to instrumental artifacts such as the fact that the two profilers have quite different observing volumes and the S-band profiler saturates at a relatively low equivalent reflectivity of 18 dBZe at 2 km.

The results presented here have potentially important applications to turbulence measurement and precipitation measurement by Doppler radar profilers. For turbulence measurements it is possible to unambiguously measure C2n and ε from Bragg scattering. For precipitation measurement it is possible to separate more precisely turbulent backscatter from hydrometeor echoes although the full utility of this methodology will not be realized until multiple-target resolving algorithms are employed to separate these sources of backscattered power within the Doppler spectrum.

Acknowledgments

We thank the Australian BMRC for technical assistance during the MCTEX Field Campaign. We also thank Drs. Sekelsky and Firda from the University of Massachusetts and Dr. Clothiaux from The Pennsylvania State University for their assistance at Garden Point. The research reported here was supported in part by the National Science Foundation under Grant ATM-9214800, and in part by NOAA’s Office of Global Programs and DOE ARM.

REFERENCES

  • Atlas, D., R. C. Srivastava, and R. S. Sekhon, 1973: Doppler radar characteristics of precipitation at vertical incidence. Rev. Geophys. Space Phys.,11, 1–35.

  • Balsley, B. B., and K. S. Gage, 1982: On the use of radars for operational wind profiling. Bull. Amer. Meteor. Soc.,63, 1009–1018.

  • Battan, L. J., 1973: Radar Observation of the Atmosphere. University of Chicago Press, 324 pp.

  • Carter, D. A., K. S. Gage, W. L. Ecklund, W. M. Angevine, P. E. Johnston, A. C. Riddle, J. S. Wilson, and C. R. Williams, 1995:Developments in UHF lower tropospheric wind profiling at NOAA’s Aeronomy Laboratory. Radio Sci.,30, 977–1001.

  • Crane, R. K., 1980: A review of radar observations of turbulence in the lower stratosphere. Radio Sci.,15, 177–193.

  • Currier, P. E., S. K. Avery, B. B. Balsley, K. S. Gage, and W. L. Ecklund, 1992: Combined use of the 50 MHz and 915 MHz wind profilers in the estimation of raindrop size distributions. Geophys. Res. Lett.,19, 1017–1020.

  • Ecklund, W. L., D. A. Carter, and B. B. Balsley, 1988: A UHF wind profiler for the boundary layer: Brief description and initial results. J. Atmos. Oceanic Technol.,5, 432–441.

  • ——, ——, ——, P. E. Currier, J. L. Green, B. L. Weber, and K. S. Gage, 1990: Field tests of a lower tropospheric wind profiler, Radio Sci.,25, 899–906.

  • ——, K. S. Gage, and C. R. Williams, 1995: Tropical precipitation studies using a 915-MHz wind profiler. Radio Sci.,30, 1055–1064.

  • ——, C. R. Williams, P. E. Johnston, and K. S. Gage, 1999: A 3-GHz Profiler for precipitating cloud studies. J. Atmos. Oceanic Technol.,16, 309–322.

  • Gage, K. S., 1990: Radar observation of the free atmosphere: Structure and dynamics. Radar in Meteorology, D. Atlas, Ed., Amer. Meteor. Soc., 534–565.

  • ——, and B. B. Balsley, 1978: Doppler radar probing of the clear atmosphere. Bull. Amer. Meteor. Soc.,59, 1074–1093.

  • ——, and ——, 1980: On the scattering and reflection mechanisms contributing to clear air radar echoes from the troposphere, stratosphere, and mesosphere. Radio Sci.,15, 243–257.

  • ——, C. R. Williams, and W. L. Ecklund, 1994: UHF wind profilers:A new tool for diagnosing tropical convective cloud systems. Bull. Amer. Meteor. Soc.,75, 2289–2294.

  • ——, ——, and ——, 1996: Application of the 915 MHz profiler for diagnosing and classifying tropical precipitating cloud systems. Meteor. Atmos. Phys.,59, 141–151.

  • Gossard, E. E., 1988: Measuring drop-size distributions in clouds with a clear-air-sensing Doppler radar. J. Atmos. Oceanic Technol.,5, 640–649.

  • ——, and R. G. Strauch, 1981: The refractive index spectra within clouds from forward-scatter radar observations. J. Appl. Meteor.,20, 170–183.

  • ——, R. B. Chadwick, T. R. Detman, and J. E. Gaynor, 1984a: Capability of surface-based clear-air Doppler radar for monitoring meteorological structure of elevated layers. J. Climate Appl. Meteor.,23, 474–485.

  • ——, W. D. Neff, R. J. Zamora, and J. E. Gaynor, 1984b: The fine structure of elevated refractive layers: Implications for over-the-horizon propagation and radar sounding systems. Radio Sci.,19, 1523–1533.

  • ——, R. G. Strauch, and R. R. Rogers, 1990: Evolution of dropsize distribution in liquid precipitation observed by ground-based Doppler radar. J. Atmos. Oceanic Technol.,7, 815–828.

  • ——, ——, D. C. Welsh, and S. Y. Matrosov, 1992: Cloud layers, particle identification, and rain-rate profiles from ZRVf measurements by clear-air Doppler radars. J. Atmos. Oceanic Technol.,9, 108–119.

  • Hardy, K. R., and K. S. Gage, 1990: History of radar studies of the clear atmosphere. Radar in Meteorology, D. Atlas, Ed., Amer. Meteor. Soc., 130–142.

  • Hill, R. J., 1978: Spectra of fluctuations in refractivity, temperature, humidity, and the temperature-humidity cospectrum in the inertial and dissipation ranges. Radio Sci.,13, 953–961.

  • ——, and S. F. Clifford, 1978: Modified spectrum of atmospheric temperature fluctuations and its application to optical propagation. J. Opt. Soc. Amer.,68, 892–899.

  • Houze, R. A., Jr., 1993: Cloud Dynamics. Academic Press, 573 pp.

  • Keenan, T., R. Carbone, S. Rutledge, J. Wilson, G. Holland, and P. May: 1996: The Maritime Continent Thunderstorm Experiment (MCTEX): Overview and initial results. Preprints, Seventh Conf. on Mesoscale Processes, Reading, United Kingdom, 326–328.

  • Knight, C. A., and L. J. Miller, 1993: First radar echoes from cumulus clouds. Bull. Amer. Meteor. Soc.,74, 179–188.

  • Ottersten, H., 1969: Atmospheric structure and radar backscattering in clear air. Radio Sci.,4, 1179–1193.

  • Parsons, D., and Coauthors, 1994: The integrated sounding system: Description and preliminary observations from TOGA COARE. Bull. Amer. Meteor. Soc.,75, 553–567.

  • Rajopadhyaya, D. K., P. T. May, R. C. Cifelli, S. K. Avery, C. R. Williams, W. L. Ecklund, and K. S. Gage, 1998: The effect of vertical air motions on rain rates and median volume diameter determined from combined UHF and VHF wind profiler measurements and comparisons with rain gauge measurements. J. Atmos. Oceanic Technol.,15, 1306–1319.

  • Ralph, F. M., 1995: Using radar-measured radial vertical velocities to distinguish precipitation scattering from clear-air scattering. J. Atmos. Oceanic Technol.,12, 257–267.

  • Rogers, R. R., and M. K. Yau, 1989: A Short Course in Cloud Physics. 3d ed. Butterworth-Heinemann, 290 pp.

  • ——, and W. O. J. Brown, 1997: Radar observations of a major industrial fire. Bull. Amer. Meteor. Soc.,78, 803–814.

  • ——, C. A. Knight, J. D. Tuttle, W. L. Ecklund, D. A. Carter, and S. A. Ethier, 1992: Radar reflectivity of the clear air at wavelengths of 5.5 and 33 cm. Radio Sci.,27, 645–659.

  • ——, W. L. Ecklund, D. A. Carter, K. S. Gage, and S. A. Ethier, 1993a: Research applications of a boundary-layer wind profiler. Bull. Amer. Meteor. Soc.,74, 567–580.

  • ——, D. Baumgardner, S. A. Ethier, D. A. Carter, and W. L. Ecklund, 1993b: Comparison of raindrop size distributions measured by radar wind profiler and by airplane. J. Appl. Meteor.,32, 694–699.

  • Tatarskii, V. I., 1971: The effects of the turbulent atmosphere on wave propagation. National Technical Information Service, U.S. Dept. of Commerce, TT 68-50464, 472 pp. [Available from NTIS, 5285 Port Royal Rd., Springfield, VA 22161.].

  • VanZandt, T. E., J. L. Green, K. S. Gage, and W. L. Clark, 1978: Vertical profiles of refractivity turbulence structure constant: Comparison of observations by the Sunset radar with a new theoretical model. Radio Sci.,13, 819–829.

  • Wakasugi, K., A. Mizutani, M. Matsuo, S. Fukao, and S. Kato, 1986:A direct method for deriving drop-size distributions and vertical air velocities from VHF Doppler radar spectra. J. Atmos. Oceanic Technol.,3, 623–629.

  • ——, ——, ——, ——, and ——, 1987a: Further discussion on deriving drop-size distribution and vertical air velocities directly from VHF Doppler radar spectra. J. Atmos. Oceanic Technol.,4, 170–179.

  • ——, B. B. Balsley, and T. L. Clark, 1987b: The VHF Doppler radar as a tool for cloud and precipitation studies. J. Atmos. Oceanic Technol.,4, 273–280.

  • Wilczak, J. M., and Coauthors, 1995: Contamination of wind profiler data by migrating birds: Characteristics of corrupted data and potential solutions. J. Atmos. Oceanic Technol.,12, 449–467.

  • Williams, C. R., W. L. Ecklund, and K. S. Gage, 1995: Classification of precipitating clouds in the tropics using 915-MHz wind profilers. J. Atmos. Oceanic Technol.,12, 996–1012.

  • Wilson, J. W., T. M. Weckwerth, J. Vivekanandan, R. M. Wakimoto, and R. W. Russell, 1994: Boundary layer clear-air radar echoes:Origin of echoes and accuracy of derived winds. J. Atmos. Oceanic Technol.,11, 1184–1206.

Fig. 1.
Fig. 1.

Map of the Tiwi Islands in Australia, which were the site of the MCTEX Field Campaign.

Citation: Journal of the Atmospheric Sciences 56, 21; 10.1175/1520-0469(1999)056<3679:UOTPDM>2.0.CO;2

Fig. 2.
Fig. 2.

The 2835-MHz profiler shroud and antenna (left) and 920-MHz profiler antenna (right) at Garden Point, Australia, during MCTEX.

Citation: Journal of the Atmospheric Sciences 56, 21; 10.1175/1520-0469(1999)056<3679:UOTPDM>2.0.CO;2

Fig. 3.
Fig. 3.

Detection and saturation thresholds for the 2835- and 920-MHz profilers operated in the high height mode (500-m resolution).

Citation: Journal of the Atmospheric Sciences 56, 21; 10.1175/1520-0469(1999)056<3679:UOTPDM>2.0.CO;2

Fig. 4.
Fig. 4.

Sample spectra from 2 km recorded during MCTEX on 28 Nov 1995 for the 920-MHz profiler and the 2835-MHz profiler during clear conditions (top), stratiform rain (middle), and convective rain (bottom).

Citation: Journal of the Atmospheric Sciences 56, 21; 10.1175/1520-0469(1999)056<3679:UOTPDM>2.0.CO;2

Fig. 5.
Fig. 5.

Time–height sections of equivalent reflectivity for the 920-MHz profiler (top) and for the 2835-MHz profiler (middle). Time–height section of equivalent reflectivity difference (920–2835 MHz) (bottom) at Garden Point, Australia, on 3 Dec 1995.

Citation: Journal of the Atmospheric Sciences 56, 21; 10.1175/1520-0469(1999)056<3679:UOTPDM>2.0.CO;2

Fig. 6.
Fig. 6.

Frequency distributions of (a) 920-MHz equivalent reflectivity, (b) 2835-MHz equivalent reflectivity, and (c) equivalent reflectivity difference for observations at Garden Point, Australia, during MCTEX (13 Nov–7 Dec 1995). The panels on the left indicate altitude distribution of observations used in the distributions.

Citation: Journal of the Atmospheric Sciences 56, 21; 10.1175/1520-0469(1999)056<3679:UOTPDM>2.0.CO;2

Fig. 7.
Fig. 7.

Two-dimensional frequency distributions at 2 km of (a) 920-MHz equivalent reflectivity vs equivalent reflectivity difference (920–2835 MHz), (b) 920-MHz Doppler velocity vs equivalent reflectivity difference (920–2835 MHz), and (c) 920-MHz spectral width vs equivalent reflectivity difference (920–2835 MHz) for observations at Garden Point, Australia, during MCTEX (13 Nov–7 Dec 1995). The panels on the left show one-dimensional probability distributions of 920-MHz equivalent reflectivity, Doppler velocity, and spectral width, respectively; (bottom) the probability distribution of the equivalent reflectivity different.

Citation: Journal of the Atmospheric Sciences 56, 21; 10.1175/1520-0469(1999)056<3679:UOTPDM>2.0.CO;2

Fig. 8.
Fig. 8.

Scatterplot of 2835-MHz equivalent reflectivity at 2 km vs 920-MHz equivalent reflectivity at 2 km (left) and frequency distribution of 2835-MHz equivalent reflectivity at 2 km vs 920-MHz equivalent reflectivity at 2 km (right) for observations at Garden Point, Australia, during MCTEX (13 Nov–7 Dec 1995). The dashed lines indicate Rayleigh scatter and the solid lines indicate Bragg scatter.

Citation: Journal of the Atmospheric Sciences 56, 21; 10.1175/1520-0469(1999)056<3679:UOTPDM>2.0.CO;2

Fig. 9.
Fig. 9.

Scatterplot of 2835-MHz Doppler velocity at 2 km vs 920-MHz Doppler velocity at 2 km (left) and frequency distribution of 2835-MHz Doppler velocity at 2 km vs 920-MHz Doppler velocity at 2 km (right) for observations at Garden Point, Australia, during MCTEX (13 Nov–7 Dec 1995).

Citation: Journal of the Atmospheric Sciences 56, 21; 10.1175/1520-0469(1999)056<3679:UOTPDM>2.0.CO;2

Table 1.

Typical parameters used at Garden Point, Australia, during MCTEX.

Table 1.
Save