Integrative and Comparative Biology Advance Access originally published online on May 8, 2008
Integrative and Comparative Biology 2008 48(1):12-23; doi:10.1093/icb/icn021
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
This article appears in the following Integrative and Comparitive Biology issue: Aeroecology: Probing and Modeling the Aerosphere–The Next Frontier [View the issue table of contents]
Detection and discrimination of fauna in the aerosphere using Doppler weather surveillance radar
Department of Biological Sciences, Clemson University, Clemson, SC 29634 0314, USA
Correspondence: 1E-mail: sagth{at}clemson.edu
 |
Synopsis |
|---|
Organisms in the aerosphere have been detected by radar since its development in the 1940s. The national network of Doppler weather radars (WSR-88D) in the United States can readily detect birds, bats, and insects aloft. Level-II data from the radar contain information on the reflectivity and radial velocity of targets and on width of the spectrum (SD of radial velocities in a radar pulse volume). Information on reflectivity can be used to quantify density of organisms aloft and radial velocity can be used to discriminate different types of targets based on their air speeds. Spectral width can also provide some useful information when organisms with very different air speeds are aloft. Recent work with dual-polarization radar suggests that it may be useful for discriminating birds from insects in the aerosphere, but more development and biological validation are required.
|
Introduction |
|---|
Over 60 years ago birds were found to be responsible for some of the puzzling radar echoes dubbed "angels" by the British (Lack and Varley 1945
; Buss 1946) and the first confirmation that insects were the source of some echoes on radar displays was published a few years later (Crawford 1949). Since that time radar has proven to be a useful tool for the detection, monitoring, and quantification of the movements of birds (Eastwood 1967; Gauthreaux 1970; Vaughn 1985), bats (Williams et al. 1973), and insects (Schaefer 1969) in the atmosphere during both day and night at spatial scales ranging from 1–10 km (for small tracking or marine radars) to 10–200 km (for single weather-surveillance and airport-surveillance radars). The continent-wide network of weather-surveillance and airport-surveillance radars in the United States can be used to detect and monitor organisms in the aerosphere at even greater spatial scales. The pulse- or resolution-volume of the radar is the basic sampling volume of a radar, and it is determined by pulse length (pulse duration) and widths of the horizontal and vertical dimensions of the beam. As pulse volumes increase, more and different types of targets (reflectors) are detected in a single pulse volume, and identifying the types of targets in the pulse volume can be a challenge. This is particularly true for weather-surveillance radars, because they have relatively long pulse lengths (greater sensitivity) and beam widths of 1°–2.5°. With the deployment of Doppler weather radar, data on the radial velocity of targets became available, and in this article, we demonstrate that it can be used in conjunction with information on winds aloft to discriminate types of reflectors based on their speed of movement. We also discuss planned upgrades to the national network of Doppler weather-surveillance radars (e.g., azimuthal sampling of radar data will be two times more frequent; dual polarization), which will provide additional and better data for researchers to use for the discrimination of different types of targets in the aerosphere.
|
Doppler weather-surveillance radar |
|---|
The WSR-88D (Weather-Surveillance Radar-1988, Doppler)—also referred to as Next Generation Weather Radar (NEXRAD) during the planning and developmental stages—is the backbone of the national network of weather radars operated by the National Weather Service (NWS), which supports activities of NWS, Federal Aviation Administration, the Department of Defense (units at military bases), the Department of Transportation and other users, including the private sector and media in the continental and non-continental United States (Crum and Alberty 1993
; Crum et al. 1993; Klazura and Imy 1993). The national network was completed by the end of 1996 and currently contains 155 WSR-88D radars, including the US Territory of Guam and the Commonwealth of Puerto Rico. Detailed characteristics of the WSR-88D can be found in Appendix A of the volume Weather Radar Technology Beyond NEXRAD (National Research Council 2002).
Base data (Level II) from the digital signal processor of the WSR-88D are recorded at all NWS and several select continental United States DOD sites on 8 mm magnetic tape and sent to the National Climatic Data Center (NCDC) in Asheville, NC for archiving and dissemination (Crum et al. 1993). Level-II data include the moments of equivalent radar-reflectivity factor (commonly referred as reflectivity), radial Doppler velocity, and the width of Doppler spectrum in addition to information on synchronization, calibration, date, time, antennal position, and operational mode. The angles of elevation of the antenna in precipitation mode (volume coverage pattern 21) are 0.5, 1.5, 2.4, 3.4, 4.3, 6.0, 9.9, 14.6, and 19.5° above horizontal, and in clear-air mode (volume coverage pattern 32) the angles are 0.5, 1.5, 2.5, 3.5, and 4.5°. Because the lowest tilt of the WSR-88D antenna averages 0.5° above the horizontal, the base of the beam is too high to detect low-flying birds over most of the surveillance area. The beam width of the WSR-88D is 
1°, and at a distance of 30 km the beam is 486 m wide. This eliminates the possibilityof precise altitudinal measurements of targets.
The data array for level-II reflectivity is 360 radials, each containing 240 1-km range bins, and the data array for level-II radial velocity and spectrum width is 360 radials each, containing 960 range bins of 250 m. From these base data, additional computer processing generates a set of predetermined products known as Level III data (e.g., base reflectivity at different antennal angles, base velocity at different antennal angles, vertical wind profile) as defined in Federal Meteorological Handbook No. 11, Part A). Level III products are recorded at 154 of the 159 worldwide sites, and sent to NCDC for permanent storage.
The archive of digital data from the WSR-88D is stored on the NCDC Robotic Mass Storage System, commonly known as the Hierarchical Data Storage System (HDSS), and they are available from the 1990s (depending on the date the radar was commissioned) to 1 day from present. The NEXRAD Inventory Search tool can be used to view the data for completeness. We have used the NCDC HDSS Access System (HAS) web page to download data for our projects. This option is best for users needing large amounts of data in a tar archive format. Up to a week of data for multiple sites may be ordered in a single request. Downloaded Level-II data from NCDC are in a unique digital binary format. We use special software to decompress and visualize the data and to select particular range bins and radials for analysis. Several free products for visualization and analysis are available for download at the NCDC website (http://www.ncdc.noaa.gov/oa/radar/radarresources.html).
|
Detection of biological targets on the WSR-88D |
|---|
Top
SynopsisIntroductionDoppler weather-surveillance... Detection of biological targets... Base reflectivityRadial velocityWidth of the spectrumFuture developments of WSR-88DDiscussionAcknowledgmentsReferences |
|---|
During the developmental period of the WSR-88D, Larkin (1983
) explored the radar's capabilities to detect birds, and he later studied the sensitivity of the radar's algorithms to return from birds and insects (Larkin 1991). Aerial biological targets are readily detected by the WSR-88D, and many biologists have used the radar for studies of the aerofauna: bird migration (e.g., Gauthreaux and Belser 1998, 1999a, 2003a; Gauthreaux et al. 2001; Diehl and Larkin 2005), bird roosts (e.g., Russell et al. 1998; Russell and Gauthreaux 1998), bat colonies (e.g., McCracken 1996; McCracken and Westbrook 2002; Horn and Kunz, this symposium), and concentrations of insects aloft (e.g., Westbrook and Wolf 1998). Westbrook and Isard (1999) have identified "biologically-relevant, temporal and spatial scales of atmospheric motion and other atmospheric variables" that influence the abundance and dispersal of airborne biota (specifically insects, spores, pollen, fungi, and plant pathogens), and they comment that the "network of WSR-88D Doppler weather radars can measure the aerial abundance, speed, and displacement direction of concentrated biota over areas of >1000 km2".
Within 60 km of the radar, WSR-88D can be used to delimit important stopover areas for migrating birds by measuring the density of birds in the beam as they begin a migratory movement (exodus) (Gauthreaux and Belser 2003b; Diehl and Larkin 2005). Within minutes of the onset of nocturnal migration, the distribution and density of echoes in the radar beam can provide information on the geographic location of source areas of migrants on the ground (migration stopover areas), and satellite imagery can be used to identify the topography and habitat type that characterizes these areas. The same approach can be used to delimit areas where insect swarms are emerging [e.g., mayfly (Hexagenia) mating swarms—Masteller and Obert 2000].
At a larger spatial scale (that of the surveillance area of a single Doppler weather radar—out to 240 km range), WSR-88D can be used to delimit locations of post-breeding, nocturnal roost sites of birds such as Purple Martins (Progne subis) and other species. Martins flying toward the roost late in the day generally do so at low altitudes and often fly under radar coverage; however, when they depart the roost near dawn they climb high into the sky and can be easily detected by Doppler radar (Russell et al. 1998; Russell and Gauthreaux 1998). In contrast, after sunset, bats (Tadarida brasiliensis) emerge from their daytime roosts to forage aloft, and the geographic locations of the colonies are readily detected by WSR-88D (McCracken 1996).
At even larger spatial scales, 10 WSR-88D radars have been used to study regional patterns of bird migration on the northern coast of the Gulf of Mexico (Gauthreaux and Belser 1999b), and the same number of radars was used to study bird migration over the Great Lakes Region (Diehl et al. 2003). At a continental scale, the national network of WSR-88D radars (155 units) can be used to monitor continuously bird migration over the United States at different altitudes dependent on distance from the radar (Gauthreaux et al. 2003). The latter achievement is significant because it provides a means of monitoring the season-to-season and year-to-year variation in the patterns of migration at different altitudes for different geographic regions and the nation as a whole.
|
Base reflectivity |
|---|
Reflectivity is the first of three moments measured by the WSR-88D. It is a measure of the fraction of radiation reflected by a given target. In other words, it is the ratio of the radiant energy reflected to the total that is incident upon the target (Fig. 1A and B). By relating levels of reflectivity in 1° x 1 km pulse volumes to densities of biological targets aloft, one can develop a calibration curve that can be used to quantify movements of organisms of interest. Gauthreaux and Belser (1998
) related maximum relative reflectivity data (dBZ) in 0.5° elevation scans to traffic rates of bird migration (number of birds crossing 1.6 km of front per hour) gathered by moon-watching. The resulting relationship had an R2 of 0.87. After Gauthreaux and Belser (1999a) converted migration-traffic rates to density of birds (y = birds/cm3) and converted relative reflectivity (dBZ) to reflectivity (x = Z), the resulting relationship (y = 1.7335x + 53.921) had a R2 of 0.89. If reflectivity is zero, however, there should be no birds aloft, and the y-intercept should be zero. Once the y-intercept was forced through zero, the resulting relationship (y = 1.8433x) had an R2 of 0.87. A comparison of the relationships among relative reflectivity (dBZ), reflectivity (Z), and density of migrating birds (birds/cm3) is given in Table 1.
|
|
Black (2000
) related WSR-88D mean reflectivity (Z) to mean density of birds (birds/km3) measured with 3-cm radar and concluded that "the measured bird density explained much of the variance in WSR-88D reflectivity" with an R2 of 0.75. More details of Black's procedures can be found in Black and Donaldson (1998, 1999) and Diehl et al. (2003). Although there is good correlation between reflectivity and migrating bird densities aloft, not all reflectors detected by the WSR-88D are migrating birds. The existing pulse volumes of the WSR-88D not only contain migrating birds, bats, and insects but also contain foraging birds, bats, and swarming insects moving randomly in the atmosphere. One method of discriminating biological targets aloft is by their air speed, and the radial velocity moment of the WSR-88D can be used to accomplish this task and determine the sources contributing to the reflectivity measures. |
Radial velocity |
|---|
Radial velocity is the second moment measured by the WSR-88D and is the component of motion of a target toward or away from the radar (Fig. 1C and D). The mean radial velocity of targets in a pulse volume (1° x 250 m) of Level II data can be used to determine if the value is in keeping with faster flying, migrating birds and bats or with slow-flying insects, randomly foraging birds and bats, or drifting particulates like dust and smoke particles (Gauthreaux and Belser 1998
, 1999a, 2003a). By sampling a discrete altitudinal band with similar winds aloft, the variation of the mean radial velocities within pulse volumes is less likely to be influenced by variation in winds aloft. For a particular volume scan, one must first select the angle of elevation of the beam for analysis. If a higher elevation angle (1.5° or 2.5°) is selected, the radar beam will pass through the layer of biological reflectors closer to the radar (where pulse volumes will be smaller) than if the lowest elevation angle (0.5°) is selected. Smaller pulse volumes likely contain fewer biological targets and provide a better subsample of the range of mean radial velocities. We define the altitudinal band to be sampled by selecting the scan elevation and the 250-m range bins associated with that altitudinal band. For each of the selected range bins, an algorithm examines all the pulse volumes at different azimuths for that range bin and selects the pulse volume with the maximum mean inbound radial velocity plus two pulse volumes on each side. The resulting data contain information on the direction of the maximum mean inbound radial velocity and the five mean radial velocities of the five pulse volumes for each of the selected 250 m range bins. Once the mean radial velocities of all the sampled pulse volumes are plotted in a frequency distribution, the information can be used to determine the proportion of base reflectivity that can be attributed to different types of biological targets.
If frequency distributions of mean radial velocities of the five pulse volumes for each of the selected 250 m range bins are plotted for a time near sunset (or when winds aloft are measured by radiosonde) and at the time of peak reflectivity of targets aloft for the night, one can determine the relative contribution of different types of targets to the reflectivity received by the radar. This is possible because migrating birds and bats have faster air speeds than do migrating insects. In general, very small insects have low air speeds and fly only slightly faster than the wind while larger insects may have higher air speeds of up to 7–8 m/s (Schaefer 1976; Larkin 1991; Riley 1994; Dean 2003). In a radar study of insect migration and dispersal in China, Feng et al. (2004) found that most insects had airspeeds <7 m/s and the modal airspeed was between 2 and 3 m/s. In contrast, most birds fly at airspeeds >8 m/s (Bruderer and Boldt 2001). Two species of bats (Nyctalus noctula and Eptesicus serotinus) tracked with radar had ground speeds of 13 m/s in a weak wind of 1–2 m/s (Bruderer and Popa-Lisseanu 2005).
On evenings with little or no bird migration, the range of mean radial velocities in the frequency distributions near sunset and an hour later are essentially the same but the frequencies often increase as more insects are aloft (Fig. 2A). In contrast, on nights with bird migration the frequency distribution of mean radial velocities shows a dramatic increase in higher mean radial velocities after dark (Fig. 2B). By quantifying the night-to-night changes in the frequency distribution data, one can develop an index of the amount of bird migration. We are currently relating the frequency distribution data to the reflectivity calibration data that was the basis of the calibration of Gauthreaux and Belser (1998, 1999a).
|
|
Width of the spectrum |
|---|
Spectral width is the third moment measured by the WSR-88D and is a measure of the dispersion of velocities within the radar sample volume (SD of the Doppler velocities in the spectrum). If a pulse volume contains targets all flying at the same radial velocity, the spectrum width should have minimal values. According to Fang et al. (2004
), daytime fair weather without birds, stratiform rain and snow, and isolated tornadic storms produce weather signals that have the smallest volumetric median values of spectral widths (i.e., <2 m/s). If different types of biological targets flying at different radial velocities occur in a pulse volume, the spectrum width will be high (e.g., 
12 m/s). Because most insects have relatively low airspeeds (see above), when the aerosphere has mostly insect targets the spectrum-width values are low (Fig. 3A). After dark when both slower and faster flyers are aloft, the spectrum-width values increase (Fig. 3B). After sunset, on the evening of October 14–15, 2005, a major migration of birds occurred and this is reflected in both the analysis of radial-velocity pulse volumes (Fig. 2B) and spectrum-width pulse volumes (Fig. 3B). The circular (azimuthal) distribution of the highest spectrum-width values corresponds to the axis (341°–161°) of maximum mean inbound and outbound radial velocities. Figure 4 shows the circular (azimuthal) distribution of pulse volumes having the maximum mean inbound radial velocity for each 250 m range bin between 5 and 65 km. The mean inbound direction is 340.9° and the length of mean vector (r) is 0.985 with a circular standard deviation of 9.8°.
|
|
|
Future developments of WSR-88D |
|---|
Beginning in 2008 the WSR-88D radar will undergo a series of upgrades that will greatly enhance the radar's capability to provide useful information for biologists who wish to study the distribution and abundance of organisms in the aerosphere. The azimuthal resolution of all three moments of data will change from1° to 0.5°, and the range resolution of reflectivity will change from 1 to 0.25 km and match the existing resolution of radial velocity. Doppler data range will increase from 230 to 300 km, and the amount of data collected and transmitted during a volume scan will increase by a factor of
2.3. In addition to the move toward super-resolution data, the radar will be upgraded to have a dual polarization capability.
Zrnic and Ryzhkov (1998, 1999) have used a 10 cm wavelength, pencil beam (1°), Doppler weather radar that transmits vertically and horizontally polarized waves alternately to discriminate biological reflectors in the atmosphere. The polarimetric variables they used included differential reflectivity (ZDR), backscatter differential phase (
), and correlation coefficient (
hv) between orthogonally polarized returns. They found that the backscatter differential phase is sizable for insects (ranging between 5° and 40°) and that both reflectivity and differential reflectivity have strong azimuthal dependence for insects—revealing a high degree of common alignment. The differential reflectivity is small (ranging between –1 and 3 dB) for birds, and the differential phase is larger sometimes over 100°. In a more recent study, Bachmann and Zrnic (2007a) reported that spectral density of differential reflectivity (ZDR) in one radial ranged from –10 to 27 dB with a mode near 10 dB for insects and from –15 to 27 dB with a mode at 3 dB for birds. With respect to differential phase in the same radial, they reported a range of 0°–175° with a peak near 80° for insects and a range of –180°–180° with a very slight peak about 50°. Bachmann and Zrnic (2007a) reported that the maxima of differential reflectivity as a function of azimuth are between 3 and 8 dB for insects and are less than about 2.5 dB for birds.
We have recently used a dual-polarimetric Doppler radar (ARMOR—Advanced Radar for Meteorological and Operational Research) operating at C-band (5625 MHz) with a beam width of 1° operated by the Department of Atmospheric Sciences at the University of Alabama, Huntsville, AL. The radar was originally deployed in Huntsville by the NWS as a local warning radar in 1977 (WSR-74C), and in 1991 it was refurbished and upgraded to Doppler. It was upgraded to dual-polarimetry using the SIGMET Antenna Mounted Receiver in the fall of 2004. The radar provides data on reflectivity, radial velocity, differential reflectivity, differential phase, and correlation coefficient. Figure 5 shows the displays of four products (reflectivity, radial velocity, differential reflectivity, differential phase) from the ARMOR radar on the evening of October 11, 2007. The differential reflectivity product (Fig. 5C) shows pixels that range from 0.2 dB to >5 dB, but the values cannot be used to separate birds from insects. The same is true for the differential phase product (Fig. 5D). This product shows pixels with values ranging from <11° to >169°. Once again nearly all of the values are in the range of overlap for birds and insects. On this evening, we also used the Level-II radial velocity product of the WSR-88D located at Huntsville, AL and generated a frequency distribution of the mean radial velocities for the same time period as the ARMOR data (Fig. 6). It is clear from Fig. 6 that the majority of the reflectors aloft after dark are birds, based on the number of pulse volumes with high mean radial velocities.
|
|
|
Discussion |
|---|
The WSR-88D was designed as a weather radar and when designed was never intended to be used by biologists studying biological targets in the atmosphere. Despite its large pulse volumes, the radar has proven to be of great value for detecting, monitoring, and quantifying biological targets in the atmosphere over most of the continental United States, Alaska, and Hawaii, but discriminating different types of biological reflectors in the atmosphere on the basis of Level II reflectivity data has been problematical. The main issue is the relative contributions of insects, birds, and bats to the reflectivity detected by the radar receiver. Some radar biologists rely mainly on reflectivity data from the WSR-88D to assess densities of targets aloft, but a recent analysis of moderately strong clear air return recorded by operational WSR-88D radars and X-band and W-band research radars (Martin and Shapiro 2007
) on two nights (May 30, 2000 in Goodland, KS and another on May 19, 2001 in Norman, OK) serve to illustrate the problem with doing so. The research X-band and W-band radars have resolution that resolves point target echoes, and by examining the radar cross-sections and determining target density of the resolved point targets, Martin and Shapiro found that in both cases of nocturnal clear air echoes nearly all of the targets were almost certainly insects. The analysis of the dependence of the echo strength on radar wavelength also supports their conclusion.
Although more biological work is needed, we agree with the findings of Martin and Shapiro (2007) and conclude that on most occasions during the day and night during the warmer months, much of the reflectivity recorded by the WSR-88D is from migrating and foraging insects and foraging birds and bats moving in diverse directions such that mean radial velocities in large pulse volumes are very slow. If fast-moving targets are moving in opposite directions in a large pulse volume or moving in random directions, the inbound radial velocities of the targets will cancel the outbound radial velocities and a very low mean radial velocity will be recorded from the pulse volume. This can be determined by analyzing the direction of the maximum mean inbound and outbound radial velocities in smaller pulse volumes (range bins closer to the radar with higher antennal tilts). During the spring and fall when bird, bat, and insect migrations are underway and most animals are moving in nearly the same direction, the mean radial velocity of the targets in the radar pulse volume is significantly faster and more closely represents the mean ground speeds of the targets aloft (Gauthreaux and Belser 1998; Gauthreaux et al. 1998). In the future, radial and azimuthal oversampling will help resolve features that are 2–3 times smaller than the size of the beam., and with very high-resolution Doppler radar one may be able to determine the radial velocity of individual targets if they are alone in a pulse volume.
In one of the first studies to use polarimetric radar to discriminate insects and bird targets aloft, Mueller (1983) found that differential phase () and differential reflectivity (ZDR) were useful in accomplishing that task and that ZDR was higher for birds than for insects. The initial reports of Zrnic and Ryzhkov (1998, 1999) also suggested that the measurements of ZDR and from polarimetric Doppler weather radar might be useful for discriminating between bird and insect targets, but they found that ZDR was higher for insects—the opposite to Mueller's finding. Recently, there has been considerable research on this topic (e.g., Zhang et al. 2005, 2006; Bachmann and Zrnic 2007a), and some of the recent findings related to discriminating types of targets with polarimetric radar are different from some of the findings in earlier reports. One possible explanation for the differences is that ZDR values of birds depend significantly on azimuth or the orientation of the bird relative to the radar. When radar illuminates a bird, the size of the echo could be small (head-on view with closed wings) or large (broad side, wings open). Consequently, ZDR values could be small or large depending on the actual size and dimensions of the bird (e.g., small versus large, different body shapes, long versus short wings, fixed versus flapping wings). In contrast, the ZDR values of small insects have been shown in the literature to be less dependent on azimuth. Thus, the findings in the reports are based on data from different species and different mixtures of species in different proportions. It has also been shown recently that spectral densities of polarimetric variables from Doppler weather radar can be used to separate the measurements of velocity of birds from those of insects mixed within the resolution volumes (Bachmann and Zrnic 2005, 2006, 2007a, 2007b), but the method works best when time series (Level I data) are available.
Most of the research with polarimetric radar is being done by radar meteorologists, physicists, and engineers; and many of the studies are case histories that focus on one or two days or a few hours of data. Advancement in this area of research will benefit greatly from future collaborations with radar biologists, and the advancements will offer the field of aeroecology new tools to detect, identify, and quantify the fauna in the aerosphere.
|
Acknowledgments |
|---|
The assistance of K. Knupp of the Department of Atmospheric Sciences at the University of Alabama, Huntsville, AL with the ARMOR polarimetric radar is greatly appreciated. We are grateful to T.H. Kunz and N.I. Hristov for inviting us to participate in this symposium. The helpful suggestions of two anonymous reviewers greatly improved the manuscript. Our recent research with Level II data from the WSR-88D has been supported by funding from the Strategic Environmental Research and Development Program (SERDP) of the Department of Defense (CS-1439). We also wish to thank the Air Force Office of Scientific Research, through a grant to Boston University (FA9550-7-1-0449 to Kunz), for providing partial travel support to S.A.G. to participate in the Aeroecology symposium. The Society for Integrative and Comparative Biology waived registration fees for symposium speakers.
|
Footnotes |
|---|
From the symposium "Aeroecology: Probing and Modeling the Atmosphere—The Next Frontier" presented at the annual meeting of the Society for Integrative and Comparative Biology, January 2–6, 2008, at San Antonio, Texas.
|
References |
|---|
Bachmann SM, Zrnic DS. Spectral polarimetry for identifying and separating mixed biological scatterers. In: Preprints, 32nd Conference on Radar Meteorology (2005) Albuquerque, NM: American Meteorological Society. 9R-5.
Bachmann SM, Zrnic DS. Spectral polarimetric VAD separates birds from insects (wind) velocities. In: Preprints, The 4th European Radar Conference. Barcelona (2006) Spain: Meteocat. 96–100.
Bachmann S, Zrnic D. Spectral density of polarimetric variables separating biological scatterers in the VAD display. J Atmos Oceanic Technol (2007a) 24::1186–1198.[CrossRef]
Bachmann SM, Zrnic DS. Spectral densities of polarimetric variables for retrieving winds and determining scatterer types. In: Preprints, The 23rd Conference on Interactive Information Processing Systems (IIPS) for Meteorology, Oceanography, and Hydrology (2007b) San Antonio, TX: American Meteorological Society. 2–13.
Black JE. Application of weather radar to monitoring numbers of birds in migration. (2000) Extended abstract from a presentation given in a Symposium on Weather Radar Ornithology at the joint meeting of the Association of Field Ornithologists/Wilson Ornithological Society/Gulf Coast Bird Observatory, April 2000: Galveston, TX. Available from: http://www.physics.brocku.ca/People/Faculty/black/Galveston/texas.html.
Black JE, Donaldson N. A comparison of bird migration observed on a small conical radar and that observed on a large Doppler radar located near King City, Ontario, Canada in the spring of 1998. In: Brock Physics Rpt, PR-1998-3 (1998) St Catharines, ON: Brock University.
Black JE, Donaldson N. Comments on "Display of Bird movements on the WSR-88D: Patterns and quantification". Weather Forecast (1999) 14::1039–1040.[CrossRef]
Bruderer B, Boldt A. Flight characteristics of birds: I. Radar measurements of speeds. Ibis (2001) 143::178–204.[CrossRef]
Bruderer B, Popa-Lisseanu AG. Radar data on wing-beat frequencies and flight speeds of two bat species. Acta Chiropterologica (2005) 7::73–82.[CrossRef][Web of Science]
Buss IO. Bird detection by radar. Auk (1946) 63::315–318.
Crawford AB. Radar reflections in the lower atmosphere. Proc IRE (1949) 37::404–405.[CrossRef]
Crum TD, Alberty RL. The WSR-88D and the WSR-88D operational support facility. Bull Am Meteor Soc (1993) 74::1669–1687.[CrossRef]
Crum TD, Alberty RL, Burgess DW. Recording, archiving, and using WSR-88D data. Bull Am Meteor Soc (1993) 74::645–653.[CrossRef]
Dean TJ. University of Florida book of insect records. Chapter 1. In: Fastest Flyer (2003) Available from: http://ufbir.ifas.ufl.edu/chap01.htm.
Diehl RH, Larkin RP. Introduction to the WSR-88D (NEXRAD) for ornithological research. Ralph CJ, Rich TD, eds. (2005) Bird conservation implementation and integration in the Americas: Proceedings of the Third International Partners in Flight conference, March 20–24, 2002: Asilomar CA. Gen Tech Rep PSW-GTR-191. U.S. Dept of Agr, Forest Service, Pacific Southwest Research Station, Albany, CA. 876–888.
Diehl RH, Larkin RP, Black JE. Bird migration around the Great Lakes on Doppler radar: implications for habitat conservation. Auk (2003) 120::278–290.[CrossRef][Web of Science]
Eastwood E. Radar ornithology (1967) London: Methuen and Co.
Fang M, Doviak RJ, Melnikov V. Spectrum width measured by WSR-88D: error sources and statistics of various weather phenomena. J Atmos Oceanic Tech (2004) 21::888–904.[CrossRef]
Feng H, Wu K, Cheng D, Guo Y. Spring migration and summer dispersal of Loxostege sticticalis (Lepidoptera: Pyralidae) and other insects observed with radar in northern China. Environ Entomol (2004) 33::1253–1265.[Web of Science]
Gauthreaux SA. Weather radar quantification of bird migration. BioScience (1970) 20::17–20.[CrossRef][Web of Science]
Gauthreaux SA, Belser CG. Displays of bird movements on the WSR-88D: patterns and quantification. Weather Forecast (1998) 13::453–464.[CrossRef]
Gauthreaux SA, Belser CG. Reply to "Comments on display of bird migration on the WSR-88D: patterns and quantification". Weather Forecast (1999a) 14::1041–1042.[CrossRef]
Gauthreaux SA, Belser CG. Bird migration in the region of the Gulf of Mexico. Slotow RH, Adams NJ, eds. (1999b) Proc 22 Int Ornithol Congr, Durban, Johannesburg: BirdLife South Africa. 1931–1947.
Gauthreaux SA, Belser CG Jr. Bird movements on Doppler weather surveillance radar. Birding (2003a) 35::616–628.
Gauthreaux SA, Belser CG Jr. Radar ornithology and biological conservation. Auk (2003b) 120::266–277.[CrossRef][Web of Science]
Gauthreaux SA, Belser CG Jr, van Blaricom D. Using a network of WSR-88D weather surveillance radars to define patterns of bird migration at large spatial scales. In: Avian migration—Berthold P, Gwinner E, Sonnenschein E, eds. (2003) Berlin: Springer. 335–346.
Gauthreaux SA, Belser CG, Farnsworth A. How to use Doppler weather-surveillance radar to study hawk migration. In: Hawkwatching in the Americas—Bildstein KL, Klem D Jr, eds. (2001) North Wales, PA: Hawk Migration Association of North America. 146–160.
Gauthreaux SA, Mizrahi DS, Belser CG. Bird migration and bias of WSR-88D wind estimates. Weather and Forecast (1998) 13::465–481.[CrossRef]
Horn JW, Kunz TH. Bats and angel rings: qualifying flight behavior and colony dynamics using NEXRAD Doppler radar data. In: Integr Comp Biol (2008) doi: 10.1093/icb/icn037.
Klazura GE, Imy DA. A description of the initial set of analysis products available from the NEXRAD WSR-88D System. Bull Am Meteo Soc (1993) 73::1293–1311.
Lack D, Varley GC. Detection of birds by radar. Nature (1945) 156::446.
Larkin RP. Evaluation of NEXRAD for bird hazard warning. In: Report to United States Air Force, AFESC (1983) 49.
Larkin RP. Sensitivity of NEXRAD algorithms to echoes from birds and insects. In: In 25th International Conference on Radar Meteorology (1991) Paris, France: American Meteorological Society. 203–205.
Martin WJ, Shapiro A. Discrimination of bird and insect radar echoes in clear air using high-resolution radars. J Atmos Oceanic Technol (2007) 24::1215–1230.[CrossRef]
Masteller EC, Obert EC. Excitement along the shores of Lake Erie-Hexagenia -echoes from the past. Great Lakes Res Rev (2000) 5::25–36.
McCracken GF. Bats aloft: a study of high-altitude feeding. Bats (1996) 14::7–10.
McCracken GF, Westbrook JK. Bat patrol. Scientists discover that high-flying mammals are bad news for bugs. National Geographic (2002) 201::114–123.
Mueller EA. Differential reflectivity of birds and insects. In: Preprints, 21st Conference on Radar Meteorology (1983) Edmonton, AB, Canada: American Meteorological Society. 465–466.
National Research Council, Committee on Weather Radar Technology Beyond NEXRAD. Weather radar technology beyond NEXRAD (2002) Washington, DC: National Academies Press.
Riley JR. Acoustic sounding, atmospheric structure and insects. Int J Remote Sense (1994) 15::293–297.[CrossRef]
Russell KR, Gauthreaux SA. Use of weather radar to characterize movements of roosting purple martins. Wildlife Soc Bull (1998) 26::5–16.
Russell KR, Mizrahi DS, Gauthreaux SA. Large-scale mapping of purple martin pre-migratory roosts using WSR-88D weather surveillance radar. J Field Ornithol (1998) 69::316–325.
Schaefer GW. Radar studies of locust, moth and butterfly migration in the Sahara. Proc Roy Ent Soc Lond C (1969) 34::39–40.
Schaefer GW. Radar observations of insect flight. In: Insect flight. Symp No. 7. R Entomol Soc London—Rainey RC, ed. (1976) Oxford: Blackwell. 157–197.
Vaughn CR. Birds and insects as radar targets: a review. Proc IEEE (1985) 73::205–227.[CrossRef]
Williams TC, Ireland LC, Williams JM. High altitude flights of the free-tailed bat, Tadarida brasiliensis, observed with radar. J Mammal (1973) 54::807–821.[CrossRef][Web of Science]
Westbrook JK, Isard SA. Atmospheric scales of biotic dispersal. Agri Forest Meteorol (1999) 97::263–274.[CrossRef]
Westbrook JK, Wolf WW. Migratory flights of bollworms, Helicoverpa zea (Boddie), indicated by Doppler weather radar. (1998) Preprints, Second Urban Environment Symposium and 13th Conference on Biometeorology and Aerobiology, November 2–6, 1998: Albuquerque, NM. Boston: American Meteorological Society. 354–355.
Zhang P, Ryzhkov A, Zrnic D. Observations of insects and birds with a polarimetric prototype of the WSR-88D radar. In: Preprints, 32nd conference on Radar Meteorology (2005) Albuquerque, NM: American Meteorological Society. CD-ROM, 9R.6.
Zhang P, Ryzhkov A, Zrnic D. Polarimetric prototype of the WSR-88D radar observations of insects and birds. In: Preprints, 12th Conference on Aviation, Range, and Aerospace Meteor (2006) Atlanta, GA: American Meteorological Society. CD-ROM, P6.4.
Zrnic DS, Ryzhkov AV. Observations of insects and birds with a polarimetric radar. IEEE Trans Geosci Remote Sens (1998) 36::661–668.[CrossRef]
Zrnic DS, Ryzhkov AV. Polarimetry for weather surveillance radars. Bull Am Meteor Soc (1999) 80::389–406.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||