An Investigation of Bird Radar Echoes Detected by the WSR-88D Doppler Radar at WFO Binghamton, NYMichael B. Sporer, National Weather Service Binghamton, NY
Kirk A Lombardy, National Weather Service, Binghamton, NY
Joseph Conklin, Jersey Weather Service, Atlantic Highlands, NJ
Radar detection of non-meteorological echoes is not a recently discovered application (Eastwood 1967, Edwards 1959). However, with the proliferation of high quality doppler radars in association with the National Weather Service (NWS) modernization and restructuring, many new occurrences of this phenomenon have been reported. Shortly after the WSR-74S s-band radar was replaced with the more powerful and sensitive WSR-88D s-band doppler radar at the NWS office in Binghamton, NY in September of 1993, the periodic appearance of a ring shaped echo configuration was noticed during the late summer and early fall near the time of sunrise. The primary purpose of this investigation was to determine the exact source of the echo using ground truth verification and to understand how the targets were creating this unique return.
Observations by the NWS Binghamton WSR-88D show that the echo usually first appears during mid August and persists into October. The first returns are observed around the time of sunrise as a point source that subsequently expands outward in all directions to form a ring (Fig. 1). This expansion is typically well depicted by the radar velocity data as a divergence signature (Fig. 2) but very strong nocturnal inversions occasionally lead to loss of data due to range folding. There is variability in the exact echo configuration with some mornings exhibiting a nearly symmetric ring while others show more of an amorphous area with little or no symmetry. Meteorologically, the best time to observe the radar echo is the morning after good radiational cooling with little fog and/or stratus development. These synoptic conditions can lead to some anomalous propagation (AP) but the quiet weather associated with these conditions allow the radar to operate in its most sensitive clear-air mode. The radar echo in question appears to be relatively shallow since it is only detected by the lowest radar elevation scan (0.5 degrees).
Of critical importance to this investigation will be real time data gathered in the field from near the point where the echo appears to originate. After examining several occurrences, a point of origin was determined to be at an azimuth of approximately 325o and a range of approximately 60 nautical miles (nm) from the radar. This position is at the north end of Cayuga Lake which is one of several long and narrow lakes formed by glaciers that make up the Finger Lakes region of central New York. The area our team focused on was in the vicinity of a large marsh, a portion of which is designated as the Montezuma National Wildlife Refuge (NWR) (Fig. 3). This wetland location, combined with the time of year for the echo occurrence, lead to speculation that the reflectors in question were composed of migrating birds, initially thought to be some sort of water fowl. Other theories involved insects and deformation of the nocturnal inversion caused by the boundary layer over the marsh.
After coordinating with the personnel at the refuge, three field trips were arranged with one outing involving a special permit for an overnight stay. The outings were scheduled for 4-5 September, 6 September, and 14 September 1997. 35mm photographs and video were used to document any possible radar reflectors.
The overnight trip on 4-5 September got off to a promising start. The synoptic conditions allowed for radiational cooling with a clear and calm night but no fog had formed so our view would be unobstructed. We took up a position along the eastern side of the Main Pool in the refuge (Location 1 in Fig. 4) at approximately 0830Z with sunrise occurring at approximately 1033Z. Surface obstructions were not present within approximately 200 feet of our observing point. After waiting patiently until 1100Z, no significant reflectors were observed to serve as a source for a radar echo. The radar did detect a significant radar echo but it appeared to be centered approximately 4km further south in the refuge and thus out of our view. The refuge biologist indicated his belief that the echo may be related to the departure of blackbirds from their nightly roost in the many cattails of the marsh. Thus, we attempted to secure a position closer to such wetland vegetation on our next attempt.
On 6 September, a point further south along a lock on the north end of Cayuga Lake was selected for a point of observation (Location 2 in Fig. 4). However, sky conditions deteriorated faster than anticipated with a solid overcast of stratocumulus clouds, intermittent showers, and approximately a 15 mph south wind blowing up the axis of the lake. These adverse weather conditions prohibited direct observations and it was necessary to postpone future attempts until the synoptic situation became more favorable and time constraints were not as great.
The morning of 14 September appeared to offer one of the last good opportunities of the season to document the source of the radar echo. The synoptic pattern featured a large area of high pressure building into the region with the center of the high very close to central New York. This allowed for radiational cooling with clear skies and light winds. However, dew points were relatively high so some fog formation was expected. After coordinating with the personnel at Montezuma NWR and discussing our previous trips, it was suggested that a location closer to the Tschachie Pool may yield better results (Location 3 in Fig. 4). This location also offered an observation tower with a platform approximately 20 feet above the ground.
Upon our pre-dawn arrival at the observation tower we found patches of dense ground fog throughout the refuge. The platform offered a position above most of the vegetation but fog occasionally limited the visibility to less than one half mile. As sunrise approached, we began to hear wildlife stirring in the brush. The initial activity appeared to be limited to small numbers of assorted water fowl. However as the sky continued to brighten large numbers of other birds joined the activity and eventually produced a volume of squawking and cawing comparable to standing near a waterfall.
After several minutes of waiting in the first few minutes of early morning sun, with nothing more than small, isolated flocks of birds flitting about, we observed an enormous flock of blackbirds lifting up from the brush near Tschachie Pool and flying to the southwest (Fig. 5). Several minutes later, another flock containing what appeared to be thousands of birds lifted up and flew in a southerly direction, followed by yet another large flock minutes later. During this period of bird activity, the Binghamton doppler radar detected an expanding, a-symmetric echo configuration over Montezuma NWR (Figs. 6 and 7). The radar data was consistent with past observations and correlated well with the visual sighting of the birds.
In all, we observed five large flocks of blackbirds lift off and fly away from Montezuma NWR in a southerly direction. As we left the refuge, it still sounded as if there were a fair number of birds remaining in the marsh but the majority appeared to have departed. The birds in each flock remained in narrow corridors within a few feet of each other as they went off to feed.
There are several factors that contribute to the detection of these birds by the NWS Binghamton WSR-88D, the two most important being: (1) blackbird behavior patterns; and, (2) basic principles governing radar detection of non-meteorological echoes. A detailed discussion of these factors follows:
4.1 Blackbird Behavior Patterns
4.1.1 Migration (Nero 1984)
Each major migratory path, known as a flyway, remains the same for fall and spring migration. The blackbirds studied at the Montezuma NWR follow the Atlantic Flyway that originates in Canada and follows the eastern seaboard of the United States southward. The birds usually remain in a tight formation separated by no more than 4 to 8 inches as they cruise at altitudes normally 200 ft to 900 ft above ground level (AGL). Since the average blackbird may live to the age of 10 years, it is likely that individual birds follow the same flyway each year and frequent the same feeding and roosting areas along the path.
4.1.2 Blackbird Roosting (Orians 1985)
Migrating blackbirds tend to live in roosting groups of 200 or more, and are commonly found across central New York during the summer and fall migratory seasons. Although blackbirds are highly territorial during the breeding season, they become more tolerant of other species during migration. This allows large collections of groups, known as colonies, to roost together at night during the migratory season. The major benefit they receive from roosting together is protection. However, this protection is primarily derived from their preferred roost location (marshes) and not necessarily due to the size of the colony. Major roosts, such as the one investigated at the Montezuma NWR, have been known to host anywhere from one hundred thousand to over one million birds.
4.1.3 Blackbird Feeding Groups (Orians 1980)
During the early morning hours, individual feeding groups break from the colony and take off from their roosts. Visible light is the mechanism that triggers the birds to take flight. Blackbirds partially depend on the sun for orientation, and rely on this orientation to return to the same feeding areas each day. If the birds are forced to remain in their roosting place due to inclement weather, they feed on local insects and shellfish that live within the marsh.
The members of the feeding groups are determined by bird species and sex. Adult males form their own groups while females and young form others. These pairings are not exclusive however, and groups may occasionally contain both sexes and all ages. Although group species was very difficult to ascertain during the field trips to Montezuma NWR, the two primary marsh-dwelling species of blackbirds are red-wing and yellow-headed, with only red-wings being indigenous to central New York. There is no evidence to suggest that there is an individual bird who leads the feeding group in flight. However, there is a definitive dominance hierarchy regarding feeding rights once the group has arrived at its destination.
The feeding groups have distinct feeding areas and typically follow a similar route from their nightly roosting place. Blackbirds may travel over 100 km day to find a suitable feeding area. The feeding areas remain the same for a period of several days which allows each group of the colony to return to the same roosting place each night. The feeding areas change as the groups search for more productive grounds and eventually lead the colony to a new roosting place along the flyway.
4.2 Basic Principles Governing Radar Detection of Non-Meteorological Echoes
The WSR-88D doppler radar located at Binghamton, NY, represents a substantial improvement over previous weather radars. However, the WSR-88D is still bound by the same physical laws that affect the propagation of electromagnetic waves through the atmosphere. In order to create uniformity among the various types of radar data, radar equations always assume that the atmosphere is in a "standard" state regarding variables such as temperature, pressure, humidity, etc (Green 1990). Unfortunately, this can lead to inaccurate radar information during cases when the atmosphere varies from this "standard" state, which can happen on a daily or even hourly basis. The times when the atmosphere is in a "non-standard" state is when the majority of non-meteorological echoes are detected.
The WSR-88D functions by emitting pulses of microwave energy in a narrow, conical shaped beam from its antenna. The beam pulses interact with objects in the atmosphere, reflecting some of the energy back to the radar. The radar uses this returned energy to calculate data regarding the character of any reflectors encountered. Under "standard" atmospheric conditions, the beam travels out from the antenna and is refracted downward by the air at a presumed index of refraction (Fig. 8). At the same time, due to the curvature of the earth, the center of the beam becomes higher and higher above the ground. The result is that under "standard" atmospheric conditions, even at its lowest elevation angle of 0.5o the center of the radar beam is actually more than 7,400 feet AGL at a distance of only 60 nm (U.S. Dept. of Commerce 1991). The Montezuma NWR is located approximately 60 nm from the radar at Binghamton so at this range the radar estimates the center of the beam to be approximately 7,400 feet AGL over the refuge.
It is also important to note that the conical shape of the beam spreads out as the distance from the radar increases. At the 60nm range of the refuge, the beam width of 0.95o has expanded to a diameter of approximately 5,700 feet (Green 1990). Since the center of the beam is approximately 7,400 feet above the refuge with a diameter of 5,700 feet, the beam may detect reflectors between 4,550 feet and 10,250 feet AGL. If this is indeed the case, it initially does not seem possible for the radar to detect the birds at all, since blackbirds fly between 200 feet and 900 feet AGL. This is explained by the physics of radar beam propagation through the atmosphere as well as the condition of the atmosphere itself.
One type of "non-standard" atmosphere results in what is called super-refraction of the radar beam. This condition is caused by a low level temperature inversion which causes the index of refraction of the atmosphere to increase with increasing altitude (Green 1990). The increasing index of refraction causes a corresponding increase in the downward curve of the radar beam as it moves away from the radar (Fig. 8). In some cases when the inversion is particularly strong, the beam will be curved to such an extent that it encounters the ground. The resulting strong returns are commonly referred to as anomalous propagation (AP) returns.
During super-refractive conditions the actual radar beam will be considerably lower in altitude at any given distance from the radar than the radar display indicates. This will lead to the radar assigning higher altitudes to echoes at a given range than are actually present. On the morning of 14 September, the upper air soundings from both NWSFO Albany, NY and Buffalo, NY indicated that the atmosphere was likely in a super-refractive state (Fig. 9) with inversions present near the surface and between 5000 and 10,000 ft. Thus, at an elevation angle of 0.5o and a range of 60 nm, it can be surmised that the actual center of the radar beam when detecting the bird echoes was much lower than the 7400 feet AGL indicated by the radar display.
As previously mentioned, the spreading of the conical radar beam also must be taken into account. At a distance of 60 nm the beam has expanded to an approximate diameter of 5700 feet. This would mean that if the birds were flying between 200 feet and 900 feet AGL, the center of the beam would have to be at an elevation of approximately 3000 feet to3800 feet AGL for the birds to intersect the bottom of the radar beam at this range. Given the fact that the atmosphere was super-refractive and that the radar did indeed detect the birds, the data suggest that the center of the beam was no greater than 3800 feet AGL over Montezuma NWR on the morning of 14 September. It is possible that the beam was also "deformed" by the non-standard conditions but a detailed examination of radar beam structure in a non-standard environment is beyond the scope of this paper.
It is also important to realize that the WSR-88D doppler radar was primarily designed to detect liquid precipitation targets which reflect, absorb, transmit, and re-radiate the energy from the radar beam in many directions. This results in only a fraction of the beam energy being returned to the radar for analysis. When the beam encounters an object which returns more energy to the radar than a liquid precipitation target, it still depicts the relative strength of the echo according to a returned power to intensity relationship that assumes a liquid target. The echo intensity is then converted to a relative color scale for the display screen. Also, since radar reflectivity is related to the 6th power of the target diameter (U. S. Dept. of Commerce 1991), one large target can return as much energy as many smaller ones. This being the case, one would expect that a bird in flight, being much larger than a liquid droplet, would return a much stronger echo than a raindrop. However, the radar display from 14 September shows the bird echo with returns normally associated with very light precipitation or clouds. The explanation for the weaker than expected echoes lies in the problems associated with radar beam spreading.
With a beam width of more than one mile over Montezuma NWR, it is highly unlikely that the birds flying in a narrow flight zone completely fill the radar beam. This partial beam filling has a direct impact on how strongly the birds are displayed on the radar screen. If a target (i.e. flock of birds) does not completely fill the radar beam only a portion of the available transmitted energy will be returned to the radar. The radar functions under the assumption that whatever reflectors are encountered completely fill the radar beam. So if a highly reflective target was encountered but only filled a small portion of the beam, even if it returned almost all of the energy it encountered, the returned energy would be only a small part of the total transmitted energy available and would cause the target to be displayed weaker than anticipated. The assumption of complete beam filling is not accurate for the blackbirds which fly in a narrow band only a couple of hundred feet thick and is a major contributor to the echo being depicted much weaker than otherwise expected.
Although there was much speculation as to the composition of the reflectors in the ring shaped echo seen by the WSR-88D located at the NWS office in Binghamton, NY, our field verification study indicated the echo to be comprised almost exclusively of several species of blackbirds. A large proportion of red-winged blackbirds is suspected but could not be verified due to low light conditions and reduced visibility in ground fog. The appearance of the echoes during the late summer and early fall corresponds to the southerly migration of the birds along the Atlantic Flyway. The echo originated from the Montezuma NWR, a wetland habitat consistent with the known roosting preferences of blackbirds. The biological habits of blackbirds allow the echo to appear around the time of sunrise on relatively clear mornings during the migratory period. The echo dissipates quickly due to the dissipation of super-refractive atmospheric conditions after sunrise and to the decrease in bird density corresponding to the arrival of the birds at their feeding grounds. The birds exhibited flight patterns consistent with the known feeding behavior of blackbirds and correlated well with radar velocity data associated with the echo configuration.
The radar beam propagation through a super-refractive atmosphere in conjunction with beam width spreading allowed for the detection of the birds at their natural flight level. This is a much lower altitude than was estimated by the radar, which does not account for the super-refractive conditions present on the morning of 14 September. Beam width spreading also resulted in a weaker echo return than expected because the narrow flight zone of the birds (200 feet AGL to 900 feet AGL) returned only a small portion of the available transmitted energy along the 5,700 foot beam width over Montezuma NWR .
The detection of migrating blackbirds by the WSR-88D in Binghamton, NY, has the potential to provide very useful data to external users outside of the meteorological community. The radar data provides a very good idea of where the birds roost and when they become active in the morning. Their flight paths can be tracked to locate their feeding grounds. The arrival and departure of the birds can be documented annually to study migratory trends. All of this data can be used by an ornithologist to assess the status of the flock and possibly the health of the ecosystem as a whole.
Aviation interests can also benefit from the knowledge of a possible low level bird hazard near the time of sunrise. Bird flights of this type have also been known to contaminate wind profiler observations and WSR-88D VAD winds (Jungbluth et al 1995, Miller et al 1997, Wilczak et al 1995). This contamination could lead to poor numerical model initiation and subsequent model performance. We are hopeful that this study has illustrated the potential utility of radar data beyond the traditional scope of severe weather warnings.
Weather radars are extremely useful tools for analyzing remote portions of the atmosphere. It is important for meteorologists to be familiar with atmospheric conditions and their impact on radar performance to avoid misinterpretation of radar data. This is the challenge facing all weather radar operators as they attempt to warn of hazardous and life threatening conditions associated with meteorological and non-meteorological phenomena alike.
The authors would like to thank the employees of the U.S. Fish and Wildlife Service at Montezuma NWR for their extensive co-operation during the period of study. Their vast knowledge of the refuge ecosystem was a valuable asset to our research.
Eastwood, E., 1967: Radar Ornithology. Methuen & Co., Ltd., London, England, 278 pp.
Edwards, J. and E. W. Houghton, 1959: Radar echoes area polar diagrams of birds. Nature, 184, p 1059.
Green, D.G., 1990: OTF Module 1, Principles of Weather Radar. NEXRAD Operational Support Facility, Norman, OK, 73 pp.
Jungbluth, K., J. Belles, M. Schumacher, and R. Arritt, 1995: Velocity contamination of WSR- 88D and wind profiler data due to migrating birds. Preprints, 27th Conference on Radar Meteorology, Vail, CO, Amer. Meteor. Soc., 666-668.
Miller, P.A., M.F. Barth, and J.R. Smart, 1997: The Extent of Bird Contamination in the Hourly Winds Measured by the NOAA Profiler Network: Results Before and After Implementation of the New Bird Contamination Quality Control Check. Preprints, First Symposium on Integrated Observing Systems, Long Beach, CA, Amer. Meteor. Soc., 138-144.
Nero, R., 1984: Redwings. Smithsonian Institute, Washington, D.C., 160 pp.
Orians, G., 1980: Some Adaptations of Marsh Nesting Blackbirds. Princeton University Press, Princeton, NJ, 296 pp.
Orians, G., 1985: Blackbirds of the Americas. University of Washington Press, Seattle, WA, 163 pp.
U. S. Department of Commerce, 1991: Doppler radar meteorological observations. Part B: Doppler radar theory and meteorology, Federal Meteorological Handbook No. 11, NOAA/OFCM, Washington, DC, 227 pp.
Wilczak, J.M., R.G. Strauch, F.M. Ralph, B.L. Weber, D.A. Merritt, J.R. Jordan, D.E. Wolfe, L,K. Lewis, D.B. Wuertz, J.E. Gaynor, S.A. McLaughlin, R.R. Rogers, A.C. Riddle, and T.S. Dye, 1995: Contamination of wind profiler data by migrating birds: Characteristics of corrupted data and potential solutions. Journal of Amospheric and Oceanic Technology, 12, 449-467.
|National Weather Service
Binghamton Weather Forecast Office
32 Dawes Drive
Johnson City, NY 13790
Page last modified: July 12, 2007