Aviation Bird Hazard in NEXRAD Dual Polarization Weather Radar Confirmed by Visual Observations

Abstract

Acknowledgements. This research benefited from the government-sponsored NSF Unidata program and software, specifically the IDV software package provided by Unidata (www.unidata.ucar.edu/software/). We thank three anonymous reviewers who made suggestions that strongly improved the manuscript, and David Ison and Chip Wolfe of IJAAA who assisted with the inclusion of the animation in the metadata for this article.

Discrimination of Biological Scatterers in Polarimetric Weather Radar Data: Opportunities and Challenges

“…We suggest that data on validated scatterer identity, density, and body orientation with respect to the radar; spatial patterns of echo return; and natural history will advance approaches to bioscatter classification. The resulting future algorithms will enable the application of data from weather surveillance radar to assist in improving flight safety [67,84], monitoring the movements of agricultural pest and disease vectors [85], making informed conservation decisions [86,87], and providing the public with information that will enhance wildlife viewing opportunities [88].…”

“…Once dual-polarization was available, some investigators used one or more of the approaches mentioned above to identify the type of scatterer and to characterize the values of the polarimetric variables produced by that type of scatterer [36]. In most of the studies that have reported values of polarimetric variables from biological scatterers, the authors made assumptions about the identity of the type of scatterer before the analysis of the polarimetric data, but in a few cases the investigators knew the identity of the scatterer [66] or they knew about the departures of purple martins and other swallows from nighttime roosts [18,51,67,68] and the exodus of free-tailed bats from their daytime roosts [18,49,51,69]. In our study, we recorded legacy and polarimetric data from known biological scatterers and rain and generated histograms of values (Figure 1) for each type of scatterer.…”

“…Where the natural history of flying animals is well established, numerous opportunities exist to develop narrowly trained classification models as event detectors; for example, waterfowl departing known lakes and reservoirs [47] or martins, swallows, and bats departing traditional roosts [49,67,77]. Exceptionally high performing models may be able to screen for uncommon or rare events, e.g., occurrence of a taxon in an unusual geographic location or at an unlikely time.…”