Presentations

Recent Advances in Acoustic Monitoring of Birds

Event:
North American Ornithological Conference V
Date:
14-18 August , 2012
Location:
Vancouver, BC
Presenters:
Damoulas, T., Bruns, N., and Farnsworth, A.
Abstract:
Machine learning techniques for automated flight call detection. Invited presentation at the North American Ornithological Conference Symposium.

Method and application developments for migrant flight orientation profiles using polarimetric radars

Event:
American Ornithologists' Union/Cooper Ornithological Society
Date:
28 July - 2 August, 2015
Location:
Norman, OK, USA
Presenters:
Kyle Horton*; Phillip Stepanian; Benjamin Van Doren; Andrew Farnsworth; Jeffrey F. Kelly
Abstract:
Flight strategies of nocturnal migrants in reference to winds aloft have been of great interest, yet have continually yielded mixed results. Much of this attention has been devoted to understanding how migrants organize in reference to wind fields, and whether strategies of compensation or drift are more frequently observed. Migration theory supports the contention that energetic cost should be minimized while maximizing the distance covered in the direction of the intended destination (e.g. breeding grounds, wintering grounds, etc.), but how migrants optimize this strategy is unclear. The 2013 upgrade of all WSR-88D radars to dual- polarization offers a direct assessment of these behaviors. Because animal migration is characterized by the aligned flight of airborne organisms across large spatial expanses, alignments (i.e. migrant orientation) are revealed in polarimetric radar products. These patterns are especially prominent in the polarimetric radar product of co-polar correlation coefficient (ρHV). We introduce a technique for the extraction of these alignments and an assessment of migrant strategies across multiple spatial domains. Using independent radar measures of migrant heading and orientation we explore the propensity of drift and compensation at single radar sites, contrast interior and coastal sites, and explore latitudinal differences.

BirdCast perspectives: autumn nocturnal bird migration in the Northeastern U.S.

Event:
American Ornithologists' Union/Cooper Ornithological Society
Date:
28 July - 2 August, 2015
Location:
Norman, OK, USA
Presenters:
Andrew Farnsworth*; Daniel Sheldon; Benjamin Van Doren; Wesley Hochachka; Frank La Sorte; Kevin Winner; Garrett Bernstein; Steve Kelling
Abstract:
Few studies describe phenology of nocturnal bird migration at broad spatial and temporal scales. Although availability of data from the existing weather surveillance radar network in the continental US (WSR-88D) addresses a fundamental challenge of how and from where to gather data describing such scales continuously, this availability does not address another fundamental challenge of how to extract meaningful information about bird migration from these data. The BirdCast project is a collaborative effort between ornithologists and computer scientists that is tackling the latter challenge. Here, we report results of an extensive processing effort to extract information about bird migration from radar data. We describe the density, direction, and speed of nocturnal migration from 13 WSR-88Ds in the Northeastern US from 1 August to 30 November 2010 and 2011. We discuss the altitudinal distribution of nocturnally migrating birds and behaviors of these migrants relative to wind. Peak nocturnal migration occurred in early in the night and later in the fall. Migrant flight direction was generally south southwesterly. Speeds were highest early in the night and showed extensive variability later in fall. Mean flight altitude was approximately 1000m above sea level with greater variation early and late in fall. Birds drifted with winds inland, compensated for winds in coastal locations. This study highlights the power of extracting bird information from radar data with a cross-disciplinary approach. Furthermore, potential for applying our results and continuing this collaborative approach is great, ranging from providing the science to inform dynamic conservation actions requiring information about the spatio-temporal distributions of nocturnally migrating birds to informing public awareness about the magnitude of these largely unseen movements of birds.

Comparing acoustic and radar monitoring of nocturnal bird migration: Past, present, and future

Event:
American Ornithologists' Union/Cooper Ornithological Society
Date:
28 July - 2 August, 2015
Location:
Norman, OK, USA
Presenters:
Andrew Farnsworth*; Justin Salamon; Matt Robbins; Sara Keen; Juan Bello; Steve Kelling
Abstract:
Nocturnal bird migration is difficult to study, and to quantify, in a comprehensive manner. Although many technologies and techniques exist to examine and to describe these movements in great detail, each has its limitations and imperfections preventing a holistic view. As such, the simultaneous use of multiple approaches is critical for achieving more comprehensive descriptions, quantifications, and visualizations these movements. Here, we discuss the utility of simultaneous acoustic and radar monitoring of nocturnal bird migration. We review the results of previous studies that compare density of nocturnally migrating birds as detected on radar with call counts of nocturnally migrating birds. We also present results from a current pilot project comparing acoustic and radar monitoring. This pilot project used acoustic data from 10 recording stations in Ithaca and New York, NY and radar data from Binghamton and Brookhaven NY weather surveillance radars (WSR-88D). So far, comparative results generally agree with previous studies suggesting a positive but widely variable relationship between bird density and call counts, with the novel twist that these data were generated from automated acoustic and radar processing pipelines. These pilot results, together with previously published knowledge about radar and acoustic monitoring, highlights the power of pairing these methodologies and the capability of providing detailed information about the density, direction, speed, and identity of nocturnal migration. Furthermore, relative to other potential combinations, advances in the automated processing of data streams may make this approach much more cost and time efficient.
Toggle Grid