Abstract
The computerised bird monitoring system devised on the basis of MRL-5 meteorological radar enables to perform around-the-clock automated bird monitoring. The paper presents an algorithm for identifying bird signals against a background of other objects. The radar signal reflected from a flying bird changes its location in space. In case the flight direction remains unchanged, the coordinates of the corresponding radio echoes obtained as a result of successive azimuth scans form a straight line. This enables to isolate bird radio echoes from those of motionless or chaotically moving objects, such as ground clutter, clouds, precipitation, atmospheric inhomogeneities, etc. In the proposed algorithm, the coordinates thus obtained are used for estimation of vectors of individual birds and bird groups.
In order to identify a bird signal, its fluctuation pattern is analysed. A fluctuation pattern is determined by the frequency of wing-flapping that is characteristic of a certain bird species. The analysis of the spectra of signals obtained in the study indicates that the proposed algorithm enables to isolate bird radio echoes with approximately 85% confidence. The technique implying simultaneous measurements of signal fluctuation pattern and of the effective scattering area (ESA) forms a promising basis for further studies aimed at determining the species of flying birds by means of radar.