Spectral analyses of the dual polarization Doppler weather radar data.

Abstract

Echoes in clear air from biological scatterers mixed within the resolution volumes over a large region are presented. These echoes were observed with the polarimetric prototype of the forthcoming WSR-88D weather radar. The study case occurred in the evening of September 7, 2004, at the beginning of the bird migrating season. Novel polarimetric spectral analyses are used for distinguishing signatures of birds and insects in multimodal spectra. These biological scatterers were present at the same time in the radar resolution volumes over a large area. Spectral techniques for (1) data censoring, (2) wind retrieval and (3) estimation of intrinsic values/functions of polarimetric variables for different types of scatterers are presented. The technique for data censoring in the frequency domain allows detection of weak signals. Censoring is performed on the level of spectral densities, allowing exposure of contributions to the spectrum from multiple types of scatterers. The spectral techniques for wind retrieval allow simultaneous estimation of wind from the data that are severely contaminated by migrating birds, and assessment of bird migration parameters. The intrinsic polarimetric signatures associated with the variety of scatterers can be evaluated using presented methodology. Algorithms for echo classification can be built on these. The possibilities of spectral processing using parametric estimation techniques are explored for resolving contributions to the Doppler spectrum from the three types of scatterers: passive wind tracers, actively flying insects and birds. A combination of parametric and non-parametric polarimetric spectral analyses is used to estimate the small bias introduced to the wind velocity by actively flying insects.