Turbulence detection by Doppler spectrum shape analysis

This paper briefly describes new methodology of turbulence detection and classification. Proposed method is based on spectrum shape analysis and does not require traditional calculations. Any image recognition technique can be used for shape processing. This allows reduction of calculation time and can be successfully used in warning system. Some results of spectrum shape analysis with use of neural network are shown in comparison with classical Doppler spectrum width estimation. Real data are used to check the developed approach.