Parametric criteria for the detection of noise-like signals

A parametric criteria for the detection of noise-like weather returns is presented. Assuming only a pulse doppler radar with synchronous linear phase detection for the physical system, probability theory is applied to spectral observations in a detailed assessment of signal existence and doppler frequency content. Three probability considerations are addressed. Two of the considerations specify the likelihood of a weather signal existing, having frequency content fi, and exceeding a specified spectral power at frequency f i. The third consideration investigates false-alarm probability or the likelihood of system and environmental noise being mistaken as a signal when detection thresholds are optimized. Computer programs for the computation of signal and noise probabilities are presented.