Automatic supression of anomalous propagation clutter for noncoherent weather raders

Pulse to pulse radar echo fluctuations have been used to distinguish rain from anomalous propagation clutter.
For 1 deg. sectors at a fixed range 14 echo samples from successive pulses are tested for clutter. This test has a clutter detection rate of only 45 %, but in each 1 deg/2 km polar radar pixel 8 range bins can be tested. The 8 tests are shown to be nearly independent and their combined result allows 85% of the clutter pixels to be detected. Finally the pattern of suspect pixels in the completed radar picture is analysed on a scale of about 10 km. The total result is an average removal efficiency of about 98 %.
Only echoes of very light precipitation are erroneously suppressed by this processing: at most 5% of light rain pixels or 15% of light snow pixels are removed.
For on-line application this method requires a fast