Abstract
Various nonatmospheric signals contaminate radar wind profiler data, introducing bias into the moments and wind velocity estimation. This study applies a fuzzy logic–based method to Doppler velocity spectra to identify and eliminate the clutter echoes. This method uses mathematical analyses and a fuzzy inference system applied to each Doppler velocity spectrum to separate the atmospheric signals from the clutter. After eliminating the clutter, an adaptive algorithm is used to estimate mean Doppler velocities accurately. This combination of techniques is applied to the spectral data obtained by the newly developed 53-MHz active phased array radar located at the National Atmospheric Research Laboratory (NARL), Gadanki, India (13.5°N, 79°E). Winds derived using the conventional method and the method developed for this study are compared with those obtained by collocated GPS radiosonde. The comparison shows that the present method derives the winds more accurately compared to the conventional method.