Open AccessMedian Filter Effects on Radar Wind Profiler Spectral Noise Statistics
Author(s) -
Timothy L. Wilfong,
E. M. Lau,
B. L. Weber,
D. A. Merritt,
Scott A. McLaughlin
Publication year - 2014
Publication title -
journal of atmospheric and oceanic technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.774
H-Index - 124
eISSN - 1520-0426
pISSN - 0739-0572
DOI - 10.1175/jtech-d-14-00070.1
Subject(s) - wind profiler , radar , noise (video) , doppler effect , doppler radar , environmental science , monte carlo method , remote sensing , signal (programming language) , spectral line , physics , filter (signal processing) , acoustics , meteorology , statistics , computer science , mathematics , geology , telecommunications , astronomy , artificial intelligence , image (mathematics) , computer vision , programming language
Radar wind profiler (RWP) systems observe radar returns from refractive index fluctuations due to clear-air turbulence. The Doppler spectra used to compute the moments of the returned signal always include noise from various sources and may contain multiple signals. A critical first step in detecting signals is the objective determination of the noise level in each spectrum. Several spectra may be averaged to improve signal detection. In addition to or instead of a mean, a median may be applied to successive spectra in order to reject transient interference. Monte Carlo simulations were used to examine the effects of the median versus the mean on the objective noise determination. When a median is used, it was found the noise statistics calculations must be slightly modified.