Reducing the effects of noise on atmospheric imaging radars using multilag correlation

Authors

  • K. D. Le,

    1. Atmospheric Radar Research Center, University of Oklahoma, Norman, Oklahoma, USA
    2. School of Electrical and Computer Engineering, University of Oklahoma, Norman, Oklahoma, USA
    Search for more papers by this author
  • R. D. Palmer,

    1. Atmospheric Radar Research Center, University of Oklahoma, Norman, Oklahoma, USA
    2. School of Meteorology, University of Oklahoma, Norman, Oklahoma, USA
    Search for more papers by this author
  • B. L. Cheong,

    1. Atmospheric Radar Research Center, University of Oklahoma, Norman, Oklahoma, USA
    Search for more papers by this author
  • T.-Y. Yu,

    1. Atmospheric Radar Research Center, University of Oklahoma, Norman, Oklahoma, USA
    2. School of Electrical and Computer Engineering, University of Oklahoma, Norman, Oklahoma, USA
    Search for more papers by this author
  • G. Zhang,

    1. Atmospheric Radar Research Center, University of Oklahoma, Norman, Oklahoma, USA
    2. School of Meteorology, University of Oklahoma, Norman, Oklahoma, USA
    Search for more papers by this author
  • S. M. Torres

    1. Atmospheric Radar Research Center, University of Oklahoma, Norman, Oklahoma, USA
    2. Cooperative Institute for Mesoscale Meteorological Studies, University of Oklahoma, Norman, Oklahoma, USA
    Search for more papers by this author

Abstract

[1] Atmospheric imaging radars offer the capability to scrutinize structures within the illuminated volume at high temporal and spatial resolutions. The retrieval of the mean signal power using an imaging radar is obtained by subtracting the noise power from the covariance function at lag zero. The results obtained at low signal-to-noise ratio (SNR) are problematic when the noise power is unsuccessfully estimated, and are difficult to interpret when adaptive weights are used because of the temporally varying noise power. In this paper, a processing technique that improves the retrieval of the mean signal power by exploiting the temporal correlation difference between the desired signal and system noise is presented. Simulations of its performance are presented for the special case of a Gaussian received spectrum for variations in SNR, normalized spectrum width, and number of time series samples. The technique is also applied to real data collected with the Turbulent Eddy Profiler, a vertically pointing phased array radar developed at the University of Massachusetts, between 1435 and 1457 UTC 15 June 2003. Even though the performance of this technique in terms of its variance and bias depends on the SNR, spectrum width, and number of time series samples, results from both simulations and real data are promising as an enhanced mean signal power in the low SNR regions is obtained.