An operational use of ground clutter to monitor radar performance (“ground clutter—a radar's best friend”)

An operational use of ground clutter to monitor radar performance

(“ground clutter - a radar's best friend”)

 

Zlatko R. Vukovic¹, Jim M.C. Young¹, and Norman Donaldson²

¹National Radar Program, Radar and Upper Air Division, Weather and Environmental Monitoring Directorate, Meteorological Service of Canada

²Cloud Physics and Severe Weather Section, Meteorological Research Division, Atmospheric Science and Technology Directorate, Science and Technology Branch

Abstract

Most of the time ground clutter contaminates radar data and filtering is required to remove such a “contamination”.  From an operational perspective, ground clutter can be used as a reference signal for monitoring the stability of several variables such as changes in transmitted/received power and/or antenna beam pointing accuracy (azimuth/elevation) from scan to scan. There are different approaches to using ground clutter as a monitoring tool. The success of a tool will depend upon (1) how many points were used for the ground clutter, (2) methodology of determining when there is clear weather to monitor the ground clutter signal, and (3) what percentage of time during the year it can be used.

This paper presents an algorithm that first identifies ground clutter by seasonally analyzing a radar bin cumulative distribution function (BCDF) of combined precipitation and clutter reflectivity obtained on an hourly basis. The radar bins that exceed 95% of BCDF for a season were considered as ground clutter bins. Secondly, in the operational, real-time regime, after each scan the number of bins (N_all) with reflectivity greater than the arbitrary threshold (15 dBT) was obtained and compared with the reference number of bins for relatively clear weather (N_clear). Whenever N_all was less than N_clear the ground clutter monitoring was declared for possible use. The reference number N_clear was obtained by analyzing seasonal data when distribution and time series of mean reflectivity from all bins distinctly indicated that it was clear weather within an acceptable certainty. The threshold range of acceptance of averaged dBT from the clutter was also obtained (dBT_min, dBT_max).

The purpose of the paper is to define a statistical methodology for obtaining ground clutter bins and to determine the criteria when they could be successfully used for calculating an average reflectivity from the clutter bins (dBT_clutter). The main monitoring parameter is dBT_clutter and whenever it is out of dBT_min-dBT_max clutter range a warning is issued.  The methodology was demonstrated in real-time data for an operational C-band radar at Exeter, Ontario, Canada (CWSO). The optimal threshold range of acceptance (dBT_min,dBT_max) regarding to the percentage of usage time  during a year and sensibility of performance anomaly detection was also explored.