Statistical quality control of wind profiler data
AssigneeThe United States of America as represented by the Secretary of Commerce
InventorsJames R. Jordan, Richard G. Strauch
Life of this patent
Estimating the term…
Abstract
The quality of wind profiling radar data is evaluated by comparing the probability density function of the power density of sets of returned samples to an exponential function. The standard deviation of individual samples from the exponential function can be evaluated to identify possible sources of contamination of the wind profiler data.
Claims20
A method for measuring wind velocities by measuring radar return signals back-scattered from air turbulence, including evaluation of the quality of the measured wind velocity, comprising the steps of:
- determining a plurality of wind velocity spectra, each wind velocity spectrum being determined by:
- obtaining a number of sets of samples of radar return data from a particular spatial location;
- and
- calculating the Fourier transforms of each of said sets of samples of radar return data to determine a corresponding wind velocity spectrum;
- and
- averaging said plurality of spectra together in order to provide an average wind velocity value for that particular spatial location;
- and
- evaluating the quality of the average wind velocity thus provided, by:
- measuring the power density of the radar return data of each set of samples;
- calculating a probability density function (PDF) with respect to the measured power density values of one or more of said sets of samples;
- comparing said calculated PDF to a stored PDF, said stored PDF exhibiting variation in power density characteristic of known classes of radar reflectors;
- and
- determining the quality of said wind velocity value responsive to said comparison step.
A method for identifying non-meteorological contribution to back-scattered radar signals as evaluated for wind profiling, comprising the steps of:
- (a) calculating the probability density function (PDF) of a measured characteristic of a number of back-scattered radar signals from a given region used in generation of a wind profiling datum;
- (b) comparing said calculated PDF to a stored PDF representative of radar signals reflected by known radar reflectors, to determine a quality assurance value indicative of the contribution of non-meteorological reflectors to the back-scattered radar signals;
- (c) repeating said steps (a) and (b) a number of times, to determine a plurality of said quality assurance values;
- and
- (d) evaluating the standard deviation of said plurality of quality assurance values, to identify said non-meteorological reflectors.
A method of evaluating the quality of wind velocities as measured using wind profiling radar techniques, comprising the steps of:
- (1) determining a wind velocity spectrum by:
- (a) obtaining a set of samples of radar return data from a particular spatial location;
- (b) calculating the Fourier transform of said set of samples, to determine a velocity spectrum corresponding thereto;
- (c) repeating said steps (a)-(b) to determine a plurality of wind velocity spectra;
- and
- (d) calculating an average wind velocity value from said plurality of wind velocity spectra;
- (2) calculating the power density of the radar return data of each set of samples;
- (3) ordering and normalizing the power densities of the radar return data of each set of samples;
- (4) calculating a probability density function (PDF) with respect to each of said ordered and normalized sets of samples;
- (5) measuring the differences between the PDFs thus determined and a PDF of a set of samples corresponding to radar return data from meteorological reflectors only, to determine the amount of non-meteorological contamination in the corresponding wind velocity spectrum;
- and
- (6) determining an average contamination value corresponding to said average wind velocity value.
Dependencies are inferred from the claim text.