Correlation Technique for Navigation of Mobile Dual-Doppler Network Data

Mobile dual-Doppler networks, such as the Doppler On Wheels (DOWs), can collect high resolution data during rapid and multiple deployments, even while moving. In order to retrieve accurate dual-Doppler wind vectors, extremely precise navigation of the data is required. Unlike stationary radars, however, the precise orientation of the radars at each deployment is not known. Since the radars do not utilize inertial navigation or similar systems, but only a lower accuracy and much less expensive flux-gate compass system, the orientation of the radars must be obtained post facto through the analysis of the data.

Identification of clutter targets such as roads and power lines can provide orientation information, as can the identification of common, sharply defined weather features. However, this process is tedious and subjective. A correlation technique has been developed to obtain the orientations more precisely and quickly. With knowledge of the precise locations of the radars (from GPS navigation), and a first guess as to the orientation of the radars, the reflectivity fields from both radars are overlaid in a common grid and a correlation coefficient calculated. The location and orientation of the radars are perturbed around this first guess result in an effort to maximize the correlation coefficient.

This technique is evaluated on several data sets including those in tornadoes and a hurricane.