Mapping the electromagnetic (E-M) field in the vicinity of HF radar antenna arrays

HF radars measure ocean currents through the Doppler shift of E-M waves back-scattered by surface gravity waves. While modern clocks and digital synthesizers yield range errors negligible compared to bandwidth-limited range resolution, azimuthal calibration is a necessity for both direction-finding and beam-forming antenna arrays. Errors in amplitudes/phases of received E-M waves are commonly surveyed by transmitters or transponders mounted on boats or unmanned aircrafts, or by echoes from ships tracked by AIS/GPS. In clean environmental settings, azimuthal errors are often negligible, but in many real-world installations reflections, multiple paths, standing waves, propagation anomalies and shoreline refraction result in radial velocity field distortions, azimuthal biases and beam-forming side-lobes.

To understand the causes of E-M phase and amplitude fluctuations in vicinity of HF radar antenna arrays, in particular in presence of reflecting structures, several procedures for mapping the E-M field from a 16 MHz monochromatic source deployed at a distance of 200 m were tested at a dune site devoid of metallic masses. Using a calibrated magnetic loop antenna on non-conducting tripod, spatial variations of E-M field strength and propagation direction were mapped. Using a pair of non-resonant active antennas at the ends of a fiberglass boom, directions tangent to E-M wavefronts were identified by minimizing the differential phase across the pair. While both methods yield results consistent with predicted E-M propagation, significant constructive/destructive interferences between direct E-M wave from the source and E-M waves reflected on shoreline ocean breakers (or Bragg-scattered at close range cells) limited the stability of amplitude/phase measurements, despite disproportionately long averaging times of several minutes.

Using a fixed non-resonant reference receive antenna, and an identical receive antenna on a mobile tripod connected by a 100-m umbilical cable, finite length equal-phase lines were mapped at spacings of 1 m and 1-cm position resolution. Fluctuations of phase lines in the propagation direction were <λ/100 at >80 m from the shore line, but increased to >λ/20 at 20 m due to ocean interferences. Similar measurements made at more industrial sites suspected to be subject to significant multiple paths will be shown at the meeting.