Disambiguation of range-Doppler radar measurements

Abstract

Observing the Moon by a monostatic synthetic-aperture radar system has the inherent problem that any given combination of range and Doppler shift in a measurement will map to two different regions on the lunar surface. This ambiguity has previously been avoided using an interferometric radar configuration or selective illumination. The objective of this thesis was to write and validate a method for disambiguating monostatic inverse synthetic-aperture radar measurements of the Moon. The method implemented in this thesis was conducted by simulating multiple range-Doppler radar maps with differing apparent rotation axes, based on data from NASA's Horizon ephemeris [13] and optical reflectivity measurements from the Lunar Reconnaissance Orbiter Camera [6]. Included in the simulation were the Hagfors scattering law to allow scaling the amount of backscattered power based on the incidence angle of the electromagnetic wave, range dependent power reduction and noise due to random surface and subsurface undulations. The disambiguation, thus estimation of the true normalized scattering cross-sections, was conducted by solving an overdetermined linear least square system of the simulated maps. The disambiguation based on three range-Doppler maps resulted in an error standard deviation of 18.56% of the average reflectivity value, which decreased by including additional maps in the procedure. The method described in this thesis was found to be unbiased and can be used with EISCAT3D when it becomes operational in the future, and existing Arecibo measurements of Venus [3].