Support: US Air Force Office of Scientific Research (AFOSR)

Period of Performance: June 17th, 2014 — June 16th, 2017

Budget: $670,000

Summary: In our work on spaceborne synthetic aperture radar (SAR) imaging (currently supported by AFOSR), we have shown that the adverse effect of the Earth’s ionosphere on the quality of the image can be attributed to the mismatch between the actual radar signal affected by the dispersion of radio waves in the ionospheric plasma, and the matched filter used for signal processing. Accordingly, to improve the image one should correct the filter, which, in turn, requires knowing the total electron content (TEC) in the ionosphere along the signal path. We have shown that the TEC can be reconstructed by probing the terrain, and hence the ionosphere, on two distinct carrier frequencies, and exploiting the resulting redundancy in the data.

The dual carrier approach can also be reformulated in the split-bandwidth framework, which simplifies its practical implementation. Moreover, we have derived specific quantitative estimates that relate the accuracy of the TEC reconstruction with the degree of improvement that it enables in the image obtained with the new corrected filter. We have also shown that the accuracy of the TEC reconstruction itself can be improved by applying an area-based image registration algorithm to the two original images obtained on two carrier frequencies. Our analysis includes not only the dispersion but also the dissipation of electromagnetic waves in the ionosphere (Ohmic losses), allows for the horizontal variation of the TEC, and differentiates between the image distortions caused by the deterministic and random part of the electron number density (the latter is due to the turbulence).

In the course of the proposed effort, we plan to address a number of important issues that still require attention. We will complete the analysis of the anisotropic case, including the gyrotropy in the plasma due to the magnetic field of the Earth and a possible anisotropy at the target. We will generalize the analysis of scattering at the target beyond the first Born approximation, to explicitly allow for surface roughness and backscattering (Bragg scattering). We will include the analysis of dispersive targets as it may be important for material identification (miSAR), and see to what extent a SAR sensor can tell between the time delay due to target dispersion and that due to having the target at a different distance. Subsequently, we will attempt to distinguish between the target dispersion and the ionosphere dispersion. We will also refine our analysis of the propagation of radio waves through the turbulent ionosphere using the structure functions instead of the correlation functions, and see what additional corrections to the matched filter may be needed to compensate for the stochastic part of the image distortions.