Improving Segmented INSAR Processing Using Presumming

Synthetic aperture radar (SAR) data constitute a shift-variant 2D process, but due to the large disparity between the time scales in the two coordinate bases (range and azimuth), the signal can be well approximated as two 1D processes. The range-Doppler (RD) method correlates the rapidly-varying pulse echo in the range dimension, and then correlates the result in the slowly-varying azimuth dimension. New SAR processing methods have been developed in recent years. However, these new algorithms have not addressed the phase errors associated with processing long, continuous strips of interferometric SAR (INSAR) data. When processing long strips of INSAR data, the strip must be divided into segments, each with its own set of Doppler parameters. The usual rectifying operations are not adequate for removing observed phase offsets between these segments. A procedure known as presumming may reduce these errors by rejecting aliased frequencies in the azimuth spectrum. The objective of this study is to implement presumming in the existing NASA/JPL TOPSAR processor and determine whether it reduces the observed segment discontinuities in the INSAR data.