Interferometric Processing of SLC Sentinel-1 TOPS Data

Abstract

InSAR processing usually involves two successive steps: focusing and interferometry. Most public-domain InSAR processing toolboxes are capable of performing both operations with data acquired in the standard Stripmap mode, starting from raw SAR data (level 0). However , the focusing of burst-mode data, such as TOPS and ScanSAR, requires substantial modifications to standard focusing methods due to the particular spectral properties of these data. Anticipating on this potential difficulty for non-expert users, the European Space Agency has chosen to release Sentinel-1 TOPS data in a Single Look Complex format (level 1). The data are already focused using state-of-the-art processing techniques, with phase information preserved. Even so, the focusing method introduces an additional quadratic phase term in the azimuth direction. In case of a small misregistration error between a pair of images, this residual term leads to steep phase ramps in azimuth that are superimposed on the desired in-terferometric phase. Therefore, this quadratic phase term needs to be removed from the SLC data prior to interfer-ogram calculation. Here, a pre-processing method allowing for compensating this phase term and simply feeding the corrected SLC data into a standard InSAR processing chain is described. The method consists of three steps. The first step uses the metadata in order to reconstruct a continuous image in the azimuth direction, accounting for the small overlap between adjacent bursts ( stitching ). In the second step, multiplication of the images by an appropriate phase screen is performed so as to cancel the azimuthal quadratic phase term ( deramp-ing ). The deramping operation uses the metadata, as well as the azimuth time lag between the images deduced from sub-pixel image correlation, in order to determine small misregistration errors. Misregistration errors are compensated using a simple affine relation deduced from least-square fitting of the azimuth offsets. Following this second step, the azimuth phase ramps are significantly reduced in the corrected interferogram. The third step consists in refining the affine coefficients that account for the misregistration error. The refinement is achieved by differencing the backward-and forward-looking interfer-ograms, exploiting the spectral diversity in burst overlap regions ( spectral diversity ). This final step makes it possible to remove residual phase jumps across burst boundaries with the desired level of accuracy. Azimuth spectral properties of burst-mode SAR data, such as ScanSAR [1] or TOPS [2], are significantly different from those of Stripmap. In burst-mode, the system observes a series of sub-swaths by periodically steering of the antenna in the elevation direction and transmitting a short succession of pulses gathered in the so-called bursts . As a consequence, a given target is imaged only during a fraction of the equivalent stripmap synthetic aperture duration, thereby reducing its duration of illumination. Therefore, the increased swath width comes at the expense of azimuth resolution (Fig. 1). Furthermore, the Doppler centroid of targets within any burst becomes variable in azimuth. Due to the latter effect, the impulse response function of standard burst-mode focusing methods includes a residual phase term that exhibits a characteristic quadratic variation of the az-imuth phase. This phase term needs to be compensated prior to using the phase for interferometric applications, a process referred to as deramping [3, 4]. Figure 1. Schematic representation of the TOPSAR acquisition mode (from [5]). The beam is steered periodically in range, in order to cover several sub-swaths, and progressively in azimuth, aft to fore, in order to increase the azimuth bandwidth, expand the footprint and decrease scalloping.