Linear Combinations for Differential Radar Interferometry

Synthetic aperture radar interferometry (InSAR) has been widely used over the past two decades for such geodetic applications as topographic mapping and ground deformation monitoring. In particular, in contrast to other geodetic techniques, differential InSAR has the capability of measuring surface deformation with a measurement precision of a few millimetres and a spatial resolution of a few tens of metres or less, over areal extents of thousands of square kilometers. However, the ground surface may deform by a large amount within a small area due to localised effects such as earthquakes, underground mining, groundwater extraction, and others, which may cause severe damage to manmade surface and underground structures. Current satellite radar interferometry, because of its single-frequency signal structure, cannot measure deformations with large horizontal gradients as they produce very dense fringe patterns. The upper limit of the deformation gradient is determined by the signal wavelength and pixel spacing. Although the longer wavelength of the radar signal is less susceptible to high deformation gradients, loss of correlation often still occurs even when L-band imagery is used In this paper, the authors propose a method to effectively monitor such large-gradient deformation using differential InSAR by linear combinations of interferograms acquired from dual-frequency sensors. Some simulated interferograms are generated through the combination of the commonly used radar wavelengths, for instance C, L₁, L₂, S and even P-band. Using appropriate linear combinations, the virtual wavelength of the combined interferogram is flexible enough to be used for different applications. From the simulated results, it is found that the linear combination technique is powerful and can improve the correlation, as well as making the phase unwrapping process much easier.