SAR interferometry and permanent scatterers technique

Kirsi Karila

 


Synthetic Aperture Radar (SAR) combines signal-processing techniques with satellite orbit information to produce a high-resolution radar image. Both amplitude and phase information are stored from the returning echoes. Interferogram is a phase difference image of two SAR-images acquired at slightly different positions. When the two images are not taken simultaneously the technique is called repeat-pass interferometry. Interferogram can be converted to a digital elevation model (DEM) with meter accuracy.

   

Besides topography also surface movement can be derived from an interferogram. In differential SAR interferometry (DINSAR) the goal is to separate phase contributions due to topography and displacement in order to get the deformation field. A digital elevation model or another SAR-image pair is needed to get the topographic phase, which will be subtracted from the deformation interferogram. In theory accuracy of a few millimeters can be obtained for the deformation velocity. Nevertheless atmospheric disturbances and orbital errors corrupt the signal and accuracy is reduced. Theoretical accuracy can be achieved by forming time series and averaging many images.

Previous studies made have shown that the problem of temporal decorrelation (change of scatterers in process of time) with repeat-pass interferometry can be solved by using permanent scatterers technique. The procedure involves identifying and exploiting stable natural reflectors (buildings, rocks etc.) called permanent scatterers (PS) using long temporal series of interferometric SAR images. Only PS-pixels are used for motion detection. In numerous previous studies DINSAR and PS have been successfully applied to study earthquakes, volcanoes, glaciers, urban subsidence and landslides (usually) in sparsely vegetated areas.


The objectives of the project is to
1)    verify whether DINSAR and PS techniques are feasible for land uplift determination, and furthermore, to find anomalies in the uplift patterns.
2)    verify is it possible to detect land subsidence in vegetated, northern areas where subsidence areas are known to be small.
3)    find new geodetic applications for INSAR and PS techniques

In co-operation with Finnish Geodetic Institute.