Beyond single interferograms: inference of subsurface processes with subtle deformation signatures from time series InSAR
Howard A. Zebker, Paul Segall Stanford University
InSAR, an exceptional technique for visualizing and quantifying crustal deformation patterns, has largely been limited in application to arid environments. Subtle phenomena such as interseismic strain accumulation or pre-eruptive deformation of volcanic edifices can produce surface signatures that are undetected in single radar interferograms because of i) temporal decorrelation noise due to vegetation and ii) propagation through a spatially-variable troposphere. We propose to view these small signals using data acquired by Envisat, a new satellite launched by ESA in March 2002. This new radar incorporates electronic beam steering, which permits multiple observations of a given study site during a single orbit cycle. Furthermore, rapid beam switching enables a ScanSAR capability for imaging wide swaths. Although the satellite is to be orbited in a 35 day repeat orbit similar to that of ERS-1 and ?2, ScanSAR can yield much denser temporal data, with each point on the ground imaged four times during each cycle. We propose here to develop algorithms and methods to produce routine ScanSAR interferograms resulting in dense deformation time series measurements. We also will develop filters based on spatio-temporal models of expected signals and error sources to coax deformations from within the noisy background.
