Geodetic Network Design for InSAR

Ground deformation can be monitored with subcentimetric precision from space, using interferometric synthetic aperture radar (InSAR). This technique can sometimes be limited by a low density of naturally occurring phase-coherent radar targets. Measurement densification may be achieved through improvements in processing algorithms and new satellites with better revisit times, but there can still exist areas where very few coherent targets are detected, e.g., in vegetated nonurbanized areas. For third-party end users of InSAR survey results, there is currently no systematic method to determine a priori whether these coherent targets have adequate spatial distribution to estimate the parameters of their interest. We propose such a method, along with a practical solution for measurement densification, i.e., deployment of coherent target devices such as corner reflectors or transponders. We propose a generic network design methodology that does the following: 1) determines whether the naturally occurring InSAR measurements are adequate; 2) finds the minimum number of additional devices (if required); and 3) finds their optimal ground locations. The method digests, as inputs, the expected locations and quality of existing coherent targets, the quality of the devices being deployed, and, if available, any prior knowledge of the deformation signal. At the core of the method is a comparison of different covariance matrices of the final parameters of interest with a criterion matrix (i.e., the desired idealized covariance matrix), using a predefined metric. The resulting network is optimized with respect to precision, reliability, and cost criteria. Simulated data sets and a subsidence case study in the Netherlands are used to demonstrate this method.