Assessing the Limits of Troposphere Mitigation Techniques for Synthetic Aperture Radar Interferometry

Synthetic Aperture Radar Interferometry (InSAR) is sensitive to range change (i.e. surface deformation, topography) as well as variations in the troposphere. For surface deformation studies the tropospheric delay is a nuisance that has to be removed for accurate analysis. In this study we investigate the limitations of current tropospheric mitigation techniques for InSAR, through time-series analysis, GPS zenith delay, multi-squint InSAR method, and high-resolution weather simulation, in the context of future SAR missions. Current SAR missions often rely on single-pass interferometry, but future bistatic architectures are being developed to solve for deformation and tropospheric phase delay simultaneously. NASA's Surface Deformation and Change mission study has several architectures operating in this mode, and similarly ESA's HARMONY and ASI's Platino-1 missions propose this capability. In this study we analyze the performance of mitigation strategies. We find out that for a given interferometric deformation accuracy of a few mm/yr in line of sight direction, our preliminary results indicate we need to have a time series with 15 scenes to conduct a time-series inversion. Relying on GPS would require a station density of about 10 stations per km square. Similarly a weather model solution would require a resolution of 2 km with 3 hourly updates. Moreover, the feasibility and practical conditions of the multi-squint InSAR method are discussed.