A high-resolution weather "hindcasting" system to model the atmosphere at the time of SAR scene acquisitions is being established to mitigate the impact of atmospheric water vapor on InSAR deformation maps. Phase delay screen are generated from the atmospheric analyses and used to correct interferometric images, improve InSAR stacking results, and advance the PSInSAR deformation mapping technique by incorporating them as constraints. The technique is being developed by focusing on two study areas, representing a typical range of meteorological conditions, geophysical signal sources and radar reflecting environments: Mount St Helens volcano in Washington, and Los Angeles in southern California. The development of this novel approach for estimating and reducing atmospheric noise will increase the effectiveness of InSAR as an operational monitoring tool, extend the possibilities for investigations using InSAR, and improve the resolution of the technique.
The differential delay caused by changes in the distribution of water vapor in the atmosphere is the biggest source of noise for the SAR interferometry (InSAR) technique. Atmospheric effects can range over all wavelengths, with amplitudes of several centimeters or even greater. Stacking of a large numbers of SAR scenes boosts the InSAR signal-to-noise ratio, however this requires large amounts of data, and atmospheric artifacts may still remain. The exciting new generation of InSAR time series analysis and processing techniques based on "persistent" scatterers (PSInSAR) are also limited by the difficulty of unambiguously removing atmospheric water vapor anomalies from interferograms. By directly modeling the atmosphere at the time of each acquisition independent estimates of the line-of-sight atmospheric phase delay are produced. These can be removed from InSAR images, improving the signal to noise ratio for geodetic signals, permitting the use of fewer scenes for analysis, reducing the necessary data volume and improving the resolution of non-linear events. They also provide an independent means of confirming the atmospheric phase screens estimates in PSInSAR processing and improve the PSInSAR deformation mapping technique by assisting in the objective separation of geodetic and atmospheric signals. A related problem that will also be mitigated derives from the common practice of refining orbital parameter for the SAR satellite when processing InSAR data. Orbital errors are manifested in interferograms as linear phase-ramps, however long wavelength weather patterns may induce similar features. The ambiguity between these two sources of artifact could lead to inappropriate corrective steps in the processing, possibly masking real long wavelength ground motion signals. The weather analyses produced are being made publicly available via GeoEarthScope and UNAVCO in order to further research efforts by other groups working on InSAR processing techniques.
