S1TBX-SE-5 Ionospheric estimation and correction using a split-bandwidth approach


The main interferometric phase components are related to topography, line-of-sight deformation, tropospheric, and ionospheric disturbances. The latter is directly related to the total electron content (TEC) integrated between the satellite and a target (along the line of sight) as the satellite passes by. If there is a linear trend in ionospheric TEC within the ionospheric (azimuth) resolution cell, each target’s phase history will exhibit an azimuth ramp. As a result, scatterers will be imaged with an azimuth shift, which will be automatically corrected during coregistration. However, a dispersive phase screen remains, which is inversely proportional to the system’s frequency. Since all other (nondispersive) phase components are directly proportional to the system’s frequency, the range spectrum can be split (using the split-spectrum method) and both (dispersive and nondispersive) phase components can be estimated. The following workflow is proposed [R-1]:1. Create two sets of SLCs (interferometric pairs) by applying two (range) band-pass filters to the original SLCs. For maximum estimation accuracy, the subbands’ bandwidth is generally chosen to be one third of the total range bandwidth.2. Find coregistration offsets based on full-bandwidth interferometric pair.3. Resample band-pass filtered SLCs separately based on coregistration offsets.4. Generate interferogram for each new data set.5. Estimate interferometric phases (by applying early multilooking).6. Unwrap and filter (by applying late multilooking) unwrapped interferometric phases.7. Invert system of equations relating both interferometric phases (each one corresponding to a different subband) and dispersive and non-dispersive phase components.






Lisa House

Fix versions