| Synthetic Aperture Radar Interferometry (InSAR) is a newly developed satellite observation technology which is applied in studies of hydrosphere, atmosphere, topography and earth surface changes caused by natural or anthropogenic activities. The technology is capable of retrieving accurate geophysical parameters with multiple air-/satellite-based SAR images through establishing interferometric geometry where the phase measurements precisely reflect the geometry between spaceborne platforms and earth surface and highly sensitive to its variation. Due to these unique advantages, interferometric technology has been widely applied in surveying the ground topography and detecting tiny dynamic changes of ground surface in the last two decades.;However, the applicability of such technology is severely affected by differential atmospheric delay induced by inhomogeneity of air refractivity. Previous studies show that the water vapour with strong variation in both spatial and temporal domain dominates the atmospheric artifacts in interferometric phase measurements. It is the problem that we were trying to solve. In this research, we aim at determining and compensating atmospheric signal in SAR interferograms. Compared with previous works, this work studies the problem in a new perspective that the information of water vapor was extracted from Atmospheric Phase Screen (APS) obtained by Permanent Scatterer SAR Interferometry (PSInSAR), and was comparatively studied with synchronized water vapor data, including GPS observations, MERIS images and MM5 simulated products.;The main contributive work in this research includes following aspects:;Firstly, a water vapor component model was proposed for comparison between SAR and non-SAR water vapors. Besides the typical mixing turbulent and stratification terms, the spatial liner trend and ground feature related stationary term has been accounted for in mixed water vapor. Based on this model, a logical strategy of differentiation between spatial linear trend and height dependent stratification, and between stationary term and turbulence signal, was developed.;Secondly, point-based Precipitable Water Vapor (PWV) from SAR APS and GPS meteorology are compared based on the proposed model in order to assess the precision of water vapor signal obtained from SAR. Two implementation methods, a differential and a pseudo absolute mode, were proposed to build the comparison links between SAR differential water vapor and GPS absolute water vapor.;Thirdly, the spatial statistical properties of water vapor components have been investigated by analyzing water vapor signal obtained from SAR APS, synchronous MERIS near infrared images, MM5 Integrated Water Vapor (IWV) in differential comparison mode and in different spatial scales. Furthermore, in a demonstration example, absolute water vapor signal in fine scale was recovered from differential APS maps with MERIS at master date.;By introducing these ideas and data analysis methods, this thesis provides an insight on water vapor signal from Radar Interferometric images. This insight would be firstly significant toward final solution of atmospheric correction in SAR interferometry. While, the water vapor study at small scale is not only beneficial for hydrological study and regional weather (e.g., rainfall) predication, but also promising in meteorological applications in future. In addition, this water vapor study can be extended in improving of atmospheric error modeling for satellite observing technologies, especially in microwave ranging way, such as GNSS, coastal satellite altimetry and VIBL. |
