Mitigation of atmospheric Water-Vapour effects on spaceborne interferometric SAR imaging through the MM5 Numerical model

Synthetic Aperture Radar (InSAR) imaging is a well established technique to derive useful products for several land applications. One of the major limitations of InSAR is due to atmospheric e® ects, and in particular to high water vapor variability. In this work, we make an experimental analysis to research the capability of Numerical Weather Prediction (NWP) models as MM5 to produce high resolution (1 km{500 m) maps of Integrated Water Vapor (IWV) in the atmosphere to mitigate the well-known disturbances that a® ect the radar signal while traveling from the sensor to the ground and back. Experiments have been conducted over the area surrounding Rome using ERS data acquired during the three days phase in '94 and using Envisat data acquired in recent years. By means of the PS technique SAR data have been processed and the Atmospheric Phase Screen (APS) of Slave images with respect to a reference Master have been extracted. MM5 provides realistic water vapor distribution elds that can be converted into electromagnetic slant delays. PSInSAR APS's have then been compared to MM5 IWV maps revealing interesting results. MM5 IWV maps have a much lower resolution than PSInSAR APS's: the turbulent term of the atmospheric vapor field cannot be well resolved by MM5, at least with the low resolution ECMWF inputs. However, the vapor distribution term that depends on the local topography has been found quite in accordance. In this work, we will present experimental results as well as discussions over the adopted processing strategy.