[en] High resolution multi-frequency synthetic aperture radar (SAR) imagery, available since early 2008, brings all weather capability and day/night operability in support of safeguards verification. Today, a combined approach of high resolution optical and radar imagery in monitoring exercise would enable looking at any area of interest on daily basis. One of the challenges is the co-registration of SAR images acquired with different acquisition mode and also with different optical images. We show in this paper the on-going research work to find a general co-register method and an automatic tool to detect changes. Before having an operational co-register tool, a method to find automatically tie points between SAR images acquired with different acquisition mode and with optical images has to be developed. Concerning an automatic change detection method we can conclude that the study of the Harmonic mean, Geometric mean and Arithmetic mean, enables several applications like change detection for SAR imagery. Thus, we developed the MAGMA (Method for Arithmetic and Geometric Means Analysis) change detection method. As shown in this paper, the MAGMA method improves the Maximum Likelihood techniques like GLRT, using Information-Theory concepts to detect changes between SAR amplitude images. The major improvement consists in a lower false detection rate, especially in low amplitude areas. The second improvement consists in a better location of the changes in clearly delimited areas, which enables precise interpretations. Results presented here reveal the potential of high resolution radar imagery for a baseline description of some sites, change detection based on repeat pass imagery acquisitions and site specific constraints in coherent change detection due to cover conditions. (A.C.)

[en] Full text: Differential radar imagery of images regularity acquired by synthetic radar aperture instrument onboard commercial satellite plateforms (ENVISAT, ERS, RADARSAT, etc.) allows to monitor evolution of ground subsidences with an accuracy typically better than few millimeters and independent measurement every about 20 m. The technique of SAR interferometry is described and illustrated on some test cases. Interferograms are computed from the phase difference of SAR images matched with an sub-pixel accuracy. The phase difference includes information about the distance between orbits, atmospheric phase delays, topography and the component of the ground displacement along the line of sight of the satellite. The displacement of the ground is estimated with an accuracy better than the radar wavelength (5.6 cm in C band) using standard Digital Elevation model (SRTM) because the sensitivity to displacement of the interferogram is far better than their sensitivity to topography. An end-to-end operational software has been developed at CEA. It allows to compute whole archives of images and thus provides information about the evolution of sites with time within the past decade. Especially a procedure has been developed that allows to estimate subsidence rates from images presenting a non uniform date distribution. The characteristics of the software are presented. Interferograms also include residual atmospheric artefact resulting from the interaction of the radar beam with both the ionospheric plasma and the atmospheric water vapor. The atmospheric delays may prevent accurate estimation of subsidence at some places. A filter is presented that use both the meteorological meso-scale model MM5 and radiometric measurements acquired simultaneously with the radar images. The meso-scale model allows to estimate the atmospheric delays at wavelengths larger than about 2km. Radiometric images acquired within the water vapor absorption bands provide an estimate of the atmospheric delays at high spatial resolution below cloud free areas. Some results over sites of interest are presented and discussed. (author)

[en] Limes (Land/Sea Integrated Monitoring for European Security) is a FP6-funded project which aims at developing satellite-based services for a range of security-related applications such as maritime, land and border surveillance and emergency response. Limes started in December 2006 and will run until early 2010. Most of the development work has now been concluded and was tested in a number of service demonstrations. Limes contains a work package focused on Treaty Monitoring, which has the objective to provide an integrated platform supporting the non proliferation image analyst in verifying treaty compliance. The main aspects addressed by the work package are: increased automation of the image processing work flow, in particular in the areas of object-based change analysis, 3 D information extraction and processing of radar imagery. Improved information management using a GIS based platform capable of integrating information from multiple sources and time-frames, including satellite imagery, site models, open source information, reports, etc. The Treaty Monitoring work package carried out two service demonstrations in 2008 and 2009 using the nuclear site Olkiluoto (Finland), which hosts a nuclear facility under construction. The demonstration scenario was the monitoring of construction activities using different types of satellite imagery as well as Open Source information. The demonstration and platform validation was performed at the European Satellite Centre (EUSC) and the results were presented to a number of potential users including IAEA and D G-TREN. The paper presents the achievements of the Treaty Monitoring work package and in particular the results of the platform demonstrations.