In this paper, a new model-based technique for the compensation of severe height-dependent atmospheric artifacts, using ground-based synthetic aperture radar (SAR) data over mountainous regions, is proposed. The method presented represents an extension of already existing techniques, but now taking into account the effect of steep topography in the atmospheric phase screen compensation process. In addition, the technique is adapted to work with polarimetric SAR data, showing, in that case, a noticeable improvement in the compensation process. The method is validated in the mountainous environment of El Forn de Canillo, located in the Andorran Pyrenees, where there is a slow-moving landslide that nowadays is being reactivated coinciding with strong rain episodes. In this framework, ten zero-baseline fully polarimetric data sets have been acquired at X-band during a one-year measurement campaign (October 2010-October 2011) with the GB-SAR sensor developed at the Universitat Politècnica de Catalunya. First, the impact of the severe atmospheric fluctuations among multitemporal GB-SAR measurements is carefully studied and analyzed. Hence, the need to correctly estimate and compensate the resulting phase differences when retrieving interferometric information is put forward in the frame of differential-SAR-interferometry applications.

The particle sizes in large rockfill structures such as dams prevent laboratory testing. The prediction of field behaviour requires the development of models that integrate size effects. A distinct element method model was developed in which grains are characterised by aggregations of a maximum of 14 elementary spherical particles (the resulting particle shape approaches real geometries and allows a reasonable breakage evolution) and the particle breakage criterion involves the subcritical propagation of fissures in the grain. Time effects are included through the velocity of crack propagation, a function of stress state and defect size, which is introduced as a random set of varying lengths. The model was used to simulate the stress–strain response, the evolution of grain size distribution and creep behaviour under oedometric conditions. The model has been used to simulate size effects in the range 0·28–56·0 cm of initial particle size (uniform distributions were tested). Compressibility and creep were partially validated by comparing calculations with test results covering a reduced range of particles. The paper presents the evolution of short-term compressibility and creep indices in terms of particle size. The model is a useful and novel tool with which to extrapolate laboratory results from scaled grain size distributions to prototype dimensions.