Results from two years of the CryoSat Validation Experiment (CryoVEx) over sea ice in the western Arctic Ocean are presented. The estimation of freeboard, the height of sea ice floating above the water level, is one the main goals of the CryoSat-2 mission of the European Space Agency (ESA) in order to investigate sea ice volume changes on an Arctic wide scale. Freeboard retrieval requires precise radar range measurements to the ice surface, therefore we investigate the penetration of the K_u-Band radar waves into the overlying snow cover as well as the effects of sub-footprint-scale surface roughness using airborne radar and laser altimeters. We find regional variable penetration of the radar signal at late spring conditions, where the difference of the radar and the reference laser range measurement never agrees with the expected snow thickness. In addition, a rough surface can lead to biases of the airborne validation dataset, since the radar overestimates the amount of open water and thin ice as well the freeboard of heavy ice deformation zones.