Satellite altimetry is quite unique to provide contemporary measurements of three main variables of the ocean / atmosphere interface, i.e the sea surface height, the significant wave height, and the mean square slope. At scales less than 200 km, where upper meso- and sub-meso scale turbulence takes a significant part in the variability of the ocean/atmosphere coupled system, standard along-track (1Hz) measurements may be impacted by different sources of noise, instrumental and geophysical, and possible numerical sampling issues. Use of these measurements is further complicated by differences between altimeters instrument and processing characteristics. Here, we report statistical analysis in this scale range using an empirical mode decomposition for 2015 as year of reference. The proposed decomposition helps to evaluate the distributed noise characteristics of the all three measured variables for Jason-2, Cryosat2, and SARAL altimeters. As demonstrated, it first helps at making the different altimeter measurements more consistent. Accordingly, the geophysical co-variance of the three variables at these short scales can then be more robustly analysed, especially to reveal and interpret persistent small-scale ocean-atmosphere features.