Scientific data collected with modern sensors or dedicated detectors exceed very often the perimeter of the initial scientific design in different application domains. These data including experiments and simulations are obtained more and more frequently with large material and human efforts. For instance high energy physics and astrophysics experiments involve multi-annual developments. Hence, the preservation of big data sets produced is of permanent concern and has been addressed in various disciplines at different levels. However, the challenge of digital preservation of scientific data lies in the need to preserve not only the dataset itself but also the ability it has to deliver knowledge to future user community. A real scientific research asset allows future users to reanalyze the data within new contexts. In fact, the data should be preserved long term such that the access and the re-use are made possible and lead to an enhancement of the initial investment. It is therefore of outmost importance to pursue coherent and vigorous approaches to preserve the scientific data at long term.