Highlighted Selections from:

Lost Science: Protecting Data Through Improved Archiving


DOI: 10.1002/2013EO370006

Simmons, Karen E. “Lost Science: Protecting Data Through Improved Archiving.” Eos, Transactions American Geophysical Union 94.37 (2013): 323-324. Print.

p.323: Unfortunately, memories are not science data, and those 50 years have not been as kind to the scientific information collected by these pioneering missions. -- Highlighted apr 28, 2014

p.323: When I received a copy of the file of the binary data that NSSDC had converted in 1970 from our 7-track tape to an IBM-compatible format, the file conversion appeared to have lost some of the floating-point data. Still needing the original data, I went to reclaim the Mariner punch cards from the storage facility where they were being held. That facility, however, was being closed, and the contents were being sent to the trash. By a quirk of timing, I was just able to reclaim the Mariner 6 and 7 cards, along with data from Mariner 9 (1971 to 1972), Pioneer Venus (1978 to 1992), and Voyager (1977 to 1989). Thus began work, with a NASA Data Restoration grant, to recover, reanalyze, and rearchive data from these missions. During this research, a number of difficulties were encountered that illustrate important data stewardship concerns. These challenges apply not only to NASA’s decades-long archiving commitment but also to data creation and curation efforts across many scientific disciplines. -- Highlighted apr 28, 2014

p.323: Science observation designs, whether planetary or otherwise, describe the science to be achieved by detailing parameters such as instrument configuration, timing and sampling, target object, and pointing strategy. These documents often contain additional graphic products that visualize the observation and provide essential information for interpreting the science data. Unfortunately, missions have traditionally coded their design products in proprietary or unique software. These files, even if saved, cannot be read once the project ends. It would be easy to archive scanned documents, with available metadata, as PDF-formatted products. -- Highlighted apr 28, 2014

p.324: Scientists’ and engineers’ notebooks, correspondence and documents, photographs, video presentations and press release articles, key equipment, and logo paraphernalia? All of these should arouse interest. The value that historians of science and technology may recognize in these objects may not be obvious to a science team. -- Highlighted apr 28, 2014

p.324: Although these examples come from NASA missions, they illustrate data stewardship concerns resulting from the growth of big data as well as the presence of born-digital data. Many U.S. and international scientific and public agencies are engaged in codifying issues of data curation, especially the knowledge and skill requirements needed in the kind of multi- disciplinary data stewardship that allows users to find things inside, and outside, their own discipline. For archiving to be successful, there must be interplay between data creators, data users, and archivists. But, in the end, it is imperative that each data creator understand that how something is saved is just as important as what is being saved. -- Highlighted apr 28, 2014