Abstract

Astronomy, like bird-watching, belongs to everyone. Throughout its history, projects have attempted to survey and monitor the skies, providing invaluable baselines for more future studies. In particular, the distant past becomes even more relevant as a marker of slow changes and subtle drifts. Astronomy owns several million archived plates; they include spectra as well as images. Unfortunately, plans to digitize them have not yet achieved sufficient attention and funding apart from the DASCH project at Harvard Observatory and a new initiative in Shanghai. 

The sky is ours to observe, to measure and to study. For millennia humans have recorded observations in some form, though not as rigorously or as permanently as is possible today. 

1 Astronomy’s Observational Data

Projects that record and monitor the skies today provide invaluable baselines for many future studies, in particular for learning how objects change with time (the time-domain). We learn much more by examining changes than we do from a single image. 

While new surveys enable us to examine and analyze the properties of variability with high precision, the distant past becomes even more relevant when it can offer unique evidence of slow changes and subtle drifts. Those changes and drifts reveal new science which short-term electronic data—however precise—cannot do. 

Up to 30 or 40 years ago astronomical surveys were recorded photographically, the oldest all-sky survey being the French Carte du Ciel (CdC). Commenced in 1887, the CdC involved partnerships with numerous observatories around the globe. CdC collections can be located in many places today. Their long time-base renders them especially valuable for studies of the time-domain. Plates that were exposed for inclusion in those surveys retain their value as baselines, and even increase in value for time-domain work. 

Astronomy owns, and can locate, several million archived plates. They contain spectra as well as images. Given the unique contributions which those archived data can add to modern research, it is somewhat surprising that observatories have not already made digital versions of them for public access. Plans to digitize them have been discussed, but have not yet gained the kind of attention and priority that would raise their status high enough to garner adequate funding for comprehensive digitizing programmes. It is also true that digitizing to the necessary levels of finesse was not possible 30 or so years ago when electronic recording first became established. But the sophistication of electronic data management and data archiving has now made huge strides forward, while the matter of photographic archives has tended to be left aside. Two exceptions are DASCH (Harvard) and a new initiative in China (Shanghai). 

Once astronomy’s heritage plates have been transformed into electronic formats, the observations can be used or re-purposed for a variety of projects. By comparing new (digital) data with older observations, we can uncover new information which can then be applied to refine or challenge theories—and astrophysics makes progress. None of that progress will be realized if the old data are left unused and unuseable. 

Coupled with this untapped wealth of historical data are issues of Open Data, Big Data, data sharing, ownership, archiving, preservation and digitization. For modern (born-digital) data, these same data management issues apply.

2 Data in Astronomy

Astronomers need to be careful when talking about data. What precisely are we referring to? (1) Pure observations (recordings of object or event), (2) facts divined from those pure observations, (3) theory or hypothesis supported by the observations, or (4) a model of the observations (heaven help us!). “Data” is the term used frequently and (as often) unthinkingly by scientist, manager, policy-maker, journalist and layman alike. Here, I use “‘data” to mean the pure, basic, unadulterated observation or record, and irrespective of medium. Nowadays the medium is mostly a digital image, observed with a CCD. 

“Legacy” and “Heritage” are used interchangeably, and rather loosely, to describe pure observations that are archived, preserved and maintained for the purpose of future or comparative research. The naming itself is misleading, and can be used to mislead deliberately. As with a car that rises in value inexplicably when it degrades from “old” to “veteran” to “vintage”, scientific data just as mysteriously change in status from “old” to “heritage” just by reassessing their status, and not due to any inherent property except ageing. 

Astronomy has a wealth of heritage data—observations inherited from a great many observers using a large number of telescopes of various sizes and capabilities. Astronomers are loath to throw anything away, for the very good reason that all objects in the cosmos are varying, on some timescale, between highly rapid (microsecond times, too fast to discern by eye) to glacial scales of thousands of centuries. Surveys, though relatively recent, are superb sources by which such changes can be suspected and perhaps confirmed. Those changes tell us much more than can be determined from a single observation. The changes may be periodic (e.g., in velocity), indicating that the object is paired with at least one other object in a binary or multiple system; they may be irregular, as when the object is having minor explosions on its surface, or we may see one major flash and then nothing further for aeons (cataclysmics). 

But those data are of little value unless we can access them so as to show comparisons with modern data and thus get some handle on the nature, and then on the length, of the changes. 

Up to about 30–40 years ago, astronomy’s data were recorded on photographic plates. Those plates are highly durable, and the emulsions neither fade nor decay if properly processed after exposure. We know of some 5 to 10 million plates (many are direct images, but many more are actual spectra of single objects), distributed in 30–40 observatories around the world. So if they are such precious resources, WHY ARE THEY NOT ALREADY DIGITIZED AND THUS MADE RE-USEABLE BY OTHERS? 

The timing was partly to blame. There were good scanners (microphotometers) in the last century but technology was primitive compared to what can be achieved today, mostly because of handling the output by computers; it was almost a case of inventing your computer first before tackling the scanning of plates, and of course astronomers are only human and wanted to find out how far they could push the new digital technologies of CCD detectors once the mass of storage devices had also been developed. Going back to work on photographic plates was considered unadventurous and unglamorous, and no-one with an eye on a future in research wanted that kind of label. CCDs can detect fainter objects than could photographic plates — though their fields are minuscule compared to the acreage of a large photographic plate. 

3 New Initiatives

But there is now a new atmosphere within the departments of data, archives and legacies, and modern technologies are making it possible, and technically not difficult, to scan, store, retrieve and re-use “data” (mostly written documents or images) for use by anyone and for whatever purpose. NOW is therefore the time for astronomers to strike, and we can do so much more effectively with the cooperation of trained librarians, archivists and users of machine learning. Astronomers want to (and need to) get on with re search or teaching, not labour with scanning programmes, and—equally—librarians and archivists need the leadership of astronomers to train, prioritize and FUND programmes that rescue astronomy’s heritage data. The ultimate beneficiaries will be astronomers themselves, of course, but also other scientists, laymen, and ultimately society as a whole.