TRANSLATE ENGLISH to Français, Deutsch, Italiano, Português, Español! ^Explanation

DIGITAL MAP DATA: ARCHIVING AND LEGAL DEPOSIT IMPLICATIONS FOR U.K. COPYRIGHT MAP LIBRARIES
James D. Elliot, British Library, Map Library, London, United Kingdom

© LIBER and author

1. Introduction

The primary function of a copyright library is to preserve and make available the national printed record - at least, until recently, this was the generally accepted idea of their duties in the United Kingdom. This traditional view has recently been challenged by the sheer speed of technological change, and the profound effect such changes are having on the supply of cartographic materials nationwide.

2. The conventional picture (i) - the Ordnance Survey
The authoritative record of topographical change in the United Kingdom has traditionally been met by the maps produced by the Ordnance Survey.This was first established as the Trigonometrical Survey in 1791, and owed its origins to the survey operations in 1787 designed to connect Britain and France. Since 1858, the O.S. has published a comprehensive series of large scale (1:2500) plans covering the entire country, with urban areas mapped at even larger scales of up to 1:500. Following the recommendations of the Davidson Committee in 1943, the sheet lines of the large-scale plans were recast on the basis of the National Grid, and three basic scales were published from 1945 onwards. These were: 1:1250, for urban areas; 1:2500 for cultivated rural areas; and 1:10,000 for all areas, including marsh, forest, mountain and moorland (this is thus the largest-scale map published which covers the entire country). The number of sheets covering the U.K. are as follows: 1:1250 - 56,972, published in km squares; 1:2500 - 160,104, published as 1 or 2 km squares; and 1:10,000 - 3,648, published in 5 km squares. In addition, new editions are published, the frequency of these varying with the degree of topographic change in any given area. In general, the O.S. issue a new chart paper copy every 300 units of change. A house typically represents one unit of change, a new road perhaps 20 units - the definition appears to be somewhat elastic. Also, the O.S. has since 1979 issued interim revisions of 1:1250 and 1:2500 plans on 24 reduction aperture cards known as SIMS (Survey Information on Microfilm Service). One SIM is issued every 50 units of change.

3. The conventional picture (ii) - legal deposit
Traditionally, the copyright libraries have received one copy of each of these sheets as published by the O.S. on conventional chart paper. These libraries comprise the British Library, the National Library of Scotland, the National Library of Wales, the Bodleian Library, Oxford, Cambridge University Library and Trinity College, Dublin. Section 15 of the Copyright Act of 1911 provides that every publisher has to deliver, at his own expense and within one month of publication, a copy of every book published in the U.K. to the British Museum (later amended by statutory instrument to the British Library), and within one year to the other libraries.
These legal deposit provisions have not been substantially altered by the subsequent Copyright Acts of 1956 and 1988. "Book" in the context of the Act was defined by subsection 7 of the 1911 act as "every part of a book, pamphlet, sheet of letterpress, sheet of music, map, plan, chart or table that is separately published" (i.e. issued to the public with the authorization of the copyright owner). In the case of the British Library Map Library, the O.S. supplies annually 75% of our total copyright map intake, some 5000 - 6000 sheets. In addition, we receive some 8,000 - 14,000 SIMs annually. Until now, this is how a definitive record of topographical change in the U.K. has been maintained for over 150 years.

4. The wind of change: the "retreat from print"
For some years the OS had been producing digital maps on computer tape, principally for the benefit of local and public authorities in large urban areas. The data offered the relevant customer a limited ability to produce map images which were independent of sheet-lines upon which data could be superimposed, but was not structured and therefore could not be manipulated or transformed in any way.
In 1985 the O.S. announced a major investigative project, the TDB (Topographic DataBase) study. The study concluded that the O.S. should move towards a fully-structured database of Britain's entire topographic record, and to create from this what it described as a "digital archive". The wider objectives of the TDB were to: allow the O.S. and other digital map users to merge topographic and other map data so as to develop a topographic information system and to provide a national basis for GIS (Geographic Information Systems); to reduce the cost of large-scale map production; and, to achieve this, to phase out the supply of existing chart paper copies and to introduce a new system of disseminating cartographic data. This new system envisaged that a map would only be produced from the database when a customer requested it. This last factor is obviously of crucial importance to us map librarians. In detail, what the O.S. are doing is to move over to an automated map production facility in which output can be directed to laser or electrostatic plotters to produce maps on paper, plastic, film or fiche. The design and contents of the map can be transformed immediately by software to vary line gauges, symbols, colour, detail, area covered and scale. Above all, this data will be supplied as CPCD (Customer Plots from Current Data) - in which map production is customised according to the needs of a particular user.
For copyright librarians, the immediate consequence is profound. It will become increasingly possible for the O.S. to issue large-scale mapping only in response to a once-off customer request. Consequently there will no longer be any need from an O.S. point of view to publish this mapping in the conventional sense of the word. As legal deposit only covers "published" works, the supply of large-scale mapping is likely to dry up at some point in the near to medium-term future.
The problem is compounded by the current U.K. political and economic climate. Unlike the U.S.A., for example, there is no automatic right of free access to information produced by government departments. More importantly, U.K. Treasury guidelines require the O.S. and other agencies "to charge what the market will bear for data of commercial value". The O.S. specifically "is required to earn revenue and limit expenditure in order to meet objectives set by government" (Ordnance Survey plan, 1988/91), and is expected to move to full cost-recovery from its current base of 59%. The existing production methods for the large-scale plans are hopelessly uneconomic, and sales of chart-paper plans are extremely small. In other words, not only is there no formal obligation on the O.S. to continue to make available for public use the national record of topographic change, but the O.S. themselves are apparently faced with a financial disincentive for doing so. Clearly, this is not a happy picture, so what is being done about it?

5. Initiatives and implications
It is perhaps fortunate in the light of technological developments at the O.S. that a certain amount of conciousness-raising had already taken place amongst the U.K. cartographic and information community. In 1983 a House of Lords Select Committee on Science and Technology, chaired by Lord Shackleton and entitled Remote sensing and digital mapping, addressed the various problems related to spatial information in digital form, and the library and archive world was invited to put forward views on matters not formerly considered at all - that was the preservation of permanent information. In response to the Shackleton Report, the Government produced a White Paper (Cmnd 9320) in 1984. This document recognised "the need for a retrospective archive of digital maps for the UK and remote sensing images". As far as digital maps were concerned, it envisaged an extension to digital databases of the present voluntary arrangement whereby the O.S. deposit SIMS - the microform maps referred to earlier - in legal deposit libraries. The report concluded that "problems of selectivity, standards, confidentiality, accessibility of methods of data storage would need to be investigated".
One consequence of this report was the setting up of yet another Committee, this time under Lord Chorley, in 1985. Its brief was "to advise the Secretary of State for the Environment within two years on the future handling of geographic information in the United Kingdom taking account of modern developments in information technology and of market need". This report, entitled Handling Geographic Information was published by the Department of the Environment in 1987. This report highlighted certain key issues for copyright librarians, as follows:
   para. 5.41: A great deal of geographic information, as with other
   information, is required permanently...so that it can be used, for example,
   for historical or "time series" studies. Permanent or archival data has
   traditionally been kept in a variety of paper forms - print, manuscript or
   maps - by bodies such as the British Library, the Public Records offices and
   Companies House. The archiving of digital data by such bodies is not well
   established and faces a number of problems:

   1. There is no legal requirement to deposit digital data.
   2. In addition, the development of digital storage facilities will require
   considerable investment.
    3. A further problem is that of selecting digital data for archiving:
with increased amounts of digital data being generated there is a need to be selective in the data archived.

You will note that in these paragraphs there was no specific mention of the Ordnance Survey's data. Much space, however, was taken in Chorley discussing this, and much subsequent criticism both of the report and of the Governments response to it in 1988 was directed at the excessive focus on large-scale mapping at the expense of wider issues of GIS. The Government in its reply did at least

"recognise the need to keep permanent archives of key digital data sets. The British Library, the British Committee for Map Information and Catalogue Systems, Ordnance Survey and ESRC (Economic and Social Research Council) Data Archive will consider the problems involved in establishing these archives in conjunction with DoE and DTI".

Paradoxically, one of Chorley's main recommendations was that the O.S. digitisation programme be accelerated - thus bringing nearer the evil day when our supply of chart paper maps might be expected to dwindle. Current predictions are that the urban areas 1250 maps will be fully digitised by March 1992, and that 62% of the entire basic scales mapping will be completed by the end of 1993.

6. Defining the problems
In late 1987 I and my colleagues in the Map Library were asked to look at the financial and technical implications for the British Library of the digitisation of the national cartographic record. BRICMICS had indeed been considering certain issues since the early 1980's, and our report was presented in May 1988. Collectively, four key problem areas were identified:

6.1. Retention of historical data
The Ordnance Survey's plan is for one single gigantic database of cartographic data for the whole of the United Kingdom. Within this, individual elements of the landscape would be feature-coded and identified by national grid coordinates. Simultaneously, the O.S. were developing their Digital Field Update System (DFUS), which allowed the electronic survey and data capture of any new landscape change. The revised data could be - and is being - loaded direct into the database within 24 hours, thus potentially providing an "instant mapping" service. We were concerned that such new data would overwrite and thus obliterate for posterity the old feature on the database. After representations to the O.S. from BRICMICS and the Royal Society, it emerged that update information could be date-coded and stored on the database along with the superseded data. As data storage media are constantly increasing in size and falling in cost, physical storage was unlikely to be too difficult. The O.S., however, could not undertake to retain all the information in perpetuity. At least, then, the question of a permanent record has a partial answer.

6.2. Legal deposit of digital data
This remains an open question. The Government's 1985 white paper on Intellectual property and innovation concluded that legal deposit needed to be considered separately from copyright, and the Office of Arts and Libraries was given responsibility in April 1986 to consider the scope and timing of such a review. So far, no review has yet taken place, and it seems that any results might be 5 to 10 years hence. It would also be unwise for us to await the outcome of such a review with too much optimism, as any such extension could conceivably only cover published digital datasets.

6.3. The cost of digital data
This, of course, is closely related to the technical issues discussed below. We calculated the cost of the entire database at market rates as being some 27,000,000 - a little more than half the entire British Library's grant-in-aid! To date, a possible solution appears to be an extension of the existing agreement on the deposit of SIMS mentioned in the 1984 White Paper. To handle this, however, would cost approximately 104,000 a year once formal publication of large-scale mapping had ceased.

6.4. Technical options

6.4.1. Storing the database
On purely technical grounds, storage of O.S. digital data seemed feasible; we estimated that the entire O.S. database would amount to some 88 Gb of data. A gigabyte, by the way, is 1,000,000,000 bytes. By comparison, the 360-volume British Library Catalogue contains about 1 Gb. Alternatively, 88Gb would be enough to store 616,000 romantic novels!
It would be theoretically feasible to mount this on a WORM (Write Once Read Many) disk system, using 40-50 of these high-storage discs and a "jukebox" system for storage and retrieval. Costs, logistics, and preservation considerations, however, rapidly ruled this out, as the installation would require a fully-equipped minicomputer installation with tape drives, database managers and probably a programmer as well. This would cost some 150,000 - 200,000 merely to set up even if we had accommodation for it, which we do not. A computerised data archive replicating the O.S. was thus out of the question on financial grounds alone.
The preservation of data in digital form is also highly problematic. The U.S. National Archives and Records Agency's subcommittee on machine-readable records, reporting to the U.S. Archivist in 1984, identified three main difficulties:-
1) the relatively short life of magnetic and optical recording media, normally no more than 20-30 years.
2) rapid changes in software (quoting the fact that Western Electric's UNIX operating system had been offered in over thirty different versions in the previous decade).
3) the extremely short life of hardware, most computer equipment being obsolescent within 10-20 years.

They concluded that any digital archive would sooner or later be locked into a cycle of eternal file conversion and regeneration. Nevertheless the issue of database archiving continues to be reviewed; in Britain the Public Record Office has been carrying out trials based on digital optical disc storage for the last three years, and is due to report on them shortly. The ESRC data archive at the University of Essex has also been successfully operating a digital archive for social sciences data since 1967, but requires sophisticated transformation software as well as a fully-equipped and staffed computer installation.

6.4.2. On-line access
Direct on-line connection will be a theoretical possibility in 5-10 years from now, though the O.S. themselves have rather different ideas as to what public (i.e. non-O.S.) use might mean. They envisage a scenario in which on-line access would be maintained through their retail agencies, the point being that access to the data would be a chargeable service. There is a potential conflict of interest here between the scholar's traditional expectation of free access to information and the commercial objectives of the map producer. One idea is that libraries could become licensed by the O.S. to offer the database on a fee-paying basis in much the same way that commercial agencies would be expected to do. Quite apart from the ethical and organisational questions this would raise within the library environment, the O.S. would need to be satisfied about copyright and revenue protection. For these reasons, this option is currently a remote one.

6.4.3. Computer-output microfilm
The most feasible solution from technical, political and financial points of view is COM (Computer-Output Microfilm). We have established with the O.S. that it is technically feasible to output to COM a block of digital data - either a 1250 or 2500 sheet equivalent - at an interval of every 50 units of change per block as a trigger point. The O.S. have also suggested that direct COM output from the TDB form a definite part of the O.S. plans.
COM has several advantages, not least its relative cheapness. Costs of mastering the expected annual output of fiche would still require additional funding in the order of 104,000 a year for two permanent-quality silver halide sets and one diazo set for use in the students room. They could be interfiled with our existing stock of SIM fiche in National Grid number order, and are relatively easy to store - ten cabinet can store the entire O.S. database of some 220,000 map sheet with room left over. In terms of preservation, too, this offers the best option. Silver-halide microfilm, provided it is processed and stored to ANSI or equivalent standards, are believed to last as long as the best quality rag paper, and as the NARA Subcommittee on Preservation concluded, "it is an essentially risk-free solution requiring no research and development breakthroughs... it is compatible with future electronic information-handling technology". The great drawback of COM, of course, is that it cannot be transformed or manipulated. General agreement, though has been established between the copyright map librarians that this would form the best interim solution in terms of cost, logistics and preservation. COM also offer scope for an extension of the existing voluntary agreement by which SIMs are currently deposited. Meanwhile, the more complex issues of the permanent preservation of digital data continue to be pursued..

7. Beyond the Ordnance Survey: other digital media
Other mapping agencies, notably the Ordnance Survey of Northern Ireland, seem set to follow the O.S. example. The relatively simple administrative structure of the province offers here the prospect of a fully-integrated geographic information system (GIS) which includes both public utility and administrative functions of the Northern Ireland Office (local councils are only responsible for relatively minor functions such as park and cemetery maintenance). The Admiralty Hydrographic Office, at least, has no immediate plans to automate its output of maritime charts.
A specific category of electronic media worthy of mention is the CD-ROM publication, again not covered under existing terms of legal deposit. The last three years have seen a small but steadily growing number of CD-ROM atlases on the market. Examples include the Electromap World Atlas, 1989; the Geodisc US Atlas, a complete 1:2 million vector geographic database for the U.S.A., including road, rail, waterway and federal lands data; Mundocart-CD, published in 1988 by Chadwick-Healey (Australia) Ltd (£8,000); others are in preparation, including the CD Atlas de France and the 1981 small area statistics for England, Wales and Scotland, both by Chadwick-Healey. We have not, so far, felt the need to purchase any CD-ROM atlases, both because of their high cost and because the information content is readily available in conventional atlas form. This, however, could change if there were significant reader demand, or if, as seems likely, primary sources of cartographic data are published only on CD-ROM.
CD production technology is currently relatively cheap and straightforward; archivally, although the CD is virtually indestructible, technology here is changing very rapidly, and presents many of the problems in miniature associated with the preservation of digital data generally.

8. Towards a solution
I do not wish to give the impression that us map curators have played a passive part in all this - something like the heroine tied to the railway line by the evil treasury economist whilst the express train of digital data bears down upon us with, apparently, no sign of a John Wayne coming to the rescue.
It is true that map librarians in the U.K. are a relatively small and weak community. They lack substantial funds and are often minor parts of larger organisations. Consequently, they are dependent to a large extent on their powers of persuasion and their ability to recruit allies. The truth is that a great deal of work has taken place over the last 4-5 years, and we have placed the issue of permanent preservation of digital data firmly on the agenda before our less specialist colleagues had even thought about it. Meetings formal and informal have been held with the O.S., who have been highly cooperative, and between the copyright librarians through the auspices of BRICMICS (British Committee for Map Information and Catalogue Systems), which represents the interests of the U.K. map archive and library community. The major issues I have outlined are all ones which can only be resolved by government departments, and these take some time to move themselves into action. At least an awareness of the problems now exists, as well as some potential solutions. Further developments are under way, instigated by BRICMICS, which will place the resolution of these issues within a firm institutional framework. The first is a seminar which we hope will be arranged under the auspices of the Royal Society. It will specifically discuss the archiving of digital data, and the O.S. have agreed to participate. The second is the establishment of a working group with the formal backing of the Office of Arts and Libraries, the Government Department responsible for the British Library. This will consult with the O.S. as to the legal, technical and financial matters relating to the archiving of data in digital form. It is expected that both will take place soon.
It is clear nevertheless that we are still a long way from the end of the road on this issue. If there are general lessons to be learned from our experiences, they are these: act in concert at national level, for map librarians will achieve very little on an individual basis; maintain and develop contacts with your own national survey organisations, and develop structures where this can be done formally; recruit and convince allies both horizontally - with other professional groups and bodies with interests in the same field, - and vertically, within the administrations of the bodies of which you are a part. Above all, impress upon those you deal with that your country's record of topographic change - whether in conventional or digital form - is a major historical resource that will be of immense benefit to future generations, not only of historians, but of lawyers, solicitors, civil engineers, local government officials and many other groups besides.