Digital Preservation is the managed set of activities required to ensure that digital information remains understandable, authentic, accessible and usable for as long as it is needed. Unlike ordinary backup, preservation is concerned with long-term continuity rather than simply recovering from short-term hardware failure or accidental deletion. Digital objects can become inaccessible because storage media deteriorate, file formats become obsolete, software disappears, encryption keys are lost, metadata is separated from content or organisational systems cease to operate.
Preservation therefore addresses not only the bits themselves but also the information necessary to interpret those bits correctly. It is relevant to libraries, archives, museums, governments, universities, businesses, scientific organisations and individuals holding valuable digital records. The preserved object may be a document, database, image, audio recording, video, website, software package, email collection, research dataset or complex digital system with multiple dependencies.
Digital Preservation involves several related but distinct preservation strategies. Bit preservation seeks to maintain the underlying digital sequence accurately through redundant storage, integrity checking, replication and controlled management of storage systems. Logical preservation seeks to ensure that the preserved object remains interpretable as technologies change. Format migration converts content from an obsolete or vulnerable format into a newer format while attempting to preserve important characteristics.
Emulation recreates or simulates the behaviour of an older computing environment so that original software can continue to operate. Encapsulation preserves an object together with information describing its dependencies and interpretation. Normalisation may convert diverse incoming materials into a smaller set of preferred archival formats to simplify long-term management. No single strategy works for every digital object, and preservation programmes often combine several approaches according to the object's characteristics, value and technical dependencies.
The development of Digital Preservation became increasingly important as organisations moved from paper and analogue media to digital information. Early digital records were often stored on magnetic tapes, disks and proprietary systems whose formats and hardware could become difficult to access after relatively short periods. Libraries and archives subsequently developed systematic approaches to format identification, metadata, migration and trusted repositories. The concept expanded as born-digital records became central to government, science, commerce and cultural heritage.
The emergence of web archives created additional challenges because websites can depend on scripts, external resources, databases and dynamic services that are difficult to capture as static files. Digital preservation standards and reference models were developed to provide common terminology and organisational guidance, including the Open Archival Information System reference model. Contemporary preservation increasingly incorporates cloud infrastructure, automated validation, format registries, scalable storage and specialised preservation software.
Preservation metadata records information needed to manage and interpret digital objects over time. It can document provenance, technical characteristics, fixity information, preservation actions, rights and relationships among files. Checksums or cryptographic hashes provide a mechanism for detecting unintended changes to stored content. Replication places copies in multiple storage locations or systems so that the loss of one copy does not destroy the collection. Storage architectures may use geographically separated copies, different media or different administrative domains to reduce correlated failure.
Versioning can preserve successive states of an object rather than overwriting earlier information. Audit trails document who performed preservation actions and when. These controls are particularly important for records whose authenticity has legal, historical or scientific significance because preservation must provide evidence that an object has not been silently altered and that subsequent transformations are documented rather than concealed.
Digital Preservation is closely connected with authenticity, provenance, integrity and reproducibility. A preserved file is not necessarily useful if users cannot establish where it came from, whether it has been altered or what processing has occurred since its creation. Provenance records can document the origin and chain of custody of an object, while fixity checks detect changes at the bit level.
Authenticity is broader than bit-for-bit identity because a legitimate preservation action such as format migration deliberately changes the representation while attempting to retain the object's significant properties. Scientific preservation additionally requires sufficient contextual information to reproduce or interpret research results, potentially including datasets, software, parameters and documentation. For complex digital objects, relationships between components can be as important as the individual files. Preservation therefore treats digital information as an evidential and technological context rather than merely a collection of isolated binary files.
Digital Preservation faces distinctive problems with databases, software, interactive media and other complex objects. A static text document may be relatively straightforward to migrate, while an interactive website can depend on browser behaviour, JavaScript libraries, server-side code, remote APIs and databases. Proprietary software formats can require unavailable applications or operating systems. Encryption can make preservation impossible if keys are lost, while digital rights management can restrict copying or transformation that preservation may require.
Large scientific datasets create storage and transfer challenges, and rapidly changing formats can make manual preservation workflows impractical. Artificial intelligence creates new preservation questions because models may depend on training data, software libraries, hardware accelerators and rapidly changing execution environments. Effective programmes therefore assess preservation risks at the level of files, formats, software, hardware, metadata, legal permissions and organisational capabilities rather than assuming that storing a copy is sufficient.
Trusted Digital Preservation depends on organisational policies as much as technical infrastructure. Institutions need to determine which materials have long-term value, how long they must be retained, which formats are acceptable, who is responsible for preservation and how resources will be allocated. Preservation repositories need documented procedures for ingest, validation, storage, monitoring, access and disposal where authorised. Standards and audit frameworks can help organisations demonstrate that preservation processes are reliable.
Costs include storage, network transfer, hardware or cloud services, software maintenance, specialist staff and periodic migration or validation. Legal requirements may establish retention periods or evidential standards, while privacy laws may constrain indefinite retention of personal information. Preservation can therefore involve a tension between keeping information available for future use and avoiding unnecessary retention of information that should legitimately be deleted.
Digital Preservation differs fundamentally from backup, disaster recovery and ordinary archival storage. Backup generally provides recoverability from operational loss over relatively short periods and may routinely overwrite older copies, whereas preservation aims to maintain intelligibility and authenticity over technological change. Disaster recovery focuses on restoring services after disruption, while preservation focuses on maintaining valuable information over time. An archive can contain physical or digital materials, but digital preservation specifically addresses the technical and organisational problems of keeping digital information usable.
The long-term significance of Digital Preservation is that digital information can disappear despite remaining physically stored: a file may survive while the software, format knowledge, metadata or cryptographic keys required to interpret it are lost. Preservation therefore requires continuous stewardship, monitoring and adaptation rather than a one-time act of copying data.
