Abstract
The complexities of the intricate geospatial resources and formats make preservation and distribution of GIS data difficult even among experts. The proliferation of, for instance, KML, Internet map services, etc, reflects the need for sharing geodata but a comprehensive solution when having to deal with data and metadata of a certain complexity is not currently provided. Original geospatial data is usually divided into several parts to record its different aspects (spatial and thematic features, etc), plus additional files containing, metadata, symbolization specifications and tables, etc; these parts are encoded in different formats, both standard and proprietary. To simplify data access, software providers encourage the use of an additional element that we call generically “map project”, and this contains links to other parts (local or remote). Consequently, in order to distribute the data and metadata refereed by the map in a complete way, or to apply the Open Archival Information System (OAIS) standard to preserve it for the future, we need to face the multipart problem. This paper proposes a package allowing the distribution of real (comprehensive although diverse and complex) GIS data over the Internet and for data preservation. This proposal, complemented with the right tools, hides but keeps the multipart structure, so providing a simpler but professional user experience. Several packaging strategies are reviewed in the paper, and a solution based on ISO 29500-2 standard is chosen. The solution also considers the adoption of the recent Open Geospatial Consortium Web Services common standard (OGC OWS) context document as map part, and as a way for also combining data files with geospatial services. Finally, and by using adequate strategies, different GIS implementations can use several parts of the package and ignore the rest: a philosophy that has proven useful (e.g. in TIFF).
Highlights
The usage of GIS often requires only the last updated version of the data and metadata, other studies may require historical data and/or time series to be included in analyses as urbanization dynamics, environmental and climate change, land cover change, etc
This gap became more obvious in the first drafts of the newly proposed ISO 19165 “Geospatial data and metadata preservation” standard
We analyzed some geospatial file formats such as netCDF, KML and European Space Agency (ESA) SAFE; even when they allow for the integration of data and metadata in a single format, they require transforming data structures into their internal data models, making them not appropriate for data preservation
Summary
The usage of GIS often requires only the last updated version of the data and metadata, other studies may require historical data and/or time series to be included in analyses as urbanization dynamics, environmental and climate change, land cover change, etc. The importance of relating geometric data to thematic attributes and data dictionaries, metadata (including data quality information, lineage, etc), symbolization and web services in a seamless environment has been recognized (Horak et al, 2010; Morris and Tuttle, 2008) These components are often stored in separated parts (allowing, for example, that a data dictionary can be used from several datasets) packaging should support their integrated treatment. In 2014 the format description document database of the US Congress Library contained 334 formats, of which 34 are geospatially related and 9 are composed by more than one part (http://www.digitalpreservation.gov:8081/formats) This figure grows if we take into consideration that metadata and symbolization instructions are usually included in separated parts (Kraak and Ormeling, 2003). Afterwards, the paper describes the chosen solution and how it is adapted to the geospatial data needs illustrated by a reference implementation
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
