Since the inception of the Internet infrastructure, modern data management practices have been trying to avoid the duplication and hosting of data on multiple servers - particularly to avoid the internal storage of data owned and managed by others. The recent evolution of cloud technology continues this development for text and video files. With open source GIS (where open source means that the software code is freely available for development and use) and standards for Web mapping services, Web feature services and Web catalogue services, why is the geospatial industry not moving faster toward truly distributed sharing of map information on the Web? The standard continues to be “data sucking” where systems routinely harvest external data sources and download the data onto a local service. Usually this is done in the name of efficiency, reliability and performance in terms of delivering the data; however with today’s existing infrastructure, standards and computing power this is no longer needed.
The Coastal and Oceans Information Network - COINAtlantic - is an initiative based at Dalhousie University, Halifax, Nova Scotia, Canada that has worked for over 20 years to provide information to researchers and managers for dealing with coastal and oceans issues on the east coast of Canada. It recognizes that required information comes from many different organizations and that no one organization could – or should – manage all of the disparate types of information, from coastline, bathymetry, fisheries areas, infrastructure, natural resources, etc. This need drives the implementation of an easy-to-use geospatial information system built on OGC standards for use by non-specialists.
COINAtlantic has implemented a system that is truly distributed in that it harvests no data, stores no geospatial information in a warehouse or portal, yet provides clients with the ability to effectively search, locate and present geospatial information from anywhere on the Internet that is published as Web mapping services (WMS) or Keyhole Markup Language (KML) OGC standards.
The COINAtlantic Chain for Information Access shown in Figure 1 is the philosophy underpinning the development of the COINAtlantic spatial tools, and can be described as a five-step process. The goal of this process is to provide tools that allow users with little technical knowledge, using only a computer and an Internet connection, to find and access spatial information, data and maps published by others. COINAtlantic has achieved success through the development of the COINAtlantic Geocontent Generator (CGG) and the COINAtlantic Search Utility (CSU) that address the key steps in the chain.
Figure 1: The COINAtlantic process. Geospatial information shared through open standards such as WMS and KML are discovered through standard Internet text search engines, and displayed in the COINAtlantic Search Utility.
One of the most important steps to improve success in this approach is to improve the initial availability of products in step one of the chain: geospatial information in WMS or KML format that is distributed using standard OGC techniques. COINAtlantic works with its members and other relevant organizations to make this type of information available for use on the Internet either through use of OGC WMS or, for the less technical, using the CGG. When metadata text is embedded in the WMS and KML files (step two), and the metadata text is then indexed by search engines such as Google (step three), it is then ready for searching through a Web interface: the CSU (step four). The CSU then allows the user to view, overlay and output information, with processing a function planned for the future (step five).
Figure 2 shows the COINAtlantic concept graphically as a means of connecting users with the various well-managed silos published and maintained by the information providers. The CSU is an online tool that builds on the power of existing text search engines to allow users to find relevant data wherever it is published.
Figure 2: Managing information “silos.” Data administrators publish spatial data to the Internet through OGC standards, using descriptive metadata that can be effectively searched through external applications.
Figure 3 shows a screen grab of search results presented by the CSU.
Figure 3: Display of remote KML files, in this case Nova Scotia lighthouses. The Google Search API was used to discover this remote KML file.
Figure 4: COINAtlantic Search Utility architecture. Only metadata pointing to external remote data is stored locally in a database. No local data is used in the application.
As of October 2014, after 18 months of CSU operation, the indexed database has 1,486 URL links to remote geospatial files. There are links to 546 WMS services and 940 KML files. Entries are added to the database as a result of users searching with the CSU for geospatial information. Each entry has been parsed for searchable text that is useful for the COINAtlantic user community. Other implementations of this architecture will generate index lists relevant to other directed user communities.
The advantage of this approach is that it does not duplicate data from data providers. It makes use of the data that is published by providers in the format and scale of their choosing. The providers have full control over how they wish their data to be provided.
There is no need for users to develop and support additional server capacity to duplicate the data that is readily available online.
The system makes use of existing standards which have advantages that have not been fully realized in any other system/portal. In this way COINAtlantic is demonstrating the ultimate benefits of implementing OGC standards. This will help to drive users to properly implement the standards to the fullest extent, such as proper implementation of legends. This is an advantage for publishers who have invested in full implementation of the standards, but it also helps to identify services that require some additional work to bring them up to the standards. Ultimately tools and architectures such as the COINAtlantic implementation will help to promote a much deeper and reliable Internet universe of geospatial information.
As with any data management system, one of the continuing challenges for the search function is the existence of useful metadata. Adequate and appropriate text keywords must be included in the metadata files. COINAtlantic work is exploring the use of OGC Web catalogue services for enhancing search effectiveness – and thus providing users with improved search results in the CSU.
In conclusion, we have an eloquent, simple implementation of open source tools to search, find and ultimately use geospatial data, for use by non-specialists who are presently lost in an ocean of data.