Geographic information is the core of both cartography/mapping and GI science/GI systems applications. Geographic information can be studied from various points of view: modelling, storing, processing and semantics. Geographic information represents the natural and manmade, tangible and intangible world. Two main established models are used to represent various phenomena: discrete objects and field models. In addition to precise, crisp data, GI can also be imprecise and imprecise information needs special modelling approaches. Thus, it is important to consider research into imprecise geospatial data models, such as fuzzy models and rough sets.
Geographic information needs to be stored and handled as data in databases. The main methods of storage involve raster and vector organization. Spatial databases tend to be huge and spatial queries need to be supported by adequate spatial indexing. Some solutions already exist – like quadtrees and R-trees – but the topic is still relevant for further study in the context of GI. The dimensionality of spatial data – two-, three- and sometime four-dimensional in nature – adds to the complexity of handling such data. Advanced indexing methods exist but they need to be applied to the context of GI (see also the section on ‘Geospatial analysis and modelling’).
Databases need to be continuously updated and the techniques for updating are problematic. Basically two main approaches exist:
- Continuous updating, usually used when maps are derived from larger scale maps (e.g. detailed municipal large scale maps), and supplemented by other updating methods such as field-based methods.
- Updating based on digital images by using change detection methods or replacing maps entirely by newly interpreted ones. Thus research is needed to address incremental updating and versioning of vector format geographic databases and updating of map databases by using digital images and change detection methods on images.
Geographical databases themselves are huge, and via the Internet one can reach even more information in integrated databases than is possible to manage. Using new methods of spatial data mining and visual data mining users can create new information and knowledge from the stored data. Satellite images as well as other gridded data products can also be mined and novel information and knowledge can be extracted from them by image mining and automated knowledge extraction.
Satellite data and orthophotos are often used without interpretation as additional information in image maps. When combining interpreted, usually vector, and noninterpreted, usually raster, information together, problems of scales and granularities appear.
The distribution of geospatial data across the Internet is becoming widespread, but there are many barriers to simple and effective access to geospatial data. Open geospatial consortium standards for serving data (www.opengeospatial.org ) are designed to assist, but they are not universally applied: there are implications of the contemporary geobrowser (e.g. Google Earth) model for cartographers to address, in handling, compiling and presenting geospatial data.
The semantics of GI links research to various application fields with related taxonomies of concepts. Ontology is an approach that aims to produce a common framework for different terminologies. Toponymy is related to GI in the sense of semantics as well. These topics affect attribute tagging, name (including geographical name) determination and processing flow lines in geodatabases.