This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Data in our day-to-day activities are produced continuously which are being collected by remote sensing satellites. In almost every aspect of our daily activities, databases are generated; business transaction, managing inventories, research and survey works, purchasing of goods from different stores among other can linked to database. 'These databases for example may hold individual address, postal code which may be used as a spatial reference to link the new data to other data in Geographical Information Systems (GIS)' (Heywood, et a. 2006). Both the spatial and alphanumeric data are handled in GIS. By definition GIS is a 'system of hardware and software use for storage, retrieval, mapping and analysis, and displaying of geographic data (Burrough, 1986) which can be regarded simply as merging of cartography, statistical analysis, data modelling, indexing and query processing with database technology. Database can be define as 'any manual or computer based set of procedures used to store and manipulate geographically referenced data which are mostly in digital form' (Aronoff, 1989). The design and implementation of DBMS by computer scientist was based on different databases including among others; hierarchical, network, entity-relation, relational, object-oriented and object/relational models.
The term database was coined from military information system. It is originated in the late 1950s and early 1960s by the development of International Business Machines (IBM) Corporation (www.snee.com accessed 25.10.10). History of GIS can be dated back in 1960 developed by government agencies to tackle the difficult issues in environmental and natural resources such as Canada Geographic Information System (CGIS). GIS databases have seen high rate of growth and development over years playing various roles in computer-assisted cartography, computer aided design, remote sensing, computer graphics and special statistics. Databases have been playing various roles in electronic computing over a long period of time. Speed and flexibility are the targets for the majority of older systems contrary to the modern ones that have wide range of database uses. Databases at that time generate difficulties in data management as multiple users accessed it. Therefore, Database Management Systems (DBMS) was developed to secure and prevent loss of data and have gained wide acceptance (Codd, 1982).
In the early years, each application has its own master file making it difficult for multiple applications to be shared. The concept of data collection was adopted to solve the problem so that the system will efficiently work for multiple applications (www.snee.com accessed 25.10.10). Due to its simple nature Punch Card Systems find it difficult to perform one-to-many mapping. Although disk storage became cost effective by mid-to-late 1970 which lower punch card system usage, its influences lives on through many standard convention and file. Hierarchical database model came to existence as the first database model in the early 1960 in which data is organized/represented in form of hierarchy (tree like structure), thereby fading away punch card systems that recorded data in form of 'flat file' (www.geog.ubc.ca accessed 5.11.10). The hierarchical data structure can be represented in parent/child relationship in which each parent can have many children while each child has only one parent. This mapping can be termed as one-to-many relationship (1:N mapping). Example of hierarchical database model is National Identity card organized by a country, within the country by state, within the state by city, and within the city by registration area. Among other, the advantages of hierarchical model include; may contain many associations that have like records in different file, files records can be added and deleted easily, quick retrieval of data. Disadvantages may include; there is data redundancy therefore making the data difficult to be located in the database, pointer path restrict access, inflexibility of the model made it difficult to define new linkages, example to define new relationship between objects. Hierarchical database model has often been used in GIS technology to store spatial data, but its inflexibility to carry out some functions like many-to-many mapping made it unsuccessful to adapt to GIS application. Introduction of Codd's relational model also contributed for Hierarchical data model to lost traction. Network Database System was design to permit modelling of many-to-many relationship where by child entity can have more than one parent (1:M Mapping). Network model was define in the year 1971 by the Conference on Data System Language (CODASYL) which based on mathematical set theory. Example of this model is network relationship found in Land Parcels, Land Zone and Ownership by individual or company. Advantages of Network Model include among other the following; enables different types of mapping between the parent and child entities, unnecessary redundancy in data is avoided, relatively flexible and efficient than hierarchical model therefore can handle complex spatial data and navigation is made easier to search for data. Disadvantages include among other the following; high skills and broader knowledge is required to understand the query language, the model is time consuming and difficult to set the data correctly at the implementation stage. Due to greater flexibility of relational database systems, Network systems are widespread in GIS technology' (www.geog.ubc.ca accessed 5.11.10).
Relational Database Management System RDBMS was developed by International Business Machines (IBM) researcher E.F. Codd in the year 1970. Relational database model is based on relational algebra that enables data to be described in their natural form only. Relational database comprises a set of table in which the data is organized in series of two-dimensional table each of which contains record of one entity. Structures in relational model are very simple and have proven to be flexible and useful in a wide area of GIS applications, storing over 95% of data in DBMS (Longley, et al. 2008). The columns have a distinctive name, and each of the entries in a single column must be drawn from the same domain (Heywood, et al. 2006). Standard Query Language (SQL) was implemented in 1979 by IMB to facilitate the commands of relational databases. Th SQL has made relational databases easier to understand. The model allows multiple queries on the data and using key fields for more than one table. Despite the flexibility of relational databases, it has not fully developed to handle complex multi-dimensional spatial data (Heywood et al, 2006). Multiuser geodatabase shared between ArcGIS and the attributes table that can be linked to spatial data using unique ID in RDBMS software is an example of relational database in GIS application. Some of the RDBMS in GIS among others include INFO in ARC/INFO that uses hybrid data model, EMPRESS in system/9.
Hybrid data Model was based to support attributes data or graphical entity. To implement this model, digital cartography i.e. coordinates and topological data required are stored in a standard commercial relational type DBMS such as INFO, ORACLE, INGRES (Healy, 1991) in a separate files as well as the attributes table that carry attributes data for each entity. Among others, types of hybrid systems may include CAD-based (INTERGRAPH IGDS/DMRS, AUTOCARD, PC DBMS and MICROSTATION-32) Vector/Topological (ARC/INFO, GEOVISION, and INTERGRAPH MICROSTATION GIS) and Quadtree-based (SPANS). The hybrid systems have successfully gained acceptance in GIS market and is expected to exist in the future and systems like SPAN may also linked to RDMS. Integrated data models on the other hand are usually based on vector-topological system which is describe as spatial system approach. The model contains all the standard data structures without additional programming. The entities or map coordinates are stored as relational tables coupled with table for topological data within the same database. The tables can also be accessed using relational join when stored separately but require high performance hardware (Healy, 1991). Because integrated systems such as system9 are newer and therefore has not fully gained acceptance to commercial GIS market but in the near future due to its significant role in minimizing problems in database input and output. Spatial Query Languages is also an approach used in solving complex GIS function such as using range of operators as described by Guptill (1980). These operators can be implemented in three different approaches using relational DBMS system to perform GIS application. These approaches are described by Healey (1991). Examples of Spatial query language are those in KGIS by Ingram and Phillips (1987), System 9 by Charles, Moon and Tulip (1987) and TIGRIS by Herring (1989).
Object-Oriented Database Model (OODBMS) was developed to tackle the weaknesses of RDBMS. OODBMS deals with representation of data in object form and can be seen as an extension of integrated system incorporation with spatial query language (Healy, 1991). It was developed by Commonwealth Scientific and Industrial Research Organisation (CSIRO) to retrieve, display and enable better management of the complex interrelationships between geographical objects and their properties. The model has flexibility to execute complex queries within GIS application. The database store and organise data (that is group attributes) into a form "objects" that people recognize and provide object-oriented query tools. The attributes are grouped in a given object form into a unit that can be stored or retrieved by its natural name. Example of commercial OODBMS include among others, VERSANT, OBJECTSTORE, OBJECTIVITY/DB and GEMSTONE/S OBJECT. Object/Relational Databases was developed by RDBMS vendors. Some OODBMS object storage features where added to relational database to produce the hybrid object-relational model. The model is more of RDBMS with capability of handling objects. The model has inherited the performance management features from RDBMS to determine object behaviour such as query interface, interpolation algorithms and flexibility of OODBMS to describe objects attribute such as colour, size among other. Example Object/Relational database software include Microsoft SQL Server, Oracle, IBM DB2 and Informix Dynamic Server among other.
In conclusion, it is worth to pointing out that, databases play vital role in terms of manipulating and accessing GIS data. Databases create good atmosphere to load, analyse, manage, display, query and store different kind of geographical data. The changing role of database in GIS technology varies with time from old systems like hierarchical model to most recent ones like object/relational model. There does not yet appear to be a single data model for geographic features which is superior in every aspect to all other data models. The use of databases is subjected to GIS application in question. Relational database has gained acceptance in GIS technology with over 95% of DBMS data stored in it but have fallen short regarding to geospatial application. And some people are in the view that data should be represented in a form that can easily be described. Therefore, they sees object-oriented model as the best option. With the re-establishing cordial relationship between relational and object-oriented database in the near future, it will offer the best possible way forward.