What is a geospatial database?

The term ‘geospatial’ refers to interdependent resources like maps, imagery, datasets, tools, and procedures that tie every event, feature, or entity to a location, and use this information for various applications. To easily understand location, data must be represented using standard parameters such as position in a coordinate system, place name, or street address.

A geospatial database is optimized for storing and querying data that represents objects defined in a geometric space, such as vector data and raster data. With data volume growing exponentially, a geospatial database provides the best manageability and security to analyze large, complex, heterogeneous spatial data.

Geospatial database platforms provide specialized management, processing, and analysis engines required for complex geospatial data. The scalability and performance of such systems are two key factors for success, along with providing development and integration support.

For interoperability, geospatial database platforms support standards defined by the Open Geospatial Consortium (OGC), which provide a unified framework and web services—Web Feature Services (WFS) for vector data, Web Coverage Service (WCS) for raster data, and Catalog Services (CSW) used to locate, manage, maintain distributed geospatial data applications and services.

Geographical Information System (GIS) is a tool on top of a geospatial database to edit and maintain geospatial data. GIS support geospatial objects, which are organized in layers that can be overlaid both visually and logically.

Geospatial analysis is about understanding complex interactions based on geographic relationships— answering questions based on where people, assets, and resources are located. Geospatial insights enable users to provide better customer service, optimize workforce, locate retail or distribution centers, manage assets, perform situational analysis, and evaluate sales and marketing campaigns, among many examples.

Fundamentals of geospatial data

‘Geospatial data’ refers to information about features, objects, and classes on Earth’s surface or even in space. Geospatial data is typically large, stored in complex data types, and require specialized indexing, querying, processing, and analysis algorithms.

Geospatial data represents:

• Simple 2D and 3D vector geometric objects such as points, lines, and polygons
• Complex raster data such as imagery and gridded data

Geospatial data is made up of geometries and their cartographic representations, called ‘attributes’. Geometries can be points, lines, polygons, and collections of these elements.

• Points are location coordinates with attached attribute tables, and can for example represent residences, store locations, or mobile phone locations.
• Lines have starting points, end points, and in the case of curves, several midpoints, and an attribute table. This is how road networks are represented in navigation systems, using connected lines and nodes with information on speed limits and wait times at intersections.
• Polygons are area units, with borders set as lines that have attribute tables.

These geometries can have attributes like color, line thickness that are cartographic (for display) and other attributes like population (inside of polygons), or items that can be measured or scaled.

Both geometry and attribute data are connected through a relational database management system like Oracle’s spatial database. The database management system can power the most demanding geospatial processes with the highest performance, scalability, and security. They also provide easy integration with other GIS and nonGIS applications, resulting in lower development efforts.

Geospatial raster data is a complex set of information gathered from Landsat satellite enhanced Thematic Mapper (ETM+) sensors, which record light, infrared reflectance value, and their position in the grid. Location data such as color, height of a digital innovation model, and several variables is attached to every grid cell. Examples include thematic maps, digital elevation model/ digital surface model (DEM/DSM), remote sensing (RS) images, photogrammetric photos, scanned maps, geophysical images, and geological maps.

Raster data types are large and have a very different data structure compared to vector data types. Raster data sets can grow very quickly, resulting in huge volumes of geospatial information that require data management systems such as Oracle’s spatial database.

In addition, point clouds are a complex 3D data type created from light detection and ranging (LiDAR) applications. A point cloud refers to a type of geometry for storing large amounts of data that represents a 3D shape or feature. Each point has its own set of X, Y, and Z coordinates along with other attributes. Point clouds are often created by methods used in photogrammetry or remote sensing by LiDAR applications.

The integration of fundamentally different types of data is one of the central tasks of geospatial data analysis. A vital tool in geospatial data analysis is data visualization, through maps. Maps are usually created from remote sensing data—the fields, forests, and more become digitized attributes given to polygons, and are then colored appropriately.

Data categories may include, but are not limited to:

• Administrative and political boundaries
• Agriculture and farming
• Atmosphere and climate
• Biology and ecology
• Business and economics
• Cadastral
• Cultural, society, and demographic
• Elevation and derived products
• Environment and conservation
• Facilities and structures
• Geological and geophysical
• Human health and disease
• Imagery and base maps
• Inland water resources
• Locations and geodetic networks
• Military
• Oceans and estuaries
• Transportation networks
• Utilities and communication

Geospatial data use cases

In today’s hyperconnected world, where every object has a digital footprint and is part of a global network, location and location-based information becomes critical for analysis, management, administration, and governance. Location intelligence helps us to know where events, activities, individuals, streets, or buildings are, enabling us to develop applications that track location of objects of interest. They have a wide application in many private and public sector organizations, for a variety of functions, such as:

• Operations and billing (see example)
• Real estate and property management (see example)
• Taxation (see example)
• Disaster and emergency management
• Market and customer analysis (see example)
• Tourism (see example)
• Risk management (see example)
• Asset tracking and analysis (see example)
• Telematics and fleet management (see example)
• Land/environment management (see example)
• Public health and disease outbreaks (see example)

Geospatial database challenges

• Lack of integration of spatial data into business processes

  GIS systems are often dedicated, specialized systems that are disconnected from business systems, which leads to increased training, operations, and maintenance costs. Delivering location-related information to the applications is a manual effort that is labor-intensive, time- consuming, error-prone, and most often not scalable for large infrastructure projects. Due to the lack of required integrations, applications can’t use the full value of geospatial information.

• Interoperability

  There is a growing need to integrate maps and data to provide valuable location-based information to/from applications. Organizations may however start using different solutions for various projects. This leads to more than one GIS or mapping component in an organization, also raising concerns about data privacy and data residency.

• Heterogenous data

  Integrated analysis is difficult, as different kinds of data are held in files or specialty data stores, and each needs a specialized skill set. When integrating geospatial data, it is crucial to have an agreement on the definition and use of metadata across an organization. Often, finding the appropriate dataset is challenging as metadata are either incomplete or not accessible/searchable, and datasets are semantically inconsistent, i.e., identical terms do not necessarily mean the same thing.

• Scalability

  Scalability has become a requirement to effectively process ever growing amounts of geospatial data for commercial applications requiring location information, such as sensor data, GPS streaming data, and 3D data.

• Application-level integration

  Due to missing integration between mapping systems and business systems, customers can usually not leverage centralized location information across decision-support systems.

How geospatial database works

• Data ingestion

  Filter and ingest spatial (shape, size, and location) and nonspatial attribute (name, length, area, volume, population, other) data from various data sources (multivariate data). The dataset could consist of a huge number of dedicated domain specific file formats from various data sources, and a lot of time is spent converting these varying data types.

• Data enrichment

  Enrich data with spatial attributes, such as address geocoding and place names for downstream analytics. Change text data to numerical data and normalize all other numerical data. Data enrichment enables users to process less structured geographic data so that the information can be categorized, compared, filtered, and associated with other structured data to perform spatial and text analysis.

• Geospatial processing

  Develop spatial analysis workflows, and combine attribute data with geometric datasets, preparing the data for spatial analysis and mapping.

• Interactive analysis

  Visualize data on interactive maps along with other contextual layers. Navigate and explore the map, viewing, zooming, panning, finding patterns, and querying/filtering by attribute.

• Sharing and publication of results

  Integrate spatial content and analysis results via REST, GeoJSON, and OGC web services.

Best practices for managing and working with geospatial data

• Achieve operational, strategic, and developer benefits by combining geospatial data with all other enterprise data, as seen in a converged database.
• Improve performance by processing where data resides. Use functionality available in database for data integration, enrichment, analysis, and machine learning.
• Enable enterprise grade security and governance with a proven data management platform designed with data security in mind.
• Leverage scalability and performance of an enterprise data management platform with high- availability features to support growing data volume and an increasing demand.
• Make a future proof investment by choosing an open geospatial platform with a possibility to combine components from across systems and vendors.
• Benefit from cloud by choosing a platform that enables both building low-code applications in the cloud as well as easy lift-and-shift to the cloud.

• Learn about converged database
• Learn how to use spatial with a step-by-step workshop

Related products

Oracle Autonomous Database

World’s first self-driving database

Oracle Database

World's leading converged, multi-model database management system

Oracle Exadata

Unmatched Oracle Database performance, scale, and availability

Oracle Autonomous Data Warehouse

Data warehousing without complexity