Press 'o' to toggle the slide overview and 'f' for full-screen mode.

Choose the theme in which to view this presentation:

Black - White - League - Sky - Beige - Simple
Serif - Blood - Night - Moon - Solarized



Copyright © John Lindsay, 2015

Spatial Data Structures

• The spatial data models (i.e. raster vs vector) only exist as conceptual constructs of the way that we model and interact with geographic entities.
• A spatial data structure refers to the way data stored on a computer are organized.
• There are many different organization structures for either vector or raster data.

Desirable properties of a data structure

1. Efficiency - databases should be efficient both in terms of storage requirements and speed of use
2. Flexibility - it should be easy to perform a wide variety of operations
3. Topology - relative spatial relationships should be present implicitly or explicitly in the data structure

Adapted from: D. Peuquet, Cartographica, Vol. 21(4), 1984.

Desirable properties of a data structure

1. Completeness - it should include all necessary features and relationships
2. Robustness - it should be able to accommodate special situations such as islands or lakes
3. Compatibility – it should be easily read by GIS software, or easily convertible to another format

Adapted from: D. Peuquet, Cartographica, Vol. 21(4), 1984.

Georelational Vector Data Structures

• Geometries and attributes stored separately in a split system.
• Geometries are stored in a graphic file and attributes in a relational database.
• Attribute data is linked to map features by the use of a Key or ID field.
• Each spatial object can have any number of attributes associated with it.
• Vectors can be related to one another through common keys in their attribute tables.

Georelational Vector Data Structures

• Almost all vector data structures are georelational.
• But this only describes the organization of geometry and attribute type data.
• We also need to consider how to structure the geometry data itself.
• One major consideration is whether or not to use a topological data structure.

Chain Topology

Shapefiles

• The shapefile is a standard, non-topological data format used in ESRI products (and many others!).
• Geometries stored in .shp files, attributes stored in .dbf files, other data stored in .shx and .prj files.
• Although originally associated with ArcView, the shapefile emerged in the 1990s as a ubiquitous vector standard for data transfer and analysis despite its limitations.
• Shapefile treats a point as a pair of x-, y-coordinates, a line as a series of points, and a polygon as a series of line segments in a loop but no files describe the spatial relationships between these geometric objects.

Raster Data Structures

• The real beauty of the raster model is the relative simplicity of its data structures.
• All raster data structures are essentially an array of values with metadata used to indicate the number of rows and columns and the location of the data edges.
• There is no need to store the location of each grid cell!
• There are ASCII (text-based) and binary raster formats.
• Rasters are essentially just image data and the GeoTIFF format has emerged as the de-facto standard, although not to the same extent as the shapefile.

Raster Data Structures

• Notice that each grid cell contains a single attribute.
• The easiest way of storing multi-spec or multi-band data is to store each attribute in a separate image file.
• There are however 'multi-band' raster data formats (including the GeoTIFF).

Raster Data Compression

• Raster data structures may also incorporate compression schemes to reduce file storage requirements.
• One of the problems with using the raster data structure is that it results in very large files for even moderate-sized grids
• Data compression methods are designed to reduce the redundancy in a raster data set.
• Compression methods are either lossless or lossy.

Raster Data Compression

• Common raster data compression methods include:
  • Run-length encoding
  • Quadtree structures
  • Wavelet compression

Run-length Encoding

Quadtree Raster Representation

How well could each of the following rasters be compressed?