Video - Categories, Sub-Types and Properties of Spatial Data

Catalogue number: Catalogue number: 89200005

Issue number: 2020001

Release date: February 17, 2020

QGIS Demo 2a

Categories, Sub-Types and Properties of Spatial Data - Video transcript

(The Statistics Canada symbol and Canada wordmark appear on screen with the title: "Demo 2a - Categories, Sub-Types and Properties of Spatial Data")

Hello everyone. So we've installed QGIS, but have no data to use.

So in today's tutorial we'll cover two items. First we'll expand some definitions, exploring the properties, main categories and subtypes of spatial data. Then we'll discuss procedures for downloading datasets online from Open Maps, which is the integrated federal archive for geospatial data.

The understanding of different spatial data types, their traits and applications in GIS developed in the first part of this video will provide the knowledge and skills to navigate the platform, as well as other online spatial data archives, and download datasets relevant to your own areas of expertise.

So in the previous video we introduced spatial data as data referenced to specific geographic locations through coordinates. Today we'll discuss the three listed properties that make datasets spatial, and explore the main categories of spatial data, vectors and rasters, as well as their characteristics, subtypes and common file formats.

The figure on the right nicely depicts some of the properties and subtypes of these two spatial data categories and their integrated use in capturing complex real-world processes and features.

The first component that makes a dataset spatial is the map projection, which defines how the Earth's 3D surface is transformed to a 2D representation. All projections distort features in some way with differing projections created to preserve certain characteristics such as distance, direction, angle, area, or shape. Projections can be thought of as a light source projecting features onto a 3D shape which is unfurled for a 2D view.

The figure shows the three main projection families which are: Cylindrical, Conical, and Planar.

The geodetic datum is a series of key base-points that define the position of select features, with WGS84 approximating mean sea level. They act as known points which are then used to develop the Coordinate Reference System and link features to the projected surface. There are horizontal and vertical datums, which define the shape and position of features in 2D and 3D space. We'll focus on horizontal datums.

So they have been developed for different scales and locations, from local to global levels. Often there is a trade-off in the area they cover and the accuracy of the fit for the specific area of interest, as is shown in the figure. WGS84 or the World Geodetic Systemis the best fit for the entire world as a whole, while the North American Datumbetter fits its specific location. Datums can also become deprecated with NAD27 being replaced by NAD83.

So Coordinate Reference Systems are comprised of meridians and parallels, or vertical and horizontal lines, which create a gridded network and enable the location of features to be defined anywhere on the projected surface through XY coordinates. There are two main types of coordinate systems: Geographic and Projected.

So, geographic coordinate systems generally cover larger areas with coordinates in angular units like decimal degrees or degrees-minutes-seconds.

Projected Coordinate Systems are derived from geographic ones and generally cover smaller extents, but are designed for planar representation of features. As such projected systems are used for spatial analysis such as overlaying layers or adding spatial measures to a dataset. The coordinates often use linear units such as metres.

The figure shows that the appearance of features can change significantly depending on the applied projection, datum and CRS. When using multiple layers in GIS it is best practice to ensure that these properties are uniform. This is particularly important for analysis using multiple layers, where differing properties can result in erroneous measures, overlays or processing tools failing.

So the first spatial data category we'll discuss is vector data, which is probably more accessible to new GIS users. The QGIS User Manual defines vector data as "table data with geometry".

Vector data is used to depict discrete features using three broad geometry types: points, lines and polygons which are shown in the figures below through rail stations and grain elevators, roads and rivers, and land-cover boundaries respectively. Additional variables - numeric or text - are stored in the attribute table, which can be used for analysis and visualization. Common file formats of vector data include shapefiles and file geodatabases. File geodatabases may contain multiple layers of any geometry type – and in certain cases must be exported to a new layer for editing and analysis. Geopackage, the default format in QGIS, combines elements of these two file formats.

Conversely, raster data often depicts continuous data. There are three broad types: single-band, composite and thematic. Single-band shows variations of a single variable like elevation, slope, precipitation, or Hillshade and Aspect – as shown on the left side of the figure. Composite rasters refer to remotely sensed data like satellite imagery, as shown with the Landsat 8 image, where different bands can be combined for analysis and visualization. Thematic rasters are often used to release land-cover classifications derived from composite imagery, like the Annual Crop Inventory on the right.

Unlike vector data, the visualization of raster data is scale-dependent, varying with its cell or pixel size, which is tied to its spatial resolution. Raster data is strictly numeric with each cell assigned a number, corresponding to variations in a specific variable or attribute.

Rasters are generally in image file formats, the most common formats being GeoTIFs (.tif), and JPEG2000 (.jp2). Raster datasets underscore the finer-grain data that is publically available, with resolutions of 15 to 30 metres being common. We will use a combination of vector and raster datasets throughout the subsequent training materials.

So that concludes the first part of this video, expanding our understanding of the properties that make a dataset spatial and the main categories, subtypes and traits of spatial data. Stay tuned for the second half of this tutorial where we'll show the procedures for navigating and accessing spatial datasets from Open Maps.

(Canada wordmark appears.)

Date modified: