Introductions + Why GIS?

Rachel Franklin

Newcastle University

Introductions

Why GIS for economic geography and regional science?
- Research example: depopulation
The basics of GIS
- Why do we need a GIS?
- What is a GIS?
- Visualizing data in a GIS
- Finding spatial data
A few more examples

Understanding local population change requires a contextual approach, such as the historical patterns of change and the experiences of surrounding regions
Our typology approach, which groups the patterns of population changes for both individual areas and their neighboring regions over a span of 70 years, provides a detailed understanding of how depopulation has spatially unfolded in the United States
GIS underpins this analysis, even if methods are more geographic data science

How does a variable or phenomenon vary over space (e.g. income)? (Can be answered with a map)
How does distance to university influence innovation or entrepreneurship? (Can be answered using traditional statistical software, with spatial variables created using GIS)
What is the spatial distribution of crime in a given city? (Can be answered with maps and descriptive spatial statistics)
How are business success and neighborhood characteristics related? (Can be answered with spatial statistics and models)

Data visualization – for example, exploratory analysis or cartography
Explore geographic coincidence – for example, how many new firms in a region
Calculate geometries – for example, length of road in an area, or the center of a polygon
Identification of spatial patterns – for example, clusters or hot spots
Finding spatial relationships – for example, measuring exposure, accessibility, or market areas
Calculating spatial variables – for example, distance or proximity
Estimating values or density – for example, interpolation

Source: Clark, 2011

Source: krygier.owu.edu

A GIS doesn’t save information as maps
- Rather, the GIS stores information on object location and relation to other objects and in space
Things to be aware of:
- Your spatial data model
- The relationships your objects have with each other in the real world
  - e.g., street intersections or shared regional borders
  - Referred to as “topology”

Features are stored as discrete points, lines or polygons
- All are combinations of nodes and/or vertices
Locations are recorded as X,Y coordinates

Each feature – point, line, or polygon – is dynamically linked to the “attribute table” or set of variables
- For example, you could have a dataset of cities, represented as points, and the attribute table might contain characteristics of the city – population, median household income, etc.

Sub-divides a study area into square pixels – rows and columns
Only need the location of the upper left-hand corner and all other locations are implicit, assuming you know your pixel size
Only one value is recorded for each pixel
- For example, temperature, precipitation, or land use type

So long as the x, y location is georeferenced (placed on the surface of the earth), we can locate each other pixel, too

Best for continuous data
Analysis can sometimes be faster
And some types of analysis require raster data
“Yes, raster is faster, but raster is vaster, and vector just seems more correcter.” – Dana Tomlin
Remember: pixel size (or resolution) matters!

Source: Olivia Iles class project

Source: Sonya Gurwitt class project