GEOG 323 Spatial Data and Spatial Properties

Kyra's Lecture on Spatial Data and Spatial Properties
Emphasis on clear, slower speech and audience interaction with repeated questions for clarity in recordings.
Research opportunity: Navigating unfamiliar cities in an immersive 360-degree environment; participants earn a $25 voucher and a brain signal map.
Class representative needed to mediate between students and academics; details to be confirmed by Friday.

Tobler's Law: things that are nearer are more similar.
Spatial dependence: the similarity of things being nearer alongside the dependency between them due to interaction or shared influences.
Spatial analysis transforms Tobler's Law into quantifiable assessments to identify patterns and clusters.

Example 1: Average salary for medical doctors in New Zealand is statistical, not spatial.
Example 2: Calculating the mean population center considers locations and attributes, making it spatial analysis.
Mean population center: the average location of all people in a given area.
- Influenced by population density; shifts with population changes.
Converting statistical to spatial analysis:
- Break New Zealand into regions to see salary clusters.
- Analyze doctor salaries globally to identify continental patterns.

Data about real-world phenomena with a location and attributes.
Can be presented on a map.
Examples:
- Air pollution or traffic representation on road networks.
- Electricity or data network.
- Bridges over rivers.
- Data about species.
- Virtual events like reported crimes.
Also known as geographic or geospatial information.
Identifies features, events, and boundaries on Earth, stored as coordinates and topology.

Topography: Drawing the space, representing the relief and landscape (mountains, elevations).
Topology: The science of where things are in space and their relationships, including adjacency and connectivity.
Importance of Topology:
- Determining neighbors and shared boundaries (e.g., land parcels sharing a fence).
- Ensuring connectivity in road networks.
- Essential for spatial analysis involving proximity and borders.
Spatial data requires topology for sensible analysis.

Basic types: points, lines, and polygons.
Polylines: Connections of different lines for realistic line structures (e.g., roads).
Raster surfaces: Satellite imagery with pixels representing the real world.
Network topologies: Networks of goods, electricity, etc.
Spatial and non-spatial attributes are linked; location is crucial.

Natural Domain:
- Imagery (satellite, drone).
- Elevation data (point clouds, digital elevation models).
- Soil types, temperatures.
Human Domain:
- Land ownership.
- Street networks.
- Land use, demographics (income, family size, health).
Hybrid Situations:
- Infrastructure like bridges (human-made) over lakes (natural).

Analysis of origin and destination data combined with real-time flight information.
Reveals predefined routes for airplanes to avoid collisions.
Circular routes near airports due to traffic and control tower instructions.
Data can be used to analyze passenger demographics, travel patterns, and for security purposes.
Visualization is the first step in data exploration.

Space and location are essential in daily life.
Examples:
- Choosing a house based on neighborhood characteristics.
- Tracking disease propagation.
- Starting a business in a suitable location.

Converting allocation descriptors (addresses) into XY points for GIS.
Sources: New Zealand Post, Ministry of Health, online services (e.g., Google Earth).
Issues:
- Place name ambiguity: Multiple places with the same name.
- Incorrect or incomplete addresses.
- Abbreviations not recognized.
- Localized names not in official databases.
- Typos.
- Single vs. multiple field addresses.
- Open space addresses: Ambiguity in location.
Lab Focus:
- Working with different data types and formats.
- Identifying and resolving geo coding issues.

John Snow's map of cholera deaths in London linked the outbreak to contaminated water pumps.
Visualizing data helps identify patterns and potential causes.
Businesses use geo demographics to find suitable locations (e.g., coffee shops).

Chicago Crime Portal:
- Maps of narcotics rates and homicides show a strong correlation.
- Raises questions about causality: Are drug takers more vulnerable, or do they commit more violence?
Oakland Suburbs Income Analysis:
- Maps from 2001 and 2013 show changes in low-income areas.
- Problem: Definition of low income changed, making comparison difficult.
- Issue: Small populations in large areas (e.g., volcanoes) can skew data representation.

Event Type (Points):
- Example: Cancer cases in Minnesota show concentrations in central areas.
- Requires further investigation into factors like pollution or population density.
Continuous Data:
- Raster surfaces representing concentrations (pollution, minerals, noise) or values (elevation).
- Pixel values can take any number depending on measurement accuracy.
Zonal Data:
- Polygons representing aerial units (regions, territorial authorities).
- Used to collect data about people and other attributes within those areas.
Space Interaction:
- Interaction between goods, data, and information.

Analysis of data and processes operating in space to describe or explain their behavior and relationships.
Aims:
- Detect patterns.
- Examine relationships (e.g., air pollution and asthma).
Transforms data into information through analytical methods.
Location is very important and may change the analysis.

Methods covered in the course: regression, spatial analysis, spatial correlation.
Aims:
- Identify, describe, and understand patterns (e.g., rainfall across New Zealand).
- Understand processes that create those patterns (e.g., water cycle).
Can work in both directions: predict phenomena or infer causes from patterns.

Fields using spatial data analysis: epidemiology, criminology, education, market analysis.
When to use it:
- Suspect different phenomena across space (crime).
- Observe relationships (asthma and air pollution).
- Have or can obtain spatial reference.