Chapter 01

What is Spatial Data?

From the infinite complexity of reality to the structured precision of digital models. Learn how we turn the world into data.

At a Glance

Prereqs: Chapter 00 Time: 20 min read + 20 min practice Deliverable: Raster vs vector decision note

Learning outcomes

  • Explain the difference between vector and raster models.
  • Describe how resolution affects what you can see and measure.
  • Choose the right model for a real GIS question and justify your choice.

Key terms

vector, raster, pixel, resolution, extent, topology

Stop & check

  1. Which data model is best for utility pipelines: raster or vector?

    Answer: Vector (lines).

    Why: Pipelines are discrete objects with connectivity and attributes.

    Common misconception: Aerial imagery shows pipelines, but the network itself is modeled as vector geometry.

  2. If you increase raster resolution, what happens to pixel size?

    Answer: Pixel size decreases.

    Why: Higher resolution means smaller ground area per pixel.

    Common misconception: "Higher" resolution does not mean bigger pixels; it changes measurement detail.

Try it (5 minutes)

  1. Toggle between raster and vector layers and zoom until you see pixels.
  2. Write one sentence: what stays sharp at any scale, and why?

Lab (Two Tracks)

Both tracks produce the same deliverable: a 6-8 sentence decision note about which model fits a chosen problem.

Desktop GIS Track (ArcGIS Pro / QGIS)

Add imagery + a vector layer (roads/parcels). Compare what questions each layer supports and include one screenshot.

Remote Sensing Track (Google Earth Engine)

Load a public image (Sentinel-2 or Landsat) and a vector AOI. Describe what is pixel-based vs feature-based in your map.

Common mistakes

  • Confusing map scale with raster resolution.
  • Assuming all rasters are photographs (many rasters store elevation, temperature, indices).
  • Forgetting CRS/units when measuring distances/areas.

Further reading: https://www.ucgis.org/site/gis-t-body-of-knowledge

The Two Pillars of GIS

In the digital realm, geographic information is predominantly stored in one of two formats: Vector or Raster. Choosing the right one is the fundamental decision of every GIS project.

Vector vs Raster Comparison

Figure 1.1: The fundamental dichotomy of spatial data.

Vector Model

Represents the world as discrete objects. Think of this as "connect the dots" or architectural drafting.

  • Points: Exact X,Y coordinates (e.g., a specific tree).
  • Lines: Connected sequences of points (e.g., a section of Interstate 10).
  • Polygons: Closed areas (e.g., the boundary of Travis County).
Best for features with clear, distinct boundaries.

Raster Model

Represents the world as a continuous grid of cells. Think of this as a digital photograph or a TV screen. In its simplest form, a raster is just a matrix of rows and columns.

Raster Cell Structure

Key Characteristics

Each cell (or pixel) holds a value representing one of three things:

  • Continuous Data: Values that change gradually, like elevation (DEMs) or temperature.
  • Thematic (Discrete) Data: Categorical values, like Land Use (1=Forest, 2=Water).
  • Pictures: Scanned maps or aerial photos where value = color.

Cell Location & Coordinates

A raster is essentially a giant spreadsheet. The location of every cell is defined by its Row and Column coordinates. In a real GIS, we map these Cartesian (x,y) coordinates to real-world locations (Latitude/Longitude) so the image sits in the right place on Earth.

Raster Coordinate System

Why use Raster?

  • Advanced Analysis: Essential for math-heavy modeling (e.g., surface runoff, suitability analysis).
  • Continuous Surfaces: The only way to truly represent things that exist everywhere (like air pressure).
  • Trade-offs: Rasters can become huge files (adding detail = 4x file size). They also suffer from "mixed pixels" where one cell must represent multiple features.
Rule of Thumb: If the data changes continuously across the landscape (elevation, rainfall), use Raster. If it has distinct edges (property lines, roads), use Vector.

Data Capture: Digitizing

Before we had satellite downloads, we had to manually "trace" paper maps into the computer. This implies mounting a map to a digitizing tablet and clicking points along lines to create vector coordinates.

Manual Digitizing Tablet

Figure: Manual Digitizing circa 1990s. This is still done today for historical maps!

Manual GIS: The Legacy of Dr. John Snow

GIS is not a new concept; it is just a new technology. Long before computers, geographers performed "Manual GIS" using transparent overlays and cognitive spatial analysis.

John Snow Cholera Map

Figure 1.3: Dr. John Snow's Cholera Map of London's Soho (1854).

In 1854, a cholera outbreak struck London's Soho district. Dr. John Snow, a physician, didn't just treat patients; he mapped them.

By plotting the location of deaths (points) and water pumps (points) on a base map, he noticed a deadly cluster around the Broad Street Pump. He famously convinced officials to remove the handle of the pump, ending the outbreak.

The First Spatial Analysis: Snow didn't just make a map; he used overlay (deaths + pumps) and proximity (clustering) to solve a problem. This is the essence of GIS.
🛑 Ethics Check: If Dr. Snow did this today, he might violate privacy laws (like HIPAA). Mapping individual patients' homes is dangerous. Modern GIS often aggregates points into larger polygons (like zip codes) to protect identity.

Interactive Explorer: Choosing Your Model

Use the buttons below to toggle between a Vector representation (streets and parcels) and a Raster representation (satellite imagery) of Austin, Texas.

Notice: When you zoom in far on the raster, you eventually see the "grain" or pixels. The vector data stays perfectly sharp at any scale.

Interactive: The Data Model Decision Matrix

As a GIS Analyst, your first job is often choosing the right format. Click the scenarios below to see if they should be Vector or Raster.

Select a scenario above to see the expert recommendation.

Summary of Big Ideas

  • Spatial data is information that has a specific location on Earth's surface.
  • The Vector model uses discrete geometry (points, lines, polygons) to represent features.
  • The Raster model uses a grid of equal-sized cells to represent continuous phenomena.
  • The GIS Sandwich is the conceptual stacking of these layers to perform analysis.

Chapter 01 Checkpoint

Test your knowledge of spatial data models.

1. Which data model would be best for representing a network of utility pipelines?

2. If you increase the resolution of a raster dataset, what happens to the cell size?

3. Which model is best suited for representing continuous elevation across a mountain range?

4. Which geometric primitive would best represent a solitary oak tree on a city park map?

5. Dr. John Snow's 1854 analysis was groundbreaking because he:

Chapter Glossary

Geometric Primitive The three basic shapes used to represent vector features: points, lines, and polygons.
Resolution (Raster) In raster GIS, the dimension of the area on the ground represented by a single pixel (e.g., 30 meters).
Topology The mathematical study of spatial relationships (overlap, adjacency, connectivity) that ensures data integrity.
← Chapter 00: Welcome Chapter 02: Map Design →

BoK Alignment

Topics in the UCGIS GIS&T Body of Knowledge that support this chapter.