Amogh N P
✦ In loving memory of Amogh N P — Architect · Designer · Visionary ✦
A split computer screen comparing two GIS data structures — on one side crisp vector polygons of city plots and roads, on the other a pixelated raster grid of terrain elevation in colour, vector versus raster.
Unit IIGIS Modelling in Urban & Regional Planning

Spatial Data Input

Vector vs raster — the two ways to hold the world.

≈ 50 min + lab task

Before you can analyse, you must get the world INTO the GIS — and there are two rival ways to hold it. Learn the difference between spatial data (the geometry) and non-spatial data (the attributes); and the two great data structures — VECTOR (points, lines, polygons) and RASTER (a grid of cells). Learn how data is entered — by digitizer and scanner — held in files and databases, and compressed. Try the data-model explorer.

Learning objectives

By the end of this lesson, you will be able to — mapped to the course outcomes for GIS Modelling in Urban & Regional Planning:

1
CO2 · Understand

Define a GIS planning problem and the data it needs.

2
CO2 · Understand

Distinguish spatial from non-spatial data.

3
CO2 · Analyse

Compare the vector and raster data structures.

4
CO2 · Apply

Explain data entry by digitizer and scanner, and data compression.

Vector vs raster

Spatial data & the two structures

GIS data has a where (geometry) and a what (attributes); vector holds discrete features as points/lines/polygons, raster holds the world as a grid of cells — neither is better, match it to the data.[1, 2]

Two ways to hold the world VECTOR — points, lines, polygons discrete, precise — plots, roads RASTER — a grid of cells continuous — elevation, satellite Neither is better — match the structure to the data and the task. 'Vector is always better than raster' is a myth — raster wins for continuous surfaces and overlay analysis.
DiagramVector stores features as points, lines and polygons; raster stores the world as a grid of cells

The where and the what

GIS data has two halves. SPATIAL (geometric) data is the WHERE — the coordinates that locate a feature on the earth. NON-SPATIAL (attribute) data is the WHAT — the table of facts about it (a plot's owner, area, land use, value). The power of GIS is that it BINDS them: select a plot on the map and read its data, or query the data and see which plots light up. Defining a GIS planning PROBLEM means deciding the QUESTION first, then what spatial and attribute data you need to answer it. MISCONCEPTION→correct: 'GIS data is just the map' — half of every GIS dataset is the attribute TABLE; without it the geometry is mute.[1]

The where and the what SPATIAL — the where geometry / coordinates ATTRIBUTE — the what ownerarealand usevalue bound Select a plot and read its data, or query the data and see which plots light up. 'GIS data is just the map' is a myth — half of every dataset is the attribute TABLE; without it geometry is mute.
DiagramGIS data has two halves — spatial geometry, the where, and non-spatial attribute table, the what, bound together
Interactive

Compare the structures

Pick vector or raster and read what it is, what it is best for, and its limit — and see why each suits different planning data.

Vector or raster · pick one

Vector

Features as POINTS, LINES and POLYGONS with exact coordinates — a well as a point, a road as a line, a plot as a polygon.

Best for: Discrete features with precise boundaries — plots, roads, buildings, administrative areas.

Limit: Awkward for continuous surfaces (elevation, temperature) and heavier to analyse cell-by-cell.

Neither is better — match the structure to the data: vector for discrete features, raster for continuous surfaces.

Digitize, scan, store

Getting data in

Data enters by digitizing, scanning, GPS and imagery, each georeferenced to align; it lives in files and databases and is often compressed — and the format and organisation matter.[1, 3]

Getting data in digitize (trace) scan a paper map GPS survey satellite imagery the GIS(all georeferenced) 'A scanned map is GIS data' is a myth — it is a picture until georeferenced and digitized into features.
DiagramSpatial data enters a GIS by digitizing, scanning, GPS and satellite imagery, each georeferenced to align with other layers

Digitize, scan, import

Spatial data enters a GIS several ways. DIGITIZING traces features into vector form — historically on a DIGITIZER tablet, now usually ON-SCREEN over an image. SCANNING turns a paper map into a raster image (then georeferenced, and optionally digitized). Data also arrives from GPS surveys, satellite/aerial imagery, and existing digital datasets. Each must be GEOREFERENCED to the right coordinate system to align with other layers. MISCONCEPTION→correct: 'a scanned map is GIS data' — a scan is just a picture until it is georeferenced and (for vector work) digitized into features with attributes; capture is the slow, costly, error-prone heart of GIS.[1]

Spatial data

At a glance

AspectDetailNote
VectorPoints / lines / polygonsDiscrete, precise — plots, roads
RasterGrid of cellsContinuous, imagery — elevation, satellite
Spatial dataThe geometry (where)On the earth
Attribute dataThe table (what)Half of every dataset
CaptureDigitize / scan / GPS / imageryThe slow, costly heart
Vocabulary

Key terms

Spatial / non-spatial data

The geometry (where) and the attribute table (what).

Vector

Points, lines, polygons with exact coordinates — discrete features.

Raster

A grid of equal cells, each a value — continuous surfaces, imagery.

Digitizing

Tracing features into vector form, on a tablet or on-screen.

Scanning / georeferencing

Turning a paper map into a raster and tying it to coordinates.

Data formats

Shapefile, geodatabase, GeoTIFF, .dbf — what software can read and share.

Apply it

Lab task

For each of these, say whether you would store it as vector or raster and why: city plots, ground elevation, the road network, a satellite image, rainfall. Then describe the steps to turn a scanned paper master plan into a usable GIS layer.

Check your understanding

Self-assessment

1. Continuous data such as ground elevation or rainfall is best held as —

2. In GIS, the attribute data is —

3. A scanned paper map becomes usable GIS data only after it is —

In a nutshell

Recap

Define the planning question first, then the spatial (geometry) and non-spatial (attribute) data needed to answer it.
Vector holds discrete features as points, lines and polygons; raster holds the world as a grid of cells.
Neither structure is better — raster suits continuous surfaces and analysis, vector suits precise discrete features.
Data enters by digitizing, scanning, GPS and imagery, and must be georeferenced to align with other layers.
GIS data lives in files and databases (shapefile, geodatabase, GeoTIFF, .dbf), often compressed; format and organisation matter.
The evidence

References & further reading

  1. [1]ESRI, Understanding GIS — spatial vs attribute data, vector and raster, data capture.
  2. [2]Maguire, Batty & Goodchild, GIS, Spatial Analysis and Modeling — data structures.
  3. [3]OGC / QGIS documentation — GIS file and data formats, compression and databases.

Further reading

  • ESRI — Understanding GIS.
  • Maguire, Batty & Goodchild — GIS, Spatial Analysis and Modeling.
  • C. D. Tomlin — Geographic Information Systems and Cartographic Modeling.

Sources gathered and fact-checked June 2026. Published values vary by source, sample and method — treat as indicative and confirm against the cited standard before structural use.