‹ LibraryBasicChapter 2/18
Chapter 02

Understanding Geographic Data

Geographic data forms the foundation of all geospatial analysis — understanding data types, formats, sources, and quality is essential.

QGISogr2ogr (GDAL)PostGISGeoJSON.ioMapshaper
10M+OSM Contributors↗ Largest crowdsourced map
+200%GeoPackage Adoption↗ Replacing shapefiles
50+ PBFree Satellite Data↗ Landsat + Sentinel

Theory & Foundations

Geographic data describes real-world features using geometry (where) and attributes (what). The two primary models — raster and vector — serve different analytical purposes.

Vector data represents discrete features as points (wells, cities), lines (roads, rivers), and polygons (lakes, districts). Each feature carries attribute data in a table, enabling SQL-like queries and spatial joins.

Raster data represents continuous phenomena as a grid of cells (pixels). Each cell stores a value — elevation, temperature, reflectance. Resolution determines the cell size on the ground: 10m Sentinel-2, 30m Landsat, 90m SRTM DEM.

Common formats: Shapefile (.shp — legacy, max 2GB), GeoPackage (.gpkg — modern, no size limit), GeoJSON (.json — web-friendly), GeoTIFF (.tif — georeferenced raster), Cloud-Optimized GeoTIFF (COG — cloud-native).

Data quality dimensions: Positional accuracy (how close to true location), Attribute accuracy (correctness of properties), Completeness (missing features), Temporal accuracy (currency of data), and Logical consistency (valid topology).

In-Depth Coverage

Vector Data Deep Dive

Vector features model discrete real-world objects with precise boundaries.

  • Points: GPS waypoints, weather stations, city centers
  • Lines: Roads, rivers, utility networks, flight paths
  • Polygons: District boundaries, land parcels, lakes, forests
  • Multi-geometries: Features with multiple parts (archipelagos)
💡 Pro Tip: Prefer GeoPackage over Shapefile for new projects — no field name limits, no file size cap, single file.

Raster Data Deep Dive

Raster grids represent continuous fields across space.

  • DEM/DTM: Elevation surfaces for terrain analysis
  • Satellite imagery: Multi-band reflectance data
  • Land cover: Classified categorical rasters
  • Climate grids: Temperature, precipitation at regular intervals

Open Data Sources

Free geospatial data is abundant from government and international organizations.

  • Natural Earth: Global vector and raster basemaps
  • OpenStreetMap: Crowdsourced global features
  • USGS Earth Explorer: Landsat and elevation data
  • Copernicus Open Access Hub: Sentinel satellite data
  • WorldPop: Gridded population estimates

Key Techniques

Format conversionData cleaningTopology validationSpatial joinsMetadata creation

Data Visualization

Geospatial Data Format Usage (%)
GeoPackage: 35Shapefile: 30GeoJSON: 20GeoTIFF: 10CSV/KML: 5
  • GeoPackage: 35%
  • Shapefile: 30%
  • GeoJSON: 20%
  • GeoTIFF: 10%
  • CSV/KML: 5%
Data Quality Dimensions
  1. 📍PositionalHow close features are to true locations
  2. AttributeCorrectness of feature properties
  3. 📊CompletenessCoverage of all real features
  4. TemporalHow current the data is

Practical Workflow

  1. Identify data needs for your analysis
  2. Search open data portals and archives
  3. Download in appropriate format
  4. Inspect metadata and CRS information
  5. Check data quality and completeness
  6. Reproject to project CRS if needed
  7. Clean: fix geometries, remove duplicates
  8. Document lineage and processing steps

Software & Tools

ogr2ogrFreeCLI Tool — Convert between 200+ vector formats
MapshaperFreeWeb Tool — Simplify and edit vector data online

Video Tutorials

Real-World Application

OpenStreetMap has 10+ million contributors mapping every road, building, and feature globally — the world's largest crowdsourced geographic database. During disasters, HOT activates rapid mapping campaigns.

Real-World Projects

OpenStreetMap Humanitarian Mapping📍 Global

Volunteers map crisis-affected areas rapidly after disasters.

Impact: 100K+ buildings mapped within 48 hours during major events.

Case Study

USGS maintains the National Map providing seamless elevation, hydrography, boundaries, and land cover data for the entire United States at multiple resolutions.

Problem-Based Learning

P1 Combine census data with administrative boundaries for population density mapping

Census data in CSV with district names, boundary data in shapefile with different naming conventions.

Solution: Clean both datasets, standardize district names, perform table join in GIS, calculate population density per km², create choropleth map.

QGISPython (pandas)Natural Earth boundaries
✅ Population density map revealed 3 underserved districts, guiding $5M health infrastructure investment.

Knowledge Check

1. Which format should replace shapefiles?
2. What is the max file size of a shapefile component?

Further Reading