GIS data is any information tied to a specific location on Earth, organized so that software can map it, measure it, and analyze it. Every piece of GIS data has two parts: a spatial component that defines where something is (coordinates, a shape, a boundary) and attribute data that describes what’s there. A city, for example, is a point or polygon on a map, but its GIS data also includes its population, land use patterns, school districts, and public transit options.
This combination of “where” and “what” is what makes GIS data different from a simple spreadsheet or a photograph. It lets you ask geographic questions: Which neighborhoods flood most often? Where should a new fire station go? How has forest cover changed over 20 years?
The Two Core Models: Vector and Raster
GIS data comes in two fundamental formats, and understanding the difference helps you make sense of nearly everything else about how it works.
Vector data represents the world as discrete shapes: points, lines, and polygons. A fire hydrant is a point. A road or stream is a line. A county boundary, soil type, or land parcel is a polygon. Vector data is ideal for anything with a clear, defined edge. Each shape sits in a database row alongside its attributes, so a road centerline might carry columns for speed limit, surface type, number of lanes, and last maintenance date. More specialized vector types exist too, including 3D objects used to model buildings and multipoint collections that can hold billions of laser-scanned points.
Raster data represents the world as a grid of uniformly spaced cells, the same way a digital photo is made of pixels. Each cell holds a single value: a color, a temperature, an elevation. Satellite imagery is raster data. So are digital elevation models, which store height values instead of colors. Some rasters carry multiple bands of information. A satellite image typically has at least three bands (red, green, blue) that combine into the color images you’d recognize from Google Earth. What makes rasters distinctive is that they cover a continuous rectangular area, making them well suited for phenomena that don’t have hard boundaries, like temperature gradients, rainfall, or air quality.
How GIS Data Gets Collected
The data feeding into a GIS comes from a surprisingly wide range of sources. GPS receivers record precise coordinates in the field. Traditional land surveying establishes high-accuracy reference points. Satellites like Landsat capture multispectral imagery of the entire planet on a regular cycle. Aerial photography, often collected through programs like the National Agriculture Imagery Program (NAIP), provides detailed views of smaller areas.
LiDAR, or light detection and ranging, deserves special mention. An airborne laser scanner fires rapid pulses of light at the ground from an aircraft and measures how long each pulse takes to bounce back. Combined with GPS positioning and instruments tracking the aircraft’s orientation, this creates extraordinarily detailed 3D point clouds of terrain, forests, and buildings. LiDAR has become essential for floodplain mapping, forest inventory, and infrastructure planning because it can “see through” tree canopy to measure ground elevation underneath.
Existing records also become GIS data through digitization. Tax parcel maps, census tables, utility records, and address databases all carry location information that can be geocoded and layered into a GIS.
Layers and Overlay Analysis
GIS organizes information into thematic layers, each representing one category of data: roads on one layer, zoning boundaries on another, water features on a third. This layering concept is central to how analysts extract meaning from spatial data.
Overlay analysis combines two or more layers to answer questions that neither layer could answer alone. Pima County, Arizona, for instance, overlays property parcels with layers for zoning, supervisor districts, school districts, and subdivisions. Each overlay transfers attributes from one layer to another, so a single parcel ends up tagged with its zoning code, its elected representative, and its school district, all derived automatically from its location. You can overlay ZIP code boundaries onto address points to assign postal codes, or combine soil maps with slope data and flood zones to determine which land is suitable for development. The power of GIS data lies largely in this ability to stack and intersect different types of information based on shared geography.
Coordinate Reference Systems
For layers to line up correctly, every dataset needs to speak the same spatial language. That’s the role of a coordinate reference system, or CRS. A geographic CRS places features on the globe using latitude and longitude. A projected CRS takes that information and flattens it onto a two-dimensional surface (your screen or a printed map), defining how distances and areas are measured in units like meters or feet.
This matters in practice because datasets from different sources often use different reference systems. If you load a soil map built on one system and a parcel boundary built on another, they may not align. Most GIS software can reproject data on the fly, but mixing systems without noticing is one of the most common sources of error in spatial analysis.
Common File Formats
If you download GIS data from a government portal or research archive, you’ll encounter a handful of standard formats:
- Shapefile (.shp): The most widely used vector format, developed by Esri but compatible with virtually all GIS software. A “shapefile” is actually a bundle of at least three files (.shp, .shx, .dbf) that must stay together in the same folder to work.
- GeoJSON: A lightweight vector format based on JSON, popular in web mapping applications because it’s easy to read and transmit over the internet.
- KML: Originally developed for Google Earth, used to share geographic annotations and visualizations.
- GeoTIFF: A raster image format that embeds geographic coordinates directly in the file header, so the image knows where it belongs on the map.
- DEM: A raster format specifically for elevation data, where each cell stores a height value rather than a color.
Metadata: Data About the Data
Reliable GIS work depends on knowing where a dataset came from, when it was collected, how accurate it is, and what coordinate system it uses. That’s what metadata provides. The international standard for geospatial metadata, ISO 19115, defines both mandatory and optional elements for documenting datasets, covering everything from geographic extent and collection date to accuracy assessments and language codes. In the United States, the Federal Geographic Data Committee (FGDC) has maintained its own complementary standard.
Metadata may sound bureaucratic, but it prevents real problems. Without it, you might combine a road network last updated in 2024 with a land use layer from 2005 and draw conclusions from the mismatch without realizing the data reflects different time periods.
Real-World Applications
GIS data touches far more industries than most people realize. In transportation, state departments of transportation use it to map crash locations, manage road assets, plan emergency response routes, and coordinate 911 systems. The Federal Highway Administration has documented how agencies across the country rely on GIS to maintain a single shared road network that serves both navigation and emergency dispatch.
Urban planners use GIS layers to evaluate where new housing, transit lines, or parks should go based on demographics, traffic, environmental constraints, and existing infrastructure. Environmental scientists track deforestation, map wetlands, and model wildfire risk by combining satellite raster imagery with vector layers of vegetation type and topography. Logistics companies route deliveries using GIS-optimized networks. Public health agencies map disease outbreaks to allocate resources.
The GIS industry as a whole was valued at roughly $14.5 billion in 2025 and is projected to more than double to $31.8 billion by 2031. Much of that growth is driven by the integration of artificial intelligence with geospatial data. Deep learning models now automate tasks that once required painstaking manual work: extracting building footprints from satellite images, detecting vehicles or ships in aerial photography, mapping flood extents, and delineating agricultural field boundaries. Object detection and image segmentation have become the most common AI applications in the geospatial field, turning raw imagery into usable vector data at a scale that would be impossible by hand.