Geospatial technologies provide location-based data that ties real-world information to specific positions on Earth. This includes everything from precise coordinates and elevation measurements to land cover classifications, surface imagery, and descriptive attributes like population density or soil type. The data falls into several distinct categories depending on the technology collecting it and how it gets stored.
Vector Data: Points, Lines, and Polygons
Vector data uses geometric shapes to represent features on Earth’s surface. A city might appear as a point, a highway as a line, and a lake as a polygon. Each shape carries exact coordinate boundaries, making vector data highly precise for mapping discrete features. Urban area boundaries, road centerlines, and water bodies are all stored this way.
Every vector feature links to an attribute table, a kind of spreadsheet that holds descriptive, non-spatial information about that feature. A polygon representing a county, for example, might carry fields for population, median income, land area, and tax rate. A line representing a river might include flow rate, water quality readings, and name. This pairing of geometry with descriptive data is foundational to how geographic information systems work. For each layer, there’s typically a one-to-one relationship between a geographic feature and a row in its attribute table.
Raster Data: Grid-Based Imagery and Classifications
Raster data divides the world into a grid of cells, where each cell holds a single value. Satellite imagery is the most familiar example. One of the most common raster datasets is land cover classification derived from satellite sensors, where each pixel is assigned a category like forest, water, urban, or cropland.
Commercial satellites now capture imagery at resolutions between 30 centimeters and 1.2 meters per pixel, detailed enough to distinguish individual vehicles or assess building damage after a disaster. The tradeoff between vector and raster captures an old saying in the field: “raster is faster, but vector is corrector.” Raster data processes quickly over large areas, while vector data preserves geometric precision.
Multispectral and Hyperspectral Imagery
Satellite and airborne sensors collect far more than visible light photographs. Multispectral sensors capture roughly 5 to 30 bands of the electromagnetic spectrum, each about 100 nanometers wide. These bands span visible light, near-infrared, and thermal wavelengths, letting analysts detect things invisible to the human eye: stressed vegetation, water turbidity, heat signatures from urban surfaces.
Hyperspectral sensors go further, capturing hundreds of contiguous, narrow spectral bands. Where multispectral imaging can classify materials on Earth’s surface (identifying that something is vegetation, for instance), hyperspectral imaging can characterize and even quantify that material, distinguishing between crop species or identifying specific minerals in exposed rock. This level of detail matters for precision agriculture, mineral exploration, and environmental monitoring.
Elevation and 3D Surface Data
LiDAR (light detection and ranging) systems fire laser pulses toward the ground and measure the time it takes for each pulse to bounce back. The result is a point cloud: a dense collection of points, each with X, Y, and Z coordinates representing its precise three-dimensional position. Modern systems record up to five returns from a single pulse, because the laser can bounce off a tree canopy, pass through gaps in the leaves, hit a lower branch, and eventually reach the ground. Each return is flagged with its return number (1 through 5), letting analysts separate vegetation from terrain.
Beyond coordinates, each point carries an intensity value measuring the strength of the returning laser signal. Intensity data can distinguish between materials (asphalt versus grass, for example) and can even produce images resembling black-and-white aerial photographs.
This raw point cloud data gets processed into three types of elevation models. A digital terrain model represents the bare ground surface, including natural features like ridges and stream channels but stripping away buildings and vegetation. A digital surface model captures the tops of everything on the surface, including rooftops, tree canopies, and power lines. A digital elevation model is a gridded raster that serves as a general term encompassing both concepts. The practical difference: if you drew a line across a neighborhood, the terrain model would follow the ground while the surface model would trace over the rooftops and treetops.
Position, Velocity, and Time Data
GNSS receivers (the technology behind GPS and similar satellite navigation systems) output a specific set of data fields. The standard format used across most receivers includes latitude, longitude, altitude, speed, heading, date and time, fix quality (how reliable the position is), and satellite signal information. A basic position fix delivers 2D coordinates plus a timestamp, while a full fix adds altitude and accuracy estimates.
This data feeds into nearly every location-aware application, from navigation apps to fleet tracking to surveying equipment. The receiver also reports how many satellites it’s communicating with and the dilution of precision, a measure of how satellite geometry affects accuracy at that moment.
Real-Time Sensor and Movement Data
Geospatial technologies increasingly provide dynamic, time-stamped data rather than static snapshots. Networks of IoT sensors collect spatial and temporal information across agriculture monitoring, environmental monitoring, smart city infrastructure, industrial systems, and health tracking. On a university campus, for instance, researchers have used networks of beam-crossing sensors and camera systems to generate continuous data on how pedestrians and vehicles move through space at different times of day and week.
These sensor feeds produce movement trajectories, traffic intensity counts, dwell times, and object classifications (person, car, bicycle) all tagged with location and time. When analyzed with spatial statistics and data mining tools, these streams reveal usage patterns that static maps never could: which pathways get congested at noon, which plazas sit empty on weekends, where vehicles consistently slow down. The same principle applies to delivery fleet telematics, air quality sensor networks, and mobile device location signals.
Common File Formats for Geospatial Data
The data types above get stored and shared in standardized formats. For vector data, the Library of Congress recommends OGC GeoPackage and GeoJSON as preferred formats, though GeoJSON can have scalability issues with very large datasets. For raster data and georeferenced images, GeoTIFF remains the widely adopted standard. GeoPackage can also store combined vector and raster data in a single file, making it increasingly popular for projects that need both data types together.
These formats matter because they determine whether your data will work across different software platforms and remain accessible years from now. A dataset in a proprietary format may lock you into one vendor’s tools, while open standards like GeoPackage ensure broad compatibility.