A raster map is a digital map built from a grid of small square cells (pixels), where each cell holds a single value representing something about that location, like elevation, temperature, or land cover type. Think of it like a spreadsheet laid over the Earth’s surface: every cell sits in a specific row and column, and the number stored in that cell tells you what’s happening at that spot on the ground.
Raster maps are one of the two fundamental ways geographic data is stored and displayed in mapping software. The other is vector data, which uses points, lines, and shapes. Each approach has strengths, and understanding how raster maps work helps you choose the right one for a given task.
How Raster Data Is Structured
A raster is a matrix of pixels organized into rows and columns. Each pixel contains a value representing whatever the map is designed to show: a spectral value from a satellite sensor, a category like “forest” or “road,” a measurement like rainfall percentage, or a height above sea level. Pixel values can be positive or negative, whole numbers, or decimals. When no data exists for a particular cell, it’s assigned a special “NoData” value so software knows to skip it rather than treat it as zero.
The key technical property of any raster map is its spatial resolution, which is simply the ground area each cell covers. If each cell represents a 5-by-5 meter patch of land, the resolution is 5 meters. Smaller cells capture finer detail but produce larger files. Larger cells keep file sizes manageable but blur small features. Choosing cell size is always a tradeoff between the detail you need and the storage and processing power you have available.
Discrete vs. Continuous Raster Data
Raster maps fall into two broad categories depending on what they represent. Discrete rasters use whole numbers as codes for categories. A land-use map, for instance, might assign the number 1 to forest, 2 to grassland, 3 to urban areas, and so on. Many cells share the same code, and the values don’t fall on a spectrum.
Continuous rasters represent phenomena that change gradually across space. An elevation map is the classic example: one cell might store a value of 564.3 meters while the cell right next to it stores 565.1 meters. Temperature, rainfall, and air pollution maps work the same way. These use decimal (floating-point) values because the data isn’t neatly divided into categories.
Common Uses for Raster Maps
Satellite and aerial imagery are the most familiar raster datasets. Every photograph taken from space or from an aircraft is stored as a grid of pixels, with each pixel recording reflected light across one or more wavelengths. By combining different wavelength bands, analysts can highlight features like vegetation health, water bodies, or urban development.
Digital elevation models (DEMs) are another widespread application. These raster surfaces store height values for every cell and are used to calculate slope, aspect, drainage patterns, and watershed boundaries. Modeling water flow over land, for example, is a task that can only realistically be done with raster data.
Weather and climate maps rely heavily on raster formats as well. Temperature grids, precipitation maps, and atmospheric pressure surfaces are all stored as continuous rasters, letting meteorologists analyze patterns across large regions cell by cell.
How Raster Maps Know Where They Are
A raw image file is just rows and columns of pixel values with no connection to real-world geography. To place that grid on the Earth’s surface, raster maps need georeferencing information: a set of parameters that converts pixel positions into map coordinates.
Some file formats store this information directly in the image header. Others rely on a small companion text file called a world file. The world file shares the same name as the image (so an image called “mytown.tif” would have a world file called “mytown.tfw”) and contains six numbers. These numbers define the pixel size in real-world units, any rotation, and the geographic coordinates of the upper-left pixel. Every time mapping software opens the image, it reads these parameters to position the raster correctly on the map.
Raster vs. Vector Maps
Vector data stores geography as points, lines, and polygons defined by precise coordinates. A house might be a point, a road a line, a property boundary a polygon. Vector maps look clean at any zoom level and store geographic features with high positional accuracy because they aren’t locked to a grid.
Raster maps, by contrast, can look pixelated when you zoom in too far, and they struggle to represent crisp linear features like roads or boundaries. They also tend to produce large files because a value must be recorded for every single cell, even in areas with little variation.
Where raster data excels is in representing surfaces that change continuously, like elevation or temperature, and in performing cell-by-cell mathematical analysis. Vector data handles that kind of continuous surface poorly. In practice, most mapping projects use both formats together, choosing whichever one fits the specific layer of information being mapped.
Raster Analysis and Map Algebra
One of the biggest advantages of raster data is how naturally it lends itself to math. Map algebra treats raster layers as variables in algebraic expressions, performing calculations pixel by pixel. You can add two rasters together (combining rainfall and runoff, for instance), subtract one from another (finding the difference between two elevation surfaces), or multiply a surface by a constant.
Beyond basic arithmetic, you can run comparisons. Asking “where is elevation greater than 1,000 meters?” produces a new raster where every cell above 1,000 meters is marked as true (1) and every cell below is false (0). Boolean operators let you combine conditions: find all cells that are above 1,000 meters AND classified as forest, for example. These operations run quickly because the grid structure makes it straightforward for software to match up corresponding cells across layers.
Common File Formats
GeoTIFF is the most widely used raster format. It stores image data as a standard TIFF file with geographic metadata embedded in the header or in a companion world file (.tfw). GeoTIFFs are supported by virtually every mapping application.
ESRI Grid is a format developed by Esri (the company behind ArcGIS) that stores raster data across a directory of files, including boundary metadata, header information, cell statistics, and an attribute table for integer grids. ASCII Grid (.asc) is a plaintext version of this format, easy to inspect in a text editor but large in file size.
Managing File Size
Because raster datasets record a value for every cell, file sizes grow quickly, especially at high resolutions or over large areas. Compression helps. Lossless methods like LZW and Deflate reduce file size without discarding any data. In one benchmark test, an uncompressed 293 MB TIFF shrank to 179 MB with LZW compression and 162 MB with Deflate.
Lossy methods sacrifice some data fidelity for dramatically smaller files. JPEG compression applied to the same 293 MB file brought it down to just 21 MB. JPEG 2000 achieved 73 MB at default settings. Lossy compression works well for imagery where slight color shifts are acceptable, but it’s a poor choice for analytical rasters like elevation models where every cell value matters.