rasteralgebra

Rasteralgebra is a formal framework for performing mathematical, logical, and statistical operations on raster data. In raster data models, a raster is a grid of cells (pixels) where each cell holds a value or a set of values, such as a single band or multiple spectral bands. Rasteralgebra treats rasters as operands in algebraic expressions and defines operators and functions for deriving new rasters.

Operations are commonly categorized as local (cellwise) operations, neighborhood operations, and global or zonal aggregations. Local operators apply to corresponding cells across rasters, for example adding two rasters, computing a ratio between bands, or applying logical predicates. Multiband rasters enable per-cell expressions that combine values from different bands. Neighborhood operators (focal) perform calculations using a moving window, such as mean, median, or range within a defined neighborhood. Global operations summarize over the entire raster or within zones, for instance calculating the total area meeting a criterion or computing zonal statistics using a mask.

Nodata handling and alignment are important aspects of rasteralgebra. Operations assume grids are aligned in terms of extent, cell size, and coordinate system, and special care is taken to handle missing values. Many implementations provide an expression language for composing complex formulas, a concept commonly called map algebra in GIS contexts.

Applications of rasteralgebra span environmental modeling, remote sensing, and image processing. It is used for index calculations (such as vegetation or moisture indices), suitability and risk analyses, hydrological modeling (flow accumulation, watershed delineation), change detection, and feature extraction. Software platforms in geographic information systems typically support raster algebra through built-in operators, functions, and scripting interfaces.