Multidimensional Map Algebra (MMA)

Multidimensional Map Algebra (MMA) is a data processing language for 2D, 2D+time, 3D, and 4D spatial and spatio-temporal raster data.  It encompasses and extends conventional map algebra.  Like conventional map algebra, MMA functions are grouped into local, focal, zonal, and global categories.  Each function takes as input one or more 2D, 2D+time, 3D, or 4D data layers and returns a data layer (or, in the case of zonal functions, a table).  Functions may be strung together in complex expressions to encode the MMA equivalent of a conventional 'cartographic model'.  

The initial prototype implementation was carried out in the programming language IDL (Research Systems, Inc.) and applied to case studies analyzing the variation in remotely sensed vegetation in southern Africa and the southeastern U.S. as it relates to El Nino/Southern Oscillation over different  land covers.  Current implementation efforts have followed an object-oriented design strategy (departing from the conventional map algebra) and are being carried out in JAVA.  The ultimate aim is to develop open source, interoperable software that can facilitate research in a number of scientific spatio-temporal computing applications, such as the analysis of time series of satellite imagery and geocomputational simulation output.  This research is funded by NASA grant #NAG5-12598.