Pattern Analysis

Abstract

Pattern refers to the rules governing the arrangement of physical events or occurrences in space and time. Depending on the dynamic or static nature of the events, patterns can be spatial, temporal, and spatio-temporal. Patterns can be identified from geoscientific datasets by visualization and mathematical techniques collectively known as pattern analysis. These techniques are chosen depending on the geospatial representation of the events, e.g., as discrete points, lines, and polygons, or as continuous field values (raster or images). Pattern analysis of single type of data (univariate) can indicate whether the events tend to occur close to (clustered) or away from (dispersed) each other, while multivariate pattern analysis can also reveal how the attributes of the events are inter-related.