Towards a temporal extension of spatial allocation modeling

Abstract

Spatial allocation models represent decision processes of allocating discrete spatial units (e.g., land parcels and grid cells) to particular uses under the assumption that attributes of those units are unchanged over time and all decisions are made at one time. Given forecasts of spatial attributes, this assumption may be dropped, though spatial and temporal considerations will be highly interwoven and difficult to articulate in simple algebraic terms. This paper describes a systematic approach to analyzing and resolving this complexity in formulating time-dependent spatial allocation models. The implication of the paper is that geographic information systems can serve not only as devices to store and manage spatio-temporal data, but also as platforms to build spatiotemporal decision models. © Springer-Verlag; 2004.