Multizonal Model of Industrial Activities Based on Journey‐Time Model

Abstract

A large‐scale regional‐simulation model constructed to forecast industrial activities in disaggregated regions is described. Since the disaggregated subregions (zones) are very close geographically, mutual influences among them cannot be ignored in forecasting future activities in each zone. Based on state‐variable representation, the variation of employment by industrial sectors taken as state variables that represent industrial activities in each zone is described. It is shown that the representation can be simplified to a linear equation with respect to unknown parameters by introducing a previously defined potential, if some reasonable assumptions are postulated. For the purpose of making a positive study, the Kansai region (in the western part of Japan), disaggregated into 120 zones, is taken as an example. First, a submodel is constructed to estimate journey time between zones, which is necessary to calculate the values of various employment and population potentials in each zone. Second, taking employment divided into 46 sectors as the state variables, unknown parameters of the linearized state equation for each of 46 industrial sectors are identified and an employment submodel is constructed. Finally, it is confirmed through the final test that the constructed multizonal‐simulation model has sufficient validity.