Author(s)

T. Osaragi

Abstract

Several Markov chain models for land-use conversion have been proposed and analysed for dynamics in land-use and environmental planning. However, some problems have been encountered in the method for estimating the land-use transition probability. A new estimation method to determine the land-use transition probability is proposed in this paper. The salient features of this method are as follows: (1) In previous studies, the land-use transition probability matrix was typically estimated using time series raster data by counting the number of cells based on the land-use categories. However, the transition probability should be estimated by counting the number of building lots since building lots and not cells are the spatial unit for land-use conversion. A method to estimate the number of building lots using raster data is proposed in this paper. (2) There exist several building lots where the land-use does not vary stochastically. Therefore, building lots in the steady state should be excluded during the estimation process of the land-use transition probability. A method to estimate the number of building lots in the steady state is proposed here. The effectiveness of the proposed methods and some new findings on land-use conversion are demonstrated using numerical examples. Keywords: land-use, Markov chain, transition probability, lot, steady state, raster data. 1 Introduction It is important as well as necessary to analyze land-use conversion in order to evaluate the land-use policies of the past and plan for land-use in the future.

Keywords

land-use, Markov chain, transition probability, lot, steady state, raster data.