A New Spatial Multiple Discrete-Continuous Modeling Approach to Land Use Change Analysis

摘要: This report formulates a multiple discrete-continuous probit (MDCP) land-use model within spatially explicit economic structural framework for change decisions. The spatial MDCP is capable of predicting both the type and intensity urban development patterns over large geographic areas, while also explicitly acknowledging proximity-based dependencies in these patterns. At methodological level, focuses on specifying estimating that allows dependent variable to exist discrete states with an associated each state. formulation accommodates dependencies, as well heterogeneity heteroscedasticity, variable, should be applicable wide variety fields where social between decision agents (or observation units) lead spillover effects choices states). A simulation exercise undertaken evaluate ability proposed maximum approximate composite marginal likelihood (MACML) approach recover parameters from cross-sectional model. results show MACML does recovering parameters. An empirical demonstration using city Austin parcel level land use data.