Development of a parallel computing-based Futureland model for multiple land-use simulation: a case study in Shanghai

Abstract

Predicting future urban land-use scenarios and assessing their encroachment on ecological land is important for sustainability policy formulation. This study develops a new Futureland model using cellular automata with user-defined transformation rules, graphical neighborhood configuration and matrixed transformation cost to simulate multiple land-use changes. Compared with existing models, Futureland applies different factors to different land-use types in a single simulation experiment. Futureland was developed using Geospatial Data Abstraction Library and parallel computing, which significantly improve the implementation efficiency. The case study of Shanghai 2010–2020 illustrates an overall accuracy of 86.6% and a Kappa simulation of 0.79. The land-use scenarios for 2020–2035 were projected under greenspace planning constraints using Futureland. The results indicate that Shanghai’s urban sprawl will gradually slow down by 2030, and the increased urban areas will be mainly in the urban fringes and suburban regions. Futureland can help decision-makers to manage future land-use and optimize urban planning policies.

Keywords

• land-use modeling
• cellular automata
• generalized logistic regression
• ecological land
• futureland