A geographical direction-based approach for capturing the local variation of urban expansion in the application of CA-Markov model

Predictive modelling and its application in land change science has been considerably advanced, however, further performance improvement of the existing models is undergoing. CA-Markov is one of the frequently used methods for this purpose, while a few studies have investigated the possibility of improving it with respect to capturing heterogeneous directional variation of change instead of a global approach. The main objective of this study is to improve the simulation performance of CA-Markov model by developing a direction-based approach so that the predictive changes vary along different directions resulting in more accurate predictions. The proposed approach was tested on Babol city of Iran for three timestamps of 1985, 2001 and 2015. Thereafter, the performance accuracy of the global and direction-based approaches was compared. For doing so, Pearson Chi-square statistic, Shannon's entropy, and degree-of-goodness index were used to check the spatial structure of urban expansion. The findings reveal that our proposed approach outperforms better than the global approach confirming the suitability of our approach for presence and futuristic urban planning. Our conclusions draw attention of the respective stakeholders and decision makers to take our findings into consideration and also take preparedness measures for the speculative consequences. •We propose a novel direction-based approach for land change simulation while employing CA-Markov.•We analyse trends of urban growth using Pearson Chi-square statistic, Shannon's entropy model, and degree-of-goodness.•We compare the performance accuracy of the direction-based CA-Markov against the conventional CA-Markov.•We discuss the usefulness of direction-based CA-Markov model for local urban planners and decision makers.