Improving the accuracy and reliability of land use/land cover simulation by the integration of Markov cellular automata and landform-based models

Land use/land cover (LULC) is one of the important variables affecting human life and the physical environment. Modelling of change in LULC is an important tool for environmental management and for supporting spatial planning in environmentally important areas. In this study, a new approach was proposed to improve the accuracy and reliability of LULC simulation by integrating Markov cellular automata (Markov-CA) and landform-based models. Landform characteristics, positions and patterns influence LULC changes that are important in understanding the effects of environmental change and other physical factors. The results of this study showed that integration of Markov-CA and landform-based models increased correct rejection as a component of agreement and reduced incorrect hits and false alarms as components of disagreement for the percentage of the study area in each resolution (multiple of native pixel size). Correctly simulated hits as a component of agreement change also increased, even though nine of the 18 pairs of three-map comparisons showed a decline in this aspect. Meanwhile, misses as a component of disagreement change simulated as persistence also increased, although six of the 18 pairs of data showed a decline. Based on the overall three-map comparison analysis, there was an increase in the figure of merit (FOM) values after the Markov-CA and landform-based models were integrated, although six of the 18 pairs of data indicated a decrease in FOM values. This indicates improved results after integration of Markov-CA and landform-based models.