Understanding the spatiotemporal dynamics of urbanization and predicting future growth is now essential for sustainable urban planning and policy making. This study explores future urban expansion in the rapidly growing region of eastern lowland Nepal. We used the hybrid cellular automata-Markov (CA-Markov) model, which utilizes historical land use and land cover (LULC) maps and several biophysical change driver variables to predict urban expansion for the years 2026 and 2036. Transitional area matrices were generated based on historical LULC data from 1996 to 2006, from 2006 to 2016, and from 1996 to 2016. The approach was validated by cross comparing the actual and simulated maps for 2016. Evaluation gave satisfactory values of Kno (0.89), Kstandard (0.84), and Klocation (0.89) which verifies the accuracy of the model. Hence, the CA-Markov model was utilized to simulate the LULC map for the years 2026 and 2036. The study area experienced rapid peri/urban expansion and sharp decline in area of cultivated land during 1989-2016. Built-up area increased by 110.90 km2 over a period of 27 years at the loss of 87.59 km2 cultivated land. Simulation analysis indicates that urban expansion will continue with urban cover increasing to 230 km2 (8.95%) and 318.51 km2 (12.45%) by 2026 and 2036, respectively, with corresponding declines in cultivated land to 1453.83 km2 (56.86%) and 1374.93 km2 (53.77%) for the same years. The alarming increase in urban areas coupled with loss of cultivated land will have negative implications for food security and environmental equilibrium in the region.
Keywords: CA-Markov; Developing countries; Food security; Spatial models; Urban expansion.