Simulating built-up expansion in west Delhi using a neural network coupled agent based prioritised growth model

The expansion of built-up areas is a complex phenomenon shaped by a range of spatial and aspatial factors that vary across space and time. Most of the previous studies have simulated land use patterns without considering the impact of futuristic development policies on land use. To address this gap, the study proposes a neural network coupled agent based prioritised growth model applied to the West region of Delhi. The model incorporates micro agents representing private developers who make land development decisions based on a cell’s transition potential from non-built-up to built-up state, calculated by the neural network model. Macro agents, representing government planning agencies, enforce development constraints and provide incentives for development on a non-built-up cell through planned interventions. Simulations for 2021 demonstrate improved accuracy (kappa 0.85) with planned interventions compared to without any planned interventions (kappa 0.83), referred to as a business-as-usual scenario. The model also simulates land use for 2041 under these two scenarios. The resulting change in spatial growth under these two scenarios is visualised through a change map, which identifies areas of gain and loss in the built-up area as growth patterns shift from a business-as-usual scenario to a planned growth scenario. This model offers a useful tool for planners to understand where future growth is expected and how to channelise the growth through strategic planning interventions.