Road Network Centrality based Model to Simulate Population Distribution.

Abstract

This paper presents a model to predict population distribution based on road network centrality. The proposed model predicts population distribution based on accessibility. The concept of network centrality is used to measure the accessibility of the location. The cases study was conducted in the Colombo Metropolitan Area, Sri Lanka. The study developed a model to predict population density by utilizing closeness centrality and the residential land use ratio. The case study result recorded an accepted level of predictability and accuracy (R²= 0.870 and MdAPE < 30%). Further, the model maintains a high accuracy with small calibration sample. The methods demonstrated in this research is effective tool for both transport and urban planners to efficient allocation of road infrastructure and urban development investment.