Modeling complex human systems: An adaptable framework of urban food deserts

Abstract

Food Deserts (FDs) are an expansive problem with cascading effects on quality of life and health outcomes. Simulation modeling of food access within FDs may help decision-makers and urban planners cost-effectively test solutions to food accessibility, and determine which interventions are best able to mitigate negative effects on FD residents. This work devises an adaptive framework that can inexpensively simulate FDs and assess which policy and built environment disruptors—e.g., expanded bus routes, increased pedestrian safety infrastructure, new grocery stores—generate the greatest increase in food access. This framework combines Agent-Based Modeling, GIS, and Discrete Event Simulation to determine the percent of FD residents without food access (as defined by the USDA) before and after a disruptor is introduced. To assess its functionality, the framework was applied to a case study region in Austin, Texas. Results indicate the modeled disruptors improved food access between 2–45%, depending on the type of disruptor, its location, and the distance residents are willing to walk. Notably, results demonstrate how applying this framework allows one to (1) inexpensively test proposed solutions to food-access issues before large-scale capital investments are made; (2) identify emergent behavior that influence disruptors’ efficacy; and (3) identify unique, area-specific solutions that achieve more sustainable improvements in food access.