An agent-based framework for high-resolution modeling domestic water use under urbanization and climate change

Abstract:

Many cities in the world are experiencing rapid urbanization in recent years. Urban water demand is continually increasing under urbanization in the context of climate change, which increases the challenge of water resource management in highly urbanized or rapidly urbanizing areas. Understanding urban water use and its key impacting factors are important for developing mitigation measures for water use under climate change and urbanization. In this study, we developed an agent-based model (ABM) with two groups of agents to estimate the domestic water use. The first group is the government agent that determines the income growth rate, adjusts water prices, and promotes water-efficient appliances. The other group is residential agents who consume water. To better capture the impact of urbanization and climate change on water use, the utility function of residential agents was further divided into base water use related to economic condition and seasonal water use that is sensitive to climate conditions. Moreover, a bass diffusion model was proposed and integrated into the ABM to consider the diffusion of water-efficient appliances. We implemented our ABM in four selected cities, Beijing and Tianjin, China, and Erie and San Diego, US to compare its performance. Results show that our ABM can capture the spatiotemporal pattern of domestic water use in different regions. Residents in the central urban area consume more water than residents in the suburbs in China, but it is opposite in the US. The trend of per capita domestic use shows a “U” shape in Beijing and Tianjin, while per capita domestic use continues to grow in Erie. The overall pattern of domestic water use is mainly influenced by the base water use in Beijing and Erie, but in Tianjin, seasonal water use is dominant. The growth of income and water-efficient appliances are two significant factors affecting domestic water use. The proposed modeling framework is transferrable to other regions to develop strategies for mitigating domestic water use.