Cities are facing increasing water stress due to climate change and population growth. By recovering dynamic local water sources, decentralized water technologies can be promising alternatives to mitigate the pressure of water scarcity and strengthen the reliability of urban water supply. Social preferences on decentralized water technologies are critical to the level of mitigation. While past studies evaluated the critical factors that explain household preference, there is a limited understanding of the spatial deployment and diffusion of decentralized water technologies in the urban built environment and its impact on the water system. In this study, we first developed a spatial agent-based model (ABM) to simulate the adoption of rainwater harvesting (RWH) and greywater recycling (GWR) systems by single-family homes in the city of Boston. We used a system dynamics model (SDM) to quantify the impact of decentralized water supply on reservoir water availability, hydropower generation, and carbon emission of the water system. The change in carbon emission was modeled to influence household adoption choices in ABM. We validated our model integration by comparing the simulated reservoir elevations with historical records and early adoptions with validated installations. Our ABM results reveal a much higher adoption ratio and earlier diffusion of RWH compared to GWR. Spatial disparity emerges in adopting two technologies. Our SDM results indicate tradeoffs in water availability and hydropower generation as water supply decreases. The utilization of increased water availability should be explored to produce more hydropower. Moreover, we did not find significant carbon emission reduction of the water system due to the high carbon intensity of GWR as compared to the central system. Reducing the impact of GWR energy consumption is critical to the benefits of large-scale implementations of decentralized water technologies. Overall, our integrated framework provides systematic solutions for planning and evaluating decentralization at the human-infrastructure-environment nexus.