A DIGITAL-TWIN-INSPIRED SOCIO-ECOLOGICAL FRAMEWORK FOR URBAN STORMWATER ADAPTIVE MANAGEMENT

Abstract

Urban stormwater systems face increasing challenges from climate extremes, urbanization, and aging infrastructure. Stormwater management thus stands at a crossroad pressured by complexity, uncertainty, and the urgent need for adaptation. Digital twins offer a promising new pathway for building resilience, as they create dynamic, continuous updating virtual models of physical entities, enabling real-time, predictive management with adapting capacities. While already prominent in fields like manufacturing and engineering, digital twin work in stormwater management remains in its infancy, particularly from an integrated social-ecological-technical system perspective. In response, this dissertation prototypes a digital-twin-inspired stormwater management framework grounded in GIS-mapping, field-scale sensing, hydrologic simulation, and social feedback crowdsourcing.

This dissertation consists of a series of pilot empirical studies, reaching toward broader methodological and conceptual insights. It begins by refining outfall-scale stormwater catchment mapping techniques that integrate stormwater infrastructure, establishing the spatial foundation for stormwater behavior understanding. Followed by the development of an Internet-of-Things network that links in-situ sensing, analyzing, and visualizing, building the temporal foundation for stormwater monitoring. Leveraging both, stormwater flow is simulated using PCSWMM across rainfall scenarios and catchment inputs. This modeling phase confirms the functional interchangeability of catchment delineation techniques through statistical comparisons and results in robust, ground-truth-calibrated models at the outfall scale. To introduce a social dimension into this digital-twin-inspired framework, the study incorporates a novel approach using social media mining coupled with sentiment and thematic analysis to explore public attitudes to stormwater. Findings reveal that visible local management efforts help mitigate negative public sentiment toward extreme precipitation, while passive greenery shows weaker relationships. Looking forward, I synthesize these layers into a conceptual framework that reimagines stormwater digital twins as a complex adaptive system that connects data assimilation, cross-domain feedback, and public engagement pathways. By grounding technical construction of digital twin layers in adaptive principles and SETS integration, this dissertation contributes to chart a path toward stormwater systems that are not only more data-informed, but more responsive, inclusive, and resilient in the face of change.