Enhancing Stormwater Resilience Using AI and Digital Tools

Introduction and Objective Stormwater systems typically operate passively, meaning they primary rely on gravity and static controls to collect and treat stormwater runoff. Artificial intelligence (AI), stormwater modeling improvements, and data availability have created the opportunity to automate data collection, link hydrologic and hydraulic models to provide results in real time, and significantly enhance visualization of results. This paper will provide examples of how forecast data and sensor data are being used to revolutionize stormwater system operations through the creation of digital twins that enhance stormwater and flood resilience. Approach and Solution Dynamic approaches to water resources management will be key in sustaining the future of our cities and counties. Climate change, pollution, rapid growth, and aging infrastructure pose unprecedented challenges. Meanwhile, urban, agricultural, and natural areas suffer from devastating floods, droughts, and poor water quality. Making decisions to manage our changing world require new, digital dynamic approaches. In an ever-evolving world, where technology shapes our lives in profound ways, digital applications emerge as powerful allies in safeguarding our most critical resources: water and urban spaces. In this paper, we will explore how stormwater system modeling coupled with digital platforms support decision-making and contribute to a sustainable future by protecting cities from floods and other natural disasters. National and international case studies will be used to illustrate how technology can help us visualize the impacts of upcoming weather events on local infrastructure and what we can learn from them. The paper will describe the creation of an Actionable Stormwater Platform (ASP) that integrates an existing watershed stormwater model with real-time and forecasted weather data, providing predictive flood analytics and assessing the impact of historical and upcoming storms on the city's stormwater system. The paper will also describe the benefits of an ASP and decision-support-platforms in general. Some of these results of the application include the following: Cost Savings Improved Resilience Improved Performance and Insights Adaptability The Need for a Tailored System The ASP workflow uses forecasted rainfall data from sources such as the National Oceanic and Atmospheric Administration (NOAA), the United Stated Geological Survey (USGS), and local available rain and tide gauges to run the stormwater model. Forecasted inundation and water levels maps are generated from model runs and published on a cloud-based dashboard. This workflow is automated and repeated at regular intervals of time (e.g., 6 hours, 12 hours, 24 hours) through public Application Programming Interfaces (API) and Microsoft Azure. Model results, maps and dashboards are updated accordingly, allowing city staff to proactively organize response activities. Improved preparedness and response, along with timely flood information and advanced warnings, can boost community resilience in mitigating flooding events. International case studies will be used to illustrate how these tools have been used to manage stormwater and water resources systems. A detailed case study for the City of Portsmouth, VA, will be used to describe the development of an ASP and flood resilience strategy to protect the community from flood hazards, reduce flood risks, and enhance safety and stormwater services during storm events. For this purpose, a citywide stormwater model using the USEPA Storm Water Management Model (SWMM) was built to analyze existing drainage systems and built the workflow described above. The ASP platform provides the City of Portsmouth a cloud-based dashboard with insights on forecasts, stormwater system performance, and potential proactive measures. The dashboard can be easily accessed online and alerts can be triggered based on preset thresholds in forecast data or model results. It also helps with emergency management, storm preparation, and targeted maintenance based on real-time and forecasted weather and tide data, as well as forecasted model results. The platform provides detailed information about potential flooded areas at a street-by-street level, considering forecasted storm conditions, enabling data-driven water management decisions. Other national and international case studies will be used to describe other applications and data integration. These examples will illustrate key concepts on the use of AI to improve decision making by creating digital twins of stormwater systems that link data and models and help predict and visualize future floods and infrastructure impacts. A decision-support platform, HydroNET, will be used to illustrate applications. Finally, this paper will provide an overview of ethical and legal consideration in the application of AI and, during the presentation, we will engage the audience in a discussion on the need to innovate and the role of AI and machine learning, workforce impacts, and opportunities to develop new skills to ride the AI/digital technology wave while enhancing sustainability and improving quality of life.