New generation water quality computational toolchain

Autonomous water systems are promising to transform watershed management, but no computational toolchains existed to evaluate their potential, due to overlapping needs to model water flow, water quality, and controls. To address this need, Brooke built an open-source, computational Python package, StormReactor, which coupled the popular EPA’s Stormwater Management Model (SWMM) with a new generation water quality module. This integrated model will revolutionize how researchers and practitioners test water quality-based real-time control strategies. More details on StormReactor can be found in a recently published article in Environmental Modelling & Software.

Novel stormwater infrastructure sensing network

Brooke is developing a novel, “Internet of Things” stormwater infrastructure sensing network in partnership with the Detroit Sierra Club. A network of over twenty custom sensor nodes, built using open-source solutions, will monitor flooding in primarily low income, minority neighborhoods in Detroit, Michigan. The network will create the largest dataset of stormwater infrastructure performance to date, shedding a light on the complexities of stormwater dynamics at an unprecedented spatial and temporal scale. Insights from this network will not only improve the design and placement of future infrastructure in Detroit, but in communities around the world. In addition, the network will serve as an instructional testbed for a workforce development program, teaching Detroiters the trade skills necessary for the rising autonomous water systems sector

Real-time controlled green infrastructure

Studies have shown inconsistent nutrient removal in green infrastructure, particularly in bioretention systems. One potential solution is real-time control, which has been shown to improve pollutant removal in grey infrastructure. Since few studies have investigated this potential solution for green infrastructure, pollutant treatment in a real-time controlled bioretention cell is simulated using StormReactor. The results suggest real-time control may provide a “digital” alternative to existing, passive upgrades, like soil amendments, for boosting pollutant treatment.

Water Quality Sampling Nodes

Collaborating with the Huron River Watershed Council on a surface water quality monitoring project. Brooke developed a small, low-cost, data node to collect real-time data, make automated decisions, and allow for remote triggering of the samplers. The node sends data to a dashboard allowing a team to monitor the depth of water in the stream/creek, along with sampler information. In manual mode, the team can simply tap a “trigger” button on the dashboard to order a sample.

First-of-its-kind systems framework for autonomous watersheds

Brooke is developing a first-of-its-kind computational framework for autonomous watersheds. At the core of this vision is system-level control, where tens to hundreds of individual stormwater assets (e.g., ponds, wetlands) will coordinate in real-time to route water and promote pollutant treatment. Although autonomous control can mitigate flooding and particulate pollutants, it has not yet been evaluated for other water quality parameters. Unfortunately, the application of control theory for water quality is not yet feasible due to the nonlinearities inherent in most stormwater models. To address this limitation, Brooke is formulating a water quality control model for the system-level control of stormwater networks.