School of Civil and Environmental Engineering

Ph.D. Thesis Defense Announcement

DATA-DRIVEN STORMWATER DECISION SUPPORT USING GIS, COMMUNITY FEEDBACK, AND BAYESIAN NETWORKS

By Michelle Reckner

Advisor:

Dr. Iris Tien (CEE)

Committee Members:

Dr. Chris Chungkei Lai (CEE), Dr. Allen Hyde (HSOC), Dr. Joe F. Bozeman III (CEE), Dr. Yiyi He (SCaRP)

Date and Time: April 13th, 2026. 3:30 pm EST

Location: Guggenheim 244 (Enter through Montgomery Knight)/ https://teams.microsoft.com/meet/26363114494118?p=MzPKNGHh2wmEOL1gXy

Climate change is putting a greater demand on stormwater infrastructure through more intense and frequent rainfall events. Stormwater infrastructure capacity is also decreasing over time due to corrosion and blockages. Communities are tasked with addressing both the increased demand and capacity degradation through maintenance and additional infrastructure. There are two main types of stormwater infrastructure: grey (e.g., underground pipes), which is the more traditional and often more affordable type, and green (e.g., bioretention ponds), which are often nature-based systems that may provide additional benefits such as increased biodiversity and improving air quality in marginalized neighborhoods. There is no continuous methodology for community-wide and community informed stormwater improvement decision making. Scientific literature provides some pieces of the decision process but does not have pairable methods. This research presents compatible decision-making methods for communities to improve their stormwater systems. First, the thesis presents a Dynamic Bayesian Network (DBN) to assess current grey infrastructure systems and recommend more frequent maintenance or additional capacity. The DBN analyzes the probabilistic relationship between variables affecting stormwater infrastructure performance, highlighting factors such as pipe sediment and wetland proximity. Then, stakeholder perspectives are gathered through a GIS-based suitability assessment to determine where green or grey infrastructure is ideal for those who live there. Those perspectives are then compared to youth in the area through a community engagement education project: Youth Advocacy for Resilience to Disasters (YARDs). Lastly, the DBN and GIS engagement methods are combined to create watershed-based recommendations for stormwater infrastructure improvement. These methods are applied to Chatham County, Georgia, a vulnerable coastal community, and can be implemented in other coastal communities. This research provides a community-centered approach to prioritizing stormwater infrastructure investments through mapped results that can be implemented by both city planners and engineers.