Date of Award

Fall 2026

Abstract

Road-stream crossings (RSCs) and municipal storm sewer networks convey water across and beneath the built environment, and their failure carries consequences for flood risk, aquatic habitat, and public safety. Much of this infrastructure is aging and incompletely characterized, and it is managed under simultaneous budget, coordination, and data constraints. Replacement needs exceed available funding, responsibility is divided among agencies with differing mandates, and asset inventories are maintained for operational rather than analytical purposes. Decisions with large and long-lived consequences are therefore made without the systematic analysis that would inform them. This dissertation develops systems analyses directed at that gap, combining multi-objective optimization, stakeholder and network analysis, and automated urban drainage modeling.

In Chapter 2, a multi-objective optimization framework using Non-dominated Sorting Genetic Algorithm II (NSGA-II) is developed to prioritize RSC replacements at a watershed scale. The framework optimizes environmental, transportation, and cost objectives. The custom-seeded initialization method significantly enhanced computational efficiency. The comparisons with conventional scoring and ranking methods revealed at least 19.56% and 37.68% improvements in environmental and transportation objectives, respectively. Moreover, structural condition was identified as the most influential characteristic of RSCs for prioritization.

Chapter 3 employs a co-produced stakeholder survey (130 validated responses from 12 stakeholder groups) to identify prioritization preferences, collaboration patterns, and non-financial barriers in RSC management. Statistical analyses reveal significant variations in prioritization factors across groups, with flood vulnerability rated as the most important overarching goal and a unifying priority among stakeholders. Network analysis highlighted the central roles of agencies like the New Hampshire Departments of Transportation and Environmental Services in future collaborative initiatives. Finally, the qualitative content analysis identified non-financial barriers such as a lack of systematic prioritization framework, project complexity, and regulatory challenges.

Chapter 4 develops an automated workflow that converts an incomplete municipal separate storm sewer system (MS4) inventory and public geospatial data into a mass-conservative stormwater model, using Manchester, New Hampshire as a case study. Motivated by the capacity barriers and flood concerns identified in Chapter 3, the workflow completes network topology, reconciles vertical geometry through constrained optimization, delineates and parameterizes contributing areas, and assembles a runnable dynamic-wave input file. The committed model closes its flow-routing budget to 0.050%, and a 477-point sensitivity analysis shows the routed response is governed by the design storm and solver settings rather than by the reconciliation judgments the incomplete inventory requires. Calibration is out of scope because no observed flow record exists for this network.

Collectively, these chapters address what the decision context lacks: a way to direct limited funding toward the greatest combined benefit, an empirical account of stakeholder alignment and divergence, and a reproducible method for building hydraulic models from incomplete records. Together, they advance systematic, sustainable management of water infrastructure.

Document Type

Dissertation

First Advisor

Weiwei Mo

Second Advisor

Erin Bell

Third Advisor

Samuel Roy

Department or Program

Civil and Environmental Engineering

Degree Name

Doctor of Philosophy

Download

Share

COinS