Back to search

Cornell University

Sensor-driven flood risk monitoring in levee-protected floodplains and urban storm sewer networks

Abstract

dc:description.abstract

Flooding continues to pose a global threat to life and property and communities in both rural and urban areas face compound flood risk from riverine, pluvial, coastal flooding, or combinations of these three types. Riverine flooding is often managed with levees that reduce the frequency of low-magnitude floods and protect infrastructure in low-lying floodplains. Flooding in urban areas is often managed with stormwater collection systems, which convey water away from populated areas. Sufficiently large precipitation events can overwhelm flood defense infrastructure, especially when the infrastructure is poorly maintained or degraded from its design specifications. Levees can fail due to overtopping or erosion and stormwater collection systems can exhibit unexpected capacity reductions, for example due to blockages. In recent decades, advancements in low-cost depth, velocity, flow sensors, increases in computer modeling resolution, and new data processing algorithms have given municipalities a broad set of tools for monitoring and modeling their combined flood risk. Given the rapid development of these tools, there are still many unanswered questions about how to efficiently and effectively allocate sensors, interpret sensor and model data to characterize flood risk, and validate the outputs. This thesis addresses aspects of these three questions through case studies and simulations in a levee protected floodplain, a model of an urban stormwater system, and a laboratory-scale stormwater pipe. Specifically, the studies address monitoring of flood propagation in real time, detection of stormwater system sections that are at risk of overflow, and diagnosis of blockages and retention basin silting in early stages of formation. The first study explores installing a sensitivity-based flow depth sensor network across a floodplain to calibrate flow resistance parameters for a levee breach flood model in real time. The sparse network of sensors is sufficient to accurately estimate the parameters and produce an accurate flood forecast with enough lead time to guide downstream emergency preparation. The following three studies focus on sensor applications to urban stormwater system monitoring for emergent flood risk. The first study in this vein optimizes the placement of depth and flow sensors in a large (~1000 manhole) sewer system to ensure complete detection of pipe blockage driven overflows. Optimized networks consist of sensors in manholes with large fields of view, meaning they experience backup from blockages far downstream. Results from numerical simulations show that the optimized sensor networks outperform randomly assigned sensor networks in blockage-induced overflow detection. In the second study, fire hydrant flushing is used as a known impulse to a section of stormwater system and the resulting impulse-response hydrograph measured downstream is used to detect blockages and retention basin silting. A calibration-optimization procedure estimates the parameters of a simple stormwater system numerical model to also diagnose the location and severity of the blockages and retention basin silting. Blockages and retention basin silting of varying severity were accurately diagnosed with the procedure under increasing noise contamination, though localization and severity characterization suffered at high noise levels. For the final study, blockage detection from downstream depth sensor measurements is investigated with experiments in a laboratory-scale stormwater system. A hydrograph recession rate metric was able to accurately detect and locate in time a 40% and 60% pipe diameter constriction from a sensor placed 60 diameters downstream of the blockage. This result introduces a novel sewer data analysis technique with the potential to augment existing data-driven blockage detection methods. Taken together, this thesis builds new methodologies for implementation of sensor-based flood risk management, but also highlights the need for developing solutions tailored to the needs of specific municipalities due to the diverse range of flood risks and no one-size-fits-all approach.

Degree

thesis:*
Name thesis:degree_name
Ph. D., Civil and Environmental Engineering
Level thesis:degree_level
Doctor of Philosophy
Discipline thesis:degree_discipline
Civil and Environmental Engineering
Grantor
Cornell University
Year dc:date.issued
2025

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Young, Alexander
Chair dc:contributor.chair
  • Albertson, John
Committee members dc:contributor.committeemember
  • Nozick, Linda
  • Steinschneider, Scott

Subjects

dc:subject × 5

Rights

dc:rights
Statement dc:rights
  • Attribution-NonCommercial-NoDerivatives 4.0 International
Language dc:language.iso
en

Identifiers

dc:identifier.*
Dc Identifier Bibid
17422386
Dc Identifier Other
ProQuest Submission ID: 15175
ProQuest Publication ID: 32169271

Chain of custody

source
Harvested from
Cornell University
Base URL
ecommons.cornell.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Young, Alexander. Sensor-driven flood risk monitoring in levee-protected floodplains and urban storm sewer networks. Doctor of Philosophy thesis, Cornell University, 2025. https://hdl.handle.net/1813/120923