Date of Award
1-1-2025
Degree Type
Dissertation
Degree Name
Doctor of Philosophy in Biological and Environmental Sciences
Specialization
Environmental and Earth Sciences
Department
Geosciences
First Advisor
Thomas B. Boving
Abstract
Effective reservoir management is essential for mitigating downstream flooding, especially with the increasing frequency and intensity of extreme weather events impacting major drinking water reservoirs such as the Scituate Reservoir in Rhode Island. These reservoirs provide water for large populations but can be overwhelmed by intense precipitation, flooding and land cover changes. Climate change is expected to increase precipitation in the study area located in New England, highlighting the need for improved reservoir systems, especially given the aging dams and higher risk of failures. Additionally, land cover changes, like the widespread defoliation caused by the Gypsy Moth infestations, affect water availability and reservoir operations. This study developed an integrated model to assess how the Scituate Reservoir responds to extreme weather and land cover changes in the future.
A hydrologic modeling (SWAT-Soil and Water Assessment tool) approach was used, and three highly parameterized models were developed to evaluate the spatial and temporal relationships between stream flows and factors such as surface/subsurface runoff processes, flow timing, and snow accumulation on reservoir elevation. These models were calibrated using multi-site calibration and parameterization techniques. Models incorporating surface/subsurface runoff processes (FSP) or a combination of FSP and snow accumulation parameters (COMP) captured to high flow conditions, while the snow-specific model (SSP) alone was overestimated to capture high flow events. All model combinations predicted low flow conditions with an acceptable overestimation of 3%.
Next, a coupled modeling approach was used to assess the impact of extreme flood events on the reservoir under CMIP6 climate scenarios (RCP4.5 and RCP8.5). The SWAT model, calibrated and validated using PRISM climate data, land cover, soil data, and multi-site streamflow gauges (N), provided inflow data for a OASIS (Operational Analysis and Simulation of Integrated Systems) reservoir management model. Simulations predicted a 10–15% rise in mean 3D30Y (3-day runoff volume during 30-year runoff simulation) April runoff volumes from baseline, with extreme events showing up to a 150% increase. The 7D30Y (7-day elevation during 30-year simulation) elevation projections indicated heightened risks of reservoir overspill in April, particularly from 2020 onwards. These findings emphasize the utility of integrated modeling in devising adaptive reservoir operating rules.
The SWAT hydrologic model, integrated with a re-calibrated and re-validated OASIS reservoir management model from 2015 to 2018, was employed to analyze the relationship between Leaf Area Index (LAI), evapotranspiration (ET), and reservoir elevation during pre-defoliation, defoliation, and post-defoliation periods. MODIS remote sensing data provides inputs for capturing vegetation dynamics. During active defoliation, reservoir inflow dynamics changed but ET remained stable despite LAI fluctuations, indicating a negligible effect of diminished vegetation on ET. Overall, these models suggest that defoliation events have minimal impact on the reservoir elevation, as the amount of excess water due to diminished ET is very small relative to this reservoir’s total capacity.
Recommended Citation
Paul, Supria, "DEVELOPMENT OF A SCITUATE RESERVOIR MITIGATION MODEL FOR ASSESSING FLOOD CONTROL DURING EXTREME SCENARIOS" (2025). Open Access Dissertations. Paper 4467.
https://digitalcommons.uri.edu/oa_diss/4467
Included in
Terms of Use
All rights reserved under copyright.