Document Type
Poster
Start Date
15-10-2024 5:00 PM
End Date
15-10-2024 8:00 PM
Abstract
Statistical characterization of reanalysis datasets during over 300 hydrologic dam incidents between 2003 and 2022 will create a detailed typology of weather systems associated with dam overtopping in the eastern United States. Dam overtopping poses significant risks to infrastructure and public safety, necessitating a comprehensive understanding of the multi-scale atmospheric conditions that lead to such events. To better account for the natural flow of water to the affected dams, we will adopt a watershed-focused Principal Component Analysis (PCA) on regional atmospheric data collected from ERA5 alongside USGS streamflow and Stage IV precipitation observations to enhance understanding of high-risk weather conditions. PCA will be employed to reduce the dimensionality of our data and identify key components that capture significant flow characteristics, such as geopotential height, temperature, and stability at multiple levels, as well as those previously found as important for extreme precipitation and convective support, such as Integrated Vapor Transport (IVT) and Convective Available Potential Energy (CAPE). We expect PCA to highlight how characteristics of dominant meteorological features, such as moisture advection, frontal evolution, and storm propagation, contribute to dam overtopping events. This initial study focuses on the Northeast watershed region, encompassing New Jersey, New York, New Hampshire, and Pennsylvania, to offer insights into the atmospheric dynamics of extreme precipitation and subsequent dam overtopping events in this area. Spatial analysis will then examine the geographical distribution of these characteristics and their correlation with dam overtopping incidents within the Northeast region. We anticipate our results from this region to be useful for comparisons with predictive models and to aid in the development of effective risk management strategies for dam safety. This focused study also provides a solid foundation for applying clustering algorithms and creating composite mapping to detail key variables for each weather type within this watershed.
DOI
10.5703/1288284317810
Included in
Atmospheric Sciences Commons, Climate Commons, Meteorology Commons
Typology of Atmospheric Conditions Leading to Dam Overtopping in the Eastern US
Statistical characterization of reanalysis datasets during over 300 hydrologic dam incidents between 2003 and 2022 will create a detailed typology of weather systems associated with dam overtopping in the eastern United States. Dam overtopping poses significant risks to infrastructure and public safety, necessitating a comprehensive understanding of the multi-scale atmospheric conditions that lead to such events. To better account for the natural flow of water to the affected dams, we will adopt a watershed-focused Principal Component Analysis (PCA) on regional atmospheric data collected from ERA5 alongside USGS streamflow and Stage IV precipitation observations to enhance understanding of high-risk weather conditions. PCA will be employed to reduce the dimensionality of our data and identify key components that capture significant flow characteristics, such as geopotential height, temperature, and stability at multiple levels, as well as those previously found as important for extreme precipitation and convective support, such as Integrated Vapor Transport (IVT) and Convective Available Potential Energy (CAPE). We expect PCA to highlight how characteristics of dominant meteorological features, such as moisture advection, frontal evolution, and storm propagation, contribute to dam overtopping events. This initial study focuses on the Northeast watershed region, encompassing New Jersey, New York, New Hampshire, and Pennsylvania, to offer insights into the atmospheric dynamics of extreme precipitation and subsequent dam overtopping events in this area. Spatial analysis will then examine the geographical distribution of these characteristics and their correlation with dam overtopping incidents within the Northeast region. We anticipate our results from this region to be useful for comparisons with predictive models and to aid in the development of effective risk management strategies for dam safety. This focused study also provides a solid foundation for applying clustering algorithms and creating composite mapping to detail key variables for each weather type within this watershed.