A Data-Driven Decision Support Tool for Anticipating Tropical Storm Impacts to the United States Power Grid
Not Available
Engineering topics
Publications and source records attributed to Pasqualini, Donatella.
Not Available
Abstract Realistic simulations of complex systems are fundamental for climate and environmental studies. Large computer systems are often not sufficient to run sophisticated computational models for large numbers of different input settings. Statistical surrogate models, or emulators, are key tools enabling fast exploration of the simulator input space. Gaussian processes have become standard for computer simulator emulation. However, they require careful implementation to scale appropriately, motivating alternative methods more recently introduced. We present a comparison study of surrogates of the Sea, Lake, and Overland Surges from Hurricanes (SLOSH) simulator—the simulator of choice for government agencies—using four emulation approaches: BASS; BART; SEPIA; and RobustGaSP. SEPIA and RobustGaSP use Gaussian processes, BASS implements adaptive splines, and BART is based on ensembles of regression trees. We describe the four models and compare them in terms of computation time and predictive metrics. These surrogates use proven and distinct methodologies, are available through accessible software, and quantify prediction uncertainty. Our data cover millions of response values. We find that SEPIA and RobustGaSP provide exceptional predictive power, but cannot scale to emulate experiments as large as the one considered in this paper as effectively as BASS and BART.
Abstract Coastal saltwater intrusion (SWI) is one key factor that affects the hydrology, ecology, and biogeochemistry of coastal ecosystems. Future climate change, especially intensified sea level rise (SLR), is expected to trigger SWI to encroach on coastal freshwater aquifers more intensively. Numerous studies have investigated decadal/century scale SWI under SLR by assuming a static coastal landscape topography. However, coastal landscapes are highly dynamic in response to SLR, and the impact of coastal landscape evolution on SWI has received very little attention. Thus, this study used a coastal marsh landscape as an example and investigated how coastal marsh evolution affects future SWI with a physically‐based coastal hydro‐eco‐geomorphologic model, Advanced Terrestrial Simulator. Our numerical experiments showed that it is very likely that the marsh elevation increases with future SLR due to sediment deposition, and a depression zone is formed due to different marsh accretion rates between the ocean boundary and the inland. We found that marsh accretion may significantly reduce the surface saltwater inflow at the ocean boundary, and the evolved topographic depression zone may prolong the residence time of surface ponded saltwater, affecting subsurface salinity distribution differently. We also predicted that marshlands may become more sensitive to upland freshwater supply under future SLR, compared with previous predictions without marsh evolution. This study demonstrates the importance of coastal evolution to coastal freshwater‐saltwater interaction. The eco‐geomorphologic effect may not be ignored when evaluating coastal SWI under SLR at decadal or century scales.
Explore the source record for details and available documents.
Not Available
Vegetation drag is a fundamental quantity directly affecting results for both long- and short-term coastal marsh and geomorphological studies. The vegetation drag in coastal marshland has been modeled by various two-dimensional (2D) and threedimensional (3D) numerical parameterizations. 2D parameterizations treat coastal marshes as bottom roughness elements, while 3D parameterizations resolve the vertically-variable vegetation drag through the water column. However, differences in tidal propagation arising from different drag parameterizations within a single model are largely unknown, and clear guidance on parameterization selection is still missing. In this study, we implemented four vegetation drag parameterizations into the Model for Prediction Across Scales-Ocean (MPAS-O), which include 1) a 2D parameterization using land-cover type-determined Manning’s n (2DLM); 2) a 2D parameterization using vegetation-determined Manning’s n (2DVM); 3) a 3D parameterization for stiff vegetation (3DSV); and 4) a 3D parameterization for flexible vegetation (3DFV). Estimates of the flow resistance effects from these parameterizations were compared using a series of idealized tidal propagation simulations. Given the same tidal condition, flooding depth and flooding distance are the largest in the 2DLM simulations and the smallest in the 3DSV simulations. 2DVM results are the closest to the 2DLM results. 3DFV results are the closest to the average of 2DVM, 3DSV, and 3DFV results. 2DVM and 3DSV results are the least and most sensitive to the vegetation aboveground biomass, respectively. Based on the input data requirement and computational efficiency of each parameterization, a comparison summary is provided to help inform parameterization selection for specific applications. Here, the effects of these parameterizations on coastal geomorphology are further discussed, and the results demonstrate that estimates of the long-term evolution of coastal marshes and coastal morphology depend upon the selection of the vegetation drag parameterization
This Master’s Capstone project will present a comparison of statistical models for computer simulation studies. The four models included in this comparison study were chosen for both their proven and diverse methodologies. We will present a case study on hurricane flood data in the Delaware bay which highlights the strengths and weaknesses of each model when applied to a very large spatial field. As computers have gotten faster, we have become interested in modeling increasingly large spatial fields in both size and resolution. Statistical algorithms that are able to efficiently handle these fields have never been more important. We therefore find this comparison to be extremely topical.
A growing number of coastal eco-geomorphologic modeling studies have been conducted to understand coastal marsh evolution under sea-level rise (SLR). Although these models quantify marsh topographic change as a function of sedimentation and erosion, their representations of vegetation dynamics that control organic sedimentation differ. How vegetation dynamic schemes contribute to simulation outcomes is not well quantified. Additionally, the sensitivity of modeling outcomes to parameter selection in the available formulations has not been rigorously tested to date, especially under the influence of an accelerating SLR. In this paper, we used a coastal eco-geomorphologic model with different vegetation dynamic schemes to investigate the eco-geomorphologic feedbacks of coastal marshes and parametric sensitivity under SLR scenarios. We found that marsh platform relief increased with sea level rise rate. The simulations with different vegetation schemes exhibited different spatial-temporal variations in elevation and biomass. The nonlinear Spartina scheme presented the most resilient prediction with generally the highest marsh accretion and vegetation biomass, and the least elevation relief under SLR. But the linear Spartina scheme predicts the lowest unvegetated-vegetated ratio. We also found that vegetation-related parameters and sediment diffusivity, which were not well measured or discussed in previous studies, were identified as some of the most critical parameters. Additionally, the model sensitivity to vegetation-related parameters increased with SLR rates. The identified most sensitive parameters may inform how to appropriately choose modeling representations of key processes and parameters for different coastal marsh landscapes under SLR, and demonstrate the importance of future field measurements of these key parameters.
Sediment settling velocities are commonly estimated from analytical or process-based approaches. These approaches have theoretical constraints due to the incompletely resolved settling physics. A parametric data-driven approach was recently proposed without theoretical constraints, but it is limited by its mathematical assumptions. To overcome these limitations, here we apply a machine learning algorithm to an aggregated sediment settling experimental database and develops a nonparametric data-driven model to estimate the noncohesive sediment settling velocity in water. A cross-comparison against five process-based equations and a parametric data-driven equation demonstrates the higher accuracy and better consistency of the new model in estimating sediment settling velocities under various physical regimes. The new model also shows an easily implemented self-update capability by assimilating theoretical data derived from the process-based equations. The updated model, incorporating experimental and theoretical data of sediment settling processes, further improves the accuracy and reduces the uncertainty in estimating sediment settling velocities. This study demonstrates the capability of machine learning in sediment transport study and illustrates an alternative framework for other hydraulic engineering challenges.
Explore the source record for details and available documents.
Transmission substations are key assets in bulk electric power systems. The loss of a substation can remove many electrical transmission buses from the transmission network, resulting in power flow congestion and forcing generation redispatch or load shedding to avoid subsequent, potentially cascading, failures. In addition, load-serving transmission substations are source points of electrical power into lower voltage sub-transmission and distribution networks. The loss of these substations may directly result in outages of connected networks.
Electric distribution networks are the final leg of the network that moves electric power from generating stations to end users. The vast majority of the circuits in these networks are above ground, where they are exposed to direct damage from strong winds, broken tree limbs, toppled trees, and flying debris. These systems are also susceptible to winter storms. The combined stress of the weight of ice (icing), the increased wind resistance of the conductors, and broken tree limbs can damage lines, poles, and support structures.
Electrical distribution networks are the final leg of the network that moves electric power from generating stations to end users. The vast majority of the circuits in these networks are above-ground and exposed to direct damage from strong winds, broken tree limbs, toppled trees, and flying debris. These systems are also susceptible to ice storms. The combined stress of the weight of ice (icing), the increased wind resistance of the conductors, and broken tree limbs can damage lines, poles, and support structures.
This report describes the application of an approach for determining grid modernization investments that can best improve the resilience of communities. Under the direction of the US Department of Energy's Grid Modernization Laboratory Consortium, Sandia National Laboratories (Sandia) and Los Alamos National Laboratory (Los Alamos) collaborated with community stakeholders in New Orleans, Louisiana on grid modernization strategies for resilience. Past disruptions to the electric grid in New Orleans have contributed to an inability to provide citizens with adequate access to a wide range of infrastructure services. Using a performance-based resilience metric, Sandia and Los Alamos performed analysis on how to improve access to infrastructure services across New Orleans after a major disruption using a system of resilience nodes. Resilience nodes rely on a combination of urban planning with grid investment planning for resilience in order to design clustered infrastructure assets with highly resilient electrical supply. Results of the analysis led to suggestion of 22 draft resilience node locations that can provide a wide range of infrastructure services equitably to New Orleans citizens. This report serves as a proof-of-concept for the Urban Resilience Planning Process, and describes several gaps that should be overcome in order to integrate resilience planning between electric utilities and local governments.
Designing policies for water systems management requires the capability to assess the economic impacts of water availability and to effectively couple water withdrawals by human activities with natural hydrologic dynamics. At the core of any scientific approach to these issues there is the estimation of water withdrawals by industrial sectors in the form of water coefficients, which are measurements of the quantity of water withdrawn per dollar of GDP or output. Here, we focus on the contiguous United States and on the estimation of water coefficients for regional scale analyses. We first compare an established methodology for the estimation of national water coefficients with a parametric one we propose. Second, we introduce a method to estimate water coefficients at the level of ecological regions and we discuss how they reduce possible biases in regional analyses of water systems. Finally, we discuss advantages and limits of regional water coefficients.
Extreme weather-related disturbances, such as hurricanes, are a leading cause of grid outages historically. Although physical asset hardening is perhaps the most common way to mitigate the impacts of severe weather, operational strategies may be deployed to limit the extent of societal and economic losses associated with weather-related physical damage.1 The purpose of this study is to examine bulk power-system operational strategies that can be deployed to mitigate the impact of severe weather disruptions caused by hurricanes, thereby increasing grid resilience to maintain continuity of critical infrastructure during extreme weather. To estimate the impacts of resilient grid operational strategies, Los Alamos National Laboratory (LANL) developed a framework for hurricane probabilistic risk analysis (PRA). The probabilistic nature of this framework allows us to estimate the probability distribution of likely impacts, as opposed to the worst-case impacts. The project scope does not include strategies that are not operations related, such as transmission system hardening (e.g., undergrounding, transmission tower reinforcement and substation flood protection) and solutions in the distribution network.
In support of the goals of Department of Homeland Security’s (DHS) National Protection and Programs Directorate and the Federal Emergency Management Agency, the DHS Office of Science and Technology is seeking to develop and/or modify consequence assessment tools to enable drinking water systems owner/operators to estimate the societal and economic consequences of drinking water disruption due to the threats and hazards. This work will expand the breadth of consequence estimation methods and tools using the best-available data describing water distribution infrastructure, owner/assetlevel economic losses, regional-scale economic activity, and health. In addition, this project will deploy the consequence methodology and capability within a Web-based platform. This report is intended to support DHS effort providing a review literature review of existing assessment tools of water and wastewater systems consequences to disruptions. The review includes tools that assess water systems resilience, vulnerability, and risk. This will help to understand gaps and limitations of these tools in order to plan for the development of the next-generation consequences tool for water and waste water systems disruption.
This manuscript brie y describes a statistical ap- proach to generate synthetic tropical cyclone tracks to be used in risk evaluations. The Synthetic Hur- ricane Generator (SynHurG) model allows model- ing hurricane risk in the United States supporting decision makers and implementations of adaptation strategies to extreme weather. In the literature there are mainly two approaches to model hurricane hazard for risk prediction: deterministic-statistical approaches, where the storm key physical parameters are calculated using physi- cal complex climate models and the tracks are usually determined statistically from historical data; and sta- tistical approaches, where both variables and tracks are estimated stochastically using historical records. SynHurG falls in the second category adopting a pure stochastic approach.