Seizo Tanaka

Hurricane Gustav (2008) Waves and Storm Surge: Hindcast, Synoptic Analysis, and Validation in Southern Louisiana

AbstractHurricane Gustav (2008) made landfall in southern Louisiana on 1 September 2008 with its eye never closer than 75 km to New Orleans, but its waves and storm surge threatened to flood the city. Easterly tropical-storm-strength winds impacted the region east of the Mississippi River for 12–15 h, allowing for early surge to develop up to 3.5 m there and enter the river and the city’s navigation canals. During landfall, winds shifted from easterly to southerly, resulting in late surge development and propagation over more than 70 km of marshes on the river’s west bank, over more than 40 km of Caernarvon marsh on the east bank, and into Lake Pontchartrain to the north. Wind waves with estimated significant heights of 15 m developed in the deep Gulf of Mexico but were reduced in size once they reached the continental shelf. The barrier islands further dissipated the waves, and locally generated seas existed behind these effective breaking zones.The hardening and innovative deployment of gauges since Hur...

Performance of the Unstructured-Mesh, SWAN+ADCIRC Model in Computing Hurricane Waves and Surge

Coupling wave and circulation models is vital in order to define shelf, nearshore and inland hydrodynamics during a hurricane. The intricacies of the inland floodplain domain, level of required mesh resolution and physics make these complex computations very cycle-intensive. Nonetheless, fast wall-clock times are important, especially when forecasting an incoming hurricane.We examine the performance of the unstructured-mesh, SWAN+ADCIRC wave and circulation model applied to a high-resolution, 5M-vertex, finite-element SL16 mesh of the Gulf of Mexico and Louisiana. This multi-process, multi-scale modeling system has been integrated by utilizing inter-model communication that is intra-core. The modeling system is validated through hindcasts of Hurricanes Katrina and Rita (2005), Gustav and Ike (2008) and comprehensive comparisons to wave and water level measurements throughout the region. The performance is tested on a variety of platforms, via the examination of output file requirements and management, and the establishment of wall-clock times and scalability using up to 9,216 cores. Hindcasts of waves and storm surge can be computed efficiently, by solving for as many as 2.3⋅1012 unknowns per day of simulation, in as little as 10 minutes of wall-clock time.

Physics-based urban earthquake simulation enhanced by 10.7 BlnDOF × 30 K time-step unstructured FE non-linear seismic wave simulation

With the aim of dramatically improving the reliability of urban earthquake response analyses, we developed an unstructured 3-D finite-element-based MPI-OpenMP hybrid seismic wave amplification simulation code, GAMERA. On the K computer, GAMERA was able to achieve a size-up efficiency of 87.1% up to the full K computer. Next, we applied GAMERA to a physics-based urban earthquake response analysis for Tokyo. Using 294,912 CPU cores of the K computer for 11 h, 32 min, we analyzed the 3-D non-linear ground motion of a 10.7 BlnDOF problem with 30 K time steps. Finally, we analyzed the stochastic response of 13,275 building structures in the domain considering uncertainty in structural parameters using 3 h, 56 min of 80,000 CPU cores of the K computer. Although a large amount of computer resources is needed presently, such analyses can change the quality of disaster estimations and are expected to become standard in the future.

Rapid Assessment of Wave and Surge Risk during Landfalling Hurricanes: Probabilistic Approach

AbstractA probabilistic framework is presented for evaluation of hurricane wave and surge risk with particular emphasis on real-time automated estimation for hurricanes approaching landfall. This framework has two fundamental components. The first is the development of a surrogate model for the rapid evaluation of hurricane waves, water levels, and run-up based on a small number of parameters describing each hurricane: hurricane landfall location and heading, central pressure, forward speed, and radius of maximum winds. This surrogate model is developed using a response surface methodology fed by information from hundreds of precomputed, high-resolution Simulating Waves Nearshore (SWAN) + Advanced Circulation Model for Oceanic, Coastal and Estuarine Waters (ADCIRC) and One-Dimensional Boussinesq Model (BOUSS-1D) runs. For a specific set of hurricane parameters (i.e., a specific landfalling hurricane), the surrogate model is able to evaluate the maximum wave height, water level, and run-up during the storm...

Simulating Hurricane Storm Surge in the Lower Mississippi River under Varying Flow Conditions

AbstractHurricanes in southeastern Louisiana develop significant surges within the lower Mississippi River. Storms with strong sustained easterly winds push water into shallow Breton Sound, overtop the river’s east bank south of Pointe a la Hache, Louisiana, penetrate into the river, and are confined by levees on the west bank. The main channel’s width and depth allow surge to propagate rapidly and efficiently up river. This work refines the high-resolution, unstructured mesh, wave current Simulating Waves Nearshore + Advanced Circulation (SWAN+ADCIRC) SL16 model to simulate river flow and hurricane-driven surge within the Mississippi River. A river velocity regime–based variation in bottom friction and a temporally variable riverine flow-driven radiation boundary condition are essential to accurately model these processes for high and/or time-varying flows. The coupled modeling system is validated for riverine flow stage relationships, flow distributions within the distributary systems, tides, and Hurric...

Implicit nonlinear wave simulation with 1.08T DOF and 0.270T unstructured finite elements to enhance comprehensive earthquake simulation

This paper presents a new heroic computing method for unstructured, low-order, finite-element, implicit nonlinear wave simulation: 1.97 PFLOPS (18.6% of peak) was attained on the full K computer when solving a 1.08T degrees-of-freedom (DOF) and 0.270T-element problem. This is 40.1 times more DOF and elements, a 2.68-fold improvement in peak performance, and 3.67 times faster in time-to-solution compared to the SC14 Gordon Bell finalists state-of-the-art simulation. The method scales up to the full K computer with 663,552 CPU cores with 96.6% sizeup efficiency, enabling solving of a 1.08T DOF problem in 29.7 s per time step. Using such heroic computing, we solved a practical problem involving an area 23.7 times larger than the state-of-the-art, and conducted a comprehensive earthquake simulation by combining earthquake wave propagation analysis and evacuation analysis. Application at such scale is a groundbreaking accomplishment and is expected to change the quality of earthquake disaster estimation and contribute to society.

HPC Enhanced Large Urban Area Evacuation Simulations with Vision based Autonomously Navigating Multi Agents.

Abstract An evacuation simulation code based on Multi Agent Systems (MAS), with moderately complex agents in 2D grid envi- ronments, is developed. The main objective of this code is to estimate the effectiveness of the measures taken to smoothen and speedup the evacuation process of a large urban area, in time critical events like tsunami. A vision based autonomous navigation algorithm, which enables the agents to move through an urban environment and reach a far visible destination, is implemented. This simple algorithm enables a visitor agent to navigate through urban area and reach a destination which is several kilometers away. The navigation algorithm is verified comparing the simulated evacuation time and the paths taken by individual agents with those of theoretical. Further, a parallel computing extension is developed for studying mass evacuation of large areas; vision based autonomous navigation is computationally intensive. Several strategies like communication hiding, dynamic load balancing, etc. are implemented to attain high parallel scalability. Preliminary tests on the K-computer attained strong scalability above 94% at least up to 2048 CPU cores, with 2 million agents.

Surge Generation Mechanisms in the Lower Mississippi River and Discharge Dependency

AbstractThe Lower Mississippi River protrudes into the Gulf of Mexico, and manmade levees line only the west bank for 55 km of the Lower Plaquemines section. Historically, sustained easterly winds from hurricanes have directed surge across Breton Sound, into the Mississippi River and against its west bank levee, allowing for surge to build and then propagate efficiently upriver and thus increase water levels past New Orleans. This case study applies a new and extensively validated basin- to channel-scale, high-resolution, unstructured-mesh ADvanced CIRCulation model to simulate a suite of historical and hypothetical storms under low to high river discharges. The results show that during hurricanes, (1) total water levels in the lower river south of Pointe a La Hache are only weakly dependent on river flow, and easterly wind-driven storm surge is generated on top of existing ambient strongly flow-dependent river stages, so the surge that propagates upriver reduces with increasing river flow; (2) natural le...

ALE finite element method for FSI problems with free surface using mesh re-generation method based on background mesh

This paper presents an arbitrary Lagrangian–Eulerian (ALE) finite element method for fluid-structure interaction (FSI) problems with a free-surface using the mesh re-generation method. The mesh re-generation method using the background mesh is introduced to improve the applicability to the complicated FSI problem with large deformed interface. The incompressible Navier–Stokes equation based on ALE description is used as the governing equation of fluid. The stabilization method based on the SUPG/PSPG method is employed. The presented method is applied to several numerical examples to show the validity and efficiency of the method.

A Scalable Workbench for Large Urban Area Simulations, Comprised of Resources for Behavioural Models, Interactions and Dynamic Environments

A multi-agent based large urban area evacuation simulator is developed with the aim of addressing the limitations of the present large area simulators. Environment model of sub-meter details and agents which can visually perceive it are implemented, so that complex evacuees behaviours can be included, making it possible to study scenarios beyond those covered by the existing simple models. A mathematical framework is extended to include sufficient expressiveness and an overview of the developed software is presented in the context of this framework. Further details of the agent system and available agents’ functions are presented. In order to increase the results’ reliability, a parallel tool for automatic calibration of the agent interactions according to observed human behaviours is included. Finally, demonstrative applications of the software highlighting the need of detailed modelling are presented.