GPU Acceleration of Tsunami Propagation Model

The Fukushima Daiichi nuclear disaster that occurred on March 11, 2011 was caused by the Tohoku tsunami which was itself triggered by the devastating 9.0Mw moment magnitude earthquake. The present study investigates spatial and temporal changes of Suspended Particulate Matter (SPM) content in the NorthEastern part of Japan (Pacific Ocean) using a geostationary ocean color sensor. The Geostationary Ocean Color Imager (GOCI), which is centered on the Korean peninsula but could also observe the Japanese area, is able to acquire 8 images per day, thus allowing the analysis of rapid daily changes in water mass. The analysis of GOCI data shows that SPM concentration notably increased both along the coast and within the Bay of Sendai shortly after the tsunami. Motionless patterns of SPM were observed at 2, 14, 25 and 37 km from the coast. It is shown that SPM concentration rapidly decreased one month later. The SPM concentration did not remain high the following year, contrary to what was observed for the Sumatra Tsunami in 2004. The origin of SPM is also investigated in this study. Our analysis suggests that some of the SPM originates from the resuspension of bottom sediments due to the reflection of the tsunami on the coastline that leads to the migration of marine particles towards the sea surface. The fate of the SPM concentration is then discussed based on the analysis of meteorological conditions, river discharge and tsunami wave properties.

show abstract

“…where H is the wave height, c = Jgh is the wave celerity, g the acceleration due to the gravity, h the bottom depth [42].…”

Section: Hcmentioning

confidence: 99%

Monitoring Suspended Particle Matter Using GOCI Satellite Data After the Tohoku (Japan) Tsunami in 2011

Minghelli

Lei²,

Charmasson

et al. 2019

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Finally, the 3D Navier-Storkes equations provide the most complete description, but they are significantly more complex than other modelsexamples include SAGE (Gisler et al (2006)) and the work of Abadie et al (2012). Most of these codes described above work on CPUs, and while there has been some work on GPU implementations by Satria et al (2012) and Liang et al (2009), which use structured meshes and finite differences, it is unclear whether these are used 15 in production. As far as we are aware, only Tsunami-HySEA (Macías et al (2017)), also using finite volumes, is using GPU clusters in production -that code however only supports GPUs, and is hand-written in CUDA.…”

Section: Related Workmentioning

confidence: 99%

The VOLNA-OP2 Tsunami Code (Version 1.0)

Reguly¹,

Gopinathan²,

Beck³

et al. 2018

Preprint

View full text Add to dashboard Cite

Abstract. In this paper, we present the VOLNA-OP2 tsunami model and implementation; a finite volume non-linear shallow water equations (NSWE) solver built on the OP2 domain specific language for unstructured mesh computations. VOLNA-OP2 is unique among tsunami solvers in its support for several high performance computing platforms: CPUs, the Intel Xeon Phi, and GPUs. This is achieved in a way that the scientific code is kept separate from various parallel implementations, enabling easy maintainability. It has already been used in production for several years, here we discuss how it can be integrated into 5 various workflows, such as a statistical emulator. The scalability of the code is demonstrated on three supercomputers, built with classical Xeon CPUs, the Intel Xeon Phi, and NVIDIA P100 GPUs. VOLNA-OP2 shows an ability to deliver productivity to its users, as well as performance and portability on a number of platforms.

show abstract

“…Finally, the 3D Navier-Stokes equations provide the most complete description, but they are significantly more complex than other models -examples include SAGE Gisler et al (2006) and the work of Abadie et al (2012). Most of these codes described above work on CPUs, and while there has been some work on GPU implementations by Satria et al (2012); Liang et al (2009a); Brodtkorb et al (2010); Acuña and Aoki (2009); Liang et al (2009b), which use structured meshes and finite differences or finite volumes, it is unclear whether these are used in production, and they are not open source.…”

Section: Related Workmentioning

confidence: 99%

Response to reviewers' comments

Reguly¹

2018

Preprint

View full text Add to dashboard Cite

In this paper, we present the VOLNA-OP2 tsunami model and implementation; a finite volume non-linear shallow water equations (NSWE) solver built on the OP2 domain specific language (DSL) for unstructured mesh computations. VOLNA-OP2 is unique among tsunami solvers in its support for several high performance computing platforms: CPUs, the Intel Xeon Phi, and GPUs. This is achieved in a way that the scientific code is kept separate from various parallel implementations, enabling easy maintainability. It has already been used in production for several years, here we discuss how it can be integrated into various workflows, such as a statistical emulator. The scalability of the code is demonstrated on three supercomputers, built with classical Xeon CPUs, the Intel Xeon Phi, and NVIDIA P100 GPUs. VOLNA-OP2 shows an ability to deliver productivity to its users, as well as performance and portability across a number of platforms.

show abstract

GPU Acceleration of Tsunami Propagation Model

Cited by 23 publications

References 17 publications

Monitoring Suspended Particle Matter Using GOCI Satellite Data After the Tohoku (Japan) Tsunami in 2011

Monitoring Suspended Particle Matter Using GOCI Satellite Data After the Tohoku (Japan) Tsunami in 2011

The VOLNA-OP2 Tsunami Code (Version 1.0)

Response to reviewers' comments

Contact Info

Product

Resources

About