Acceleration of flow modeling using a freeware 2D GPU-shallow water equations code

被引:2
作者
Prince Mahmood
Sajjad Haider
Hamza Farooq Gabriel
Muhammad Shahid
Zain Syed
机构
[1] School of Civil & Environmental Engineering (SCEE),Faculty of Civil Engineering
[2] National University of Sciences and Technology,undefined
[3] (NUST),undefined
[4] University of Engg. & Technology,undefined
关键词
Flow modeling; GPU; CUDA; Nullah Lai; BASEMENT;
D O I
10.1007/s12517-022-10836-6
中图分类号
学科分类号
摘要
Excessive rainfall, storm surges and dam break are some of the serious Geohazards afflicting humankind. The non-structural technique to combat floods includes flood modeling which provides a complete flood history over an area of interest in the time domain. Over the past decades, a number of 1D and 2D hydraulic models have been developed to simulate floods; amongst these, the shallow water equations figure prominently as a potent model. However, these models suffer from a serious shortcoming, that is, of long run times, particularly for high resolution and large areas. A number of approaches and strategies have been proposed to remedy this issue, e.g. using models of reduced complexity, parallelization of algorithms and use of Graphical Processing Units (GPUs) to make the codes run faster. In this study, a freeware, 2D, GPU-enhanced code BASEMENT has been applied to three different test cases related to the laboratory and field. The results showed that, for lab-based cases, the GPU accelerated model was able to achieve a significant reduction in model execution time, attaining a Speedup ratio of 80, while for the field case the ratio was 8, which showed the potential of the GPU technology for flood inundation modeling. Further, the availability of a GPU-enhanced model in the public domain is very important to several scientific disciplines, e.g. Hydraulic Engineering, Flood Control and Management, Hazard Mitigation, etc.
引用
收藏
相关论文
共 224 条
[1]  
Alho P(2008)Comparing a 1D hydraulic model with a 2D hydraulic model for the simulation of extreme glacial outburst floods Hydrol Processes: An Int J 22 1537-1547
[2]  
Aaltonen J(2021)A dual-layer MPI continuous large-scale hydrological model including Human Systems Environ Model Softw 139 105003-13
[3]  
Avesani D(2013)Graphics processing unit (GPU) programming strategies and trends in GPU computing J Parallel Distribut Comput 73 4-223
[4]  
Galletti A(2021)SW2D-GPU: a two-dimensional shallow water model accelerated by GPGPU Environ Model Softw 145 105205-16
[5]  
Piccolroaz S(2018)A review on GPU programming strategies and recent trends in GPU computing J Graphic Era Univ 6 207-11
[6]  
Bellin A(2012)A coarse-grid approach to representing building blockage effects in 2D urban flood modelling J Hydrol 426 1-214
[7]  
Majone BA(2012)Multi-layered coarse grid modelling in 2D urban flood simulations J Hydrol 470 1-522
[8]  
Brodtkorb AR(2021)Comparison of two modelling strategies for 2D large-scale flood simulations Environ Model Softw 146 105225-165
[9]  
Hagen TR(2011)Simulation of one-layer shallow water systems on multicore and CUDA architectures J Supercomput 58 206-27
[10]  
Sætra ML(2019)2D numerical simulation of unsteady flows for large scale floods prediction in real time Adv Water Resour 134 103444-3026