Abstract. The local inertial two-dimensional (2D) flow model on LISFLOOD-FP, the so-called ACCeleration (ACC) uniform grid solver, has been widely used to support fast, computationally efficient fluvial/pluvial flood simulations. This paper describes new releases, on LISFLOOD-FP 8.1, for parallelised flood simulations on the graphical processing units (GPUs) to boost efficiency of the existing parallelised ACC solver on the central processing units (CPUs) and enhance it further by enabling a new non-uniform grid version. The non-uniform solver generates its grid using the multiresolution analysis (MRA) of the multiwavelets (MWs) to a Galerkin polynomial projection of the digital elevation model (DEM). This sensibly coarsens the resolutions where the local topographic details are below an error threshold ε and allows classes of land use to be properly adapted. Both the grid generator and the adapted ACC solver on the non-uniform grid are implemented in a GPU new codebase, using the indexing of Z-order curves alongside a parallel tree traversal approach. The efficiency performance of the GPU parallelised uniform and non-uniform grid solvers is assessed for five case studies, where the accuracy of the latter is explored for ε=10-4 and 10−3 in terms of how close it can reproduce the prediction of the former. On the GPU, the uniform ACC solver is found to be 2–28 times faster than the CPU predecessor with increased number of elements on the grid, and the non-uniform solver can further increase the speed up to 320 times with increased reduction in the grid's elements and decreased variability in the resolution. LISFLOOD-FP 8.1, therefore, allows faster flood inundation modelling to be performed at both urban and catchment scales. It is openly available under the GPL v3 license, with additional documentation at https://www.seamlesswave.com/LISFLOOD8.0 (last access: 12 March 2023).
Abstract. The local inertial two-dimensional (2D) flow model on LISFLOOD-FP, so-called ACC uniform grid solver, has been widely used to support fast, computationally efficient fluvial/pluvial flood simulations. This paper describes new releases, on LISFLOOD-FP 8.1, for parallelised flood simulations on the graphical processing units (GPU) to boost efficiency of the existing parallelised ACC solver on the central processing units (CPU) and enhance it further by enabling a new non-uniform grid version. The non-uniform solver generates its grid using the multiresolution analysis (MRA) of the multiwavelets (MWs) to a Galerkin projection of the digital elevation model (DEM). This sensibly coarsens the resolutions where the local topographic details are below an error threshold ε and allows to properly adapt classes of land use. Both the grid generator and the adapted ACC solver on the non-uniform grid are implemented in a GPU new codebase, using the indexing of Z-order curves alongside a parallel tree traversal approach. The efficiency performance of the GPU parallelised uniform and non-uniform grid solvers are assessed for five case studies, where the accuracy of the latter is explored for ε = 10-4 and 10-3 in terms of how close it can reproduce the prediction of the former.
Wavelet-based grid adaptation driven by the ‘multiresolution analysis’ (MRA) of the Haar wavelet (HW) allows to devise an adaptive first-order finite volume (FV1) model (HWFV1) that can readily preserve the modelling fidelity of its reference uniform-grid FV1 counterpart. However, the MRA incurs a high computational cost as it involves ‘encoding’ (coarsening), ‘decoding’ (refining), analysing and traversing modelled data across a deep hierarchy of nested, uniform grids. GPU-parallelisation of the MRA is needed to reduce its computational cost, but its algorithmic structure (1) hinders coalesced memory access on the GPU and (2) involves an inherently sequential tree traversal problem. This work redesigns the algorithmic structure of the MRA in order to parallelise it on the GPU, addressing (1) by applying Z-order space-filling curves and (2) by adopting a parallel tree traversal algorithm. This results in a GPU-parallelised HWFV1 model (GPU-HWFV1). GPU-HWFV1 is verified against its CPU predecessor (CPU-HWFV1) and its GPU-parallelised reference uniform-grid counterpart (GPU-FV1) over five shallow water flow test cases. GPU-HWFV1 preserves the modelling fidelity of GPU-FV1 while being up to 30 times faster. Compared to CPU-HWFV1, it is up to 200 times faster, suggesting that the GPU-parallelised MRA could be used to speed up other FV1 models.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.