Preprints
https://doi.org/10.22541/essoar.176442648.85093032/v3
https://doi.org/10.22541/essoar.176442648.85093032/v3
05 Feb 2026
 | 05 Feb 2026

CaMa-Flood-GPU: A GPU-based hydrodynamic model implementation for scalable global simulations

Shengyu Kang, Jiabo Yin, and Dai Yamazaki

Abstract. Floods are among the costliest natural hazards, demanding scalable models to simulate river and floodplain dynamics at a global scale. The Catchment-based Macro-scale Floodplain (CaMa-Flood) model is a leading system for this purpose, but its CPU-based implementation is computationally demanding. This paper introduces CaMa-Flood-GPU, a fundamental refactoring of the model optimized for Graphics Processing Unit (GPU) architectures. We systematically reinterpreted its core algorithms—including river routing on irregular networks, runoff interpolation, and water depth diagnosis—into highly parallel, GPU-native operations. Key challenges were addressed by implementing scatter-add for flux updates, sparse matrix multiplication for runoff mapping, and branchless kernels for floodplain dynamics, all while preserving the original model's physical fidelity. Implemented in Python with Triton kernels and PyTorch, CaMa-Flood-GPU achieves multi-GPU scalability through optimized communication patterns that minimize synchronization overhead. The software adopts a modular structure with optional components (e.g., bifurcation routing, adaptive time stepping) and flexible data interfaces. Benchmarks demonstrate an order-of-magnitude speedup over a 192-core CPU baseline and near-linear scaling on multiple GPUs, with negligible numerical differences from the original model. This performance leap reduces simulation times for high-resolution global runs from days to hours, enabling larger ensembles and rapid scenario analysis. By providing a reproducible and efficient tool, CaMa-Flood-GPU lowers the barrier for adopting GPU acceleration in large-scale hydrology. The released implementation provides a reproducible reference for future method development.

Share

Journal article(s) based on this preprint

29 Jun 2026
CaMa-Flood-GPU: a GPU-based hydrodynamic model implementation for scalable global simulations
Shengyu Kang, Jiabo Yin, and Dai Yamazaki
Geosci. Model Dev., 19, 5623–5640, https://doi.org/10.5194/gmd-19-5623-2026,https://doi.org/10.5194/gmd-19-5623-2026, 2026
Short summary
Shengyu Kang, Jiabo Yin, and Dai Yamazaki

Interactive discussion

Status: closed

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2025-6500', Anonymous Referee #1, 18 Mar 2026
    • AC1: 'Reply on RC1', Jiabo Yin, 01 May 2026
  • RC2: 'Comment on egusphere-2025-6500', Anonymous Referee #2, 22 Mar 2026
    • AC2: 'Reply on RC2', Jiabo Yin, 01 May 2026

Peer review completion

AR – Author's response | RR – Referee report | ED – Editor decision | EF – Editorial file upload
AR by Jiabo Yin on behalf of the Authors (06 May 2026)  Author's response   Author's tracked changes   Manuscript 
ED: Publish subject to minor revisions (review by editor) (07 Jun 2026) by Thomas B. Wild
AR by Jiabo Yin on behalf of the Authors (10 Jun 2026)  Author's response   Author's tracked changes   Manuscript 
ED: Publish as is (21 Jun 2026) by Thomas B. Wild
AR by Jiabo Yin on behalf of the Authors (21 Jun 2026)  Manuscript 

Interactive discussion

Status: closed

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2025-6500', Anonymous Referee #1, 18 Mar 2026
    • AC1: 'Reply on RC1', Jiabo Yin, 01 May 2026
  • RC2: 'Comment on egusphere-2025-6500', Anonymous Referee #2, 22 Mar 2026
    • AC2: 'Reply on RC2', Jiabo Yin, 01 May 2026

Peer review completion

AR – Author's response | RR – Referee report | ED – Editor decision | EF – Editorial file upload
AR by Jiabo Yin on behalf of the Authors (06 May 2026)  Author's response   Author's tracked changes   Manuscript 
ED: Publish subject to minor revisions (review by editor) (07 Jun 2026) by Thomas B. Wild
AR by Jiabo Yin on behalf of the Authors (10 Jun 2026)  Author's response   Author's tracked changes   Manuscript 
ED: Publish as is (21 Jun 2026) by Thomas B. Wild
AR by Jiabo Yin on behalf of the Authors (21 Jun 2026)  Manuscript 

Journal article(s) based on this preprint

29 Jun 2026
CaMa-Flood-GPU: a GPU-based hydrodynamic model implementation for scalable global simulations
Shengyu Kang, Jiabo Yin, and Dai Yamazaki
Geosci. Model Dev., 19, 5623–5640, https://doi.org/10.5194/gmd-19-5623-2026,https://doi.org/10.5194/gmd-19-5623-2026, 2026
Short summary
Shengyu Kang, Jiabo Yin, and Dai Yamazaki
Shengyu Kang, Jiabo Yin, and Dai Yamazaki

Viewed

Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.

Total article views: 1,378 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
1,359 0 19 1,378 0 0
  • HTML: 1,359
  • PDF: 0
  • XML: 19
  • Total: 1,378
  • BibTeX: 0
  • EndNote: 0
Views and downloads (calculated since 05 Feb 2026)
Cumulative views and downloads (calculated since 05 Feb 2026)

Viewed (geographical distribution)

Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.

Total article views: 1,397 (including HTML, PDF, and XML) Thereof 1,397 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 18 Nov 2026
Download

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Short summary
Global floods pose serious risks, but existing models are too slow for large-scale prediction. We redesigned the CaMa-Flood model for GPUs by reformulating key components such as irregular river networks, flux updates, and floodplain dynamics into highly parallel, GPU-native algorithms. The resulting CaMa-Flood-GPU runs global simulations in hours instead of days with the same accuracy, enabling larger ensembles, improved flood-risk analysis, and better preparedness worldwide.
Share