Browsing by Issue Date, starting with "2025-09-05"
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- Containerizing and evaluating the WRF model for cloud-based HPCPublication . Gouveia, Diogo Manuel Sales; Baras, Karolina; Vieira, Rui Ricardo AlmeidaNumerical models are mathematical representations that solve algebraic or differential equations to simulate phenomena across time and space. Some models, such as the Weather Research and Forecasting (WRF) model, are applied to weather prediction, providing valuable information to individuals, companies, institutions, and governments on potential future events. However, the WRF model can become highly computationally demanding, requiring increasingly power ful hardware resources with the increase of the resolutions and complexity of the simulations. High-Performance Computing (HPC) allows users to leverage powerful instances tailored to their needs, providing the necessary hardware resources for running complex WRF simulations. Mean while, cloud computing presents an enticing option for computationally demanding programs, offering advantages such as economies of scale, high reliability, and availability. On the software side, compiling, configuring, and installing multiple libraries are essential to exploit WRF’s po tential, which can be an arduous task. This study explores a cloud-enabled approach using pre configured Docker images and compares it to on-premises, HPC, and cloud solutions. We also analyze whether cloud computing is economically and performance-wise viable, using a compa rable base architecture across multiple cloud providers, and assess if the convenience of Docker meets the requirements despite potential overhead.