High-Resolution WRF Wind Field Dataset for GHG Applications: Navigating Computational Challenges in Energy Research

Abstract

This study presents a high-resolution Weather Research and Forecasting (WRF) simulation dataset focused on generating a comprehensive 3D wind field dataset for high-resolution energy applications, specifically targeting the Permian Basin, a prominent oil and gas region with significant greenhouse gas (GHG) emissions. Winds play a crucial role in the dispersion and transport of greenhouse gas (GHG) emissions, and high resolution and accurate 3D dispersion models are still being developed for source attribution. Wind-induced turbulence enhances atmospheric mixing, resulting in a vertical distribution of GHGs in the atmosphere enabling a better observing sensing system design. Employing a nested simulation approach with grid resolutions of 900m, 300m, and 100m, we model 3D wind fields over a 3-month period from September to November 2019. To enhance realism, the High-Resolution Rapid Refresh (HRRR) model provides boundary conditions for the outermost domain, ensuring smooth interactions between the nested WRF domains. The simulations were performed on the a cluster with advanced computing resources, including infiniband inter-connects, and Volta GPUs. The handling of vast data at high resolutions required efficient data I/O and storage mechanisms. Carefully selected physics parameterizations, such as WRF Single-Moment 6-Class microphysics, RRTM longwave radiation, Dudhia shortwave radiation, Noah land surface model, and Yonsei University planetary boundary layer scheme, were incorporated to ensure accurate modeling of atmospheric processes. In conclusion, this research highlights the significance of advanced computational resources and physics parameterizations in creating a high-resolution WRF simulation dataset for wind fields over the Permian Basin. The understanding of atmospheric dynamics and wind transport processes is crucial for GHG emissions assessment and renewable energy forecasting in this energy-rich region. Navigating computational difficulties is vital for enabling high-resolution simulations that contribute to environmental monitoring and sustainable energy solutions.