Note: This talk is available over Zoom.

Title: Data-Driven Advances in Wind Energy: From Observations to Machine Learning

Abstract: The rapid expansion of wind energy—both onshore and offshore—presents unprecedented challenges and opportunities in atmospheric science, engineering, and data science. Understanding the complex interactions between wind farms and the atmosphere is essential for optimizing wind energy production and integrating it effectively into the energy system. In this seminar, I will present results from a multidisciplinary approach that combines field observations, numerical modeling, and machine learning (ML) to advance wind energy research.

Since my October 2024 JHU seminar provided an in-depth look at my research focused on field observations and numerical modeling, in this seminar I will only provide a brief summary of key insights from large-scale field campaigns, including the American WAKE ExperimeNt (AWAKEN) and the third Wind Forecast Improvement Project (WFIP3). I will briefly discuss how these observations inform numerical simulations and international benchmarking efforts, ultimately enhancing our ability to predict wind farm performance under diverse atmospheric conditions. However, as observational datasets continue to grow, there is a pressing need for advanced data-driven approaches.

The core of this seminar will therefore highlight the transformative role of ML in wind energy applications. With increasing access to atmospheric observations, ML offers novel approaches to addressing key challenges. I will start with a brief introduction to ML in Earth sciences, followed by two case studies demonstrating its impact on wind energy. The first investigates the use of random forests for improving wind speed vertical extrapolation to turbine rotor heights, revealing the importance of rigorous cross-validation techniques for realistic performance assessment. The second explores an ML-based approach to estimating modeled offshore wind resource uncertainty, leveraging floating lidar and buoy observations to enhance long-term numerical data reliability.

Bio: Dr. Nicola Bodini is a senior atmospheric scientist at the National Renewable Energy Laboratory (NREL). With a PhD in Atmospheric and Oceanic Sciences from the University of Colorado Boulder, Dr. Bodini specializes in observations of the atmospheric boundary layer and machine learning application to wind energy. He is currently involved in two large field campaigns as he serves as the science lead for the American Wake Experiment (AWAKEN), and he is the NREL Principal Investigator for the Wind Forecast Improvement Project part 3 (WFIP3). He has also recently led the creation of the 2023 National Offshore Wind data set, which is NREL’s state-of-the-art offshore wind resource assessment product.