Speaker

Gabriele Accarino

Gabriele Accarino (Columbia University) - ga2673@columbia.edu

Gabriele Accarino received his Master’s degree in Computer Engineering with honors from the University of Salento (Italy), where he presented a thesis on the application of Long Short-Term Memory (LSTM) neural networks to forecast sea level rise across different locations in the Mediterranean basin. In November 2018, he began a PhD in Environmental Sciences at the same university, focusing on the intersection of machine learning and climate science. His doctoral research explored data-driven methods for a range of applications in the weather and climate, including spatial downscaling and the detection of extreme events, with a particular emphasis on wildfires and tropical cyclones. He got also interested in the potential relationship between environmental factors and the COVID-19 spread during the pandemic in 2019 and the link of climate change and armed conflicts. Following his PhD, Gabriele was awarded a research fellowship at the Department of Engineering for Innovation at the University of Salento, where he worked on the design and development of data-driven ocean emulators using Transformer architectures. During this time, he also served as an adjunct professor, teaching courses in Information Processing Systems and High-Performance Computing.

He later joined the CMCC Foundation as a Junior Scientist in the Advanced Scientific Computing division (now ADIC), where he led the machine learning research unit. At CMCC, he contributed to and co-supervised multiple European research projects and served as a Scientific Officer for the Silvanus project.

At the end of 2024, Gabriele joined Columbia University in New York and the Learning the Earth with Artificial Intelligence and Physics (LEAP) Science and Technology Center (STC) as a Postdoctoral Research Scientist. His current research focuses on the design and development of multi-scale similarity metrics, spatio-temporal verification techniques for climate fields, and benchmarking of data-driven and physics-based models. He has been a CMCC affiliate ever since.

More Information:

• LinkedIn

Talks at this conference:

 Overview