Hi All,
Sorry for the late notice, I’ve been very busy working on this. Tomorrow (5/23) at 1000 ET / 1400 UTC I’ll give a public seminar at my workplace (Rosenstiel School at the University of Miami). It is not as widely advertised that this seminar is also a part of my interview for a tenure-track faculty position here. I’m posting this just in case anybody’s interested and got nothing better to do. Later in the talk I will pitch Fortran for machine learning in Earth Sciences and also have a slide promoting Fortran-lang and LFortran. For anybody interested, here’s the Zoom link:
A recording will be available on YouTube for replay.
Title: Advancing Earth System Prediction With Machine Learning and
State-of-the-Art Measurements
Abstract: Weather, ocean wave, and ocean circulation models have improved tremendously over the past few decades, both in terms of prediction skills and the physical processes that they resolve. They provide crucial information that reduces risk to human life and property, while also reducing costs in trillion-dollar industries that rely on weather and ocean forecasts. Historically, atmosphere and ocean models have been developed in isolation from one another. Although we have made significant strides toward unified, fully-coupled, Earth System modeling capability over the past 15 years, many coupled physical processes remain insufficiently observed, poorly understood, or simply missing from operational Earth System models. Recent advances in machine learning promise to become an essential component of model physics and data analysis workflows alike. In this talk, I will review our progress toward comprehensive Earth System modeling capability and recent laboratory and field observations that are necessary to better understand and implement Earth System model components. I will also discuss recent advances in deep learning methods to emulate and significantly accelerate components of Earth System models. Finally, I will emphasize the importance of integrating theory, observations, numerical modeling, and machine learning for the Rosenstiel School to remain at the forefront of Earth System science.
