Atmosphere, Ocean, and Earth Science

CARTHE Drifters Tamay Ozgokmen

Atmosphere, Ocean, and Earth Science

Machine learning and big data analytics are ideally suited for the classification and detection of various properties of physical systems (e.g., atmosphere, ocean, and the solid earth) with extraordinary precision in the presence of an enormous amount of disparate data. They are also highly effective tools for developing models and making predictions where data are sparse, and the uncertainty is high. For example, machine learning techniques are currently being developed to detect rare but extremely dangerous events such as tsunamis, submarine earthquakes, and volcanoes. The pairing of machine learning with Bayesian statistics combines multiple prediction tools that can provide more reliable forecasts and a detailed understanding of the uncertainties inherent in forecasting in general. As such, these tools can help solve the problem of estimating current states in physical systems where it is difficult to conduct in situ observations, such as in the deep ocean and polar ice caps, and assisting in accurate weather, climate, and solid earth predictions.

Ben Kirtman, PhD Deputy Director, Earth Systems

Ben Kirtman, PhD
Deputy Director, Institute for Data Science and Computing
Director, IDSC Atmosphere, Ocean, and Earth Science

In addition to his role at IDSC, Ben Kirtman is Professor in the Department of Atmospheric Sciences and  Director, NOAA Cooperative Institute for Marine and Atmospheric Studies. Professor Benjamin Kirtman received his BS in Applied Mathematics from the University of California-San Diego in 1987, and his MS and PhD in 1992 from the University of Maryland–College Park. From 1993-2002, Dr. Kirtman was a research scientist with the Center for Ocean-Land-Atmosphere Studies, and in 2002, joined the faculty of George Mason University as a tenured Associate Professor. In 2007, Dr. Kirtman moved to the University of Miami’s Rosenstiel School for Marine and Atmospheric Science as a full professor of meteorology and physical oceanography and Program Director for Climate & Environmental Hazards at the Center for Computational Science (now IDSC). He currently serves as the Deputy Director of the Institute for Data Science and Computing, and as Director for IDSC Data-Driven Discovery in Atmosphere, Ocean, and Earth Science. Dr. Kirtman is also a 2017-18 recipient of the UM Provost’s Award for Scholarly Activity and was awarded the Department of Atmospheric Sciences undergraduate teaching award in 2016, 2017, and 2018. In 2018, Dr. Kirtman was elected as a Fellow in the American Meteorological Society.

In 2011, he was appointed Associate Dean of Research for the Rosenstiel School. In 2008, Professor Kirtman received the Distinguished Alumnus Award from the Department of Atmospheric and Oceanic Science at the University of Maryland. Dr. Kirtman was also awarded the Stony Brook University School of Marine and Atmospheric Science (SOMAS) Robert D. Cess Distinguished Lecture in Recognition of Outstanding Contributions to Atmospheric Sciences in 2018.

Internationally, Dr. Kirtman has enjoyed a leadership role in the World Climate Research Program (WCRP) seasonal-to-interannual prediction activities. In particular, he has chaired the international CLIVAR Working Group on Seasonal to Interannual Prediction (WGSIP), and the WCRP Task Force for Seasonal Prediction (TFSP). Dr. Kirtman was a coordinating lead author for the Intergovenmental Panel on Climate Change (IPCC) working group one—the Scientific Basis.

Professor Kirtman was an Executive Editor of Climate Dynamics, and is an Associate Editor of the American Geophysical Union  Journal of Geophysical Research and has received numerous research grants from the National Science Foundation (NSF), Department of Energy (DOE), NOAANASA, and the Office of Naval Research, and he leads the North American Multi-Model Ensemble Prediction (NMME) Experiment. Professor Kirtman is the author and/or co-author of over 130 peer reviewed papers focused on understanding and predicting climate variability on time scales from days to decades.