Marat Khairoutdinov

Marat Khairoutdinov is an atmospheric scientist who builds cloud-resolving models of the climate system and co-authored ClimSim, a NeurIPS 2023 Outstanding Paper on a dataset for machine-learning climate emulation. He studied at the Moscow Institute of Physics and Technology, worked at the Central Aerological Observatory in Moscow, and earned a meteorology PhD at the University of Oklahoma in 1997. At Colorado State University he developed SAM, the System for Atmospheric Modeling, and with David Randall published preliminary results in 2001 on using a cloud-resolving model as the cloud parameterization in a climate model. He joined Stony Brook University in 2007, and by 2019 his Global SAM had been compared favorably with other global models in the DYAMOND project. He is a professor at Stony Brook's School of Marine and Atmospheric Sciences, where he leads a National Science Foundation project on the climate system with a global storm-resolving model.

Now at: Stony Brook University
Badges: Loyal

Timeline

1987–1992

Cloud models in Moscow

He studies at the Moscow Institute of Physics and Technology in the late 1980s and works at the Central Aerological Observatory in Moscow from 1987 to 1992. There he builds a numerical model of aircraft dry-ice seeding of orographic clouds, using explicit bin microphysics, and his first 3-D cloud-resolving model.
1992–1997

A meteorology PhD at Oklahoma

He arrives at the University of Oklahoma in July 1992 to work with Yefim Kogan and earns a PhD in meteorology in December 1997. His doctoral model is one of the first large-eddy simulation models with explicit bin microphysics, which he applies to drizzling marine stratocumulus clouds. With Kogan he turns its results into a bulk microphysics parameterization, published in 2000.
1998–2007

Building SAM at Colorado State

In January 1998, he joins David Randall's research group in the Department of Atmospheric Science at Colorado State University, where the model code and physics are overhauled. In 2003, the model receives its official name, SAM, for System for Atmospheric Modeling.
2001

Cloud models inside climate models

In 2001, he and David Randall publish preliminary results from using a cloud-resolving model as the cloud parameterization in the NCAR Community Climate System Model. They replace the cloud parameterizations of the Community Atmosphere Model with his high-resolution cloud model, an approach they call super-parameterization. In the resulting Multiscale Modeling Framework, a cloud-resolving model runs in every grid cell of the global model.
2007

A professorship at Stony Brook

In 2007, he accepts a professorship at Stony Brook University's School of Marine and Atmospheric Sciences, where he serves as associate professor until 2016 and as professor from 2017.
By 2019

Global SAM and DYAMOND

By 2019, his Global SAM, a global cloud-resolving model, compares favorably with other global models in the international DYAMOND project. The 40-day simulation uses more than 3 billion grid cells and 4,608 cores of NCAR's Cheyenne supercomputer.
2023

A NeurIPS Outstanding Paper

He is a co-author of ClimSim, a dataset of 5.7 billion pairs of input and output vectors from multi-scale climate simulations, built by a consortium of climate scientists and machine-learning researchers. The paper receives an Outstanding Paper award in the Datasets and Benchmarks Track at NeurIPS 2023.
As of October 2026

Global storm-resolving models

He is a professor of atmospheric science at Stony Brook University's School of Marine and Atmospheric Sciences and a founding faculty member of its Institute for Advanced Computational Science. He is principal investigator on a National Science Foundation project, running from August 2022, on the climate system with a global storm-resolving model. He also explores machine-learning emulation of surface fluxes and cloud microphysics, trained on high-resolution simulations and observational datasets.