Modern climate projections lack adequate spatial and temporal resolution due to computational constraints. A consequence is inaccurate and imprecise predictions of critical processes such as storms. Hybrid methods that combine physics with machine learning (ML) have introduced a new generation of higher fidelity climate simulators that can sidestep Moore's Law by outsourcing compute-hungry, short, high-resolution simulations to ML emulators. However, this hybrid ML-physics simulation approach requires domain-specific treatment and has been inaccessible to ML experts because of lack of training data and relevant, easy-to-use workflows. We present ClimSim, the largest-ever dataset designed for hybrid ML-physics research. It comprises multi-scale climate simulations, developed by a consortium of climate scientists and ML researchers. It consists of 5.7 billion pairs of multivariate input and output vectors that isolate the influence of locally-nested, high-resolution, high-fidelity physics on a host climate simulator's macro-scale physical state.The dataset is global in coverage, spans multiple years at high sampling frequency, and is designed such that resulting emulators are compatible with downstream coupling into operational climate simulators. We implement a range of deterministic and stochastic regression baselines to highlight the ML challenges and their scoring. The data (https://huggingface.co/datasets/LEAP/ClimSim_high-res) and code (https://leap-stc.github.io/ClimSim) are released openly to support the development of hybrid ML-physics and high-fidelity climate simulations for the benefit of science and society.

The Clouds and the Earth's Radiant Energy System (CERES) is part of NASA's Earth Observing System (EOS), CERES objectives include the following. (1) For climate change analysis, provide a continuation of the Earth Radiation Budget Experiment (ERBE) record of radiative fluxes at the top-of-the-atmosphere (TOA), analyzed using the same techniques as the existing ERBE data. (2) Double the accuracy of estimates of radiative fluxes at TOA and the Earth's surface. (3) Provide the first long-term global estimates of the radiative fluxes within the Earth's atmosphere. (4) Provide cloud property estimates collocated in space and time that are consistent with the radiative fluxes from surface to TOA. In order to accomplish these goals, CERES uses data from a combination of spaceborne instruments: CERES scanners, which are an improved version of the ERBE broadband radiometers, and collocated cloud spectral imager data on the same spacecraft. The CERES cloud and radiative flux data products should prove extremely useful in advancing the understanding of cloud-radiation interactions, particularly cloud feedback effects on the Earth's radiation balance. For this reason, the CERES data should be fundamental to the ability to understand, detect, and predict global climate change. CERES results should also be very useful for studying regional climate changes associated with deforestation, desertification, anthropogenic aerosols, and ENSO events. This overview summarizes the Release 3 version of the planned CERES data products and data analysis algorithms. These algorithms are a prototype for the system that will produce the scientific data required for studying the role of clouds and radiation in the Earth's climate system.