{"url":"/dataset/chaosbench","name":"ChaosBench","full_name":null,"description_markdown":"We propose ChaosBench, a large-scale, multi-channel, physics-based benchmark for subseasonal-to-seasonal (S2S) climate prediction. \r\nIt is framed as a high-dimensional video regression task that consists of 45-year, 60-channel observations \r\nfor validating physics-based and data-driven models, and training the latter. \r\nPhysics-based forecasts are generated from 4 national weather agencies with 44-day lead-time and serve as baselines to data-driven forecasts. \r\nOur benchmark is one of the first to incorporate physics-based metrics to ensure physically-consistent and explainable models. \r\nWe establish two tasks: full and sparse dynamics prediction. \r\n\r\n🔗: [https://leap-stc.github.io/ChaosBench/](https://leap-stc.github.io/ChaosBench/)\r\n\r\n📚: [https://arxiv.org/abs/2402.00712](https://arxiv.org/abs/2402.00712)\r\n\r\n## Getting Started\r\n**Step 1**: Clone the [ChaosBench](https://github.com/leap-stc/ChaosBench) Github repository\r\n\r\n**Step 2**: Install package dependencies\r\n```\r\ncd ChaosBench\r\npip install -r requirements.txt\r\n```\r\n\r\n**Step 3**: Initialize the data space by running\r\n```\r\ncd data/\r\nwget https://huggingface.co/datasets/LEAP/ChaosBench/resolve/main/process.sh\r\nchmod +x process.sh\r\n```\r\n**Step 5**: Download the data \r\n\r\n```\r\n# NOTE: you can also run each line one at a time to retrieve individual dataset\r\n\r\n./process.sh era5            # Required: For input ERA5 data\r\n./process.sh climatology     # Required: For climatology\r\n./process.sh ukmo            # Optional: For simulation from UKMO\r\n./process.sh ncep            # Optional: For simulation from NCEP\r\n./process.sh cma             # Optional: For simulation from CMA\r\n./process.sh ecmwf           # Optional: For simulation from ECMWF\r\n```","description_withheld":null,"homepage":"https://huggingface.co/datasets/LEAP/ChaosBench/","introduced_date":"2024-02-01","introduced_date_note":null,"introduced_by":{"paper":"/paper/chaosbench-a-multi-channel-physics-based","title":"ChaosBench: A Multi-Channel, Physics-Based Benchmark for Subseasonal-to-Seasonal Climate Prediction","first_author":"Juan Nathaniel","url":null},"license":null,"modalities":[],"tasks":[],"languages":[],"variants":["ChaosBench"],"data_loaders":[],"num_papers_in_archive":1,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[],"papers_with_a_benchmark_row":[],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":0,"samples_harvested":0,"samples_ran":0,"samples_unverified":0,"pointer_only_for_licence":0,"papers_with_no_sample_that_ran":0,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}