相关数据集
LEAP/ChaosBench
ChaosBench是一个用于改进和扩展深度天气模拟器在次季节到季节(S2S)范围内预测能力的基准项目。数据集包含ERA5、LRA5和ORAS5再分析数据,涵盖了从1979年到2023年的45年数据,用于全耦合地球系统模拟(大气-陆地-海冰)。数据集还包括来自欧洲、英国、美国和亚洲主要国家机构的物理基础预测数据,并引入了两种可微分的物理基础指标,以最小化长时间预测范围内的功率谱衰减(模糊度)。此外
Hugging Face2025-06-09 更新3550
LEAP/ClimSim_low-res
--- license: cc-by-4.0 --- Corresponding GitHub repo can be found here: https://github.com/leap-stc/ClimSim Read more: https://arxiv.org/abs/2306.08754.
Hugging Face2023-09-29 更新110
LEAP/ClimSim_high-res
--- license: cc-by-4.0 --- The corresponding GitHub repo can be found here:https://github.com/leap-stc/ClimSim Read more: https://arxiv.org/abs/2306.08754.
Hugging Face2023-09-29 更新80
LEAP/ClimSim_low-res-expanded
这是ClimSim_low-res数据集的扩展版本。每个.mlexpand.文件包含了与ClimSim_low-res中对应的.mli.文件相同的变量,并且还包括了额外的变量,如动态强迫、对流记忆、纬度的cos/sin值。
Hugging Face2024-07-21 更新120
LEAP/subsampled_low_res
Inputs and targets in this dataset are pre-normalized and scaled with .nc files found on the GitHub repo: https://github.com/leap-stc/ClimSim/tree/main/preprocessing/normalizations Read more: https:
Hugging Face2023-10-09 更新80



