全球逐月高空间分辨率CMIP6统计降尺度近地面气温数据集(MoCHAT, 1950-2100)

全球逐月高空间分辨率CMIP6统计降尺度近地面气温数据集 (MoCHAT, 1950-2100) 由NEX-GDDP-CMIP6的16个大气环流模型和高分辨率历史气象数据集WorldClim为输入，采用Delta统计学降尺度方法进行空间降尺度，其空间分辨率为1 km，时间分辨率为1月。该数据集包括16个大气环流模型和其多模式平均的最高气温、最低气温和平均气温，涵盖历史时期（1950-2014）和三个未来排放情景（2015-2100），如SSP1-2.6、SSP2-4.5和SSP5-8.5。通过与全球每日地表气象站观测数据（GSOD）和现有的高分辨率气候数据（CHELSA）进行对比，结果显示这些变量的平均绝对误差范围为1.60到2.38 K，平均偏差低于2.0 K，较原始模式结果有所改善。多模式平均（MME）对所有变量表现最好，具有最低的MAE和RMSE，因此我们建议用户可以优先考虑使用MME数据集。该数据集存储为NetCDF格式，采用World Geodetic System 1984 (WGS84) 地理坐标系，命名规则和其他详细信息请参见“MoCHAT_ReadMe.docx”以及论文”Global monthly CMIP6-downscaled high-resolution (1 km) near-surface air temperature dataset (1950–2100)“。如果您在使用MoCHAT数据集时有任何问题，请直接联系数据集制作者：雷雪雯 ([e-mail])、赵启康 ([e-mail])、孟庆岩 ([e-mail])。

The Global Monthly High Spatial Resolution CMIP6 Statistical Downscaled Near-Surface Air Temperature Dataset (MoCHAT, 1950–2100) was developed using 16 general circulation models (GCMs) from NEX-GDDP-CMIP6 and the high-resolution historical meteorological dataset WorldClim as input data, and adopted the Delta statistical downscaling method for spatial downscaling. Its spatial resolution is 1 km, with a temporal resolution of 1 month. This dataset includes maximum air temperature, minimum air temperature, and mean air temperature from the 16 GCMs as well as their multi-model ensemble (MME) mean, covering the historical period (1950–2014) and three future emission scenarios (2015–2100), namely SSP1-2.6, SSP2-4.5, and SSP5-8.5. Compared with the global daily surface meteorological station observation dataset (GSOD) and existing high-resolution climate dataset (CHELSA), the results show that the mean absolute error (MAE) of these variables ranges from 1.60 to 2.38 K, and the mean bias is below 2.0 K, which outperforms the raw model outputs. The multi-model ensemble (MME) performs best for all variables, with the lowest MAE and root mean square error (RMSE), so we recommend that users prioritize using the MME dataset. This dataset is stored in NetCDF format, using the World Geodetic System 1984 (WGS84) geographic coordinate system. For naming rules and other detailed information, please refer to "MoCHAT_ReadMe.docx" and the paper titled "Global monthly CMIP6-downscaled high-resolution (1 km) near-surface air temperature dataset (1950–2100)". If you have any questions when using the MoCHAT dataset, please contact the dataset creators directly: Xuewen Lei ([e-mail]), Qikang Zhao ([e-mail]), and Qingyan Meng ([e-mail]).

搜集汇总

背景与挑战

背景概述

该数据集是全球逐月高分辨率(1 km)近地面气温数据，涵盖1950-2100年，包括历史时期和三个未来排放情景。通过统计降尺度方法处理16个大气环流模型数据，验证显示其精度较原始模型有所提升，多模式平均(MME)表现最优。