A global historical twice-daily (daytime and nighttime) land surface temperature dataset produced by AVHRR observations from 1981 to 2021 (1981–2000)

Land surface temperature (LST) is a key variable for monitoring and evaluating global long-term climate change. However, existing satellite-based twice-daily LST products only date back to 2000, which makes it difficult to obtain robust long-term temperature variations. We developed the first global historical twice-daily LST dataset (GT-LST), with a spatial resolution of 0.05°, using Advanced Very High Resolution Radiometer Level-1b Global Area Coverage data from 1981 to 2021. Validation with in situ measurements from Surface Radiation Budget sites showed that the overall root-mean-square errors of GT-LST varied from 2.0 K to 3.9 K. Inter-comparison with a common LST product (i.e., MYD11A1) revealed that the overall root-mean-square-difference was approximately 3.2 K. More details of this dataset can be seen in <em>readme.pdf.</em> This dataset provides GT-LST product from 1981 to 2000.地表温度（Land Surface Temperature，LST）是监测与评估全球长期气候变化的关键变量。然而，当前已有的基于卫星的每日两次LST产品仅可追溯至2000年，这使得获取稳健的长期温度变化序列颇具挑战。本研究构建了首套全球历史逐日两次LST数据集（GT-LST），空间分辨率为0.05°，所用数据源为1981年至2021年的高级甚高分辨率辐射计（Advanced Very High Resolution Radiometer）1B级全球区域覆盖数据。利用地表辐射预算（Surface Radiation Budget）站点的原位观测数据对该数据集进行验证，结果显示GT-LST的总体均方根误差介于2.0 K至3.9 K之间。与主流LST产品MYD11A1进行交叉比对的结果表明，该数据集的总体均方根偏差约为3.2 K。有关该数据集的更多详细信息，请参阅<em>readme.pdf</em>。本数据集提供1981年至2000年的GT-LST产品。