Global Cropland Land Surface Temperature dataset(2017-2022)

The LST dataset contains the surface temperature data of cultivated land worldwide during the period from 2003 to 2022. The unit is in degrees Celsius, with a time resolution of daily and a spatial resolution of 1000 meters.It is produced by combing Fengyun and MODIS daily data, and meteorological station data to reconstruct real LST under cloud coverage in daily LST images, and then a regression analysis model is constructed to further improve accuracy in six natural subregions with different climatic conditions. The accuracy analysis shows that the reconstruction result is closely correlated with the in-situ measurements, with an average MAE of 1-2 K.LST数据集包含2003年至2022年期间全球耕地的地表温度数据。其数据单位为摄氏度，时间分辨率为逐日尺度，空间分辨率为1000米。该数据集通过融合风云（Fengyun）卫星逐日数据、中分辨率成像光谱仪（MODIS）逐日数据以及气象站点观测数据，对每日LST影像中的云覆盖区域进行真实地表温度重构；随后针对六个气候条件各异的自然子区域构建回归分析模型，以进一步提升数据精度。精度分析结果表明，重构结果与原位实测数据高度相关，平均绝对误差（Mean Absolute Error，MAE）为1-2开尔文。