Li Shiyu, Wan Hong, Yu Qun, Wang Xinyuan
College of Information Science and Engineering, Shandong Agricultural University, Tai'an, 271000, China.
Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing, 100094, China.
Sci Rep. 2025 Jan 3;15(1):675. doi: 10.1038/s41598-024-83944-w.
Land Surface Temperature (LST) is widely recognized as a sensitive indicator of climate change, and it plays a significant role in ecological research. The ERA5-Land LST dataset, developed and managed by the European Centre for Medium-Range Weather Forecasts (ECMWF), is extensively used for global or regional LST studies. However, its fine-scale application is limited by its low spatial resolution. Therefore, to improve the spatial resolution of ERA5-Land LST data, this study proposes an Attention Mechanism U-Net (AMUN) method, which combines data acquisition and preprocessing on the Google Earth Engine (GEE) cloud computing platform, to downscale the hourly monthly mean reanalysis LST data of ERA5-Land across China's territory from 0.1° to 0.01°. This method comprehensively considers the relationship between the LST and surface features, organically combining multiple deep learning modules, includes the Global Multi-Factor Cross-Attention (GMFCA) module, the Feature Fusion Residual Dense Block (FFRDB) connection module, and the U-Net module. In addition, the Bayesian global optimization algorithm is used to select the optimal hyperparameters of the network in order to enhance the predictive performance of the model. Finally, the downscaling accuracy of the network was evaluated through simulated data experiments and real data experiments and compared with the Random Forest (RF) method. The results show that the network proposed in this study outperforms the RF method, with RMSE reduced by approximately 32-51%. The downscaling method proposed in this study can effectively improve the accuracy of ERA5-Land LST downscaling, providing new insights for LST downscaling research.
陆地表面温度（LST）被广泛认为是气候变化的敏感指标，并且在生态研究中发挥着重要作用。由欧洲中期天气预报中心（ECMWF）开发和管理的ERA5-Land LST数据集被广泛用于全球或区域LST研究。然而，其精细尺度应用受到低空间分辨率的限制。因此，为了提高ERA5-Land LST数据的空间分辨率，本研究提出了一种注意力机制U-Net（AMUN）方法，该方法在谷歌地球引擎（GEE）云计算平台上结合数据采集和预处理，将中国境内ERA5-Land的每小时月平均再分析LST数据从0.1°降尺度到0.01°。该方法综合考虑了LST与地表特征之间的关系，有机结合了多个深度学习模块，包括全局多因素交叉注意力（GMFCA）模块、特征融合残差密集块（FFRDB）连接模块和U-Net模块。此外，使用贝叶斯全局优化算法来选择网络的最优超参数，以提高模型的预测性能。最后，通过模拟数据实验和真实数据实验评估了网络的降尺度精度，并与随机森林（RF）方法进行了比较。结果表明，本研究提出的网络优于RF方法，均方根误差（RMSE）降低了约32%-51%。本研究提出的降尺度方法能够有效提高ERA5-Land LST降尺度的精度，为LST降尺度研究提供了新的思路。
