Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Guangdong Province Key Laboratory for Climate Change and Natural Disaster Studies, School of Atmospheric Sciences, Sun Yat-sen University, Guangzhou, China.
College of Global Change and Earth System Science, Beijing Normal University, Beijing, China.
Sci Data. 2020 Jan 3;7(1):2. doi: 10.1038/s41597-019-0345-6.
Depth to bedrock influences or controls many of the Earth's physical and chemical processes. It plays important roles in soil science, geology, hydrology, land surface processes, civil engineering, and other related fields. However, information about depth to bedrock in China is very deficient, and there is no independent map of depth to bedrock in China currently. This paper describes the materials and methods to produce high-resolution (100 m) depth-to-bedrock maps of China. For different research and application needs, two sets of data are provided for users. One is the prediction by the ensemble of the random forests and gradient boosting tree models, and the other is the prediction and the uncertainty of prediction based on quantile regression forests model. In comparison with depth-to-bedrock maps of China extracted from previous global predictions, our predictions showed higher accuracy and more spatial details. These data sets can provide more accurate information for Earth system research compared with previous depth-to-bedrock maps.
基岩埋深影响或控制着地球的许多物理和化学过程。它在土壤科学、地质学、水文学、陆面过程、土木工程和其他相关领域发挥着重要作用。然而，中国基岩埋深的信息非常匮乏，目前中国没有独立的基岩埋深图。本文介绍了制作中国高分辨率（100m）基岩埋深图的材料和方法。针对不同的研究和应用需求，为用户提供了两套数据。一套是基于随机森林和梯度提升树模型集成的预测结果，另一套是基于分位数回归森林模型的预测结果及其不确定性。与从以前的全球预测中提取的中国基岩埋深图相比，我们的预测结果具有更高的准确性和更丰富的空间细节。与以前的基岩埋深图相比，这些数据集可为地球系统研究提供更准确的信息。
