Global map of clay minerals in terrestrial soils

We developed a comprehensive dataset of major soil clay minerals covering the global land surface for both topsoil (near-surface) and subsoil at different spatial resolutions. This dataset is intended for application in a variety of earth science fields that require interdisciplinary data. We gathered observational data on clay minerals through a literature survey and meta-analysis. Most observations were originally obtained by x-ray diffraction (XRD) analysis. The multitude of clay minerals that occur in soils were classified into ten groups: chlorite, gibbsite, kaolinite, mica-illite, smectite, quartz, vermiculite, non-crystalline (amorphous and short-range order minerals), iron oxide, and others. We then aggregated the clay mineral composition data on the basis of 12 soil orders. Using a global map of soil orders and additional soil datasets, we developed global maps of clay mineral abundances in topsoil and subsoil at a resolution of 2° grid cells (about 3.7 km) and, by averaging, at lower spatial resolutions (e.g., 1° grid cells). We examined uncertainties in the dataset by statistical (i.e., Monte Carlo) methods and by comparison with previous datasets. The new dataset will facilitate continental-scale studies of biogeochemistry and climatology by providing more precise soil properties related to, for example, soil adsorption and dust emission. The dataset should also find application in interdisciplinary studies in fields such as hydrology and agronomy, both as input data for model simulations and for the interpretation of observational data.

本研究构建了一套覆盖全球陆地表面的表层土壤（近地表层）与深层土壤的主要土壤黏土矿物综合数据集，涵盖多种空间分辨率。该数据集旨在服务于各类需跨学科数据支撑的地球科学领域研究。研究团队通过文献调研与荟萃分析收集黏土矿物观测数据，其中绝大多数观测结果最初通过X射线衍射（X-ray diffraction, XRD）分析获取。土壤中产出的各类黏土矿物被划分为十大类别：绿泥石（chlorite）、三水铝石（gibbsite）、高岭石（kaolinite）、云母-伊利石（mica-illite）、蒙脱石（smectite）、石英（quartz）、蛭石（vermiculite）、非晶质矿物（non-crystalline，涵盖非晶态与短程有序矿物）、氧化铁（iron oxide）及其他类别。随后，研究团队基于12个土壤纲（soil orders）对黏土矿物组成数据进行聚合整合。借助全球土壤纲分布图与配套土壤数据集，研究团队构建了表层与深层土壤黏土矿物丰度的全球分布图，其空间分辨率为2°网格单元（约3.7千米）；同时通过平均处理生成更低空间分辨率的分布图（如1°网格单元）。研究团队采用统计学方法（即蒙特卡洛（Monte Carlo）模拟）及与已有数据集的对比分析，对本数据集的不确定性开展了评估。本数据集可提供与土壤吸附、扬尘排放等相关的更为精准的土壤属性，能够助力大陆尺度的生物地球化学与气候学研究。此外，该数据集还可应用于水文学、农学等领域的跨学科研究，既可作为模型模拟的输入数据，也可用于观测数据的解译分析。