鄱阳湖流域90米分辨率土壤砂粒数据集（2010-2018年）

该数据集是鄱阳湖流域土壤砂粒分布数据集，水平空间分辨率为90m；垂直方向包括土层深度：0-5, 5-15, 15-30, 30-60。
该数据集是基于近年“我国土系调查与《中国土系志》编制项目”获得的土壤剖面样点，采用预测性土壤制图范式，利用地理信息与遥感技术对成土环境条件进行精细刻画和空间分析，研发自适应深度函数拟合方法，集成先进的集合式机器学习方法，生成了中国高分辨率系列土壤属性三维数字分布图及其不确定性分布。与现有土壤图和相关土壤数据集相比，显著提高了准确性和精细度，更好地表征了我国土壤属性的空间变异特征，可广泛应用于土壤、农业、生态、水文、气候、环境等多个学科领域，也是对“全球土壤制图科学计划”（GlobalSoilMap.net）的重要贡献。数据文件名称：
GeoTIFF格式，土壤属性名称+土层深度.tif，例如btsnd05.tif表示0-5cm深度土层的土壤砂粒值。

This dataset is a soil sand particle distribution dataset for the Poyang Lake Basin, with a horizontal spatial resolution of 90 meters. In the vertical dimension, it covers four soil layer depths: 0-5 cm, 5-15 cm, 15-30 cm, and 30-60 cm.

This dataset is developed based on soil profile sampling points collected from the recent National Soil Series Survey and Compilation Project of the Soil Series Annals of China. It adopts the predictive soil mapping paradigm, uses geographic information and remote sensing technologies to finely characterize and conduct spatial analysis on soil-forming environmental conditions, develops an adaptive deep function fitting approach, and integrates advanced ensemble machine learning methodologies to generate high-resolution three-dimensional digital distribution maps of Chinese soil series properties and their corresponding uncertainty distributions.

Compared with existing soil maps and related soil datasets, this dataset significantly improves both accuracy and spatial resolution, better characterizes the spatial variability of soil properties across China, and can be widely applied to multiple disciplines including soil science, agriculture, ecology, hydrology, climatology, and environmental science. It also represents an important contribution to the "Global Soil Mapping Science Initiative" (GlobalSoilMap.net).

The data files are stored in GeoTIFF format, following the naming convention of "soil property name + soil layer depth.tif". For example, btsnd05.tif indicates the soil sand particle content of the 0-5 cm soil layer.

搜集汇总

背景与挑战

背景概述

该数据集提供了2010-2018年鄱阳湖流域土壤砂粒的空间分布数据，水平分辨率为90米，覆盖0-5、5-15、15-30和30-60厘米四个土层深度。它基于土壤剖面调查，采用先进的预测性制图和机器学习方法生成，显著提高了土壤属性空间变异的表征精度，适用于土壤、农业、生态和水文等多学科研究。