吉林省耕地土壤有机质数据集

使用遥感技术进行土壤有机质（SOM）监测对于现代土地资源管理和环境保护至关重要。然而在土壤类型复杂、环境变量差异明显的盐碱-黑土交错区绘制 SOM 仍然具有挑战性。本研究整合先验知识，将吉林省分为盐碱区和黑土区，获取2019年至2023年吉林省裸土时期（4月至7月）的所有Sentinel-2影像。将 Sentinel-2 图像分为三个时间窗口（DOY 90-120、DOY 120-150 和 DOY 150-180），系统地评估了时间窗口、光谱指数（盐分指数、植被水分指数）、环境变量（地形和气候）以及局部回归在盐碱-黑土交错区的SOM含量制图潜力。结果表明：（1）盐碱区与黑土区SOM制图的最优时间窗口均为DOY 90-120；（2）加入盐度指数可提高盐碱区 SOM 制图精度，但会降低黑土区 SOM 制图精度，而植被水分指数可提高这两个地区的精度；（3）加入环境变量可以提高所有地区的SOM制图精度，其中黑土区地形变量相对重要，盐碱区气候变量相对重要；（4）基于盐碱区和黑土区分区的局部回归在SOM制图精度上优于整体回归，但其不确定性更高。研究表明将先验知识与多时相遥感影像的融合，可以显著提高盐碱-黑土交错区SOM制图精度，进而为不同土壤类型区域的精准管理和保护提供科学依据。

Soil Organic Matter (SOM) monitoring using remote sensing technology is critical for modern land resource management and environmental protection. However, mapping SOM in the saline-alkali and black soil ecotone, where soil types are complex and environmental variables differ significantly, remains challenging. This study integrates prior knowledge by dividing Jilin Province into saline-alkali and black soil regions, and acquires all Sentinel-2 images captured during the bare soil period (April to July) from 2019 to 2023 in Jilin Province. The Sentinel-2 images are divided into three time windows (DOY 90-120, DOY 120-150, and DOY 150-180), and the mapping potential of time windows, spectral indices (salinity index, vegetation water index), environmental covariates (topography and climate), and local regression for SOM content in the saline-alkali and black soil ecotone is systematically evaluated. The results demonstrate that: (1) The optimal time window for SOM mapping in both saline-alkali and black soil regions is DOY 90-120; (2) Incorporating the salinity index can improve the SOM mapping accuracy in the saline-alkali area but reduce that in the black soil area, while the vegetation water index enhances the accuracy of both regions; (3) Adding environmental covariates can elevate the SOM mapping accuracy of all regions, with topographic variables being relatively more important in the black soil area and climatic variables relatively more important in the saline-alkali area; (4) Local regression based on the partition of saline-alkali and black soil regions outperforms global regression in SOM mapping accuracy, yet carries higher uncertainty. This study indicates that integrating prior knowledge with multi-temporal remote sensing images can significantly improve the accuracy of SOM mapping in the saline-alkali and black soil ecotone, thereby providing a scientific basis for precise management and protection of regions with different soil types.