Key Laboratory of Agricultural Remote Sensing, Ministry of Agriculture/CAAS-CIAT Joint Laboratory in Advanced Technologies for Sustainable Agriculture-Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences, Beijing 100081, China; Inner Mongolia Academy of Agricultural & Animal Husbandry Science, Institute of Rural Economic and Information, Hohhot 010030, China.
Key Laboratory of Agricultural Remote Sensing, Ministry of Agriculture/CAAS-CIAT Joint Laboratory in Advanced Technologies for Sustainable Agriculture-Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences, Beijing 100081, China.
Sci Total Environ. 2022 Jan 20;805:150286. doi: 10.1016/j.scitotenv.2021.150286. Epub 2021 Sep 11.
Due to the unfavorable soil conditions and water resources, the cropland use pattern in the farming-pastoral ecotone in northern China is complex. The program named "Grain for Green" has accelerated the cropland change. However, the complex cropland and retired cropland are challenging to monitor with remote sensing due to their spatially dispersed and easily confused with spectrally similar land use classes such as nature grasslands and non-cropped fields. Taking farming-pastoral ecotone in the northern foot of the Yinshan Mountains as a case study, we explored a classification approach for complex cropland and retired cropland, which was introduced as a specific land use class by using multi-temporal Landsat TM and OLI images with Google Earth Engine. During 1990-2000, cropland increased with a sharper growth and increased with a slower growth from 2001 to 2010, and then decreased significantly from 2011 to 2019, to lead the cropland area in 2019 was smaller than an area in 1990. We analyzed the spatiotemporal trajectories of retired cropland in 2019 using the Land Use Change Trajectory method to evaluate its source. In our finding, approximately 77% of retired cropland was labelled as cropland before 2019; albeit, not all retired cropland was converted from cropland. Moreover, we qualitatively assessed the vegetation dynamics in the study area by utilizing the long-term NDVI-mean value to reveal that vegetation coverage has shown a continuously increasing trend. It is related to the decline of cropland and the increase of retired cropland at the same rate. Our results highlighted that the "Grain for Green" program had led the vegetation restoration in the farming-pastoral ecotone. Our approach for monitoring cropland and retired cropland can improve the understanding of the driving factors and consequences of these critical land use change trajectories.
由于土壤条件和水资源不利，中国北方农牧交错带的耕地利用格局复杂。“退耕还林还草”工程加速了耕地变化。然而，由于复杂耕地和休耕地在空间上分散，并且容易与光谱相似的土地利用类型（如自然草地和无耕地）混淆，因此遥感监测具有挑战性。以阴山北麓农牧交错带为例，我们探索了一种利用多时相 Landsat TM 和 OLI 图像和 Google Earth Engine 引入特定土地利用类别的复杂耕地和休耕地分类方法。在 1990-2000 年期间，耕地增长较快，2001-2010 年增长较慢，2011-2019 年显著减少，导致 2019 年的耕地面积小于 1990 年。我们使用土地利用变化轨迹方法分析了 2019 年休耕地的时空轨迹，以评估其来源。在我们的发现中，大约 77%的休耕地在 2019 年之前被标记为耕地；尽管如此，并非所有休耕地都是从耕地转化而来的。此外，我们利用长期 NDVI-平均值定性评估了研究区的植被动态，结果表明植被覆盖度呈持续增加趋势。这与同期耕地和休耕地的减少有关。我们的结果强调了“退耕还林还草”工程导致了农牧交错带的植被恢复。我们监测耕地和休耕地的方法可以提高对这些关键土地利用变化轨迹的驱动因素和后果的理解。
