冯俊鑫, 陈国坤, 左丽君, 温庆可, 赵晶晶, 王亦文. 基于GF-6 WFV影像和CSLE模型的山区耕地侵蚀定量评价及特征分析[J]. 农业工程学报, 2022, 38(21): 169-179. DOI: 10.11975/j.issn.1002-6819.2022.21.020
    引用本文: 冯俊鑫, 陈国坤, 左丽君, 温庆可, 赵晶晶, 王亦文. 基于GF-6 WFV影像和CSLE模型的山区耕地侵蚀定量评价及特征分析[J]. 农业工程学报, 2022, 38(21): 169-179. DOI: 10.11975/j.issn.1002-6819.2022.21.020
    Feng Junxin, Chen Guokun, Zuo Lijun, Wen Qingke, Zhao Jingjing, Wang Yiwen. Quantitative evaluation and characteristic analysis of cultivated land erosion in mountain area using GF-6 WFV and CSLE model[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2022, 38(21): 169-179. DOI: 10.11975/j.issn.1002-6819.2022.21.020
    Citation: Feng Junxin, Chen Guokun, Zuo Lijun, Wen Qingke, Zhao Jingjing, Wang Yiwen. Quantitative evaluation and characteristic analysis of cultivated land erosion in mountain area using GF-6 WFV and CSLE model[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2022, 38(21): 169-179. DOI: 10.11975/j.issn.1002-6819.2022.21.020

    基于GF-6 WFV影像和CSLE模型的山区耕地侵蚀定量评价及特征分析

    Quantitative evaluation and characteristic analysis of cultivated land erosion in mountain area using GF-6 WFV and CSLE model

    • 摘要: 坡耕地是高原山区水土流失最主要的策源地,其严重的土壤侵蚀已经威胁到山区粮食安全和农业可持续发展。该研究以云南省土壤侵蚀较为严重的文山州为例,在第三次全国国土调查数据、水利普查抽样单元和GF-6 WFV等多源遥感数据基础上,结合随机森林算法快速自动提取梯田,利用CSLE模型对山区耕地土壤侵蚀进行精细定量评估,并以空间自相关分析揭示土壤侵蚀空间分布特征。结果表明:1)文山州耕地空间异质性显著,以广南、西畴、马关一线为界,呈自西向东递减格局;坡耕地占耕地总面积的79.27%,3 929.57 km2耕地采取水土保持工程措施,占49.79%。2)全区耕地平均侵蚀模数为5 180.72 t/(km²•a),其中69.99%的耕地遭受轻度及以上侵蚀;耕地总侵蚀量达4 081.50万t,其中未采取保护措施耕地的侵蚀贡献率为79.40%,土壤侵蚀对坡度和高程的响应存在差异,具有明显垂直地带性特征,1 000~1 800 m高程带侵蚀情况严峻,3 073.65万t侵蚀量源于20°以上陡坡,占侵蚀总量的75.30%。3)侵蚀空间聚集特征鲜明,文山州南部和东部为侵蚀高值中心,低值中心见于中西部和北部。利用区域侵蚀因子自动提取产品和空间自相关分析能够快速完成侵蚀精细定量评估,可为区域水土保持资源精准分配和生态环境建设提供参考。

       

      Abstract: Slope cropland erosion is the most serious land degradation problem in plateau mountainous areas. A great threat has been posed to regional food security and sustainability of agriculture. Particularly, the Wenshan Prefecture of Yunnan Province is characterized by serious soil erosion in China. Taking the Wenshan Prefecture as an example, this study aims to quantitatively evaluate the soil erosion of the farmland and spatial features using multi-source remote sensing data, such as the Third National Land Survey, Water Conservancy Sampling Survey, and the Wide Field of View (WFV) imaging system of GaoFen-6 (GF-6) satellite. Firstly, the non-cultivated land was masked in the GF-6 image with the Third National Land Survey data. The sampling units were also used as the sample set. Secondly, object-oriented multi-scale segmentation was performed on the GF-6 image, where the Jeffries-Matusita distance was adopted to determine the preferred feature set of spectral, texture and shape. Thirdly, Random Forest (RF) algorithm was employed to quickly and automatically extract the terraces. The fine rasterization of the E factor was realized to combine the engineering measures in the sampling units. Finally, the Chinese Soil Loss Equation (CSLE) model was utilized to conduct a fine quantitative assessment of soil erosion in the mountainous arable land. A spatial autocorrelation analysis was also carried out on the soil erosion features and influence factors in the GeoDa software in order to reveal the spatial distribution pattern. The results indicated that: 1) The distribution of arable land in the Wenshan Prefecture was bounded by the Guangnan, Xichou, and Maguan, indicating a descending pattern from the west to the east. There was about 6 252.85 km2 of slope cropland in the study area, accounting for 79.27% of the total arable land. The cultivated land was 3 929.57 km2 with the soil and water conservation engineering measures, accounting for 49.79% of the total arable land area. 2) Soil erosion modulus of the cultivated land ranged from 0 to 118 540.97 t/(km2•a) with an average erosion modulus of 5 180.72 t/(km2•a). About 69.99% of the total cultivated land was suffering from serious erosion. The total annual amount of soil loss from the cultivated land was 40.815 million tons, 20.60% of which was eroded from the land with the soil conservative measures. By contrast, the soil loss from the cultivated land without any soil conservative measures was 32.407 million tons, accounting for 79.40% of the total. There was also a significant difference in the response of erosion to the slope and elevation, particularly with the vertical zonal characteristics. The area and intensity of soil erosion increased first and then decreased, as the elevation and slope increased. The high incidence areas of erosion were found between the elevation zones of 1 000-1 400 m and 1 400-1 800 m, accounting for 27.23% and 50.54% respectively. The steep slope cropland (>20°) erosion also contributed 75.30% of the total erosion. 3) The spatial distribution of cropland erosion demonstrated outstanding aggregation characteristics and strong spatial correlations with the R and LS factors, respectively. Specifically, the Maguan, Malipo, Xichou, and Funing in the south of Wenshan were the high-value centers of erosion, and, the low-value centers were found in the Wenshan, Yanshan, Qiubei, and Guangnan in the central and western regions. The fine quantitative assessment of soil erosion was achieved to automatically extract the soil erosion factors on the regional scale for the spatial autocorrelation analysis. The finding can provide a strong reference to accurately allocate the regional soil and water conservation resources, as well as the ecological environment construction.

       

    /

    返回文章
    返回