引用本文:
【打印本页】   【HTML】   【下载PDF全文】   查看/发表评论  【EndNote】   【RefMan】   【BibTex】
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 58次   下载 37 本文二维码信息
码上扫一扫!
分享到: 微信 更多
沂蒙山区不同抽样密度对土壤侵蚀因子估算精度的影响
齐斐1, 苏新宇1, 黎家作2, 胡续礼3, 刘霞1, 张春强2, 邢先双4, 丁鸣鸣5
1.江苏省水土保持与生态修复重点实验室/南方现代林业协同创新中心/南京林业大学 林学院, 江苏 南京 210037;2.淮河水利委员会淮河流域水土保持监测中心站, 安徽 蚌埠 233001;3.水利部淮河水利 委员会水土保持处, 安徽 蚌埠 233001;4.山东省水文局, 山东 济南 250012;5.南京市水务局, 江苏 南京 210036
摘要:
[目的]探讨分层系统抽样方法下不同抽样密度对土壤侵蚀因子估算精度的影响,为区域水土流失动态监测抽样方法和抽样密度的选取提供数据支撑。[方法]以沂蒙山泰山国家级重点治理区蒙阴县为对象,基于2013年SPOT5遥感影像和1:1万地形图,采用人机交互解译、野外调查、统计分析等方法,以全县土壤侵蚀因子为基准值,对1%和4%密度土壤侵蚀因子进行精度评价。[结果]①1%抽样密度下,S,E,K因子相对误差较大,分别为33.48%,23.46%,20.64%,主要受坡度、土地利用和土壤类型影响;L,B,T相对误差均小于11%;六者平均14.44%。②4%抽样密度下,E,K,B相对误差较大,分别为15.07%,13.94%和10.69%,主要受土地利用和土壤类型影响;L,S,T相对误差均小于10%;6者平均7.89%。③以栅格计算法结果为基准值,采用单元插值外推法推算全县水土流失面积,1%密度下水土流失面积比偏高19.73%,4%密度下水土流失面积比偏高11.77%。[结论]蒙阴县1%和4%密度各因子均有不同程度的精度损失,并对水土流失估算结果造成一定影响,在区域水土流失动态监测过程中可根据需求选取合适的抽样密度。
关键词:  抽样密度  精度损失  相对误差  CSLE模型
DOI:10.13961/j.cnki.stbctb.2019.02.029
分类号:S157
基金项目:全国水土流失动态监测与公告项目“淮河流域国家级重点防治区水土流失动态监测”(HWSBJ201302),“山东省省级重点治理区水土流失动态监测项目”(SWJ201601);国家自然科学基金项目(31070627);江苏省高校优势学科建设工程项目
Effects of Sampling Densities on Estimation Precision of Soil Erosion Factors in Yimeng Mountain Area
Qi Fei1, Su Xinyu1, Li Jiazuo2, Hu Xuli3, Liu Xia1, Zhang Chunqiang2, Xing Xianshuang4, Ding Mingming5
1.Jiangsu Key Laboratory of Soil and Water Conservation and Ecological Restoration, Co-innovation Center for Sustainable Forestry in Southern China, Forestry College of Nanjing Forestry University, Nanjing, Jiangsu 210037, China;2.Mornitoring Center Station of Soil and Water Conservation, Huaihe River Commission, Ministry of Water Resources, Bengbu, Anhui 233001, China;3.Soil and Water Conservation Division of Huaihe River Commission of Water Resources Ministry, Bengbu, Anhui 233001, China;4.Hydrographic Office of Shandong Province, Ji'nan, Shandong 250002, China;5.Nanjing Water Bureau, Nanjing, Jiangsu 210036, China
Abstract:
[Objective] In order to provide data support for sampling method and sampling density selection in regional dynamic monitoring of soil erosion, the influence of sampling densities on estimation precision of soil erosion factors was studied.[Methods] The paper took Mengyin County, in the Yimeng and Tai Mountains national key control areas, as the research object to calculate the precision loss of soil erosion factors estimation under 1% and 4% sampling densities. The precision loss of soil erosion factors estimation under 1% and 4% sampling densities, compared to the soil erosion factors of the county was evaluated by human-computer interaction interpretation, field investigation and statistical analysis, based on the SPOT5 remote sensing images in 2013 and 1:10 000 topographic maps.[Results] ① Under the 1% sampling density, the relative errors of S, E and K factors were 33.48%, 23.46% and 20.64% respectively, which are mainly influenced by slope, land use and soil types, while the relative errors of L, B and T factors were less than 11%. The average relative error of soil erosion factors was 14.44% in the field investigation units. ② Under the 4% sampling density, the relative errors of E, K and B factors were 15.07%, 13.94% and 10.69% respectively, which were mainly affected by land use and soil types, while the relative errors of L, S and T factors were less than 10%. The average relative error of soil erosion factors in the field survey units was 7.89%. ③ The results calculated by the element interpolation extrapolation method under the two sampling densities were higher than that of the grid calculation method. The area of soil erosion under the 1% sampling density was 19.73% higher than that of the grid calculation method, and under the 4% sampling density, it was 11.77% higher.[Conclusion] The 1% and 4% density factors of Mengyin County had different degree of precision loss, which had a certain influence on the results of soil erosion estimation. In the process of dynamic monitoring of regional soil erosion, the appropriate sampling density can be selected according to the demand.
Key words:  sampling densities  precision loss  relative error  CSLE model