基于Stacking集成学习的土壤侵蚀速率计算与主导因子分析——以三峡库区奉节县为例

doi:10.16843/j.sswc.2023.04.012

摘要
图/表
参考文献
相关文章 (15)

全文: HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要土壤侵蚀速率的计算是水土保持工作的关键之一。为提高计算精度，引入Stacking集成方法，利用其能充分融合不同机器学习模型的特点，获取高精度的土壤侵蚀速率空间分布数据并分析影响研究区土壤侵蚀速率的主导因子。基于重庆市奉节县三峡库区2018年降雨量、遥感影像等数据构建特征集，以奉节县土壤侵蚀速率真实数据作为基准，通过训练不同机器学习模型，使用精度评价指标和多样性度量来建立最优的基学习器和元学习器组合，构建Stacking模型并获取土壤侵蚀速率空间分布图，然后针对土壤侵蚀速率分布规律对其主导因子进行边际依赖性分析。结果表明：1）以轻型梯度提升机、随机森林为基学习器，线性回归器为元学习器的Stacking集成模型效果最优，平均绝对误差、均方根误差和决定系数的表现分别为252.48 t/（km²·a）、537.78 t/（km²·a）和0.868 7；2）高程、降雨量、植被覆盖、坡度、距道路距离和距水源距离对奉节县土壤侵蚀速率影响程度排序位于前6，重要性所占比例均超过9%；3）在高程200~520 m，年总降雨量高于1 250 mm，NDVI为0.24~0.27，坡度在26°~35°之间，距道路距离0~220 m，距水源地距离63~387 m的地区土壤侵蚀速率较高。综上，构建的Stacking模型能够有效融合不同模型优势，提升预测土壤侵蚀速率的精度；奉节县土壤侵蚀速率受多方面因素综合影响，总体上与高程、植被覆盖程度之间呈正相关关系，与降雨量、坡度之间呈负相关关系，较高速率的土壤侵蚀倾向于发生在降雨充沛、植被覆盖度低、距道路及水源较近的低海拔陡峭区域。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	林娜
	潘鹏
	王斌
	张迪
	冯珊珊
	潘建平

关键词 ：机器学习, 土壤侵蚀, Stacking, 优化集成, 主导因子分析

Abstract：[Background] The calculation and assessment of soil erosion is the key to soil and water conservation. In order to improve the calculation accuracy, stacking ensemble method is introduced, which can fully integrate different machine learning models to obtain high-precision spatial distribution data of soil erosion rate. At the same time, the leading factors affecting the soil erosion rate in the study area were analyzed.[Methods] Firstly, the feature dataset was constructed based on the data of 2018 rainfall, remote sensing images and others in Fengjie county, Chongqing, and the actual data of soil erosion rate in Fengjie county was used as the benchmark to train different machine learning models. Then, the accuracy evaluation index and diversity measure were used to establish the optimal combination of base-learners and meta-learner, construct the stacking integrated model, and to calculate the soil erosion rate in the whole county. Finally, the marginal dependence of the leading factors was analyzed according to the distribution law of soil erosion rate.[Results] 1) The stacking ensemble model with light gradient boosting machineand random forest as the base-learners and linearregressionas the meta-learner has the best effect. The M_AE(mean absolute error), R_MSE (root mean square error) and accuracy of R²(R-squared) are as follows:252.48 t/(km²·a), 537.78 t/(km²·a) and 0.868 7. 2) Elevation, rainfall, vegetation cover, slope, distance from the road and distance from water source were the top 6 factors influencing soil erosion rate in Fengjie county,with importance accounting for more than 9%. 3) Soil erosion rate was higher in the region with an elevation of 200-520 m, annual rainfall higher than 1 250 mm, NDVI (normalized difference vegetation index) of 0.24-0.27, slope of 26°-35°, distance from the road to 0-220 m, and distance from the water source to 63-387 m.[Conclusions] The results show that the stacking model constructed in this paper can effectively integrate different models and improve the accuracy of predicting soil erosion rate. Soil erosion rate in Fengjie county is affected by many factors.In general, soil erosion rate was positively correlated with elevation and vegetation cover degree, and negatively correlated with rainfall and slope.The higher rate of soil erosion tended to occur in steep low-elevation areas with abundant rainfall, low vegetation cover, and close proximity to roads and water sources.

Key words： machine learning soil erosion Stacking optimization and integration leading factor analysis

收稿日期: 2022-07-22

PACS:	X87
	P237

基金资助:国家重点研发计划课题"融合北斗系统的空天地一体化智能监控预警技术"（2021YFB2600603）；宁夏回族自治区重点研发计划课题"面向贺兰山生态屏障防护体系的生态环境监测及平台研发"（2022CMG02014）

通讯作者: 潘鹏(1997-),男,硕士研究生。主要研究方向:遥感图像智能处理。E-mail:1157908009@qq.com E-mail: 1157908009@qq.com

作者简介: 林娜(1981-),女,博士,副教授。主要研究方向:遥感图像智能处理。E-mail:56654072@qq.com

引用本文:

林娜, 潘鹏, 王斌, 张迪, 冯珊珊, 潘建平. 基于Stacking集成学习的土壤侵蚀速率计算与主导因子分析——以三峡库区奉节县为例[J]. 中国水土保持科学, 2023, 21(4): 100-112.
LIN Na, PAN Peng, WANG Bin, ZHANG Di, FENG Shanshan, PAN Jianping. Soil erosion rate calculation based on Stacking ensemble learning and leading factor analysis: A case study of Fengjie county in the Three Gorges Reservoir Area. SSWC, 2023, 21(4): 100-112.

链接本文:

http://journal12.magtechjournal.com/Jweb_stbc/CN/10.16843/j.sswc.2023.04.012 或 http://journal12.magtechjournal.com/Jweb_stbc/CN/Y2023/V21/I4/100

[1]	孙鸿烈.我国水土流失问题与防治对策[J].中国水利,2011(6):16. SUN Honglie. Problems and countermeasures of soil erosion in China[J]. China Water Resources, 2011(6):16.
[2]	TIRUWA D B, KHANAL B R, LAMICHHANE S, et al. Soil erosion estimation using Geographic Information System (GIS) and Revised Universal Soil Loss Equation (RUSLE) in the Siwalik Hills of Nawalparasi, Nepal[J].Journal of Water and Climate Change, 2021, 12(5):1958.
[3]	ABDUL RAZAD A Z, SAMSUDIN S H, et al. Investigating the impact of land use change on sediment yield for hydropower reservoirs through GIS application[J]. IOP Conference Series:Earth and Environmental Science, 2020, 540(1):1
[4]	张超,陈国建,李春娟,等. 基于USLE模型的重庆生态涵养发展区土壤侵蚀量估算[J]. 水土保持研究, 2017, 24(3):33. ZHANG Chao, CHEN Guojian, LI Chunjuan, et al. Estimate of soil erosion amount of the ecological conservation development area in Chongqing based on the USLE model[J]. Research of Soil and Water Conservation, 2017, 24(3):33.
[5]	温莉. 基于遥感与GIS的三峡库区重庆段土壤侵蚀动态变化研究[D]. 重庆:西南大学, 2019:4. WEN Li. Study on dynamic changes of soil erosion in the Three Gorges Reservoir Area of Chongqing using remote sensing and GIS technologies[D]. Chongqing:Southwest University, 2019:4.
[6]	葛玲玲. 基于机器学习的土壤湿度和气温遥感反演方法研究[D].南京:南京信息工程大学,2021:5. GE Lingling. Remote sensing inversion of soil moisture and temperature based on machine learning[D].Nanjing:Nanjing University of Information Science & Technology,2021:5.
[7]	黄晨璐. 近40年黄土高原土壤侵蚀时空变化及其主控因子研究[D].西安:西北大学,2021:62. HUANG Chenlu. Temporal and spatial changes of soil erosion and its main controlling factors in Loess Plateau in recent 40 years[D]. Xi'an:Northwest University, 2021:62.
[8]	SAHOUR H, GHOLAMI V, VAZIFEDAN M, et al. Machine learning applications for water-induced soil erosion modeling and mapping[J]. Soil & Tillage Research, 2021, 211:36.
[9]	RAKHOHORI B, ISMAIL M, MAHROO D, et al. Modelling and mapping of soil erosion susceptibility using machine learning in a tropical hot sub-humid environment[J]. Journal of Cleaner Production, 2022, 364:41.
[10]	郝姗姗,李梦华,马永强,等.黄土丘陵区土壤侵蚀因子敏感性分析[J].中国水土保持科学,2019,17(2):77. HAO Shanshan, LI Menghua, MA Yongqiang, et al. Significance analysis of soil erosion factors in loess hilly gully region[J].Science of Soil and Water Conservation, 2019,17(2):77.
[11]	朱青,国佳欣,郭熙,等.基于随机森林算法的土壤侵蚀影响因子研究:以赣江上游流域为例[J].水土保持通报,2020,40(2):59. ZHU Qing, GUO Jiaxin, GUO Xi, et al. Research on influencing factors of soil erosion based on random forest algorithm:A case study in upper reaches of Ganjiang River basin[J].Bulletin of Soil and Water Conservation,2020,40(2):59.
[12]	游文霞,李清清,杨楠,等.基于多异学习器融合Stacking集成学习的窃电检测[J].电力系统自动化,2022,46(24):178. YOU Wenxia, LI Qingqing, YANG Nan, et al. Electric theft detection based on multi-different learner fusion Stacking integrated learning[J]. Automation of Electric Power Systems, 2022,46(24):178.
[13]	SOLOMON E,MENBERU T, ENDALKACHEW S, et al. Integrating RUSLE model with remote sensing and GIS for evaluation soil erosion in Telkwonz Watershed, northwestern Ethiopia[J]. Remote Sensing Applications:Society and Environment, 2021, 24:16.
[14]	王萌,刘云,宋超,等.基于RUSLE模型的2000-2010年长江三峡库区土壤侵蚀评价[J].水土保持通报,2018,38(1):12. WANG Meng, LIU Yun, SONG Chao, et al. Evaluating soil erosion based on RUSLE model in Three Gorges Reservoir Area during 2000-2010[J]. Bulletin of Soil and Water Conservation,2018,38(1):12.
[15]	曾凤铃, 刘淑婧, 运剑苇, 等. 模拟降雨条件下三峡库区紫色土坡地产流入渗特征[J]. 中国水土保持科学, 2022, 20(1):107. ZENG Fengling, LIU Shujing, YUN Jianwei, et al. Runoff and infiltration characteristics of purple soil slope in the Three Gorges Reservoir region under simulated rainfall[J].Science of Soil and Water Conservation, 2022, 20(1):107.
[16]	陈羽璇, 杨勤科, 刘宝元, 等. 基于CSLE模型的珠江流域土壤侵蚀强度评价[J]. 中国水土保持科学, 2021, 19(6):86. CHEN Yuxuan, YANG Qinke, LIU Baoyuan, et al. Assessment of soil erosion intensity in Pearl River Basin based on CSLE model[J].Science of Soil and Water Conservation, 2021, 19(6):86.
[17]	ZHOU, HONG, GUAN, et al. Applications of erosion hotspots for watershed investigation in the Appalachian Hills of the United States[J]. Journal of Irrigation and Drainage Engineering, 2016, 142(3):49.
[18]	FENTA A, YASUDA H, SHIMIZU K, et al. Dynamics of soil erosion as influenced by watershed management practices:A case study of the Agula Watershed in the semi-arid highlands of northern Ethiopia.[J]. Environmental Management, 2016, 58(5):889.
[19]	何芷.基于ArcGIS的赣州市土地脆弱性分析与评价[J]. 测绘与空间地理信息, 2018, 41(7):187. HE Zhi. Land vulnerability analysis and evaluation based on ArcGlS in Ganzhou city[J]. Geomatics & Spatial Information Technology, 2018, 41(7):187.
[20]	杨迎港, 刘培, 张合兵, 等. 基于特征优选随机森林算法的GF-2影像分类[J]. 航天返回与遥感, 2022, 43(2):115. YANG Yinggang,LIU Pei, ZHANG Hebing, et al. Research on GF-2 image classification based on feature optimization random forest algorithm[J]. Spacecraft Recovery & Remote Sensing, 2022, 43(2):115.
[21]	史佳琪,张建华.基于多模型融合Stacking集成学习方式的负荷预测方法[J].中国电机工程学报,2019,39(14):4032. SHI Jiaqi, ZHANG Jianhua. Load forecasting based on multi-model by stacking ensemble learning[J]. Proceedings of the CSEE,2019,39(14):4032.
[22]	刘德军,戴庆庆,左建平,等.基于Stacking集成算法的岩爆等级预测研究[J].岩石力学与工程学报,2022,41(1):2915. LIU Dejun, DAI Qingqing, ZUO Jianping, et al. Research on rock burst grade prediction based on stacking integrated algorithm[J]. Chinese Journal of Rock Mechanics and Engineering, 2022,41(1):2915.
[23]	徐铖龙. 基于RS和GIS的土壤侵蚀对土地利用变化的响应研究[D]. 哈尔滨:黑龙江大学, 2021:56. XU Yuelong. Response of soil erosion to land use change based on RS and GIS[D]. Harbin:Heilongjiang University, 2021:56.