人工神经网络与普通克里金插值法对土壤属性空间预测精度影响研究

doi:10.19741/j.issn.1673-4831.2020.0811

生态与农村环境学报 ›› 2021, Vol. 37 ›› Issue (7): 934-942.doi: 10.19741/j.issn.1673-4831.2020.0811

人工神经网络与普通克里金插值法对土壤属性空间预测精度影响研究

谢梦姣^1,2, 王洋^2,3, 康营³, 吴志涛⁴, 陈奇乐³, 刘琦¹, 吴超玉³, 张俊梅^2,3

1. 河北农业大学国土资源学院, 河北保定 071000;
2. 华北作物改良与调控国家重点实验室, 河北保定 071000;
3. 河北农业大学资源与环境科学学院/河北省农田生态环境重点实验室, 河北保定 071000;
4. 北京邮电大学信息与通信工程学院, 北京 100876

收稿日期:2020-09-28 出版日期:2021-07-25 发布日期:2021-07-23
通讯作者: 张俊梅 E-mail:254157642@qq.com
作者简介:谢梦姣(1994-),女,山东临朐人,主要从事土地资源利用与生态环境效应研究。E-mail:xiemengjiao94@163.com
基金资助:
国家重点研发计划（2017YFD0300905-6）；华北作物改良与调控国家重点实验室开放课题（NCCIR2020ZZ-20）；河北省自然科学青年基金（C2019204351）

Accuracy Study of Spatial Predicting in Soil Attributes Based on Interpolations by Artificial Neural Network and Ordinary Kriging

XIE Meng-jiao^1,2, WANG Yang^2,3, KANG Ying³, WU Zhi-tao⁴, CHEN Qi-le³, LIU Qi¹, WU Chao-yu³, ZHANG Jun-mei^2,3

1. College of Land and Resources, Hebei Agricultural University, Baoding 071000, China;
2. State Key Laboratory of North China Crop Improvement and Regulation, Baoding 071000, China;
3. College of Resources and Environmental Science/Hebei Province Key Laboratory for Farmland Eco-environment, Hebei Agricultural University, Baoding 071000, China;
4. School of Information and Communication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China

Received:2020-09-28 Online:2021-07-25 Published:2021-07-23

摘要/Abstract

摘要： 空间插值方法与研究样区空间尺度会对土壤有机质和全氮含量空间变异特征预测精度产生重要影响。选取黄淮海北部未收割的夏玉米收获期50 m×50 m田块尺度（取80个土样）和1 000 m×1 000 m（取100个土样）农场尺度为研究样区，采用普通克里金插值法和径向基函数（RBF）人工神经网络插值法探究两种尺度研究样区土壤有机质和全氮含量空间变异特征及两种空间插值方法的预测精度。结果表明，两个研究样区土壤有机质和全氮含量分别为8.39~20.59和0.31~2.90 g·kg^-1，块金系数为0.448~0.637，表明其呈现中等程度空间变异。土壤有机质和全氮含量在农场尺度研究样区大致呈现以西北部东南对称线含量较高向两边逐渐减少的空间分布趋势，在田块尺度上呈现东北部地区含量较高且整体向西南部递减的空间分布特征。基于RBF人工神经网络插值法的田块尺度研究样区均方根误差（RMSE）、平均绝对误差（MAE）和平均相对误差（MRE）均小于农场尺度，这表明土壤有机质和全氮含量空间分布预测精度受到采样尺度的影响，田块尺度预测精度优于农场尺度。基于RBF人工神经网络插值法在同一尺度研究样区土壤有机质和全氮含量空间分布预测的各项误差均有所减小，模型拟合决定系数R²有所增加，且避免了普通克里金插值结果的"平滑效应"现象，其预测精度更高。研究结果表明，RBF人工神经网络插值法更适用于土壤有机质和全氮含量空间分布特征预测。

关键词: 土壤有机质, 土壤全氮, 预测精度, 普通克里金, RBF人工神经网络

Abstract: Precise predictions of spatial variability of soil organic matter and total nitrogen are largely depended on methods and spatial scales of the interpolation. In this study, two spatial scales including plot scale (50 m×50 m, 80 soil samples) and farm scale (1 000 m×1 000 m, 100 soil samples) were selected from nearly harvesting field of summer maize in the northern part of Huanghuaihai Plain. And two interpolating methods including ordinary Kriging and artificial neural network based on radial basis function (RBF) were compared to explore more accurate spatial variations of soil organic matter and total nitrogen content in the study area. Results of nugget to sill ratio (0.448-0.637) showed moderate spatial variations of both spatial scales with contents of soil organic matter and total nitrogen ranged from 8.39 to 20.59 g·kg^-1 and from 0.31 to 2.90 g·kg^-1, respectively. Spatial variations of soil organic matter and total nitrogen presented decreasing trends with higher contents distributed along the diagonal from northwest to southeast and lower contents distributed around the edges at farm scale, in contrast, higher and lower contents respectively located in northeast and southwest parts at plot scale. Concerning root mean square error (RMSE), average absolute error (MAE) and average relative error (MRE) of interpolation by artificial neural network based on RBF, smaller values indicated better predictions of spatial variations at plot scale rather than at farm scale. Without smoothing effect, increased R² and decreased errors proved better performance of artificial neural network based on RBF rather than ordinary Kriging in modeling spatial variations of soil organic matter and total nitrogen at each spatial scale. It indicates that artificial neural network based on RBF is the better interpolating method for predicting spatial variations of soil organic matter and total nitrogen in this study.

Key words: soil organic carbon, soil total nitrogen, prediction accuracy, ordinary Kriging, RBF neural network

中图分类号:

S159

谢梦姣, 王洋, 康营, 吴志涛, 陈奇乐, 刘琦, 吴超玉, 张俊梅. 人工神经网络与普通克里金插值法对土壤属性空间预测精度影响研究[J]. 生态与农村环境学报, 2021, 37(7): 934-942.

XIE Meng-jiao, WANG Yang, KANG Ying, WU Zhi-tao, CHEN Qi-le, LIU Qi, WU Chao-yu, ZHANG Jun-mei. Accuracy Study of Spatial Predicting in Soil Attributes Based on Interpolations by Artificial Neural Network and Ordinary Kriging[J]. Journal of Ecology and Rural Environment, 2021, 37(7): 934-942.

参考文献 15

[1]	吴乐知,蔡祖聪.中国土壤有机质含量变异性与空间尺度的关系[J].地球科学进展,2006,21(9):965-972.[WU Le-zhi,CAI Zu-cong.The Relationship between the Spatial Scale and the Variation of Soil Organic Matter in China[J].Advances in Earth Science,2006,21(9):965-972.]
[2]	陈飞香,程家昌,胡月明,等.基于RBF神经网络的土壤铬含量空间预测[J].地理科学,2013,33(1):69-74.[CHEN Fei-xiang,CHENG Jia-chang,HU Yue-ming,et al.Spatial Prediction of Soil Properties by RBF Neural Network[J].Scientia Geographica Sinica,2013,33(1):69-74.]
[3]	吴黎军,贺军亮,冯晓淼.地统计学及其在土壤生态学研究中的应用与进展[J].安徽农业科学,2009,37(25):12353-12356.[WU Li-jun,HE Jun-liang,FENG Xiao-miao.Geostatistics and Its Applications and Advance in the Studies of Soil Ecology[J].Journal of Anhui Agricultural Sciences,2009,37(25):12353-12356.]
[4]	谢梦姣,陈奇乐,张俊梅,等.短距离样点对土壤呼吸空间变异预测精度的影响[J].中国生态农业学报(中英文),2020,28(3):421-428.[XIE Meng-jiao,CHEN Qi-le,ZHANG Jun-mei,et al.Effects of Short Distance Sampling on the Prediction Accuracy of the Spatial Variability of Soil Respiration[J].Chinese Journal of Eco-agriculture,2020,28(3):421-428.]
[5]	LARK R M,MARCHANT B P.How Should a Spatial-coverage Sample Design for a Geostatistical Soil Survey Be Supplemented to Support Estimation of Spatial Covariance Parameters?[J].Geoderma,2018,319:89-99.
[6]	BHATTACHARJEE S,MITRA P,GHOSH S K.Spatial Interpolation to Predict Missing Attributes in GIS Using Semantic Kriging[J].IEEE Transactions on Geoscience and Remote Sensing,2014,52(8):4771-4780.
[7]	杨顺华,张海涛,郭龙,等.基于回归和地理加权回归Kriging的土壤有机质空间插值[J].应用生态学报,2015,26(6):1649-1656.[YANG Shun-hua,ZHANG Hai-tao,GUO Long,et al.Spatial Interpolation of Soil Organic Matter Using Regression Kriging and Geographically Weighted Regression Kriging[J].Chinese Journal of Applied Ecology,2015,26(6):1649-1656.]
[8]	郭龙,张海涛,陈家赢,等.基于协同克里格插值和地理加权回归模型的土壤属性空间预测比较[J].土壤学报,2012,49(5):1037-1042.[GUO Long,ZHANG Hai-tao,CHEN Jia-ying,et al.Comparison between Co-Kriging Model and Geographically Weighted Regression Model in Spatial Prediction of Soil Attributes[J].Acta Pedologica Sinica,2012,49(5):1037-1042.]
[9]	李艳,史舟,徐建明,等.地统计学在土壤科学中的应用及展望[J].水土保持学报,2003,17(1):178-182.[LI Yan,SHI Zhou,XU Jian-ming,et al.Utilization and Perspective of Geostatistics in Soil Sciences[J].Journal of Soil Water Conservation,2003,17(1):178-182.]
[10]	贾振宇,张俊华,丁圣彦,等.基于GIS和地统计学的黄泛区土壤磷空间变异:以周口为例[J].应用生态学报,2016,27(4):1211-1220.[JIA Zhen-yu,ZHANG Jun-hua,DING Sheng-yan,et al.Spatial Variation of Soil Phosphorus in Flooded Area of the Yellow River Based on GIS and Geo-statistical Methods:A Case Study in Zhoukou City,Henan,China[J].Chinese Journal of Applied Ecology,2016,27(4):1211-1220.]
[11]	马泉来,王小玉,赵曼宇,等.黑土区小流域土壤氮磷生态化学计量空间分异特征[J].生态与农村环境学报,2020,36(10):1325-1332.[MA Quan-lai,WANG Xiao-yu,ZHAO Man-yu,et al.Spatial Variability of Ecological Stoichiometry of Soil Nitrogen and Phosphorus in a Mollisol Watershed of China[J].Journal of Ecology and Rural Environment,2020,36(10):1325-1332.]
[12]	李启权,王昌全,岳天祥,等.基于定性和定量辅助变量的土壤有机质空间分布预测:以四川三台县为例[J].地理科学进展,2014,33(2):259-269.[LI Qi-quan,WANG Chang-quan,YUE Tian-xiang,et al.Prediction of Distribution of Soil Organic Matter Based on Qualitative and Quantitative Auxiliary Variables:A Case Study in Santai County in Sichuan Province[J].Progress in Geography,2014,33(2):259-269.]
[13]	许珊,邹滨,王敏,等.PM2.5浓度空间估算的神经网络与克里格方法对比[J/OL].武汉大学学报(信息科学版),2020,45(10):1-8[2020-05-14].https://doi.org/10.13203/j.whugis20180482.[XU Shan,ZOU Bin,WANG Min,et al.Performance Comparison of Artificial Neural Network and Kriging in Spatial Estimation of PM2.5 Concentration[J/OL].Geomatics and Information Science of Wuhan University,2020,45(10):1-8[2020-05-14].https://doi.org/10.13203/j.whugis20180482.]
[14]	梁旭光,王福林,赵红磊,等.基于RBF神经网络的大豆种植密度和施肥量优化[J].大豆科学,2020,39(3):406-413.[LIANG Xu-guang,WANG Fu-lin,ZHAO Hong-lei,et al.Optimization of Soybean Planting Density and Fertilizer Application Rate Based on RBF Neural Network[J].Soybean Science,2020,39(3):406-413.]
[15]	杨海荣.基于RBF人工神经网络的空间插值[J].长沙交通学院学报,2006,22(1):68-71.[YANG Hai-rong.A Space Interpolation Method Based on RBF Artificial Neural Network[J].Journal of Changsha Communications University,2006,22(1):68-71.]

人工神经网络与普通克里金插值法对土壤属性空间预测精度影响研究

Accuracy Study of Spatial Predicting in Soil Attributes Based on Interpolations by Artificial Neural Network and Ordinary Kriging

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献 15

相关文章 12

编辑推荐

Metrics

本文评价

[1]	彭倩,朱慧可,钟寰,贺美. 腐殖酸对汞污染稻田中甲基汞行为的影响[J]. 生态与农村环境学报, 2015, 31(5): 748-752.
[2]	李姗姗, 曹广超, 石平超, 蒋刚, 袁杰. 青岛城区土壤重金属元素空间分布及其现状评价[J]. 生态与农村环境学报, 2015, 31(1): 112-117.
[3]	李玮, 乔玉强, 陈欢, 杜世州, 赵竹, 曹承富. 秸秆还田配施氮肥对砂姜黑土有机质组分与碳库管理指数的影响[J]. 生态与农村环境学报, 2014, 30(4): 475-480.
[4]	毛亮, 高扬, 曹杰君, 陈晓燕, 周培, 张春华, 靳治国, 施婉君. 不同土地利用方式下Cd、Pb复合污染对土壤酶活性的影响[J]. 生态与农村环境学报, 2010, 26(2): 167-173.
[5]	马金妍, 石冰, 王开运, 巩晋楠, 侯颖. 崇明东滩湿地围垦区芦苇生物量影响因素初探[J]. 生态与农村环境学报, 2009, 25(4): 100-102.
[6]	张秀玲, 李君剑, 石福臣. 速生杨人工林对土壤碳氮含量及微生物生物量的影响[J]. 生态与农村环境学报, 2008, 24(2): 32-35.
[7]	陆琦, 马克明, 张洁瑜, 卢涛, 倪红伟. 三江平原退化湿地和农田土壤养分的比较研究[J]. 生态与农村环境学报, 2007, 23(2): 23-28.
[8]	穆琳, 张继宏, 关连珠. 施肥与地膜覆盖对土壤有机质平衡的影响[J]. 生态与农村环境学报, 1998, 14(2): 20-23,28.
[9]	石长松. 沙滩梨园生态经济效益初论[J]. 生态与农村环境学报, 1992, 8(1): 29-32.
[10]	秦祖平, 徐琪. 两种稻麦轮作制下土壤性状的比较研究[J]. 生态与农村环境学报, 1988, 4(2): 16-20.
[11]	杨景辉. 铜矿开采和冶炼造成的土壤污染及其改良途径[J]. 生态与农村环境学报, 1986, 2(2): 30-34.
[12]	魏开湄. 农药在土壤环境中的行为和归宿(一)[J]. 生态与农村环境学报, 1985, 1(3): 41-46.