引用本文:高凤杰,吴啸,师华定,等.基于贝叶斯最大熵的黑土区小流域土壤有机质空间分布预测[J].环境科学研究,2019,32(8):1365-1373.
GAO Fengjie,WU Xiao,SHI Huading,et al.Prediction of Spatial Distribution of Soil Organic Matter in a Mollisol Watershed of China based on BME Method[J].Reserrch of Environmental Science,2019,32(8):1365-1373.]
【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 105次   下载 108 本文二维码信息
码上扫一扫!
分享到: 微信 更多
基于贝叶斯最大熵的黑土区小流域土壤有机质空间分布预测
高凤杰1, 吴啸1, 师华定2, 鞠铁男1, 王鑫1, 高东晶1, 刘媚媚1
1. 东北农业大学资源与环境学院, 黑龙江 哈尔滨 150030;2. 中国环境科学研究院土壤与固体废物环境研究所, 北京 100012
摘要:
进行w(SOM)空间预测研究,对掌握区域w(SOM)空间分布现状、实施精准农业以及保护区域生态环境都有着重要意义.以土地利用类型为辅助变量,将不同土地利用类型所对应w(SOM)的概率分布作为“软数据”,采用BME(贝叶斯最大熵)方法对我国东北黑土丘陵区海沟河小流域表层w(SOM)的分布情况进行空间预测,并与以w(TN)和土地利用类型为辅助变量的CK(协同克里格)方法进行比较,探讨两种方法的可行性与精度.结果表明:我国东北黑土丘陵区海沟河小流域表层w(SOM)平均值为24.04 g/kg,空间变异程度为中等.w(SOM)与w(TN)呈极显著正相关,与土地利用类型存在较强的相关性,不同土地利用类型w(SOM)差异明显,w(TN)与土地利用类型能够用来辅助w(SOM)的空间分布插值.相较于CK方法,BME方法能更好地利用“软数据”进行空间插值,使对w(SOM)预测结果的平均误差(ME)、平均绝对误差(MAE)和均方根误差(RMSE)均有所降低,精度大幅提高,空间插值效果明显优于CK方法.研究显示,研究区w(SOM)以阶梯状自东向西依次递减分布,在南北方向上变化不大,空间变化特征较为明显,BME方法利用“软数据”插值后的结果能较好地反映研究区w(SOM)空间分布的实际情况.
关键词:  土壤有机质  软数据  协同克里格  贝叶斯最大熵  空间预测
DOI:10.13198/j.issn.1001-6929.2018.11.16
分类号:X142
基金项目:国家重点研发计划项目(No.2016YFD0201009);国家自然科学基金项目(No.31700407)
Prediction of Spatial Distribution of Soil Organic Matter in a Mollisol Watershed of China based on BME Method
GAO Fengjie1, WU Xiao1, SHI Huading2, JU Tienan1, WANG Xin1, GAO Dongjing1, LIU Meimei1
1. College of Resource and Environment, Northeast Agricultural University, Harbin 150030, China;2. Institute of Soil and Solid Waste, Chinese Research Academy of Environmental Sciences, Beijing 100012, China
Abstract:
The research of spatial prediction of soil organic matter (SOM) is of great important for mastering the current situation of spatial distribution of SOM content, implementing precision agriculture and protecting the regional ecological environment. With land use types as the auxiliary variable, the probability distribution of different land use types corresponding to SOM content was used as ‘soft data’, predicting the spatial distribution of SOM content in the surface soil of Haigouhe Watershed by using the Bayesian maximum entropy (BME) method, and compared with the co-Kriging (CK) which used soil total nitrogen and land use type as a supplementary variable, then discussed feasibility and precision of the two methods. The results show that the mean SOM content was 24.04 g/kg for the whole watershed with a degree of variation at a moderate level. SOM content was significantly positively correlated with total nitrogen, and had a strong correlation with land use types, the SOM content of different land use types was obviously different. Total nitrogen and land use types could be used to assist the spatial distribution interpolation of SOM content. Compared with the CK method, the BME method could make good use of ‘soft data’ for spatial interpolation, so that the mean error (ME), mean absolute error (MAE) and mean square root error (RMSE) of the prediction results of SOM content were reduced, the accuracy was greatly improved, and the spatial interpolation effect was obviously better than CK method. The research shows that the SOM content in the study area decreases stepwise from east to west, the content of north and south direction changed little, and the spatial variation characteristics were obvious. The results of BME interpolation using ‘soft data’ can better reflect the actual situation of the spatial distribution of SOM content in the study area.
Key words:  soil organic matter  soft data  Bayesian maximum entropy  cooperative Kriging  spatial prediction