引用本文:刘新路,彭 杰,冯春晖,吴家林,罗德芳,齐 威.基于电磁感应仪数据的南疆棉田土壤电导率反演模型研究[J].土壤学报,2020,57(3):646-655. DOI:10.11766/trxb201902190068
LIU Xinlu,PENG Jie,FENG Chunhui,WU Jialin,LUO Defang,QI Wei.Research on the inversion model for soil conductivity of cotton field using EM38-MK2 data in southern Xinjiang[J].Acta Pedologica Sinica,2020,57(3):646-655. DOI:10.11766/trxb201902190068
【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 441次   下载 340 本文二维码信息
码上扫一扫!
分享到: 微信 更多
基于电磁感应仪数据的南疆棉田土壤电导率反演模型研究
刘新路, 彭杰, 冯春晖, 吴家林, 罗德芳, 齐威
塔里木大学植物科学学院
摘要:
土壤盐渍化是南疆棉田优质高产的主要障碍因素,定额灌溉可有效改良土壤盐渍化,EM38-MK2可快速、动态获取土壤盐渍化数据,适时监测土壤盐渍化,为棉田定额灌溉提供数字依据。针对电磁感应仪(EM38-MK2)的发射圈和接收圈设计了不同采样方案,在同一条田内采集了6个不同时期的土壤表观电导率数据及相应的剖面土样,分析了不同土壤采集方案及土壤含水量对表观电导率模型精度的影响,对比了以单一时期数据建模的局部模型和6个时期整体数据建模的全局模型的反演精度。结果表明,由单点采集电磁感应仪发射线圈位置土样所构建的模型精度更高,稳定性更好,能有效减少由采样带来的误差。当土壤含水量低于10%时,表观电导率与实测电导率之间的相关性较低,决定系数为0.58,反演模型只具备粗略估计实测电导率的能力;当土壤含水量高于10%时,表观电导率与实测电导率具有很好的相关性,决定系数达到0.80以上,反演模型具有较好的预测能力。EMH+EMV多测定模式下表观电导率与实测电导率之间的模型精度高于EMH或EMV单一测定模式,确立ECh0.375 +(ECh0.75 +EC v0.75)/2+ECv1.5为建模因子,有效提高了反演模型的精度。不同深度土层的局部模型反演精度均高于全局模型精度,局部模型的RPD均大于2.0,具有较好的预测能力。
关键词:  土壤盐分  土壤表观电导率  土壤电导率  反演模型  棉田
基金项目:国家重点研发计划项目(2018YFE0107000,2016FYC0501407)
Research on the inversion model for soil conductivity of cotton field using EM38-MK2 data in southern Xinjiang
LIU Xinlu,PENG Jie,FENG Chunhui,WU Jialin,LUO Defang,QI Wei
College of Plant Science, Tarim University
Abstract:
【Objective】 Soil salinization is the major obstacle restraining high quality and high yield of cotton in South Xinjiang, and rated irrigation can effectively alleviate soil salinization. EM38-MK2 can be used to obtain real-time dynamical soil salinization data, monitor in-situ soil salinization in time, and provide a digital basis for rated irrigation in cotton fields. 【Method】 In this study, different soil sampling schemes were designed in line with the designing of the transmitting coil and receiving coil of the electromagnetic sensor(EM38-MK2), and soil apparent electrical conductivities (ECa) and soil samples from corresponding soil profiles were collected in the same field in six different time periods for analysis of, effects of soil sampling scheme and soil moisture content on accuracy of the apparent EC model, and specification of factors for building apparent EC models for different soil layers. Moreover, comparison was done of the local model based on data of a single period with the model based on overall data of the six periods in inversion accuracy.【Result】 Results show that the model based on single point sampling at the site of the transmitting coil of the EM38-MK2 was higher in sensoring accuracy and stability than the others and hence could effectively minimize the error caused by sampling. When soil moisture content was less than 10%, the correlation between apparent EC and measured EC was low, with determination coefficient being 0.58 and the inversion model could only roughly estimate EC; When it was more than 10%, apparent EC displayed a good relationship with measured EC, with determination coefficient being 0.80 and the inversion model could make an estimation of EC close to the measured one. The EMH+EMV model was built on the equation of ECh0.375 +(ECh0.75 +EC v0.75)/2+ECv1.5 defined as modeling factor and much higher than the EMH or EMV model in inversion accuracy with predicted apparent EC being very approximate to measured EC. The local models for predicting ECs in various soil layers different in depth were all higher than the overall one in inversion accuracy and their RPDs were more than 2.0, which indicate that they have high prediction ability. 【Conclusion】 All the findings in the study demonstrate that sampling under the transmitting coil is conducive to stability of the sampling, which may guide future utilization of the equipment, and open up new ways of thinking and server as reference for monitoring soil salinization dynamically with the electromagnetic sensor technology on the field scale.
Key words:  Soil salinization  Soil apparent electrical conductivity (ECa)  Soil electrical conductivity  Inversion models  Cotton field