欢迎访问《生态学杂志》官方网站,今天是 分享到:

生态学杂志

• 研究报告 • 上一篇    下一篇

利用高光谱遥感预测小麦籽粒蛋白质产量

冯伟1,2;朱艳2;田永超2;曹卫星2;郭天财1;王晨阳1   

  1. 1河南农业大学国家小麦工程技术研究中心, 郑州 450002; 2南京农业大学江苏省信息农业高技术研究重点实验室, 南京 210095
  • 收稿日期:2007-09-14 修回日期:1900-01-01 出版日期:2008-06-10 发布日期:2008-06-10

Prediction of wheat grain protein yield by canopy hyperspectal remote sensing.

FENG Wei1,2;ZHU Yan2;TIAN Yong-chao2;CAO Wei-xing2;GUOTian-cai1;WANG Cheng-yang1   

  1. 1National Engineering Research Center for Wheat, Henan Agricultural Uni
    versity, Zhengzhou 450002, China;2HiTech Key Laboratory of Information
    Agriculture of Jiangsu Province, Nanjing Agricultural University, Nanjing 210095
    , China
  • Received:2007-09-14 Revised:1900-01-01 Online:2008-06-10 Published:2008-06-10

摘要: 2003—2006年连续3年采用不同小麦品种在不同施氮水平下进行大田试验,于小麦不同生育期采集田间冠层高光谱数据并测定植株氮素含量。根据特征光谱参数-叶片氮素营养籽粒蛋白质产量这一技术路径,以叶片氮素营养为连接点将模型链接,建立基于开花期高光谱参数的小麦籽粒蛋白质产量预报模型。结果表明:开花期叶片氮含量、氮积累量以及花后叶片氮转运量均能够较好地反映成熟期籽粒蛋白质产量状况;对叶片氮含量和氮积累量的光谱反演,在不同品种、氮素水平和年度间可以使用统一的光谱参数,其中利用红边位置(REPle)和修正型ND705(mND705)可以较好表达叶片氮含量的动态变化,以红蓝边面积比(SDr/SDb)和742 nm处一阶微分(FD742)为变量建立叶片氮积累量监测模型效果较好;经独立试验数据的检验表明,以参数REPle、SDr/SDb和FD742为变量建立成熟期籽粒蛋白质产量预报模型均给出理想的检验结果,模型测试精度R2分别为0.854、0.803和0.795,相对误差RE分别为16.4%、18.2%和149%;利用开花期关键特征光谱指数可以有效地评价小麦成熟期籽粒蛋白质产量状况。

关键词: 放牧羊草草原, 围栏羊草草原, 土壤呼吸

Abstract: Three years (2003-2006) field experiment was conducted with different wheat varieties and nitrogen (N) supply, and timecourse measurements were made on the canopy hyperspectral reflectance, plant dry mass and its N content. Based on the technique route of characteristic spectral parameters-leaf N nutrition-grain protein yield, and with leaf N nutrition as a link, the prediction models of wheat grain protein yield based on the canopy hyperspectal parameters at anthesis were constructed. The results showed that the wheat grain protein yield at maturity increased with increasing N supply. Plant N nutritional status, such as leaf N content (LNC) and leaf N accumulation (LNA) at anthesis and leaf N transportation (LNT) after anthesis, could well indicate the grain protein yield at maturity. The regression analysis between existing vegetation indices and leaf N index indicated that some key spectral parameters could accurately estimate the changes in leaf N status across a broad ranges of growth stages, N supply and growing seasons, with unified spectral parameters for each wheat cultivar, such as REPle and mND705 for leaf N content, and SDr/SDb and FD742 for leaf N accumulation. The testing of the prediction models with independent dataset indicated that the spectral indices of REPle, SDr/SDb and FD742 could give accurate grain protein yield estimation, and the R2 of REPle, SDr/SDb and FD742 was 0.854, 0.803 and 0.795, with the RE of 16.4%, 18.2% and 14.9%, respectively. Overall, the wheat grain protein yield at maturity could be predicted by the key characteristic spectral indices at anthesis.

Key words: Grazing Leymus chinensis grassland, Fenced Leymus chinensis grassland, Soil respiration