高温胁迫下水稻产量的高光谱估测研究  

Estimation of Rice Yield under High Temperature Stress by Hyper-spectral Remote Sensing

作　　者：谢晓金[1,2] 李映雪[1] 李秉柏[2] 申双和[1] 程高峰[2] 

机构地区：[1]南京信息工程大学应用气象学院,江苏南京210044 [2]江苏省农业科学院资源与环境研究所,江苏南京210014

出　　处：《中国水稻科学》2010年第2期196-202,共7页Chinese Journal of Rice Science

基　　金：国家科技支撑计划资助项目(2006BAD04B04)

摘　　要：为了定量分析不同生育期冠层反射光谱参数与水稻产量及产量构成要素的相互关系,确定能够准确预测高温胁迫下水稻籽粒产量的敏感光谱参数,通过盆栽试验,测定了孕穗期4种温度胁迫处理后的2个水稻品种不同生育期冠层高光谱反射率以及成熟后的理论产量、实际产量、穗数、每穗粒数、千粒重、穗长、穗干质量和结实率。结果表明,相对于蜡熟期的光谱参数,抽穗期和灌浆期的光谱参数与理论产量、实际产量、穗数、每穗粒数、千粒重、穗长、穗干质量以及结实率的相关性都较高,均达到显著水平。此两个时期可以作为预测水稻产量的关键时期。其中,差值植被指数DVI[810,A(450,560,680)]、垂直植被指数PVI(810,680)、红边幅值Dλred和红边峰值面积可以同时预测成熟水稻的理论产量和实际产量。而差值植被指数DVI(810,450)和DVI(810,560)、垂直植被指数PVI(810,680)和Dλred可以同时预测成熟水稻的穗数、每穗粒数和千粒重。相对于灌浆期的模型,抽穗期的模型能较可靠地监测水稻产量。To determine the relationships between the canopy spectral reflectance indices during different growth stages and grain yield and its components of rice,and to seek the sensitive spectral parameters for exactly estimating rice grain yield under high temperature stress,a pot experiment was conducted with two rice cultivars under four high-temperature stresses at the booting stage.The canopy hyper-spectral reflectance at different growth stages after heading,theoretical yield,actual yield,grain number per panicle,grain weight,panicle length,panicle weight and seed setting rate after maturity were measured.The coefficients of correlations between the spectral indices and the theoretical yield,actual yield,panicle number,grain number per panicle,grain weight,panicle length,panicle weight and seed setting rate at the heading and filling stages were significantly higher than those at the ripening stage,and the indices at the heading and filling stages could be key to predicting rice yield.In these spectral indices,the difference vegetation index DVI ／[810,A（450,560,680）／],perpendicular vegetation index PVI （810,680）,peak value of red edge and area of the red edge peak could be used for simultaneously estimating the theoretical and actual yields of matured rice.Besides,DVI （810,450） and DVI（810,560）,PVI （810,680） and peak value of red edge could be used for simultaneously estimating the panicle number,grain number per panicle and grain weight of matured rice.The model based on the indices at the heading stage could monitor rice yield more reliably than that based on the indices at the filling stage.

关 键 词：水稻 高温胁迫 冠层高光谱反射率 产量预测 产量构成因素 模型 遥感 

分 类 号：S511[农业科学—作物学]