摘要:利用ASD地物非成像高光谱仪,获取2个加工番茄品种4水平施氮量和3种配置种植方式6个关键生育时期冠层的反射光谱数据,通过计算得到归一化植被指数(NDVI)、比值植被指数(RVI)、修改型二次土壤调节植被指数( MSAVI2)和红边归一化植被指数(RENDVI),并分别与其冠层叶绿素密度(CH.D)、叶面积指数(LAI)、地上鲜生物量(AFBM)和地上干生物量(ADBM)进行相关分析,经检验,相关系数均达到1%的极显著水平.其中RENDVI与CH.D的线性相关模型,RVI与LAI的幂指数函数模型的相关性最好(RRENDVI-CH.D=0.8034**,RRVI-LAI=0.8703**,n=54,α=1%),用上述2个相关模型方程分别估算加工番茄CH.D和LAI,实测值与估测值之间均呈极显著的线性相关关系(R实测CH.D-估测CH.D=0.8113**,R实测LAI-估测LAI=0.8546**,n=54,α=1%),估算精度分别为85.5%和86.3%.试验结果表明,用高光谱植被指数,可以对加工番茄冠层CH.D、LAI、AFBM和ADBM进行遥感估算,实现对加工番茄生长状况的实时、无损、非接触和定量的高光谱监测研究.%Hyperspectral reflectance data of the two processing tomato cultivars Li Geer 87 - 5 and Tun He No. 8 under four nitrogen fertilization treatments and three different plant patterns were recorded in a field experiment by the ASD Fieldspec non-imaging spectroradiometer at six main growth stages. Calculation of reflectance spectrum data obtained the normalized difference vegetation index (NDVI),ratio vegetation index (RVI),the second modified soil adjusted vegetation index (MSAVI2) and red edge normalized difference vegetation index (RENDVI). Regression analysis techniques were then performed to establish function models among the four indices and the measured chlorophyll density(CH. D) ,leaf area index(LAI),aboveground fresh biomass (AFBM) and aboveground dry biomass (ADBM) .respectively. The results showed that the four vegetation indices are positively significant correlated with the four physiological parameters at 1% level. Among them,RENDVI and CH. D,RVI and LAI had the strongest linear relationship and the strongest power exponential function(RRENDVI-CH.D = 0. 8034** >RRVFLAI= 0. 8703** ,n= 54, α=l%) .respectively. Based on their strongest regression functions to predict CH. D and LAI,respectively. There were significant correlation at 1% between tested CH. D and estimated CH. D,tested LAI and estimated LAI (RMeasured CH.D-Estimated CH.D=0.8113** ,RMeasured LAI-Eatimated LAI=0. 8546** ,n=54,α=l%).The regression function accuracies were 85. 5% and 86. 3 %, respectively. The study examined that CH. D,LAI, AFBM and ADBM of processing tomato canopy can be estimated by vegetation indices,and hyperspectral remote sensing can be achieved instant, nondestructive, un-touched and quantitative monitoring growth status of processing tomato.