CAJ | 학술논문

以锡林郭勒草原为试验区,采集草地冠层高光谱反射率数据和对应样方的地上生物量数据,综合分析了高光谱主要变换形式(高光谱特征变量、小波能量系数和植被指数)与草地地上生物量间的相关关系.研究结果表明：在选取的13个高光谱特征变量中,草地地上生物量与红谷的吸收深度Rd、红边区一阶微分总和与蓝边区一阶微分总的比值SDr/SDb的相关系数最高(0.86)；在11个小波能量系数中,第四小波能量系数与草地地上生物量的相关性最强,相关系数为0.85；在选用的8种常用植被指数中,ARVI 与生物量之间的相关性最强(R=0.874),而DVI的相关性最差(R=0.578).在对不同近红外、红光和蓝光波段进行组合而获得的植被指数中,ARVI(R945,R620,R430)模型对草地地上生物量的预测具有最好的准确性,训练样本和验证样本的拟合方程R值分别为0.882和0.865.
이석림곽륵초원위시험구,채집초지관층고광보반사솔수거화대응양방적지상생물량수거,종합분석료고광보주요변환형식(고광보특정변량、소파능량계수화식피지수)여초지지상생물량간적상관관계.연구결과표명：재선취적13개고광보특정변량중,초지지상생물량여홍곡적흡수심도Rd、홍변구일계미분총화여람변구일계미분총적비치SDr/SDb적상관계수최고(0.86)；재11개소파능량계수중,제사소파능량계수여초지지상생물량적상관성최강,상관계수위0.85；재선용적8충상용식피지수중,ARVI 여생물량지간적상관성최강(R=0.874),이DVI적상관성최차(R=0.578).재대불동근홍외、홍광화람광파단진행조합이획득적식피지수중,ARVI(R945,R620,R430)모형대초지지상생물량적예측구유최호적준학성,훈련양본화험증양본적의합방정R치분별위0.882화0.865.
Vegetation aboveground biomass and corresponding canopy hyper-spectral reflectance curves were collected in the Xilingol grassland ecosystem,and comprehensive correlation analysis was conducted between biomass and hyper-spectral conversion types:hyper-spectral feature variables,wavelet energy coefficients,vegetation indexes.The results indicated that hyper-spectral feature variable Rd and SDr/SDb was highly correlated with grassland biomass with a R value of 0.86 in the all selected 13 hyper-spectral feature variables;in the 1 1 wavelet energy coefficients,the fourth energy coefficient was highly correlated with biomass with a R value of 0.85;in the selected 8 common vegetation indexes,ARVI(R=0.874)has the highest correlation with biomass and DVI has the lowest R value(R=0.578).In the vegetation index derived from combination of different near infrared band,red band and blue band,the ARVI(R945 ,R620 , R430 )model has the best prediction for biomass with a R value of 0.882 and 0.865 for training data sets and testing data sets,respectively.