论文部分内容阅读
将不同时相的柑橘树冠层光谱数据和氮、磷、钾等农学参数与实测产量数据作回归分析,构建不同时相的柑橘估产模型,并对不同时相模型进行比较分析,探索最佳遥感估产模型.结果表明:以柑橘4个生育时期(萌芽期、开花期、成熟期、休眠期)光谱特征点的反射峰深度、绿峰归一化差值指数、红边位置等光谱变量和氮、磷、钾等农学参数组合构建的复合回归模型估产效果最佳,经过柑橘实测产量的检验表明,2010年(R2=0.965,Re=2.540)和2011年(R2=0.975,Re=2.163)都达到极显著相关,说明该估产模型具有可靠性.同时,也证明了反射峰深度、绿峰归一化差值指数、红边位置等光谱特征变量和氮、磷、钾等农学参数与柑橘产量的重要关系,及其对柑橘估产研究的重要作用.
The canopy spectral data and the nitrogen, phosphorus, potassium and other agronomic parameters at different phases were analyzed by regression analysis with the measured yield data to construct the citrus production estimation model at different phases, and the comparative analysis of different time-phase models to explore the best remote sensing Yield model.The results showed that the spectral reflectance peak depth, the normalized difference index of green peak, the position of the red edge and other spectral variables such as spectral characteristics of nitrogen and nitrogen in the four growth stages of the citrus (germination, flowering, maturity and dormancy) (P <0.05). The compound regression model constructed by the combination of agronomic parameters such as phosphorus, potassium, etc. showed the best estimation effect. The test results of citrus measured yield showed that in 2010 (R2 = 0.965, Re = 2.540) and 2011 (R2 = 0.975, Re = 2.163) The results show that the estimated yield model is reliable.At the same time, the spectral characteristic variables such as reflectance peak depth, green peak normalized difference index and red edge position, and other agronomic parameters such as N, P, K and citrus yield The important relationship between the citrus and its assessment of the important role.