论文部分内容阅读
土壤理化性质影响土壤质量,直接决定作物的产量,极易受到灌溉的影响。选择新疆典型绿洲———渭干河-库车河三角洲绿洲作为靶区,利用土壤光谱反射率预测土壤的电导率、pH值。首先,对土壤光谱反射率做变换,得到18种形式的反射率;其次,对18种形式的反射率与土壤电导率、pH值进行相关与回归分析,得到预测方程;最后,验证预测方程的精度,并确定最佳方程。结果显示:可以用土壤的光谱反射率预测土壤电导率、pH值,土壤电导率的预测方程为反射率的一阶导数微分形式,均方根误差为0.184;土壤pH值的预测方程为倒数的二阶导数微分形式,均方根误差为0.278。快速预测土壤电导率、pH值可以为土壤质量的评价提供数据基础,有利于正确有效地指导农业生产。
Soil physical and chemical properties affect soil quality, directly determine the crop yield, easily affected by irrigation. The typical oasis in Xinjiang --- Weigan River - Kuqa River delta oasis was chosen as target area, soil electrical conductivity and pH value were predicted by spectral reflectance of soil. First of all, the spectral reflectance of the soil is transformed to obtain 18 forms of reflectance. Secondly, correlation and regression analysis of the 18 forms of reflectance with the soil conductivity and pH are conducted to obtain the predictive equation. Finally, the predictive equation Accuracy and determine the best equation. The results show that the prediction equation of soil electrical conductivity, soil pH and soil conductivity can be predicted by the spectral reflectance of soil. The root mean square error of reflectance is 0.184. The prediction equation of soil pH is reciprocal The derivative of the second derivative has a root mean square error of 0.278. Rapid prediction of soil conductivity and pH value can provide a data basis for the evaluation of soil quality and help to guide agricultural production correctly and effectively.