论文部分内容阅读
高光谱遥感技术因其便捷快速、节约成本、非破坏性和准确度高的特点,被广泛应用于资源环境、军事、大气遥感和测绘等领域。本研究在化学实验(碳库分离)的基础上,结合高光谱遥感技术,利用统计分析工具SAS分析家模块将三库分离数据与光谱反射率数据进行多元逐步回归分析,筛选出敏感波段,从而建立起了一种森林土壤惰性碳和缓效性碳的快速估算模型。研究发现,无截距模型的显著性较为明显,解释能力和预测能力较好。利用光谱求平均和平滑去噪方法处理过的光谱反射率数据通过逐步回归筛选出来的敏感波段组合对森林土壤惰性碳和缓效性碳含量有较强的预测能力。可以利用高光谱技术来进行森林土壤惰性碳含量和缓效性碳含量的快速估算与监测。
Hyperspectral remote sensing technology is widely used in resources and environment, military, atmospheric remote sensing and surveying and mapping for its convenience, speed, cost saving, non-destructiveness and high accuracy. Based on the chemical experiments (carbon sequestration), combined with hyperspectral remote sensing technology, the SAS analytical module was used to conduct multiple stepwise regression analysis of the data and spectral reflectance data of the three reservoirs to screen the sensitive bands A rapid estimation model of inert carbon and slow-acting carbon in forest soils was established. The study found that the significant non-intercept model is obvious, good ability to explain and predict. Spectral reflectance data processed by spectral averaging and smoothing denoising methods have strong predictive ability to predict the amount of inert and slow-acting carbon in forest soils through the combination of sensitive bands screened by stepwise regression. Hyperspectral techniques can be used to rapidly estimate and monitor the inert and carbon content of forest soils.