基于多光谱与高光谱遥感数据的冬小麦叶面积指数反演比较

发布时间：2018-03-09 20:52

本文选题：植被　切入点：遥感　出处：《农业工程学报》2016年03期 　论文类型：期刊论文

【摘要】：近年来,高光谱遥感数据广泛应用于农作物叶面积指数(LAI)反演。与常用的多光谱遥感数据相比,高光谱数据能否提高农作物LAI反演的精度和稳定性还存在争议。针对这一问题,该研究利用实测冬小麦冠层高光谱反射率数据,构造了不同光谱分辨率和波段组合的5种光谱数据。基于ACRM(a two-layer canopy reflectance model)模型、2套参数化方案及上述5种光谱数据,对冬小麦LAI进行反演,分析光谱分辨率、高光谱数据波段选择、模型参数不确定性3方面因素对LAI反演精度与稳定性的影响。研究结果表明:当波段选择适宜、模型参数不确定性较小且光谱数据分辨率较高时,LAI反演精度与稳定性更高,提高光谱分辨率对LAI反演精度的改进作用随光谱分辨率的升高而降低;反之,当高光谱数据波段选择不当或者模型参数不确定性较大时,提高光谱数据的分辨率并未提高LAI反演精度。该研究解释了"高光谱遥感数据能否提高植被参数反演精度"问题,为进一步发挥高光谱数据在农作物LAI反演中的潜力提供了科学参考。
[Abstract]:In recent years, hyperspectral remote sensing data have been widely used in crop leaf area index (Lai) inversion. Compared with conventional multispectral remote sensing data, whether hyperspectral data can improve the precision and stability of crop LAI inversion is controversial. In this study, five spectral data of different spectral resolution and band combination were constructed by using the measured data of canopy hyperspectral reflectance of winter wheat. Based on ACRM(a two-layer canopy reflectance model, two sets of parameterized schemes and the above five spectral data were constructed. The effects of spectral resolution, band selection of hyperspectral data and uncertainty of model parameters on the inversion accuracy and stability of winter wheat LAI were analyzed. When the model parameter uncertainty is low and the spectral data resolution is high, the accuracy and stability of Lai inversion are higher. The improvement of spectral resolution on LAI inversion accuracy decreases with the increase of spectral resolution. When the band selection of hyperspectral data is improper or the model parameters are uncertain, improving the resolution of spectral data does not improve the precision of LAI inversion. This study explains whether hyperspectral remote sensing data can improve the precision of vegetation parameter inversion. It provides a scientific reference for further exerting the potential of hyperspectral data in crop LAI inversion.
【作者单位】：农业部农业信息技术重点实验室/中国农业科学院农业资源与农业区划研究所;华中师范大学城市与环境科学学院;
【基金】：测绘地理信息公益性行业科研专项(201512028) 国家自然科学基金项目(41271112)
【分类号】：S512.11;S127

【参考文献】