基于高分辨率遥感影像的玉米田叶面积指数反演

根据玉米田样地生物物理参量的野外实测数据,对PROSAIL模型输入参数的取值范围进行率定。通过PROSAIL模型参数敏感性分析,确定不同的输入参数设置方案,模拟不同的叶面积指数、叶倾角、叶绿素含量对应的玉米冠层反射率,建立叶面积指数的缨帽三角分布模式,从而获得玉米田红光-近红外波段反射率-LAI查找表,选取宁夏中卫市WV-3高分辨率遥感影像对玉米种植区域LAI进行反演。通过与实测数据比较,分析了PROSAIL模型在高分辨率遥感影像农作物LAI反演方面的适用性,为高分辨率遥感影像反演农作物LAI提供了方法参考。结果表明,PROSAIL模型输入参数的范围率定与不同设置方案的确定是有必要的,并且运用该查找表从WV-3影像反演的LAI与实测数据较一致,查找表均方根误差为0.47,LAI反演均方根误差为0.24。研究表明,该方法在利用WV-3遥感影像进行玉米田LAI反演中具有较强的适用性,能够进行准确有效的大面积叶面积指数遥感反演。

Inversion of maize field leaf area index based on high-resolution remote sensing image

According to the field data of maize field plots, the range of input parameters of PROSAIL model was determined. Through the PROSAIL model parameters sensitivity analysis, the different input parameters were determined for setting scheme. We simulated different maize canopy reflectance with different leaf area index, leaf inclination and chlorophyll content, and the tasseled-cap triangle distribution model of leaf area index was established to obtain red-near-infrared band reflectance-LAI look-up table of maize field. The LAI of maize planting area was retrieved by high resolution remote sensing image of WV-3 in Zhongwei City, Ningxia. The applicability of the PROSAIL model in LAI inversion of high resolution remote sensing images was analyzed by comparing with the measured data, which provided a reference for high resolution remote sensing image inversion of crop LAI. The results showed that it was necessary to determine the range of the input parameters and the different setting scheme. The LAI of WV-3 image was consistent with the measured data using the look-up table. The mean square error of the look-up table was 0.47, and the mean square error of simulated LAI was 0.24. This study showed that the methods has strong applicability in maize field LAI inversion using WV-3 remote sensing image, and can be used for accurate and effective leaf area index remote sensing inversion in large area.