尺度问题是森林叶面积指数(leaf area index,LAI)遥感反演的研究热点.以亚热带次生林与人工林为主要研究对象,利用典型样地LAI实测值和环境卫星(HJ1B-CCD2)数据进行LAI反演,通过升尺度获得1 000 m分辨率的平均叶面积指数(LAIa,1 000)和表观叶面积指数(LAIp,1 000),并与中分辨率成像光谱仪(MODIS)LAI产品进行比对,分析地表异质性和植被长势差异引起的MODIS LAI尺度效应.结果表明,MODIS LAI与LAIa,1 000和LAIp,1 000相关性较好(R2分别为0.302 1和0.505 7),MODIS LAI能够反映亚热带森林叶面积指数空间分布特征,但是MODIS LAI会出现明显的低估,MODIS LAI与LAIp,1 000、LAIa,1 000的平均差值分别为0.99、1.63利用混合像元内的植被面积比例和植被长势方差量纠正MODIS LAI的尺度效应,可以更好地描述森林实际生长情况.
The scaling issues have been paid more and more attention for the leaf area index (LAI) derived from remote sensing. In this paper, the subtropical secondary forestry and artificial forest was selected as the study area to investigate the scale effect of MODIS LAI products caused by the spatial heterogeneity and the difference of vegetation growth. Firstly, the measured LAI of the typical sampling plots was transformed into the average LAI (LAIa,1000 )and apparent LAI (LAIp,looo )at 1 km resolution scale with the aid of HJ1B-CCD2 data. Then, LAIa,1000 and LAIp,1000 was used to validate MODIS LAI at 1 km resolution. The results show that MODIS LAI presents strong correlations with LAIa.1000 or LAIp,1000 , suggesting MODIS LAI can reflect the spatial distribution characteristics of LAI in subtropi- cal forest, but the average difference between MODIS LAI and LAIa,1000 , LAIp,1000 is 0.99 and 1.63, respectively, showing a serious underestimation with MODIS LAI. In order to better describe the true growth of subtropical forest, this scale effect of MODIS LAI can be corrected by using the ratio of vege- tation area and the variance of vegetation growth in the mixed pixel.