基于MODIS数据的玉米植被参数估算方法的对比分析

Comparison of Methods for Estimating Corn Canopy Parameters Based on MODIS Data

基于实测数据建立了FPAR、LAI的植被指数估算模型(NDVI、RVI、NDWI),并将其应用于MODIS BRDF数据对德惠地区玉米FPAR、LAI进行估算,然后将MODIS 15A2 FPAR/LAI产品值分别与BRDF估算值、地面实测值进行对比分析。主要得出以下结论:植被指数NDVI、RVI都能较好地用于实测数据和MODIS BRDF数据的FPAR、LAI估算;NDWI虽然在实测数据中估算玉米FPAR、LAI的效果优于NDVI、RVI,但其应用于MODIS BRDF数据估算FPAR、LAI时,效果却较差。BRDF数据估算FPAR与MODIS 15A2 FPAR值的关系因生长时期不同而异,在玉米生长前期,前者高于后者,而生长后期两者却较相近;BRDF估算LAI值一直都高于MODIS 15A2LAI产品值。生长季前期,MOD15A2 FPAR、LAI值接近实测值,而在后期却高于实测值。通过分析也表明,玉米苗期MODIS 15A2 FPAR数值变化范围较小,产品算法对实际FPAR变化尚不够敏感,这可能是影响MODIS FPAR产品精度的一个原因。

Based on the field measured data of corn,this paper established FPAR and LAI estimated models of the vegetation indices（NDVI,RVI,NDWI）,which were used to calculate FPAR and LAI with the MODIS BRDF （Bi-directional Reflectance Distribution Function） product of two periods. MODIS 15A2 FPAR/ LAI,FPAR/LAI estimated by the vegetation indices models and BRDF data, and field measured FPAR/ LAI data were compared. It can be concluded that NDVI, RVI were efficient to estimate FPAR and LAI for measured data and remote sensing data,NDWI was good for measured data but not for remote sensing data for atmospheric effect. The relation between FPAR/LAI estimated by BRDF data and the MODIS product was dependant on the growth periods,the vegetation indices estimated FPAR and LAI were larger than the product in the prophase of corn growing, but had the close value in the anaphase. The MOD15A2 FPAR and LAI are close to the measured data in the prophase, but smaller than the field data in the anaphase. The results showed that MODIS 15A2 FPAR variation was not sensitive enough for the field FPAR changes, which maybe a factor of restricting FPAR product precision.